Alexander Matveev 019877253b

[Bugfix] multi-step + flashinfer: ensure cuda graph compatible (#8427)

2024-09-12 21:01:50 +00:00

..

backends

[Bugfix] multi-step + flashinfer: ensure cuda graph compatible (#8427)

2024-09-12 21:01:50 +00:00

ops

[Core/Bugfix] Add FP8 K/V Scale and dtype conversion for prefix/prefill Triton Kernel (#7208)

2024-08-12 22:47:41 +00:00

__init__.py

[Core] Add AttentionState abstraction (#7663)

2024-08-20 18:50:45 +00:00

layer.py

[Core] Subclass ModelRunner to support cross-attention & encoder sequences (towards eventual encoder/decoder model support) (#4942)

2024-08-06 16:51:47 -04:00

selector.py

[Core][Kernels] Enable FP8 KV Cache with Flashinfer backend. + BugFix for kv_cache_dtype=auto (#7985)

2024-08-29 14:53:11 -04:00