ViperEkura c6a82a5029 refactor: align linear backward dtype with weight ...

- cast gradients and inputs to weight.dtype instead of hardcoded bf16
- single code path covers bf16 and fp32 models, no branch needed
- gradient dtype now matches the leaf parameter dtype exactly

2026-08-14 01:01:58 +08:00

..

lib

refactor: move compiled kernel .so files into extension/lib

2026-07-31 15:36:32 +08:00

__init__.py

refactor : use factory for attention backends

2026-08-05 15:37:22 +08:00

attention_backend.py

perf: flatten paged prefill tile dispatch

2026-08-09 23:12:53 +08:00

attention_ops.py

perf: flatten paged prefill tile dispatch

2026-08-09 23:12:53 +08:00

fp8_dispatch.py

refactor: align linear backward dtype with weight

2026-08-14 01:01:58 +08:00

fp8_ops.py

feat: add fp8 training via cublasLt dispatch

2026-08-14 00:39:49 +08:00

loader.py

feat: add fp8 training via cublasLt dispatch

2026-08-14 00:39:49 +08:00

rotary_backend.py

refactor: unify rotary embedding interface and update docs

2026-07-31 16:52:25 +08:00

rotary_ops.py

refactor: unify rotary embedding interface and update docs

2026-07-31 16:52:25 +08:00