Daniël de Kok 52e48739a5 Remove vLLM dependency for CUDA (#2751) ... * Remove vLLM dependency for CUDA This change adds `attention-kernels` as a dependency for paged attention and cache reshaping. With that, we don't use vLLM anywhere for CUDA. Tested run (since we don't have paged attention in CI): ``` ❯ ATTENTION=paged python -m pytest integration-tests -k "llama and awq" --release [...] 5 snapshots passed. ``` * Fix clippy warning		2024-11-17 17:34:50 +01:00
..
custom_modeling	Remove vLLM dependency for CUDA (#2751)	2024-11-17 17:34:50 +01:00
__init__.py	Fix: Change model_type from ssm to mamba (#2740)	2024-11-15 13:15:36 +01:00
bloom.py	Refactor dead code - Removing all `flash_xxx.py` files. (#2166)	2024-07-05 10:29:56 +02:00
causal_lm.py	feat: prefill chunking (#2600)	2024-10-16 12:49:33 +02:00
flash_causal_lm.py	Fixing linting on main. (#2719)	2024-11-04 15:21:41 +01:00
galactica.py	feat: add ruff and resolve issue (#2262)	2024-07-26 10:29:09 -04:00
globals.py	feat: prefill chunking (#2600)	2024-10-16 12:49:33 +02:00
idefics_causal_lm.py	feat: prefill chunking (#2600)	2024-10-16 12:49:33 +02:00
mamba.py	Choosing input/total tokens automatically based on available VRAM? (#2673)	2024-10-28 04:59:49 +01:00
metadata_kernels.py	Hotfixing auto length (warmup max_s was wrong). (#2716)	2024-11-04 09:55:54 +01:00
mllama_causal_lm.py	feat: add triton kernels to decrease latency of large batches (#2687)	2024-10-25 21:10:00 +00:00
model.py	Choosing input/total tokens automatically based on available VRAM? (#2673)	2024-10-28 04:59:49 +01:00
pali_gemma.py	feat: add ruff and resolve issue (#2262)	2024-07-26 10:29:09 -04:00
seq2seq_lm.py	feat: prefill chunking (#2600)	2024-10-16 12:49:33 +02:00
types.py	feat: prefill chunking (#2600)	2024-10-16 12:49:33 +02:00
vlm_causal_lm.py	fix cuda graphs for qwen2-vl (#2708)	2024-11-01 03:05:34 +01:00