c0de128 719ccfd773 Add unit tests for get_fp8_min_max() helper function ...

Add test_fp8_min_max_helper.py with mocked unit tests that verify:
- Standard FP8 dtype uses PyTorch's finfo values
- fnuz dtype on fnuz platform (MI300) returns 224.0, not 240.0
- Standard dtype on fnuz platform uses finfo values
- fnuz dtype on non-fnuz platform uses finfo values

These tests use mocking to verify the logic without requiring
actual ROCm hardware.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Signed-off-by: c0de128 <kevin.mckay@outlook.com>

2025-12-24 13:20:26 -06:00

..

nvfp4_utils.py

…

test_allspark_gemm.py

…

test_awq_triton.py

…

test_awq.py

[Bugfix] awq_gemm: fix argument order swap (#30364)

2025-12-14 18:15:37 +08:00

test_block_fp8.py

[CI/Build][AMD] Skip quantization kernels tests that require CUTLASS or e4m3fn when not supported by platform (#30020)

2025-12-10 02:28:37 +00:00

test_block_int8.py

[Misc] Make SchedulerConfig.max_model_len init-only (#28733)

2025-11-15 01:59:31 -08:00

test_cutlass_2of4_sparse.py

…

test_cutlass_scaled_mm.py

[CI/Build][AMD] Skip quantization kernels tests that require CUTLASS or e4m3fn when not supported by platform (#30020)

2025-12-10 02:28:37 +00:00

test_cutlass_w4a8_moe.py

[CI/Build][AMD] Skip test_cutlass_w4a8_moe tests on ROCm sine they require cutlass_pack_scale_fp8 (#30508)

2025-12-12 01:02:19 +00:00

test_cutlass_w4a8.py

[CI/Build][AMD] Skip quantization kernels tests that require CUTLASS or e4m3fn when not supported by platform (#30020)

2025-12-10 02:28:37 +00:00

test_flashinfer_nvfp4_scaled_mm.py

…

test_flashinfer_scaled_mm.py

…

test_fp8_min_max_helper.py

Add unit tests for get_fp8_min_max() helper function

2025-12-24 13:20:26 -06:00

test_fp8_quant_group.py

[CI/Build][Kernel][BugFix][AMD] Fix per_token_group_quant_fp8 to use correct fp8 min/max values and update atol/rtol in test_quantfp8_group_functionality (#30292)

2025-12-12 18:41:56 -05:00

test_fp8_quant.py

…

test_ggml.py

…

test_gguf.py

…

test_gptq.py

…

test_hadacore.py

[CI/Build][AMD] Skip marlin, machete, and hadacore tests since these require _C functions not defined for ROCm (#30109)

2025-12-06 12:54:17 +08:00

test_int8_kernel.py

…

test_int8_quant.py

…

test_machete_mm.py

[CI/Build][AMD] Skip marlin, machete, and hadacore tests since these require _C functions not defined for ROCm (#30109)

2025-12-06 12:54:17 +08:00

test_marlin_gemm.py

[CI/Build][AMD] Skip marlin, machete, and hadacore tests since these require _C functions not defined for ROCm (#30109)

2025-12-06 12:54:17 +08:00

test_mxfp4_qutlass.py

…

test_nvfp4_quant.py

[Bugfix] Use latency MOE backend as default for Flashinfer and other misc fixes (#27439)

2025-11-07 04:18:39 -08:00

test_nvfp4_qutlass.py

…

test_nvfp4_scaled_mm.py

…

test_per_token_group_quant.py

…

test_rocm_skinny_gemms.py

[platform] Move get_cu_count to utils (#27005)

2025-11-13 08:48:47 +08:00

test_scaled_mm_kernel_selection.py

[ROCm] Enable Triton ScaledMM fallback + kernel selection fix (#26668)

2025-12-12 13:28:20 -05:00

test_silu_mul_nvfp4_quant.py

…

test_triton_scaled_mm.py

…