Benchmark · pre-generated, always available← back to leaderboard

Migration Readiness Report

cupy/cupy

51/ 100 ready

Manual blockers present

1. Analyze the Request:
Repo:* cupy/cupy
ROCm readiness score:* 51/100 (Manual blockers present)
Findings:* A (works as-is): 43, B (mechanical): 2044, C (manual blocker): 12
Task:* Write a 3-4 sentence advisory summary of ROCm readiness and first steps.
Constraints:* Advisory tone (never claim auto-fixed), concrete, concise, technically accurate, 3-4 sentences.

2. Analyze the Data:
* Score 51/100 indicates a moderate readiness, but heavily skewed by mechanical issues.
* Bucket A (43): Small number of files/kernels work as-is.
* Bucket B (2044): Huge number of mechanical replacements needed (e.g., cuda -> hip, header changes, API renames). CuPy actually has a HIP backend, but maybe this scan was against the CUDA codebase without considering the HIP backend, or it's evaluating the raw CUDA code. Regardless, the scan shows 2044 mechanical blockers.
* Bucket C (12): 12 manual blockers. These are the critical path items (e.g., unsupported intrinsics, complex kernel logic, missing ROCm equivalents).

3. Drafting the Summary:
Sentence

43Works as-is

2044Mechanical change

12Manual blocker

Findings

2099

Python files

743

Files scanned

1187

Custom CUDA kernels

detected

Detected but out of scope (not analyzed): C, C/C++ header

Findings by file

2099 findings · 621 files

cupy_backends/cuda/libs/__init__.py· 1

cuda-python low-level driver API

from cuda import pathfinder as _pathfinder

Fireworks AI:9

cuda.pathfinder is a cuda-python utility for discovering CUDA shared libraries (libcudart, libcublas, etc.) on the system. On ROCm there is no equivalent package; AMD libraries are typically located via ROCM_PATH or standard system paths. This import will fail on ROCm and must be replaced with ROCm-aware library discovery or guarded behind a backend conditional.

RecommendRewrite driver-API calls against the HIP runtime.

Suggested change · advisory

--- a/cupy_backends/cuda/libs/__init__.py
+++ b/cupy_backends/cuda/libs/__init__.py
@@ -9,1 +9,7 @@
-from cuda import pathfinder as _pathfinder
+# Advisory: cuda.pathfinder has no ROCm equivalent. Guard or replace
+# with ROCm library discovery (e.g., via HIP_PATH / ROCM_PATH).
+try:
+    from cuda import pathfinder as _pathfinder
+except ImportError:
+    _pathfinder = None  # ROCm path: use os.environ.get('ROCM_PATH', '/opt/rocm')

cupy/_environment.py· 8

cuda-python low-level driver API

from cuda.pathfinder import (

Fireworks AI:157

The cuda.pathfinder module is part of the cuda-python package and is used to locate CUDA toolkit installations at runtime. On ROCm, there is no direct equivalent; path discovery is typically done via the ROCM_PATH environment variable or the default /opt/rocm location. This import and any downstream usage must be replaced with ROCm-specific path resolution logic.

RecommendRewrite driver-API calls against the HIP runtime.

Suggested change · advisory

--- a/cupy/_environment.py
+++ b/cupy/_environment.py
@@ -157,1 +157,1 @@
-from cuda.pathfinder import (
+# Advisory: cuda.pathfinder has no ROCm equivalent; replace with ROCm path
+# discovery logic, e.g. os.environ.get('ROCM_PATH', '/opt/rocm')
+# from cuda.pathfinder import (

cuda-python low-level driver API

from cuda.pathfinder import find_nvidia_binary_utility

Fireworks AI:201

This import relies on the NVIDIA-specific cuda.pathfinder package to locate NVIDIA binary utilities, which has no ROCm equivalent. On AMD Instinct targets, this pathfinding logic must be replaced with ROCm-specific tool discovery (e.g., via hipconfig or rocm-path) or guarded behind a backend conditional so the module does not fail at import time.

RecommendRewrite driver-API calls against the HIP runtime.

Suggested change · advisory

--- a/cupy/_environment.py
+++ b/cupy/_environment.py
@@ -199,7 +199,11 @@
 
 # Advisory only: replace NVIDIA-specific pathfinder with ROCm equivalent
-from cuda.pathfinder import find_nvidia_binary_utility
+try:
+    from cuda.pathfinder import find_nvidia_binary_utility
+except ImportError:
+    # ROCm migration: no cuda.pathfinder equivalent; use hipconfig/rocm-path
+    find_nvidia_binary_utility = None

cuda-python low-level driver API

from cuda.pathfinder import (

Fireworks AI:403

The cuda.pathfinder module is a cuda-python utility for locating CUDA toolkit installations and shared libraries at runtime. On ROCm there is no direct equivalent package; path discovery must be replaced with ROCm-specific logic (e.g., ROCM_PATH env var or /opt/rocm fallback). This is a runtime/environment change that affects how backend libraries are located, not just a simple import swap.

RecommendRewrite driver-API calls against the HIP runtime.

Suggested change · advisory

--- a/cupy/_environment.py
+++ b/cupy/_environment.py
@@ -403,7 +403,11 @@
- from cuda.pathfinder import (
-     _is_windows, find_cuda_include_dir, find_nvidia_dynamic_lib_dir,
- )
+ # Advisory: cuda.pathfinder has no ROCm equivalent. Replace with
+ # ROCm path discovery using ROCM_PATH env var or /opt/rocm fallback.
+ import os
+ _rocm_path = os.environ.get('ROCM_PATH', '/opt/rocm')
+ # TODO: replace find_cuda_include_dir / find_nvidia_dynamic_lib_dir
+ # with ROCm-specific include/lib path resolution under _rocm_path.

Device string "cuda"

cuda_version = config['cuda']

template:419

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

cupy_cuda_lib_path, config['cuda'], lib, version, x,

template:435

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

_get_cutensor_from_wheel(min_pypi_version, config['cuda']) +

template:441

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

_get_nccl_from_wheel(min_pypi_version, config['cuda']) +

template:446

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

cuda = config['cuda']

template:572

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

cupy/cuda/__init__.py· 51

CuPy dependency

import cupy as _cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._environment import get_cuda_path # NOQA

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._environment import get_nvcc_path # NOQA

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._environment import get_rocm_path # NOQA

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._environment import get_hipcc_path # NOQA

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import compiler # NOQA

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device # NOQA

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import function # NOQA

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import memory # NOQA

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import memory_hook # NOQA

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import memory_hooks # NOQA

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import pinned_memory # NOQA

template:19

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import profiler # NOQA

template:20

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import stream # NOQA

template:21

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import texture # NOQA

template:22

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cub # NOQA

template:38

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import thrust # NOQA

template:47

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy.cuda.jitify as jitify

template:71

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cuda-python low-level driver API

from cuda import pathfinder

Fireworks AI:78

cuda-python's pathfinder module is a CUDA-specific utility for locating CUDA shared libraries at runtime; there is no ROCm equivalent. On AMD Instinct, library discovery should use ROCM_PATH or rpath-based resolution instead, so this import must be removed or stubbed behind a HIP/ROCm code path.

RecommendRewrite driver-API calls against the HIP runtime.

Suggested change · advisory

--- a/cupy/cuda/__init__.py
+++ b/cupy/cuda/__init__.py
@@ -75,7 +75,11 @@
 
 # Advisory: `cuda.pathfinder` has no ROCm equivalent.
 # Replace with ROCM_PATH-based discovery or guard behind a CUDA-only path.
-from cuda import pathfinder
+try:
+    from cuda import pathfinder
+except ImportError:
+    # ROCm migration: no pathfinder equivalent; use ROCM_PATH env or rpath.
+    pathfinder = None

CuPy dependency

from cupy.cuda.device import Device # NOQA

template:127

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.device import get_cublas_handle # NOQA

template:128

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.device import get_device_id # NOQA

template:129

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.function import Function # NOQA

template:130

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.function import Module # NOQA

template:131

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory import alloc # NOQA

template:132

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory import BaseMemory # NOQA

template:133

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory import malloc_managed # NOQA

template:134

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory import malloc_async # NOQA

template:135

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory import ManagedMemory # NOQA

template:136

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory import Memory # NOQA

template:137

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory import MemoryAsync # NOQA

template:138

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory import MemoryPointer # NOQA

template:139

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory import MemoryPool # NOQA

template:140

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory import MemoryAsyncPool # NOQA

template:141

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory import PythonFunctionAllocator # NOQA

template:142

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory import CFunctionAllocator # NOQA

template:143

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory import set_allocator # NOQA

template:144

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory import get_allocator # NOQA

template:145

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory import UnownedMemory # NOQA

template:146

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory_hook import MemoryHook # NOQA

template:147

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.pinned_memory import alloc_pinned_memory # NOQA

template:148

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.pinned_memory import PinnedMemory # NOQA

template:149

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.pinned_memory import PinnedMemoryPointer # NOQA

template:150

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.pinned_memory import PinnedMemoryPool # NOQA

template:151

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.pinned_memory import set_pinned_memory_allocator # NOQA

template:152

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.stream import Event # NOQA

template:153

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.stream import get_current_stream # NOQA

template:154

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.stream import get_elapsed_time # NOQA

template:155

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.stream import Stream # NOQA

template:156

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.stream import ExternalStream # NOQA

template:157

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.graph import Graph # NOQA

template:158

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/cuda/compiler.py· 12

CuPy dependency

from cupy.cuda import device

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import function

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda._compiler_cache import (

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _environment

template:24

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _util

template:25

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import core

template:203

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import get_cuda_path

template:268

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import jitify

template:311

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import core

template:316

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import get_nvcc_path

template:442

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

Inline PTX assembly

options = options + ('-o', 'preprocess.ptx')

Fireworks AI:542

PTX is NVIDIA-specific IR and has no direct ROCm equivalent; the compilation pipeline must target AMDGPU ISA via clang/llvm instead of nvcc/ptxas. This option string is part of a CUDA-only preprocessing/compilation path that will not function on AMD Instinct GPUs and must be replaced with ROCm's compiler invocation (e.g., clang with -target amdgcn-amd-amdhsa).

RecommendRewrite the PTX block in HIP or a portable high-level equivalent.

Suggested change · advisory

--- a/cupy/cuda/compiler.py
+++ b/cupy/cuda/compiler.py
@@ -542,7 +542,11 @@
-        options = options + ('-o', 'preprocess.ptx')
+        # Advisory: PTX is NVIDIA-only; on ROCm use clang targeting AMDGPU.
+        # Replace PTX output with AMDGPU ISA / bitcode output path.
+        if _is_rocm():
+            options = options + ('-o', 'preprocess.amdgcn')
+        else:
+            options = options + ('-o', 'preprocess.ptx')

Inline PTX assembly

ls.add_ptr_data(ptx, 'cupy.ptx')

Fireworks AI:693

PTX is NVIDIA-specific intermediate assembly with no direct ROCm equivalent; AMD uses GCN/AMDGPU ISA or HIP IR instead. This code path in the compiler module will not function on AMD Instinct GPUs and must be redirected to a HIP/ROCm compilation pipeline or guarded behind a backend check.

RecommendRewrite the PTX block in HIP or a portable high-level equivalent.

Suggested change · advisory

--- a/cupy/cuda/compiler.py
+++ b/cupy/cuda/compiler.py
@@ -690,7 +690,11 @@
 
-    ls.add_ptr_data(ptx, 'cupy.ptx')
+    # Advisory: PTX is NVIDIA-only; on ROCm, use HIP/AMDGPU ISA instead.
+    # Consider branching on backend (CUDA vs HIP/ROCm) here.
+    if not _is_rocm_backend():
+        ls.add_ptr_data(ptx, 'cupy.ptx')
+    else:
+        raise RuntimeError('PTX path not supported on ROCm; use HIP compilation instead')

cupyx/_runtime.py· 7

CuPy dependency

import cupy

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cuda-python low-level driver API

from cuda import pathfinder

Fireworks AI:17

The cuda.pathfinder module is part of cuda-python and is used to locate CUDA shared libraries at runtime; there is no direct ROCm equivalent. This import must be removed or replaced with ROCm-specific library discovery (e.g., standard ROCm paths under /opt/rocm or environment variables like ROCM_PATH).

RecommendRewrite driver-API calls against the HIP runtime.

Suggested change · advisory

--- a/cupyx/_runtime.py
+++ b/cupyx/_runtime.py
@@ -14,7 +14,6 @@
 
-# Advisory: cuda.pathfinder has no ROCm equivalent; remove or replace
-# with ROCm library path discovery (e.g., os.environ.get('ROCM_PATH', '/opt/rocm'))
-from cuda import pathfinder
+# from cuda import pathfinder  # Removed: no ROCm equivalent
+# TODO: replace with ROCm library discovery if runtime path lookup is needed

cuda-python low-level driver API

import cuda.bindings

Fireworks AI:150

The cuda.bindings import is the cuda-python low-level driver/runtime binding package; on ROCm the analog is hipPython's rocm.bindings module. This is a bucket-C low-level API surface, so all downstream driver calls (e.g., cuDriverGetVersion, cuDeviceGet, cuMemAlloc) must be mapped to their HIP equivalents (hipDriverGetVersion, hipDeviceGet, hipMalloc) — not just the import line. Verify the hipPython package version exposes the same binding surface before committing the change.

RecommendRewrite driver-API calls against the HIP runtime.

Suggested change · advisory

--- a/cupyx/_runtime.py
+++ b/cupyx/_runtime.py
@@ -150,1 +150,1 @@
-import cuda.bindings
+import rocm.bindings  # advisory: hipPython analog of cuda.bindings

cuda-python low-level driver API

import cuda

Fireworks AI:153

The import cuda statement pulls in cuda-python's low-level driver API bindings, which have no direct ROCm equivalent package. On ROCm, this import will fail at runtime and must either be replaced with a HIP/ROCm Python binding (e.g., rhipy or amdsmi where applicable) or guarded behind a backend-detection conditional so the module remains importable on AMD Instinct systems.

RecommendRewrite driver-API calls against the HIP runtime.

Suggested change · advisory

--- a/cupyx/_runtime.py
+++ b/cupyx/_runtime.py
@@ -153,1 +153,5 @@
-import cuda
+# Advisory: cuda-python has no ROCm equivalent; guard or replace.
+try:
+    import cuda
+except ImportError:
+    cuda = None  # ROCm: use rhipy/amdsmi bindings where needed

CuPy dependency

from cupy.cuda import cufft

template:182

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy.cuda.thrust as thrust

template:223

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy.cuda.jitify as jitify

template:268

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_include.py· 2

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cuda-python low-level driver API

from cuda.pathfinder import find_nvidia_header_directory

Fireworks AI:57

cuda.pathfinder is an NVIDIA-specific utility for locating CUDA headers and has no ROCm equivalent. On AMD Instinct, this import will fail at collection time, so the test should be guarded or skipped for ROCm builds. Advisory only — verify whether this test is relevant to the ROCm path before applying.

RecommendRewrite driver-API calls against the HIP runtime.

Suggested change · advisory

--- a/tests/cupy_tests/core_tests/test_include.py
+++ b/tests/cupy_tests/core_tests/test_include.py
@@ -54,6 +54,9 @@
 
 @pytest.mark.skipif(
+    os.environ.get('HIP_PLATFORM') == 'amd',
+    reason='cuda.pathfinder is NVIDIA-specific; no ROCm equivalent')
+@pytest.mark.skipif(
     not cuda_path_available,
     reason='cuda.pathfinder is not available')
 def test_include_cupy_complex():

tests/cupy_tests/core_tests/test_raw.py· 8

CuPy dependency

import cupy

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _util

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _accelerator

template:19

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import compiler

template:20

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import _compiler_cache

template:21

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import memory

template:22

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

Inline PTX assembly

file_path += '.ptx'

Fireworks AI:632

PTX is NVIDIA-specific intermediate assembly and is not executable on AMD Instinct GPUs; ROCm uses AMDGPU ISA / code objects (e.g., .hsaco or .co). This test path that loads a .ptx file will fail on ROCm unless a corresponding ROCm code object is provided or the test is conditionally skipped.

RecommendRewrite the PTX block in HIP or a portable high-level equivalent.

Suggested change · advisory

--- tests/cupy_tests/core_tests/test_raw.py
+++ tests/cupy_tests/core_tests/test_raw.py
@@ -629,7 +629,11 @@
         # ... existing context ...
-        file_path += '.ptx'
+        if cupy.cuda.runtime.is_hip:
+            # Advisory: PTX is NVIDIA-only; use a ROCm code object instead
+            file_path += '.hsaco'
+        else:
+            file_path += '.ptx'

cupy/__init__.py· 365

CuPy dependency

from cupy import _environment

Fireworks AI:8

CuPy is a CUDA-centric array library; this internal _environment import sets up CUDA paths and runtime detection. For ROCm, you should use the cupy-rocm build (or CuPy's HIP backend) where _environment is adapted for HIP, or replace CuPy with an ROCm-compatible alternative such as PyTorch on HIP or rocarray.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

Suggested change · advisory

--- cupy/__init__.py
+++ cupy/__init__.py
@@ -5,7 +5,11 @@
 # Advisory: CuPy's _environment module is CUDA-specific. On ROCm,
 # use the cupy-rocm build where this module is HIP-aware, or guard
 # the import so non-CUDA backends do not fail.
-from cupy import _environment
+try:
+    from cupy import _environment
+except ImportError:
+    # ROCm/HIP backend may not provide _environment; skip CUDA-specific setup
+    pass

CuPy dependency

from cupy import _version

Fireworks AI:9

CuPy is a CUDA-oriented array library, but it does have an experimental ROCm/HIP backend that can be installed via cupy-rocm-* packages. No source change is strictly required if the ROCm-compatible build of CuPy is installed in the environment; otherwise the dependency must be replaced with an alternative such as PyTorch or Numba-ROCm.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core # NOQA

Fireworks AI:18

CuPy itself ships CUDA-specific and ROCm-specific wheels; the import path from cupy import _core is unchanged, but the underlying binary must be the ROCm build (cupy-rocm) rather than the default CUDA build. No source edit is needed here—migration impact is purely environmental (install the correct wheel matching the AMD Instinct ROCm version).

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda # NOQA

Fireworks AI:30

CuPy is a CUDA-centric array library; on ROCm you must use the ROCm-enabled CuPy build (cupy-rocm) which provides a cupy.cuda compatibility shim backed by HIP. No source change is strictly required if the ROCm CuPy wheel is installed, but verify that cupy.cuda APIs your code relies on are implemented in the HIP backend.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

Suggested change · advisory

--- a/cupy/__init__.py
+++ b/cupy/__init__.py
@@ -30,1 +30,3 @@
-from cupy import cuda  # NOQA
+# Advisory: ensure cupy-rocm is installed; cupy.cuda is shimmed to HIP in ROCm builds.
+# If unavailable, fall back to a conditional import or raise a clear RuntimeError.
+from cupy import cuda  # NOQA

CuPy dependency

from cupy import fft # NOQA

Fireworks AI:50

CuPy itself supports ROCm when built against HIP (installed as cupy-rocm), so the from cupy import fft import can remain unchanged at the source level. The migration impact is purely environmental: the build/installation must target ROCm rather than CUDA, and runtime behavior of FFT routines should be validated against ROCm's hipFFT backend.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import linalg # NOQA

Fireworks AI:51

CuPy itself is a CUDA-centric library, but it has experimental ROCm/HIP support when built with CUPYINSTALLUSE_HIP=1 against AMD's ROCm stack. The import line itself likely needs no source change, but the runtime/build environment must switch to a HIP-enabled CuPy build or be replaced with an alternative (e.g., PyTorch-ROCm, Numba-ROCm, or direct PyHIP) if CuPy's ROCm backend is insufficient.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import polynomial # NOQA

Fireworks AI:52

CuPy itself must be built against ROCm/HIP rather than CUDA; the Python-level import path (cupy.polynomial) remains unchanged, but the underlying library binary and runtime backend differ. No source-level edit to this import line is needed—migration impact is at the build/install layer (install cupy-rocm or build CuPy from source with -DHIP=ON).

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import random # NOQA

Fireworks AI:53

This is an internal CuPy import within CuPy's own package; the import statement itself does not change for ROCm. CuPy can be built against ROCm/HIP, so the migration impact is at the build/installation level (installing a ROCm-enabled CuPy wheel or building from source with HIP), not a source-level edit to this line.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import ndarray # NOQA

template:56

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import ufunc # NOQA

template:57

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.basic import astype # NOQA

template:191

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.basic import empty # NOQA

template:192

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.basic import empty_like # NOQA

template:193

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.basic import eye # NOQA

template:194

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.basic import full # NOQA

template:195

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.basic import full_like # NOQA

template:196

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.basic import identity # NOQA

template:197

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.basic import ones # NOQA

template:198

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.basic import ones_like # NOQA

template:199

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.basic import zeros # NOQA

template:200

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.basic import zeros_like # NOQA

template:201

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.from_data import copy # NOQA

template:203

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.from_data import array # NOQA

template:204

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.from_data import asanyarray # NOQA

template:205

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.from_data import asarray # NOQA

template:206

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.from_data import ascontiguousarray # NOQA

template:207

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.from_data import fromfile # NOQA

template:208

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.from_data import fromfunction # NOQA

template:209

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.from_data import fromiter # NOQA

template:210

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.from_data import frombuffer # NOQA

template:211

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.from_data import fromstring # NOQA

template:212

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.from_data import loadtxt # NOQA

template:213

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.from_data import genfromtxt # NOQA

template:214

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.ranges import arange # NOQA

template:216

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.ranges import linspace # NOQA

template:217

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.ranges import logspace # NOQA

template:218

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.ranges import meshgrid # NOQA

template:219

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.ranges import mgrid # NOQA

template:220

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.ranges import ogrid # NOQA

template:221

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.matrix import diag # NOQA

template:223

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.matrix import diagflat # NOQA

template:224

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.matrix import tri # NOQA

template:225

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.matrix import tril # NOQA

template:226

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.matrix import triu # NOQA

template:227

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.matrix import vander # NOQA

template:228

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._functional.piecewise import piecewise # NOQA

template:233

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._functional.vectorize import vectorize # NOQA

template:234

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.lib._shape_base import apply_along_axis # NOQA

template:235

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.lib._shape_base import apply_over_axes # NOQA

template:236

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.lib._shape_base import put_along_axis # NOQA

template:237

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.basic import copyto # NOQA

template:242

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.shape import shape # NOQA

template:244

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.shape import ravel # NOQA

template:245

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.shape import reshape # NOQA

template:246

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.transpose import moveaxis # NOQA

template:248

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.transpose import rollaxis # NOQA

template:249

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.transpose import swapaxes # NOQA

template:250

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.transpose import transpose # NOQA

template:251

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.transpose import matrix_transpose # NOQA

template:252

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.dims import atleast_1d # NOQA

template:257

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.dims import atleast_2d # NOQA

template:258

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.dims import atleast_3d # NOQA

template:259

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.dims import broadcast # NOQA

template:260

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.dims import broadcast_arrays # NOQA

template:261

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.dims import broadcast_to # NOQA

template:262

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.dims import expand_dims # NOQA

template:263

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.dims import squeeze # NOQA

template:264

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.join import column_stack # NOQA

template:266

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.join import concatenate # NOQA

template:267

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.join import dstack # NOQA

template:268

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.join import hstack # NOQA

template:269

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.join import row_stack # NOQA

template:270

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.join import stack # NOQA

template:271

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.join import vstack # NOQA

template:272

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.kind import asarray_chkfinite # NOQA

template:277

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.kind import asfarray # NOQA

template:278

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.kind import asfortranarray # NOQA

template:279

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.kind import require # NOQA

template:280

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.split import array_split # NOQA

template:282

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.split import dsplit # NOQA

template:283

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.split import hsplit # NOQA

template:284

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.split import split # NOQA

template:285

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.split import vsplit # NOQA

template:286

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.tiling import repeat # NOQA

template:288

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.tiling import tile # NOQA

template:289

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.add_remove import delete # NOQA

template:291

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.add_remove import append # NOQA

template:292

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.add_remove import resize # NOQA

template:293

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.add_remove import unique # NOQA

template:294

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.add_remove import unique_all # NOQA

template:295

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.add_remove import unique_counts # NOQA

template:296

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.add_remove import unique_values # NOQA

template:297

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.add_remove import unique_inverse # NOQA

template:298

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.add_remove import trim_zeros # NOQA

template:300

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.rearrange import flip # NOQA

template:302

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.rearrange import fliplr # NOQA

template:303

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.rearrange import flipud # NOQA

template:304

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.rearrange import roll # NOQA

template:305

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation.rearrange import rot90 # NOQA

template:306

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._binary.elementwise import bitwise_and # NOQA

template:315

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._binary.elementwise import bitwise_or # NOQA

template:316

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._binary.elementwise import bitwise_xor # NOQA

template:317

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._binary.elementwise import bitwise_not # NOQA

template:318

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._binary.elementwise import bitwise_count # NOQA

template:319

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._binary.elementwise import invert # NOQA

template:320

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._binary.elementwise import left_shift # NOQA

template:321

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._binary.elementwise import right_shift # NOQA

template:322

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._binary.packing import packbits # NOQA

template:329

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._binary.packing import unpackbits # NOQA

template:330

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.core import min_scalar_type # NOQA

template:387

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.generate import c_ # NOQA

template:416

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.generate import indices # NOQA

template:417

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.generate import ix_ # NOQA

template:418

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.generate import mask_indices # NOQA

template:419

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.generate import tril_indices # NOQA

template:420

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.generate import tril_indices_from # NOQA

template:421

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.generate import triu_indices # NOQA

template:422

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.generate import triu_indices_from # NOQA

template:423

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.generate import r_ # NOQA

template:424

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.generate import ravel_multi_index # NOQA

template:425

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.generate import unravel_index # NOQA

template:426

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.indexing import choose # NOQA

template:428

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.indexing import compress # NOQA

template:429

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.indexing import diagonal # NOQA

template:430

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.indexing import extract # NOQA

template:431

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.indexing import select # NOQA

template:432

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.indexing import take # NOQA

template:433

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.indexing import take_along_axis # NOQA

template:434

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.insert import place # NOQA

template:436

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.insert import put # NOQA

template:437

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.insert import putmask # NOQA

template:438

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.insert import fill_diagonal # NOQA

template:439

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.insert import diag_indices # NOQA

template:440

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.insert import diag_indices_from # NOQA

template:441

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing.iterate import flatiter # NOQA

template:443

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._io.npz import load # NOQA

template:453

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._io.npz import save # NOQA

template:454

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._io.npz import savez # NOQA

template:455

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._io.npz import savez_compressed # NOQA

template:456

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._io.formatting import array_repr # NOQA

template:458

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._io.formatting import array_str # NOQA

template:459

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._io.formatting import array2string # NOQA

template:460

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._io.formatting import format_float_positional # NOQA

template:461

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._io.formatting import format_float_scientific # NOQA

template:462

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._io.text import savetxt # NOQA

template:464

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._einsum import einsum # NOQA

template:484

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._product import cross # NOQA

template:486

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._product import dot # NOQA

template:487

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._product import inner # NOQA

template:488

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._product import kron # NOQA

template:489

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._product import matmul # NOQA

template:490

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._product import outer # NOQA

template:491

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._product import tensordot # NOQA

template:492

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._product import vdot # NOQA

template:493

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._norms import trace # NOQA

template:495

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.comparison import allclose # NOQA

template:500

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.comparison import array_equal # NOQA

template:501

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.comparison import array_equiv # NOQA

template:502

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.comparison import isclose # NOQA

template:503

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.content import isfinite # NOQA

template:505

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.content import isinf # NOQA

template:506

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.content import isnan # NOQA

template:507

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.content import isneginf # NOQA

template:508

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.content import isposinf # NOQA

template:509

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.truth import in1d # NOQA

template:511

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.truth import isin # NOQA

template:512

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.type_testing import iscomplex # NOQA

template:514

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.type_testing import iscomplexobj # NOQA

template:515

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.type_testing import isfortran # NOQA

template:516

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.type_testing import isreal # NOQA

template:517

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.type_testing import isrealobj # NOQA

template:518

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.truth import in1d # NOQA

template:520

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.truth import intersect1d # NOQA

template:521

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.truth import isin # NOQA

template:522

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.truth import setdiff1d # NOQA

template:523

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.truth import setxor1d # NOQA

template:524

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.truth import union1d # NOQA

template:525

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.ops import logical_and # NOQA

template:536

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.ops import logical_not # NOQA

template:537

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.ops import logical_or # NOQA

template:538

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.ops import logical_xor # NOQA

template:539

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.comparison import equal # NOQA

template:541

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.comparison import greater # NOQA

template:542

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.comparison import greater_equal # NOQA

template:543

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.comparison import less # NOQA

template:544

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.comparison import less_equal # NOQA

template:545

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.comparison import not_equal # NOQA

template:546

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.truth import all # NOQA

template:548

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic.truth import any # NOQA

template:549

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.lib._polynomial import poly1d # NOQA

template:554

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.lib._routines_poly import poly # NOQA

template:555

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.lib._routines_poly import polyadd # NOQA

template:556

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.lib._routines_poly import polysub # NOQA

template:557

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.lib._routines_poly import polymul # NOQA

template:558

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.lib._routines_poly import polyfit # NOQA

template:559

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.lib._routines_poly import polyval # NOQA

template:560

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.lib._routines_poly import roots # NOQA

template:561

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.trigonometric import arccos # NOQA

template:566

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.trigonometric import arcsin # NOQA

template:567

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.trigonometric import arctan # NOQA

template:568

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.trigonometric import arctan2 # NOQA

template:569

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.trigonometric import cos # NOQA

template:577

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.trigonometric import deg2rad # NOQA

template:578

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.trigonometric import degrees # NOQA

template:579

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.trigonometric import hypot # NOQA

template:580

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.trigonometric import rad2deg # NOQA

template:581

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.trigonometric import radians # NOQA

template:582

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.trigonometric import sin # NOQA

template:583

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.trigonometric import tan # NOQA

template:584

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.trigonometric import unwrap # NOQA

template:585

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.hyperbolic import arccosh # NOQA

template:587

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.hyperbolic import arcsinh # NOQA

template:588

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.hyperbolic import arctanh # NOQA

template:589

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.hyperbolic import cosh # NOQA

template:590

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.hyperbolic import sinh # NOQA

template:591

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.hyperbolic import tanh # NOQA

template:592

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.rounding import around # NOQA

template:599

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.rounding import ceil # NOQA

template:600

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.rounding import fix # NOQA

template:601

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.rounding import floor # NOQA

template:602

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.rounding import rint # NOQA

template:603

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.rounding import round # NOQA

template:604

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.rounding import round_ # NOQA

template:605

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.rounding import trunc # NOQA

template:606

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.sumprod import prod # NOQA

template:608

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.sumprod import product # NOQA

template:609

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.sumprod import sum # NOQA

template:610

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.sumprod import cumprod # NOQA

template:611

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.sumprod import cumproduct # NOQA

template:612

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.sumprod import cumsum # NOQA

template:613

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.sumprod import ediff1d # NOQA

template:614

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.sumprod import nancumprod # NOQA

template:615

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.sumprod import nancumsum # NOQA

template:616

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.sumprod import nansum # NOQA

template:617

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.sumprod import nanprod # NOQA

template:618

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.sumprod import diff # NOQA

template:619

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.sumprod import gradient # NOQA

template:620

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.sumprod import trapezoid # NOQA

template:621

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.window import bartlett # NOQA

template:622

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.window import blackman # NOQA

template:623

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.window import hamming # NOQA

template:624

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.window import hanning # NOQA

template:625

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.window import kaiser # NOQA

template:626

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.explog import exp # NOQA

template:628

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.explog import exp2 # NOQA

template:629

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.explog import expm1 # NOQA

template:630

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.explog import log # NOQA

template:631

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.explog import log10 # NOQA

template:632

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.explog import log1p # NOQA

template:633

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.explog import log2 # NOQA

template:634

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.explog import logaddexp # NOQA

template:635

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.explog import logaddexp2 # NOQA

template:636

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.special import i0 # NOQA

template:638

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.special import sinc # NOQA

template:639

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.floating import copysign # NOQA

template:641

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.floating import frexp # NOQA

template:642

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.floating import ldexp # NOQA

template:643

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.floating import nextafter # NOQA

template:644

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.floating import signbit # NOQA

template:645

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.rational import gcd # NOQA

template:647

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.rational import lcm # NOQA

template:648

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import add # NOQA

template:650

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import divide # NOQA

template:651

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import divmod # NOQA

template:652

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import floor_divide # NOQA

template:653

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import float_power # NOQA

template:654

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import fmod # NOQA

template:655

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import modf # NOQA

template:656

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import multiply # NOQA

template:657

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import negative # NOQA

template:658

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import positive # NOQA

template:659

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import power # NOQA

template:660

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import reciprocal # NOQA

template:661

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import remainder # NOQA

template:662

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import remainder as mod # NOQA

template:663

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import subtract # NOQA

template:664

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import true_divide # NOQA

template:665

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import angle # NOQA

template:669

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import conjugate as conj # NOQA

template:670

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import conjugate # NOQA

template:671

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import imag # NOQA

template:672

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.arithmetic import real # NOQA

template:673

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import absolute as abs # NOQA

template:675

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import absolute # NOQA

template:676

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import cbrt # NOQA

template:677

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import clip # NOQA

template:678

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import fabs # NOQA

template:679

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import fmax # NOQA

template:680

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import fmin # NOQA

template:681

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import interp # NOQA

template:682

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import maximum # NOQA

template:683

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import minimum # NOQA

template:684

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import nan_to_num # NOQA

template:685

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import real_if_close # NOQA

template:686

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import sign # NOQA

template:687

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import heaviside # NOQA

template:688

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import sqrt # NOQA

template:689

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import square # NOQA

template:690

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import convolve # NOQA

template:691

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._misc.byte_bounds import byte_bounds # NOQA

template:696

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._misc.memory_ranges import may_share_memory # NOQA

template:697

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._misc.memory_ranges import shares_memory # NOQA

template:698

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._misc.who import who # NOQA

template:699

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._padding.pad import pad # NOQA

template:708

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.count import count_nonzero # NOQA

template:714

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.search import argmax # NOQA

template:716

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.search import argmin # NOQA

template:717

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.search import argwhere # NOQA

template:718

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.search import flatnonzero # NOQA

template:719

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.search import nanargmax # NOQA

template:720

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.search import nanargmin # NOQA

template:721

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.search import nonzero # NOQA

template:722

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.search import searchsorted # NOQA

template:723

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.search import where # NOQA

template:724

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.sort import argpartition # NOQA

template:726

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.sort import argsort # NOQA

template:727

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.sort import lexsort # NOQA

template:728

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.sort import msort # NOQA

template:729

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.sort import sort_complex # NOQA

template:730

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.sort import partition # NOQA

template:731

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting.sort import sort # NOQA

template:732

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.correlation import corrcoef # NOQA

template:737

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.correlation import cov # NOQA

template:738

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.correlation import correlate # NOQA

template:739

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.order import amax # NOQA

template:741

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.order import amax as max # NOQA

template:742

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.order import amin # NOQA

template:743

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.order import amin as min # NOQA

template:744

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.order import nanmax # NOQA

template:745

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.order import nanmin # NOQA

template:746

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.order import percentile # NOQA

template:747

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.order import ptp # NOQA

template:748

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.order import quantile # NOQA

template:749

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.meanvar import median # NOQA

template:751

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.meanvar import average # NOQA

template:752

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.meanvar import mean # NOQA

template:753

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.meanvar import std # NOQA

template:754

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.meanvar import var # NOQA

template:755

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.meanvar import nanmedian # NOQA

template:756

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.meanvar import nanmean # NOQA

template:757

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.meanvar import nanstd # NOQA

template:758

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.meanvar import nanvar # NOQA

template:759

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.histogram import bincount # NOQA

template:761

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.histogram import digitize # NOQA

template:762

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.histogram import histogram # NOQA

template:763

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.histogram import histogram2d # NOQA

template:764

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._statistics.histogram import histogramdd # NOQA

template:765

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import exceptions # NOQA

template:770

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import AxisError # NOQA

template:771

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import ComplexWarning # NOQA

template:772

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import ModuleDeprecationWarning # undocumented # NOQA

template:773

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import RankWarning # NOQA

template:774

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import TooHardError # NOQA

template:775

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import VisibleDeprecationWarning # NOQA

template:776

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import size # NOQA

template:782

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._util import clear_memo # NOQA

template:806

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._util import memoize # NOQA

template:807

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import ElementwiseKernel # NOQA

template:809

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import RawKernel # NOQA

template:810

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import RawModule # NOQA

template:811

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._reduction import ReductionKernel # NOQA

template:812

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._dtype import make_aligned_dtype # NOQA

template:814

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import fromDlpack # NOQA

template:820

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import from_dlpack # NOQA

template:821

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._kernel import _ufunc_doc_signature_formatter

template:1133

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_binary/elementwise.py· 1

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_binary/packing.py· 2

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_core/__init__.py· 75

CuPy dependency

from cupy._core import core # NOQA

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import fusion # NOQA

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal # NOQA

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._accelerator import set_elementwise_accelerators # NOQA

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._accelerator import set_reduction_accelerators # NOQA

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._accelerator import set_routine_accelerators # NOQA

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._accelerator import get_elementwise_accelerators # NOQA

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._accelerator import get_reduction_accelerators # NOQA

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._accelerator import get_routine_accelerators # NOQA

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._kernel import create_ufunc # NOQA

template:20

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._kernel import ElementwiseKernel # NOQA

template:21

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._kernel import ufunc # NOQA

template:22

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._kernel import _get_warpsize # NOQA

template:23

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._kernel import _full_mask # NOQA

template:24

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._kernel import _full_mask_hex # NOQA

template:25

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._kernel import _is_hip_7_plus # NOQA

template:26

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._reduction import create_reduction_func # NOQA

template:27

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._reduction import ReductionKernel # NOQA

template:28

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_binary import bitwise_and # NOQA

template:29

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_binary import bitwise_or # NOQA

template:30

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_binary import bitwise_xor # NOQA

template:31

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_binary import bitwise_count # NOQA

template:32

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_binary import invert # NOQA

template:33

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_binary import left_shift # NOQA

template:34

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_binary import right_shift # NOQA

template:35

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_linalg import _mat_ptrs # NOQA

template:36

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_linalg import dot # NOQA

template:37

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_linalg import get_compute_type # NOQA

template:38

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_linalg import matmul # NOQA

template:39

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_linalg import set_compute_type # NOQA

template:40

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_linalg import tensordot_core # NOQA

template:41

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_logic import create_comparison # NOQA

template:42

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_logic import equal # NOQA

template:43

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_logic import greater # NOQA

template:44

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_logic import greater_equal # NOQA

template:45

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_logic import less # NOQA

template:46

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_logic import less_equal # NOQA

template:47

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_logic import not_equal # NOQA

template:48

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_manipulation import array_split # NOQA

template:49

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_manipulation import broadcast # NOQA

template:50

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_manipulation import broadcast_to # NOQA

template:51

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_manipulation import concatenate_method # NOQA

template:52

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_manipulation import moveaxis # NOQA

template:53

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_manipulation import rollaxis # NOQA

template:54

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_manipulation import size # NOQA

template:55

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_math import absolute # NOQA

template:56

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_math import add # NOQA

template:57

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_math import angle, angle_deg # NOQA

template:58

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_math import conjugate # NOQA

template:59

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_math import divide # NOQA

template:60

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_math import floor_divide # NOQA

template:61

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_math import multiply # NOQA

template:62

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_math import negative # NOQA

template:63

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_math import positive # NOQA

template:64

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_math import power # NOQA

template:65

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_math import remainder # NOQA

template:66

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_math import sqrt # NOQA

template:67

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_math import subtract # NOQA

template:68

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_math import true_divide # NOQA

template:69

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_statistics import nanmax # NOQA

template:70

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_statistics import nanmin # NOQA

template:71

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.core import _internal_ascontiguousarray # NOQA

template:72

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.core import _internal_asfortranarray # NOQA

template:73

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.core import array # NOQA

template:74

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.core import ascontiguousarray # NOQA

template:75

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.core import asfortranarray # NOQA

template:76

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.core import divmod # NOQA

template:77

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.core import elementwise_copy # NOQA

template:78

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.core import ndarray # NOQA

template:79

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.dlpack import fromDlpack # NOQA

template:80

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.dlpack import from_dlpack # NOQA

template:81

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.internal import complete_slice # NOQA

template:82

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.internal import get_size # NOQA

template:83

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.raw import RawKernel # NOQA

template:84

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.raw import RawModule # NOQA

template:85

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_core/_fusion_interface.py· 5

CuPy dependency

from cupy._core._dtype import get_dtype

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _fusion_thread_local

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import core

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._scalar import get_typename

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_core/_fusion_op.py· 8

CuPy dependency

from cupy._core import _codeblock

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._fusion_variable import _TraceVariable

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._fusion_variable import _TraceArray

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._fusion_variable import _VariableSet

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _fusion_thread_local

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _kernel

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _reduction

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._scalar import get_typename

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_core/_fusion_optimization.py· 2

CuPy dependency

from cupy._core import _fusion_variable

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _fusion_op

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_core/_gufuncs.py· 5

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._core._routines_manipulation as _manipulation

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.internal import _normalize_axis_indices

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._dtype import get_dtype, _raise_if_invalid_cast

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_core/_ufuncs.py· 2

CuPy dependency

from cupy._core._kernel import create_ufunc

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._util import bf16_loop

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_core/include/cupy/atomics.cuh· 1

Custom CUDA kernels (.cu/.cuh)

CUDA kernel source

templatefile

Hand-written .cu/.cuh kernels compile only with nvcc. AMD's HIPify tools translate the vast majority of CUDA C++ to HIP automatically.

RecommendRun hipify-perl over the .cu/.cuh sources and build with hipcc.

cupy/_core/include/cupy/bfloat16.cuh· 1

Custom CUDA kernels (.cu/.cuh)

CUDA kernel source

templatefile

Hand-written .cu/.cuh kernels compile only with nvcc. AMD's HIPify tools translate the vast majority of CUDA C++ to HIP automatically.

RecommendRun hipify-perl over the .cu/.cuh sources and build with hipcc.

cupy/_core/include/cupy/carray.cuh· 1

Custom CUDA kernels (.cu/.cuh)

CUDA kernel source

templatefile

Hand-written .cu/.cuh kernels compile only with nvcc. AMD's HIPify tools translate the vast majority of CUDA C++ to HIP automatically.

RecommendRun hipify-perl over the .cu/.cuh sources and build with hipcc.

cupy/_core/include/cupy/complex.cuh· 1

Custom CUDA kernels (.cu/.cuh)

CUDA kernel source

templatefile

Hand-written .cu/.cuh kernels compile only with nvcc. AMD's HIPify tools translate the vast majority of CUDA C++ to HIP automatically.

RecommendRun hipify-perl over the .cu/.cuh sources and build with hipcc.

cupy/_core/include/cupy/float16.cuh· 1

Custom CUDA kernels (.cu/.cuh)

CUDA kernel source

templatefile

Hand-written .cu/.cuh kernels compile only with nvcc. AMD's HIPify tools translate the vast majority of CUDA C++ to HIP automatically.

RecommendRun hipify-perl over the .cu/.cuh sources and build with hipcc.

cupy/_core/include/cupy/hip_workaround.cuh· 1

Custom CUDA kernels (.cu/.cuh)

CUDA kernel source

templatefile

Hand-written .cu/.cuh kernels compile only with nvcc. AMD's HIPify tools translate the vast majority of CUDA C++ to HIP automatically.

RecommendRun hipify-perl over the .cu/.cuh sources and build with hipcc.

cupy/_core/include/cupy/structview.cuh· 1

Custom CUDA kernels (.cu/.cuh)

CUDA kernel source

templatefile

Hand-written .cu/.cuh kernels compile only with nvcc. AMD's HIPify tools translate the vast majority of CUDA C++ to HIP automatically.

RecommendRun hipify-perl over the .cu/.cuh sources and build with hipcc.

cupy/_core/include/cupy/type_dispatcher.cuh· 1

Custom CUDA kernels (.cu/.cuh)

CUDA kernel source

templatefile

Hand-written .cu/.cuh kernels compile only with nvcc. AMD's HIPify tools translate the vast majority of CUDA C++ to HIP automatically.

RecommendRun hipify-perl over the .cu/.cuh sources and build with hipcc.

cupy/_core/include/cupy/unstructured_void.cuh· 1

Custom CUDA kernels (.cu/.cuh)

CUDA kernel source

templatefile

Hand-written .cu/.cuh kernels compile only with nvcc. AMD's HIPify tools translate the vast majority of CUDA C++ to HIP automatically.

RecommendRun hipify-perl over the .cu/.cuh sources and build with hipcc.

cupy/_creation/basic.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.internal import _get_strides_for_order_K, _update_order_char

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.typing._types import (

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_creation/from_data.py· 2

CuPy dependency

from cupy import _core

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import fusion

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_creation/matrix.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_creation/ranges.py· 3

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._util import bf16_loop

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_functional/piecewise.py· 2

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_functional/vectorize.py· 1

CuPy dependency

from cupy import _core

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_indexing/generate.py· 3

CuPy dependency

import cupy

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation import from_data

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation import join

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_indexing/indexing.py· 2

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_indexing/insert.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_indexing/iterate.py· 3

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_io/formatting.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_io/npz.py· 1

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_io/text.py· 1

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_logic/comparison.py· 3

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic import content

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_logic/content.py· 3

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._util import bf16_loop

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_logic/ops.py· 2

CuPy dependency

from cupy._core import _kernel

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _routines_logic

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_logic/truth.py· 5

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _routines_logic as _logic

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _fusion_thread_local

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting import search as _search

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _util

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_logic/type_testing.py· 1

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_manipulation/add_remove.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_manipulation/basic.py· 5

CuPy dependency

from cupy import _core

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _fusion_interface

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import fusion

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting import search

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import stream

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_manipulation/dims.py· 3

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._core._routines_manipulation as _manipulation

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_manipulation/join.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_manipulation/kind.py· 2

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_manipulation/rearrange.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_manipulation/shape.py· 1

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_manipulation/split.py· 1

CuPy dependency

from cupy import _core

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_manipulation/tiling.py· 2

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_manipulation/transpose.py· 2

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _routines_manipulation

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_math/arithmetic.py· 3

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import fusion

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._util import bf16_loop

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_math/explog.py· 3

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math import ufunc

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._util import bf16_loop

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_math/floating.py· 3

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math import ufunc

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._util import bf16_loop

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_math/hyperbolic.py· 1

CuPy dependency

from cupy._math import ufunc

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_math/misc.py· 6

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _routines_math as _math

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import fusion

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.lib import stride_tricks

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._util import bf16_loop

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_math/rational.py· 1

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_math/rounding.py· 4

CuPy dependency

from cupy import _core

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import fusion

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math import ufunc

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._util import bf16_loop

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_math/special.py· 3

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math import ufunc

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._util import bf16_loop

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_math/sumprod.py· 4

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _routines_math as _math

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _fusion_thread_local

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_math/trigonometric.py· 5

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math import sumprod

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math import ufunc

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._util import bf16_loop

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_math/ufunc.py· 2

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._util import bf16_loop

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_math/window.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_misc/byte_bounds.py· 1

CuPy dependency

from cupy._core._memory_range import get_bound as _get_bounds

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_misc/memory_ranges.py· 4

CuPy dependency

from cupy._core import _kernel

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _memory_range

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._manipulation import join

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._sorting import search

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_misc/who.py· 1

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_padding/pad.py· 1

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_sorting/count.py· 1

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_sorting/search.py· 6

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import fusion

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _util

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _routines_indexing as _indexing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _routines_statistics as _statistics

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_sorting/sort.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import thrust

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_statistics/correlation.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_statistics/histogram.py· 6

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _accelerator

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cub

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import common

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_statistics/meanvar.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _routines_statistics as _statistics

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/_statistics/order.py· 5

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _routines_statistics as _statistics

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _fusion_thread_local

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic import content

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/cublas.py· 4

CuPy dependency

import cupy

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import _util

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/cuda/cupy_cub.cu· 1

Custom CUDA kernels (.cu/.cuh)

CUDA kernel source

templatefile

Hand-written .cu/.cuh kernels compile only with nvcc. AMD's HIPify tools translate the vast majority of CUDA C++ to HIP automatically.

RecommendRun hipify-perl over the .cu/.cuh sources and build with hipcc.

cupy/cuda/cupy_cufftXt.cu· 1

Custom CUDA kernels (.cu/.cuh)

CUDA kernel source

templatefile

Hand-written .cu/.cuh kernels compile only with nvcc. AMD's HIPify tools translate the vast majority of CUDA C++ to HIP automatically.

RecommendRun hipify-perl over the .cu/.cuh sources and build with hipcc.

cupy/cuda/cupy_thrust.cu· 1

Custom CUDA kernels (.cu/.cuh)

CUDA kernel source

templatefile

Hand-written .cu/.cuh kernels compile only with nvcc. AMD's HIPify tools translate the vast majority of CUDA C++ to HIP automatically.

RecommendRun hipify-perl over the .cu/.cuh sources and build with hipcc.

cupy/cuda/cutensor.py· 1

CuPy dependency

from cupy._environment import _preload_warning

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/cuda/memory_hooks/__init__.py· 4

CuPy dependency

from cupy.cuda.memory_hooks import debug_print # NOQA

template:1

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory_hooks import line_profile # NOQA

template:2

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory_hooks.debug_print import DebugPrintHook # NOQA

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory_hooks.line_profile import LineProfileHook # NOQA

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/cuda/memory_hooks/debug_print.py· 1

CuPy dependency

from cupy.cuda import memory_hook

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/cuda/memory_hooks/line_profile.py· 1

CuPy dependency

from cupy.cuda import memory_hook

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/cuda/nccl.py· 1

CuPy dependency

from cupy import _environment

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/fft/__init__.py· 19

CuPy dependency

from cupy.fft._config import config # NOQA

template:1

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import fft # NOQA

template:2

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import fft2 # NOQA

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import fftfreq # NOQA

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import fftn # NOQA

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import fftshift # NOQA

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import hfft # NOQA

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import ifft # NOQA

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import ifft2 # NOQA

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import ifftn # NOQA

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import ifftshift # NOQA

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import ihfft # NOQA

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import irfft # NOQA

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import irfft2 # NOQA

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import irfftn # NOQA

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import rfft # NOQA

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import rfft2 # NOQA

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import rfftfreq # NOQA

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import rfftn # NOQA

template:19

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/fft/_config.py· 3

CuPy dependency

from cupy import _util

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._cache import (get_plan_cache,

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._callback import (

template:19

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/fft/_fft.py· 20

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft import config

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._cache import get_plan_cache

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:65

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:85

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:334

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:499

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:637

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:702

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:726

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:750

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:776

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:802

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:828

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:855

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:883

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:910

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:939

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:967

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:1005

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/lib/__init__.py· 1

CuPy dependency

from cupy.lib import stride_tricks # NOQA

template:1

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/lib/_routines_poly.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import RankWarning

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/lib/_shape_base.py· 2

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/lib/stride_tricks.py· 1

CuPy dependency

import cupy as _cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/linalg/__init__.py· 22

CuPy dependency

from cupy.linalg._product import matrix_power # NOQA

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._product import linalg_cross as cross # NOQA

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._product import matmul # NOQA

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._product import matrix_transpose # NOQA

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._decomposition import cholesky # NOQA

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._decomposition import qr # NOQA

template:19

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._decomposition import svd # NOQA

template:20

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._eigenvalue import eigh # NOQA

template:25

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._eigenvalue import eig # NOQA

template:26

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._eigenvalue import eigvalsh # NOQA

template:27

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._eigenvalue import eigvals # NOQA

template:28

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._norms import norm # NOQA

template:33

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._norms import cond # NOQA

template:34

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._norms import det # NOQA

template:35

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._norms import matrix_rank # NOQA

template:36

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._norms import slogdet # NOQA

template:37

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._solve import solve # NOQA

template:42

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._solve import tensorsolve # NOQA

template:43

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._solve import lstsq # NOQA

template:44

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._solve import inv # NOQA

template:45

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._solve import pinv # NOQA

template:46

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._solve import tensorinv # NOQA

template:47

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/linalg/_decomposition.py· 4

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import _util

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/linalg/_eigenvalue.py· 5

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import _util

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _dtype

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/linalg/_einsum_cutn.py· 4

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _util

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _accelerator

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.device import Handle

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/linalg/_einsum.py· 6

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _accelerator

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _util

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._einsum_opt import _greedy_path

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._einsum_opt import _optimal_path

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg._einsum_cutn import _try_use_cutensornet

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/linalg/_norms.py· 4

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import _decomposition

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import _util

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/linalg/_product.py· 6

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._gufuncs import _GUFunc

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import _solve

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import _util

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/linalg/_solve.py· 6

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import _decomposition

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import _util

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cublas import batched_gesv, get_batched_gesv_limit

template:38

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/linalg/_util.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._util

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/polynomial/__init__.py· 2

CuPy dependency

from cupy.polynomial import polynomial # NOQA

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.polynomial import polyutils # NOQA

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/polynomial/polynomial.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/polynomial/polyutils.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/random/__init__.py· 58

CuPy dependency

import cupy as _cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._generator_api import Generator

template:38

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy.random._generator_api

template:55

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import beta # NOQA

template:63

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import binomial # NOQA

template:64

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import chisquare # NOQA

template:65

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import dirichlet # NOQA

template:66

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import exponential # NOQA

template:67

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import f # NOQA

template:68

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import gamma # NOQA

template:69

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import geometric # NOQA

template:70

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import gumbel # NOQA

template:71

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import hypergeometric # NOQA

template:72

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import laplace # NOQA

template:73

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import logistic # NOQA

template:74

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import lognormal # NOQA

template:75

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import logseries # NOQA

template:76

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import multivariate_normal # NOQA

template:77

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import negative_binomial # NOQA

template:78

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import noncentral_chisquare # NOQA

template:79

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import noncentral_f # NOQA

template:80

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import normal # NOQA

template:81

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import pareto # NOQA

template:82

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import poisson # NOQA

template:83

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import power # NOQA

template:84

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import rayleigh # NOQA

template:85

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import standard_cauchy # NOQA

template:86

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import standard_exponential # NOQA

template:87

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import standard_gamma # NOQA

template:88

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import standard_normal # NOQA

template:89

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import standard_t # NOQA

template:90

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import triangular # NOQA

template:91

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import uniform # NOQA

template:92

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import vonmises # NOQA

template:93

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import wald # NOQA

template:94

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import weibull # NOQA

template:95

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._distributions import zipf # NOQA

template:96

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._generator import get_random_state # NOQA

template:97

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._generator import RandomState # NOQA

template:98

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._generator import reset_states # NOQA

template:99

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._generator import seed # NOQA

template:100

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._generator import set_random_state # NOQA

template:101

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._permutations import permutation # NOQA

template:102

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._permutations import shuffle # NOQA

template:103

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._sample import choice # NOQA

template:104

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._sample import multinomial # NOQA

template:105

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._sample import rand # NOQA

template:106

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._sample import randint # NOQA

template:107

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._sample import randn # NOQA

template:108

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._sample import random_integers # NOQA

template:109

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._sample import random_sample # NOQA

template:110

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._sample import random_sample as random # NOQA

template:111

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._sample import random_sample as ranf # NOQA

template:112

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._sample import random_sample as sample # NOQA

template:113

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._bit_generator import BitGenerator # NOQA

template:114

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._bit_generator import XORWOW # NOQA

template:115

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._bit_generator import MRG32k3a # NOQA

template:116

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._bit_generator import Philox4x3210 # NOQA

template:117

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/random/_distributions.py· 2

CuPy dependency

from cupy.random import _generator

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _util

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/random/_generator.py· 7

CuPy dependency

import cupy

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:19

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:20

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random import _kernels

template:21

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random._generator_api import FeistelBijection

template:22

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _util

template:23

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/random/_kernels.py· 1

CuPy dependency

from cupy import _core

template:65

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/random/_permutations.py· 1

CuPy dependency

from cupy.random import _generator

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/random/_sample.py· 4

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation import basic

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random import _distributions

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random import _generator

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/random/cupy_distributions.cu· 1

Custom CUDA kernels (.cu/.cuh)

CUDA kernel source

templatefile

Hand-written .cu/.cuh kernels compile only with nvcc. AMD's HIPify tools translate the vast majority of CUDA C++ to HIP automatically.

RecommendRun hipify-perl over the .cu/.cuh sources and build with hipcc.

cupy/random/cupy_distributions.cuh· 1

Custom CUDA kernels (.cu/.cuh)

CUDA kernel source

templatefile

Hand-written .cu/.cuh kernels compile only with nvcc. AMD's HIPify tools translate the vast majority of CUDA C++ to HIP automatically.

RecommendRun hipify-perl over the .cu/.cuh sources and build with hipcc.

cupy/testing/__init__.py· 51

CuPy dependency

from cupy.testing._array import assert_allclose # NOQA

template:1

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._array import assert_array_almost_equal # NOQA

template:2

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._array import assert_array_almost_equal_nulp # NOQA

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._array import assert_array_equal # NOQA

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._array import assert_array_less # NOQA

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._array import assert_array_list_equal # NOQA

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._array import assert_array_max_ulp # NOQA

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._attr import multi_gpu # NOQA

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._attr import slow # NOQA

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._helper import assert_warns # NOQA

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._helper import numpy_satisfies # NOQA

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._helper import NumpyAliasBasicTestBase # NOQA

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._helper import NumpyAliasValuesTestBase # NOQA

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._helper import AssertFunctionIsCalled # NOQA

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._helper import shaped_arange # NOQA

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._helper import shaped_linspace # NOQA

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._helper import shaped_sparse_random # NOQA

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._helper import shaped_random # NOQA

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._helper import generate_matrix # NOQA

template:19

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._helper import shaped_reverse_arange # NOQA

template:20

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._helper import with_requires # NOQA

template:21

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._helper import installed # NOQA

template:22

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import for_all_dtypes # NOQA

template:23

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import for_all_dtypes_combination # NOQA

template:24

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import for_CF_orders # NOQA

template:25

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import for_complex_dtypes # NOQA

template:26

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import for_contiguous_axes # NOQA

template:27

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import for_dtypes # NOQA

template:28

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import for_dtypes_combination # NOQA

template:29

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import for_float_dtypes # NOQA

template:30

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import for_int_dtypes # NOQA

template:31

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import for_int_dtypes_combination # NOQA

template:32

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import for_orders # NOQA

template:33

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import for_signed_dtypes # NOQA

template:34

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import for_signed_dtypes_combination # NOQA

template:35

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import for_unsigned_dtypes # NOQA

template:36

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import for_unsigned_dtypes_combination # NOQA

template:37

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import numpy_cupy_allclose # NOQA

template:38

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import numpy_cupy_array_almost_equal # NOQA

template:39

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import numpy_cupy_array_almost_equal_nulp # NOQA

template:40

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import numpy_cupy_array_equal # NOQA

template:41

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import numpy_cupy_array_less # NOQA

template:42

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import numpy_cupy_array_list_equal # NOQA

template:43

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import numpy_cupy_array_max_ulp # NOQA

template:44

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import numpy_cupy_equal # NOQA

template:45

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import numpy_cupy_raises # NOQA

template:46

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._parameterized import parameterize # NOQA

template:47

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._parameterized import product # NOQA

template:48

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._parameterized import product_dict # NOQA

template:49

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._random import fix_random # NOQA

template:50

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._random import generate_seed # NOQA

template:51

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/testing/_array.py· 1

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/testing/_attr.py· 1

CuPy dependency

from cupy.testing._pytest_impl import is_available, check_available

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/testing/_helper.py· 3

CuPy dependency

import cupy

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._pytest_impl import is_available

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/testing/_loops.py· 5

CuPy dependency

import cupy

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import AxisError

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import _array

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import _parameterized

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._pytest_impl import is_available

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/testing/_parameterized.py· 2

CuPy dependency

from cupy.testing import _bundle

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import _pytest_impl

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/testing/_protocol_helpers.py· 2

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/testing/_pytest_impl.py· 1

CuPy dependency

import cupy.testing._parameterized

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/testing/_random.py· 1

CuPy dependency

import cupy

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/typing/__init__.py· 4

CuPy dependency

from cupy.typing._types import ArrayLike # NOQA

template:1

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.typing._types import DTypeLike # NOQA

template:2

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.typing._types import NBitBase # NOQA

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.typing._types import NDArray # NOQA

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupy/typing/_types.py· 1

CuPy dependency

from cupy._core import core

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/__init__.py· 2

CuPy dependency

from cupy._core.syncdetect import allow_synchronize # NOQA

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.syncdetect import DeviceSynchronized # NOQA

template:19

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/_gufunc.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/_pinned_array.py· 3

CuPy dependency

from cupy import cuda

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation.basic import _new_like_order_and_strides

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/_rsqrt.py· 1

CuPy dependency

from cupy._core.core import create_ufunc

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/_texture.py· 4

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import texture

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/cusparse.py· 5

CuPy dependency

import cupy as _cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _dtype

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device as _device

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import stream as _stream

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _util

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/distributed/_init.py· 1

CuPy dependency

from cupy.cuda import nccl

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/distributed/_nccl_comm.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import nccl

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/distributed/array/_array.py· 8

CuPy dependency

import cupy

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.core import ndarray

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._creation.from_data as _creation_from_data

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._core._routines_math as _math

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._core._routines_statistics as _statistics

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.device import Device

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.stream import Stream

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.stream import get_current_stream

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/distributed/array/_chunk.py· 7

CuPy dependency

from cupy._core.core import ndarray

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._creation.basic as _creation_basic

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._manipulation.dims as _manipulation_dims

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.device import Device

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.stream import Event

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.stream import Stream

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.stream import get_current_stream

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/distributed/array/_data_transfer.py· 9

CuPy dependency

from cupy._core.core import ndarray

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._creation.from_data as _creation_from_data

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._creation.basic as _creation_basic

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.device import Device

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.stream import Event

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.stream import Stream

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.stream import get_current_stream

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import nccl

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.nccl import NcclCommunicator as _Communicator

template:20

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/distributed/array/_elementwise.py· 6

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._creation.basic as _creation_basic

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.core import ndarray

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.device import Device

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.stream import Stream

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.stream import get_current_stream

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/distributed/array/_linalg.py· 1

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/distributed/array/_modes.py· 1

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/distributed/array/_reduction.py· 1

CuPy dependency

import cupy._manipulation.dims as _manipulation_dims

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/jit/_builtin_funcs.py· 3

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

Device string "cuda"

elif env.mode == 'cuda':

template:104

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

cupyx/jit/_compile.py· 4

CuPy dependency

from cupy.exceptions import ComplexWarning

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._codeblock import CodeBlock, _CodeType

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _kernel

template:19

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._dtype import _raise_if_invalid_cast

template:20

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/jit/_cuda_typerules.py· 8

CuPy dependency

import cupy

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic import ops

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math import arithmetic

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._logic import comparison

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._binary import elementwise

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

Device string "cuda"

if mode == 'cuda':

template:106

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

if mode == 'cuda':

template:125

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

cupyx/jit/_cuda_types.py· 2

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._scalar import get_typename

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/jit/_interface.py· 7

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import core

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.compiler import _get_arch

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.compiler import _get_nvrtc_version

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.compiler import _jitify_prep

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.compiler import _NVRTCProgram

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

Device string "cuda"

def rawkernel(*, mode='cuda', device=False):

template:244

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

cupyx/jit/cg.py· 1

CuPy dependency

from cupy.cuda import runtime as _runtime

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/lapack.py· 2

CuPy dependency

import cupy as _cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device as _device

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/linalg/_solve.py· 2

CuPy dependency

from cupy.linalg import _util

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/linalg/sparse/_solve.py· 3

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import _util

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/optimizing/_optimize.py· 1

CuPy dependency

from cupy._core import _optimize_config

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/profiler/__init__.py· 1

CuPy dependency

from cupy.cuda import runtime as _runtime

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/profiler/_time_range.py· 1

CuPy dependency

from cupy import cuda

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/profiler/_time.py· 1

CuPy dependency

import cupy as _cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/__init__.py· 1

CuPy dependency

from cupy._core import ndarray as _ndarray

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/_lib/_util.py· 1

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/fft/__init__.py· 1

CuPy dependency

from cupy.fft import fftshift, ifftshift, fftfreq, rfftfreq

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/fft/_fft.py· 10

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import _fft, _default_fft_func, _swap_direction

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:96

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:130

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:229

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:267

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:309

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:344

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:452

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:494

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/fft/_fftlog.py· 1

CuPy dependency

import cupy

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/fft/_realtransforms.py· 4

CuPy dependency

import cupy

template:40

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:41

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import _cook_shape

template:42

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import AxisError

template:44

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/fftpack/_fft.py· 13

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft import config

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import (_convert_fft_type, _default_fft_func, _fft,

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._cache import get_plan_cache

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:66

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:200

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:234

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:268

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:304

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:340

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:376

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:419

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:470

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/interpolate/_bspline.py· 3

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._scalar import get_typename

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/interpolate/_bspline2.py· 2

CuPy dependency

import cupy

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import solve

template:475

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/interpolate/_cubic.py· 3

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import solve

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.misc import _legacy_sign

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/interpolate/_fitpack_repro.py· 1

CuPy dependency

import cupy

template:26

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/interpolate/_interpnd.py· 2

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._scalar import get_typename

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/interpolate/_interpolate.py· 3

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal # NOQA

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._scalar import get_typename # NOQA

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/interpolate/_ndbspline.py· 1

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/interpolate/_ndgriddata.py· 1

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/interpolate/_polyint.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/interpolate/_rbfinterp.py· 1

CuPy dependency

import cupy as cp

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/interpolate/_rgi.py· 1

CuPy dependency

import cupy as cp

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/linalg/_array_utils.py· 2

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import _util

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/linalg/_decomp_lu.py· 5

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cublas

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import _util

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/linalg/_matfuncs.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import _util

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/linalg/_solve_triangular.py· 4

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cublas

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import _util

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/linalg/_special_matrices.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/linalg/_uarray.py· 2

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy.linalg as _cp_linalg

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/ndimage/_filters_core.py· 3

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/ndimage/_filters_generic.py· 2

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _util

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/ndimage/_filters.py· 3

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/ndimage/_fourier.py· 3

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/ndimage/_interp_kernels.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._core.internal

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/ndimage/_interpolation.py· 4

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/ndimage/_measurements.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _util

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/ndimage/_morphology.py· 2

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/ndimage/_pba_2d.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

Device string "cuda"

kernel_directory = os.path.join(os.path.dirname(__file__), "cuda")

template:72

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

cupyx/scipy/ndimage/_pba_3d.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

Device string "cuda"

kernel_directory = os.path.join(os.path.dirname(__file__), "cuda")

template:80

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

cupyx/scipy/ndimage/_spline_prefilter_core.py· 1

CuPy dependency

import cupy

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/ndimage/_util.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import AxisError

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_arraytools.py· 1

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_bsplines.py· 1

CuPy dependency

import cupy

template:32

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_czt.py· 1

CuPy dependency

import cupy

template:35

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_filter_design.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.polynomial.polynomial import (

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_fir_filter_design.py· 3

CuPy dependency

from cupy.fft import fft, ifft

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import solve, lstsq, LinAlgError

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_iir_filter_conversions.py· 1

CuPy dependency

import cupy

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_iir_filter_design.py· 1

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_iir_utils.py· 3

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.internal import _normalize_axis_index

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._scalar import get_typename

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_lti_conversion.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_ltisys.py· 1

CuPy dependency

import cupy

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_max_len_seq.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_optimize.py· 1

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_peak_finding.py· 2

CuPy dependency

import cupy

template:31

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._scalar import get_typename

template:33

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_polyutils.py· 1

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_resample.py· 1

CuPy dependency

import cupy

template:34

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_savitzky_golay.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import lstsq

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_signaltools_core.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_signaltools.py· 3

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import lstsq

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_spectral_impl.py· 1

CuPy dependency

import cupy

template:32

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_spectral.py· 1

CuPy dependency

import cupy

template:31

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_splines.py· 3

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._scalar import get_typename

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core.internal import _normalize_axis_index

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_upfirdn.py· 1

CuPy dependency

import cupy

template:31

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_waveforms.py· 2

CuPy dependency

import cupy

template:30

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._scalar import get_typename

template:31

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/_wavelets.py· 1

CuPy dependency

import cupy

template:31

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/signal/windows/_windows.py· 1

CuPy dependency

import cupy

template:32

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/_base.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/_compressed.py· 4

CuPy dependency

import cupy

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _scalar

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation import basic

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/_construct.py· 1

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/_coo.py· 2

CuPy dependency

import cupy

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/_csc.py· 1

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/_csr.py· 2

CuPy dependency

import cupy

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/_data.py· 3

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import internal

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _util

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/_dia.py· 2

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/_extract.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/_index.py· 4

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/_sputils.py· 2

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._dtype import get_dtype

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/_util.py· 2

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import core

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/csgraph/_traversal.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/linalg/_eigen.py· 4

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cublas

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _dtype

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/linalg/_interface.py· 1

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/linalg/_iterative.py· 4

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cublas

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _dtype

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/linalg/_lobpcg.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy.linalg as linalg

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/linalg/_norm.py· 1

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/sparse/linalg/_solve.py· 5

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cublas

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import _util

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/spatial/_delaunay.py· 1

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/spatial/_kdtree_utils.py· 2

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._scalar import get_typename

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/spatial/_kdtree.py· 1

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/spatial/delaunay_2d/_kernels.py· 1

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/spatial/delaunay_2d/_tri.py· 1

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/spatial/distance.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/__init__.py· 2

CuPy dependency

from cupy._math.rounding import round # NOQA

template:115

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._math.special import sinc # NOQA

template:119

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_basic.py· 1

CuPy dependency

from cupy import _core

template:28

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_bessel.py· 1

CuPy dependency

from cupy import _core

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_beta.py· 1

CuPy dependency

from cupy import _core

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_binom.py· 1

CuPy dependency

from cupy import _core

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_convex_analysis.py· 1

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_digamma.py· 1

CuPy dependency

from cupy import _core

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_ellip.py· 1

CuPy dependency

from cupy import _core

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_erf.py· 1

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_exp1.py· 1

CuPy dependency

from cupy import _core # NOQA

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_expi.py· 1

CuPy dependency

from cupy import _core # NOQA

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_expn.py· 1

CuPy dependency

from cupy import _core

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_gamma.py· 1

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_gammainc.py· 1

CuPy dependency

from cupy import _core

template:22

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_gammaln.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_gammasgn.py· 1

CuPy dependency

from cupy import _core

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_lambertw.py· 1

CuPy dependency

from cupy import _core

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_loggamma.py· 1

CuPy dependency

from cupy import _core

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_logsoftmax.py· 1

CuPy dependency

import cupy as cp

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_logsumexp.py· 1

CuPy dependency

import cupy as cp

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_lpmv.py· 1

CuPy dependency

from cupy import _core

template:22

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_poch.py· 1

CuPy dependency

from cupy import _core

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_polygamma.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_sici.py· 1

CuPy dependency

from cupy import _core

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_softmax.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_sph_harm.py· 1

CuPy dependency

from cupy import _core

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_spherical_bessel.py· 2

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_statistics.py· 1

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_stats_distributions.py· 1

CuPy dependency

from cupy import _core

template:20

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_wright_bessel.py· 1

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_xlogy.py· 1

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_zeta.py· 1

CuPy dependency

from cupy import _core

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/special/_zetac.py· 1

CuPy dependency

from cupy import _core

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/stats/_distributions.py· 1

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/stats/_morestats.py· 1

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/stats/_stats_py.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/scipy/stats/_stats.py· 1

CuPy dependency

import cupy as cp

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/signal/_acoustics/_cepstrum.py· 1

CuPy dependency

import cupy

template:23

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/signal/_convolution/_convolve.py· 1

CuPy dependency

import cupy

template:28

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/signal/_filtering/_filtering.py· 3

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._scalar import get_typename

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _util

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/signal/_radartools/_beamformers.py· 1

CuPy dependency

import cupy as cp

template:23

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/signal/_radartools/_radartools.py· 1

CuPy dependency

import cupy

template:28

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

cupyx/time.py· 1

CuPy dependency

import cupy as _cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

docs/source/conf.py· 1

CuPy dependency

import cupy

template:21

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/cg/cg.py· 1

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/cufft/callback/jit_r2c_c2r_string.py· 1

CuPy dependency

import cupy as cp

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/cufft/callback/jit_string.py· 1

CuPy dependency

import cupy as cp

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/cufft/callback/jit_string2.py· 1

CuPy dependency

import cupy as cp

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/cufft/callback/legacy.py· 1

CuPy dependency

import cupy as cp

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/cusparselt/matmul.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/custom_struct/builtin_vectors.py· 1

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/custom_struct/complex_struct.py· 1

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/custom_struct/packed_matrix.py· 1

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/cutensor/contraction.py· 1

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/cutensor/elementwise_binary.py· 1

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/cutensor/elementwise_trinary.py· 1

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/cutensor/reduction.py· 1

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/finance/black_scholes.py· 1

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/finance/monte_carlo_multigpu.py· 1

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/finance/monte_carlo.py· 1

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/gemm/sgemm.cu· 1

Custom CUDA kernels (.cu/.cuh)

CUDA kernel source

templatefile

Hand-written .cu/.cuh kernels compile only with nvcc. AMD's HIPify tools translate the vast majority of CUDA C++ to HIP automatically.

RecommendRun hipify-perl over the .cu/.cuh sources and build with hipcc.

examples/gemm/sgemm.py· 1

CuPy dependency

import cupy as cp

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/gemm/utils.py· 1

CuPy dependency

import cupy as cp

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/gmm/gmm.py· 1

CuPy dependency

import cupy

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/interoperability/mpi4py_multiple_devices.py· 1

CuPy dependency

import cupy

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/jit/elmentwise_op.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/jit/reduction_atomic.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/jit/reduction_simple.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/kmeans/kmeans.py· 1

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/peer/peer_matrix.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/stream/cublas.py· 1

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/stream/cufft.py· 1

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/stream/cupy_event.py· 1

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/stream/cupy_kernel.py· 1

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/stream/cupy_memcpy.py· 1

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/stream/curand.py· 1

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/stream/cusolver.py· 1

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/stream/cusparse.py· 1

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/stream/map_reduce.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

examples/stream/thrust.py· 1

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/conftest.py· 3

CuPy dependency

import cupy as cp

template:77

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._core.numpy_allocator as ac

template:81

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.compiler import _set_kernel_cache_backend

template:95

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/_gcp_kernel_cache_backend.py· 2

CuPy dependency

from cupy.cuda.compiler import _get_cupy_cache_key

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda._compiler_cache import DiskKernelCacheBackend # noqa

template:36

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/binary_tests/test_elementwise.py· 1

CuPy dependency

from cupy import testing

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/binary_tests/test_packing.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/fusion_tests/fusion_utils.py· 3

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import _wraps_partial_xp

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/fusion_tests/test_array.py· 1

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/fusion_tests/test_example.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/fusion_tests/test_indexing.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/fusion_tests/test_kernel_cache.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/fusion_tests/test_misc.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/fusion_tests/test_optimization.py· 2

CuPy dependency

import cupy # NOQA

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/fusion_tests/test_reduction.py· 3

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import AxisError

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/fusion_tests/test_routines.py· 1

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/fusion_tests/test_ufunc.py· 1

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_array_function.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_bfloat16.py· 3

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_carray.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_core.py· 4

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import core

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._protocol_helpers import (

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_cub_reduction.py· 6

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _environment

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _accelerator

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _cub_reduction

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import memory

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_dlpack.py· 3

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_elementwise.py· 4

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_flags.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import flags

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_function.py· 5

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import core

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import compiler

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_gufuncs.py· 3

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._gufuncs import _GUFunc

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_internal.py· 1

CuPy dependency

from cupy._core import internal

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_iter.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_mdspan.py· 3

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._scalar import get_typename

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_multithreading.py· 2

CuPy dependency

import cupy

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.memory import alloc

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray_adv_indexing.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray_byteswap.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray_complex_ops.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray_contiguity.py· 1

CuPy dependency

from cupy import testing

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray_conversion.py· 3

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray_copy_and_view.py· 5

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import ComplexWarning

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _util

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._protocol_helpers import (

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray_cuda_array_interface.py· 3

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray_elementwise_op.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray_get.py· 3

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray_indexing.py· 4

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._protocol_helpers import (

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import ComplexWarning

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray_math.py· 1

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray_owndata.py· 1

CuPy dependency

from cupy import _core

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray_reduction.py· 4

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._core._accelerator as _acc

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _cub_reduction

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray_scatter.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray_ufunc.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray_unary_op.py· 2

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ndarray.py· 7

CuPy dependency

import cupy

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _util

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import get_array_module

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import AxisError

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_nep50_examples.py· 1

CuPy dependency

import cupy as cp

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_reduction.py· 6

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._core._accelerator as _acc

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _core

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import ComplexWarning, AxisError

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._protocol_helpers import (

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_scan.py· 4

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core._routines_math import _scan_for_test as scan

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_structured.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._protocol_helpers import DummyObjectWithCudaArrayInterface

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_syncdetect.py· 1

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_ufunc_methods.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/core_tests/test_userkernel.py· 4

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.texture import (ChannelFormatDescriptor, CUDAarray,

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/creation_tests/test_basic.py· 3

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import ComplexWarning

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/creation_tests/test_from_data.py· 3

CuPy dependency

import cupy

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda, testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._protocol_helpers import (

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/creation_tests/test_matrix.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/creation_tests/test_ranges.py· 2

CuPy dependency

import cupy

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/memory_hooks_tests/test_debug_print.py· 3

CuPy dependency

import cupy.cuda

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import memory

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import memory_hooks

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/memory_hooks_tests/test_line_profile.py· 2

CuPy dependency

from cupy.cuda import memory

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import memory_hooks

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_compiler_cache.py· 1

CuPy dependency

from cupy.cuda._compiler_cache import (

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_compiler.py· 2

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import compiler

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_cublas.py· 1

CuPy dependency

from cupy.cuda import cublas

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_cufft.py· 5

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft import config

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import _convert_fft_type

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_curand.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import curand

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_cusolver.py· 1

CuPy dependency

from cupy import cuda

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_cusparse.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cusparse

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_cutensor.py· 1

CuPy dependency

from cupy.cuda import cutensor

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_device.py· 3

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_driver.py· 3

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import driver

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_graph.py· 3

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_memory_hook.py· 3

CuPy dependency

import cupy.cuda

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import memory

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import memory_hook

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_memory.py· 6

CuPy dependency

import cupy.cuda

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import memory

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import stream as stream_module

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_nccl.py· 4

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import nccl

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_nvrtc.py· 1

CuPy dependency

from cupy.cuda import nvrtc

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_nvtx.py· 1

CuPy dependency

from cupy import cuda

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_pinned_memory.py· 1

CuPy dependency

from cupy.cuda import pinned_memory

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_profile.py· 1

CuPy dependency

from cupy import cuda

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_runtime.py· 3

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import driver

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_stream.py· 4

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._creation import from_data

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/cuda_tests/test_texture.py· 4

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.texture import (ChannelFormatDescriptor, CUDAarray,

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/fft_tests/test_cache.py· 6

CuPy dependency

import cupy

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft import config

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/fft_tests/test_callback.py· 4

CuPy dependency

import cupy

template:20

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:21

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cufft

template:22

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.device import get_compute_capability

template:23

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/fft_tests/test_fft.py· 10

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft import config

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import (_default_fft_func, _fft, _fftn,

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import _wraps_partial

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft import fftn

template:556

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft import fft

template:629

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft import fft

template:698

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft import rfft

template:848

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:1225

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/functional_tests/test_piecewise.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/functional_tests/test_vectorize.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/indexing_tests/test_generate.py· 5

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import driver

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._indexing import generate

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/indexing_tests/test_indexing.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/indexing_tests/test_insert.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/indexing_tests/test_iterate.py· 3

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import ComplexWarning

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/io_tests/test_base_n.py· 1

CuPy dependency

from cupy import testing

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/io_tests/test_formatting.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/io_tests/test_npz.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/io_tests/test_text.py· 1

CuPy dependency

import cupy

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/lib_tests/test_polynomial.py· 5

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import driver

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import RankWarning

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/lib_tests/test_shape_base.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import AxisError

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/lib_tests/test_strided_tricks.py· 3

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.lib import stride_tricks

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/linalg_tests/test_decomposition.py· 6

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import driver

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.linalg import _util

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import _condition

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/linalg_tests/test_eigenvalue.py· 4

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import driver

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/linalg_tests/test_einsum.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/linalg_tests/test_norms.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/linalg_tests/test_product.py· 2

CuPy dependency

import cupy

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/linalg_tests/test_solve.py· 4

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import _condition

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cublas import get_batched_gesv_limit, set_batched_gesv_limit

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/logic_tests/test_comparison.py· 2

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/logic_tests/test_content.py· 1

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/logic_tests/test_ops.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/logic_tests/test_truth.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/logic_tests/test_type_test.py· 1

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/manipulation_tests/test_add_remove.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/manipulation_tests/test_basic.py· 4

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import ComplexWarning

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/manipulation_tests/test_dims.py· 4

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import AxisError

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/manipulation_tests/test_join.py· 4

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import AxisError

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/manipulation_tests/test_kind.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/manipulation_tests/test_rearrange.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/manipulation_tests/test_shape.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/manipulation_tests/test_split.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/manipulation_tests/test_tiling.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/manipulation_tests/test_transpose.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import AxisError

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/math_tests/test_arithmetic.py· 3

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import ComplexWarning

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/math_tests/test_explog.py· 1

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/math_tests/test_floating.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/math_tests/test_hyperbolic.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/math_tests/test_matmul.py· 3

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _routines_linalg as _linalg

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/math_tests/test_misc.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/math_tests/test_rational.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/math_tests/test_rounding.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/math_tests/test_special.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/math_tests/test_sumprod.py· 6

CuPy dependency

import cupy

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._core._accelerator as _acc

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy.cuda.cutensor

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _cub_reduction

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import AxisError

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/math_tests/test_trigonometric.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/math_tests/test_window.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/misc_tests/test_byte_bounds.py· 2

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/misc_tests/test_memory_ranges.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/misc_tests/test_who.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/padding_tests/test_pad.py· 3

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import ComplexWarning

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/polynomial_tests/test_polynomial.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/polynomial_tests/test_polyutils.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/random_tests/common_distributions.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import _condition

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/random_tests/test_bit_generator.py· 3

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import random

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/random_tests/test_distributions.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random import _distributions

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/random_tests/test_generator_api.py· 4

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import random

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import _condition

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/random_tests/test_generator.py· 7

CuPy dependency

import cupy

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.random import _generator

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import _condition

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import _hypothesis

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/random_tests/test_init.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/random_tests/test_permutations.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import _condition

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/random_tests/test_random.py· 2

CuPy dependency

from cupy import random

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/random_tests/test_sample.py· 6

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import random

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import _condition

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import _hypothesis

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/sorting_tests/test_count.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/sorting_tests/test_search.py· 4

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._core._accelerator as _acc

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _cub_reduction

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/sorting_tests/test_sort.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import AxisError

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/statistics_tests/test_correlation.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/statistics_tests/test_histogram.py· 3

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _accelerator

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/statistics_tests/test_meanvar.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import AxisError

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/statistics_tests/test_order.py· 4

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._core._accelerator as _acc

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/test_cublas.py· 3

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cublas

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/test_init.py· 2

CuPy dependency

import cupy

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/test_ndim.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/test_numpy_interop.py· 2

CuPy dependency

import cupy

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/test_type_routines.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/test_typing.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/testing_tests/test_array.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/testing_tests/test_condition.py· 1

CuPy dependency

from cupy.testing import _condition

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/testing_tests/test_helper.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import AxisError

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/testing_tests/test_loops.py· 4

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import _loops

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import AxisError

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupy_tests/testing_tests/test_parameterized.py· 1

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/distributed_tests/comm_runner.py· 4

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import cuda

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import nccl

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/distributed_tests/test_array_nccl.py· 2

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/distributed_tests/test_array_no_nccl.py· 1

CuPy dependency

import cupy.cuda

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/distributed_tests/test_comm.py· 2

CuPy dependency

from cupy.cuda import nccl

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/distributed_tests/test_linalg.py· 2

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/distributed_tests/test_store.py· 2

CuPy dependency

from cupy.cuda import nccl

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import _condition

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/jit_tests/test_cooperative_groups.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/jit_tests/test_cub.py· 2

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/jit_tests/test_device_function.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/jit_tests/test_raw.py· 4

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/jit_tests/test_thrust.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/linalg_tests/sparse_tests/__init__.py· 1

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/linalg_tests/sparse_tests/test_solve.py· 3

CuPy dependency

import cupy as cp

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import _condition

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/linalg_tests/test_solve.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/profiler_tests/test_benchmark.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/profiler_tests/test_time_range.py· 1

CuPy dependency

from cupy import cuda

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/profiler_tests/test_timeit_magic.py· 1

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/fft_tests/test_fft.py· 15

CuPy dependency

import cupy as cp

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import _default_fft_func, _fftn

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:121

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:240

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:376

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:492

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:633

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:749

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:878

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:949

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:1089

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:1210

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:1342

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:1463

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/fft_tests/test_fftlog.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/fft_tests/test_realtransforms.py· 1

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/fftpack_tests/test_fftpack.py· 20

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.fft._fft import _default_fft_func, _fftn

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:90

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:166

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy

template:183

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy.fft.config as config

template:235

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy.fft.config as config

template:255

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:277

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy.fft.config as config

template:314

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy.fft.config as config

template:334

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:356

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy.fft.config as config

template:410

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy.fft.config as config

template:430

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:452

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy.fft.config as config

template:489

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy.fft.config as config

template:509

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:531

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy

template:543

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda.cufft import get_current_plan

template:618

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/interpolate_tests/test_bspline.py· 1

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/interpolate_tests/test_bspline2.py· 4

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import assert_allclose

template:654

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/interpolate_tests/test_fitpack2.py· 2

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/interpolate_tests/test_interpnd.py· 2

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/interpolate_tests/test_ndbspline.py· 3

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/interpolate_tests/test_ndgriddata.py· 2

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/interpolate_tests/test_polyint.py· 4

CuPy dependency

import cupy

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import assert_array_almost_equal

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/interpolate_tests/test_ppoly.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import assert_allclose

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/interpolate_tests/test_rbfinterp.py· 2

CuPy dependency

import cupy as cp

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/interpolate_tests/test_rgi.py· 2

CuPy dependency

import cupy as cp

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import (assert_allclose, assert_array_equal,

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/linalg_tests/test_array_utils.py· 1

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/linalg_tests/test_decomp_lu.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/linalg_tests/test_matfuncs.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/linalg_tests/test_solve_triangular.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/linalg_tests/test_special_matrices.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/linalg_tests/test_uarray.py· 1

CuPy dependency

from cupy import testing

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/ndimage_tests/test_distance_transform.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/ndimage_tests/test_filters.py· 4

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import AxisError

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/ndimage_tests/test_fourier.py· 3

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.exceptions import AxisError

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/ndimage_tests/test_interpolation.py· 3

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/ndimage_tests/test_measurements.py· 5

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import _util

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _accelerator

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/ndimage_tests/test_morphology.py· 2

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_bsplines.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_cont2discrete.py· 1

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_czt.py· 3

CuPy dependency

import cupy

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import assert_allclose

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_dltisys.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_filter_design.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import assert_array_almost_equal, assert_allclose

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_fir_filter_design.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_iir_filter_conversion.py· 3

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import assert_array_almost_equal

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_iir_filter_design.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_iir_utils.py· 4

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import driver

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_ltisys.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_max_len_seq.py· 2

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_peak_finding.py· 4

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import driver

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_polyutils.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_resample.py· 2

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_savitzky_golay.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import (assert_allclose, assert_array_equal as assert_equal,

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_signaltools.py· 4

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import driver

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_spectral.py· 4

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import driver

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_splines.py· 3

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import driver

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_upfirdn.py· 1

CuPy dependency

from cupy import testing

template:46

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_waveforms.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_wavelets.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/signal_tests/test_windows.py· 1

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/sparse_tests/csgraph_tests/test_traversal.py· 1

CuPy dependency

from cupy import testing

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/sparse_tests/test_base.py· 2

CuPy dependency

import cupy

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/sparse_tests/test_construct.py· 3

CuPy dependency

import cupy

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/sparse_tests/test_coo.py· 4

CuPy dependency

import cupy

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import driver

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/sparse_tests/test_csc.py· 4

CuPy dependency

import cupy

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import driver

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/sparse_tests/test_csr.py· 2

CuPy dependency

import cupy

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:19

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/sparse_tests/test_dia.py· 4

CuPy dependency

import cupy

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import driver

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/sparse_tests/test_extract.py· 2

CuPy dependency

import cupy

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/sparse_tests/test_index.py· 3

CuPy dependency

import cupy

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/sparse_tests/test_linalg.py· 5

CuPy dependency

import cupy

template:16

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import driver

template:19

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:20

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import _condition

template:21

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/sparse_tests/test_sparray.py· 2

CuPy dependency

import cupy

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:13

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/sparse_tests/test_sparse_int64_indices.py· 4

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:3843

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:3858

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/spatial_tests/test_delaunay.py· 2

CuPy dependency

from cupy import testing

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy # NOQA

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/spatial_tests/test_distance.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:20

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/spatial_tests/test_kdtree_cuvs.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/spatial_tests/test_kdtree.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_basic.py· 4

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import (

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import numpy_cupy_allclose

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_bessel.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_beta.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_binom.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_convex_analysis.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_digamma.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_ellipk.py· 1

CuPy dependency

from cupy import testing

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_erf.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_exp1.py· 2

CuPy dependency

from cupy import testing

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import numpy_cupy_allclose

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_expi.py· 2

CuPy dependency

from cupy import testing

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import numpy_cupy_allclose

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_expn.py· 2

CuPy dependency

from cupy import testing

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import numpy_cupy_allclose

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_exprel.py· 3

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import numpy_cupy_allclose

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_gamma.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_gammainc.py· 2

CuPy dependency

import cupy as cp

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import assert_allclose, assert_array_equal

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_gammaln.py· 1

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_lambertw.py· 1

CuPy dependency

from cupy import testing

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_log_softmax.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_logsumexp.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_polygamma.py· 1

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_sici.py· 1

CuPy dependency

from cupy import testing

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_softmax.py· 2

CuPy dependency

import cupy

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_sph_harm.py· 3

CuPy dependency

import cupy as cp

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing import numpy_cupy_allclose

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_spherical_bessel.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_statistics.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_ufunc_dispatch.py· 2

CuPy dependency

import cupy

template:4

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_wright_bessel.py· 1

CuPy dependency

from cupy import testing

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_zeta.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/special_tests/test_zetac.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/stats_tests/test_distributions.py· 2

CuPy dependency

import cupy

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/stats_tests/test_morestats.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/stats_tests/test_stats.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/scipy_tests/test_get_array_module.py· 1

CuPy dependency

from cupy import testing

template:5

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/signal_tests/acoustics_tests/test_cepstrum.py· 2

CuPy dependency

import cupy

template:26

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:27

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/signal_tests/convolution_tests/test_convolve.py· 2

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/signal_tests/filtering_tests/test_filtering.py· 2

CuPy dependency

import cupy

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/signal_tests/radartools_tests/test_radartools.py· 2

CuPy dependency

import cupy

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/test_cupyx.py· 1

CuPy dependency

from cupy import testing

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/test_cusolver.py· 3

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _routines_linalg as _linalg

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/test_cusparse.py· 4

CuPy dependency

import cupy

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:15

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import driver

template:17

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import runtime

template:18

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/test_cutensor.py· 5

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _routines_linalg as _linalg

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import device

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.cuda import cutensor as ct

template:14

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/test_lapack.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/test_optimize.py· 4

CuPy dependency

import cupy

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:11

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy._core import _accelerator

template:12

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

import cupy._core._optimize_config

template:21

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/test_pinned_array.py· 3

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy.testing._loops import _wraps_partial

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/test_rsqrt.py· 2

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/test_runtime.py· 1

CuPy dependency

import cupy

template:6

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/test_time.py· 2

CuPy dependency

import cupy

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

CuPy dependency

from cupy import testing

template:9

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/cupyx_tests/tools_tests/test_install_library.py· 5

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

Device string "cuda"

rec['cuda'] for rec in install_library.library_records[lib]]))

template:19

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

@pytest.mark.parametrize('cuda', _get_supported_cuda_versions('nccl'))

template:27

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

@pytest.mark.parametrize('cuda', _get_supported_cuda_versions('cutensor'))

template:35

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

if rec['cuda'] != cuda:

template:46

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

tests/example_tests/example_test.py· 1

CuPy dependency

import cupy

template:7

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/example_tests/test_finance.py· 1

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/example_tests/test_gmm.py· 1

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/example_tests/test_kmeans.py· 1

CuPy dependency

from cupy import testing

template:8

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/import_tests/test_import.py· 1

CuPy dependency

import cupy

template:3

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

tests/install_tests/test_cupy_builder/test_features.py· 1

CuPy dependency

import cupy

template:10

CuPy targets CUDA directly. AMD ships a ROCm build of CuPy, so this is a mechanical swap rather than a rewrite.

RecommendInstall the ROCm CuPy build (cupy-rocm-6-0) instead of cupy-cudaXXx.

.pfnci/generate.py· 6

Device string "cuda"

full_ver = self.schema['cuda'][matrix.cuda]['full_version']

template:61

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

repo = self.schema['cuda'][matrix.cuda]['repository']

template:62

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

nccl_cuda_schema = self.schema['nccl'][nccl]['cuda'][cuda]

template:261

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

for cuda, _ in value_schema.get('cuda', {}).items():

template:484

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

if cuda not in schema['cuda'].keys():

template:485

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

supports = schema[key][value].get('cuda', None)

template:553

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

cupyx/jit/cub.py· 1

Device string "cuda"

return 'hipcub' if _runtime.is_hip else 'cuda'

template:42

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

cupyx/tools/_generate_wheel_metadata.py· 2

Device string "cuda"

'cuda': cuda_version,

template:35

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

if record['cuda'] == cuda_version:

template:40

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

cupyx/tools/install_library.py· 5

Device string "cuda"

'cuda': cuda_version,

template:64

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

'cuda': cuda_version,

template:112

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

if record['cuda'] == cuda:

template:169

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

Should be one of {}.'''.format(str([x['cuda'] for x in lib_records[library]])))

template:174

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

library, record[library], record['cuda'], destination))

template:201

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

install/cupy_builder/_command.py· 4

Device string "cuda"

ctx.features['cuda'].get_version())

template:141

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

os.path.join('cupy', 'cuda', 'cupy_cufilt.c'))

template:177

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

out_dir = os.path.join(self.build_lib, 'cupy', 'cuda')

template:179

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

editable_dst = os.path.join('cupy', 'cuda', 'cupy_cufilt.dll')

template:197

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

install/cupy_builder/_compiler.py· 2

Device string "cuda"

cuda_version = self._context.features['cuda'].get_version()

template:183

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

cuda_version = self._context.features['cuda'].get_version()

template:221

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

install/cupy_builder/_features.py· 3

Device string "cuda"

'name': 'cuda',

template:164

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

'cuda',

template:282

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

self.name = 'cuda'

template:439

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

install/cupy_builder/cupy_setup_build.py· 4

Device string "cuda"

cuda_version = ctx.features['cuda'].get_version()

template:145

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

if ctx.use_hip and module['name'] == 'cuda':

template:159

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

if module['name'] == 'cuda':

template:220

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

Device string "cuda"

if 'cuda' in ret:

template:228

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.

setup.py· 1

Device string "cuda"

cuda_major = ctx.features["cuda"].get_version() // 1000

template:84

The literal device string "cuda" resolves to the active AMD GPU on ROCm PyTorch. .to("cuda") / device="cuda" need no edits.

RecommendNo change required on ROCm PyTorch.