GAOXinyu 0cb595ad94 [bugfix] handle_x not define when using checkpoint_lvl = 2 (#502) ... when using checkpoint_lvl=2, we all_gather_raw(x) without async_op=True. So we don't need to wait for handle. Just skip.		2023-08-29 23:46:10 -07:00
..
layers	Run isort and black on python files	2023-08-18 14:22:11 -07:00
losses	Run isort and black on python files	2023-08-18 14:22:11 -07:00
models	[GPT] Generalize last_token_only arg to num_last_tokens	2023-08-26 20:47:53 -07:00
modules	Change causal for CrossAttention in mha.py to align to bottom right	2023-08-26 12:57:33 -07:00
ops	[bugfix] handle_x not define when using checkpoint_lvl = 2 (#502)	2023-08-29 23:46:10 -07:00
utils	[Gen] Minor fix to modify logits for top_p	2023-08-29 14:29:06 -07:00
__init__.py	Update Cutlass to v3.2.0	2023-08-27 23:47:28 -07:00
bert_padding.py	add unpad_input_for_concatenated_sequences (#499)	2023-08-29 02:23:56 -07:00
flash_attn_interface.py	Change causal mask to be aligned to bottom-right instead of top-left	2023-08-24 23:41:07 -07:00
flash_attn_triton_og.py	Run isort and black on python files	2023-08-18 14:22:11 -07:00
flash_attn_triton.py	Run isort and black on python files	2023-08-18 14:22:11 -07:00
flash_blocksparse_attention.py	Run isort and black on python files	2023-08-18 14:22:11 -07:00
flash_blocksparse_attn_interface.py	Run isort and black on python files	2023-08-18 14:22:11 -07:00
fused_softmax.py	Run isort and black on python files	2023-08-18 14:22:11 -07:00
pyproject.toml	Move pyproject.toml to flash-attn and tests dir to avoid PEP 517	2023-08-25 15:05:28 -07:00