RunFMe 332eabf309 Fix batched generation for prompts of different lengths (#2216) ... * fix ignoring of attention mask after prefill stage in decoding * update naming to avoid confusion --------- Co-authored-by: Неизвестный Пользователь722497 <dolegosmirnov@sberbank.ru>		2025-06-02 03:59:10 -07:00
..
__init__.py	vLLM Windows CUDA support [tested] (#2158)	2025-05-12 05:33:42 -07:00
_utils.py	Make replacement logic conscise	2025-06-01 05:57:43 +00:00
cohere.py	Bug fixes (#1516)	2025-01-07 04:23:14 -08:00
dpo.py	Update dpo.py	2025-02-13 14:59:42 -08:00
gemma.py	Prelim release	2025-03-01 00:13:11 -08:00
gemma2.py	Prelim release	2025-03-01 00:13:11 -08:00
granite.py	Update granite.py	2025-03-01 00:20:34 -08:00
llama.py	Fix batched generation for prompts of different lengths (#2216)	2025-06-02 03:59:10 -07:00
llama4.py	Qwen 3	2025-05-02 03:09:44 -07:00
loader.py	Fix Whisper, ModernBERT (#2565)	2025-05-17 05:11:50 -07:00
loader_utils.py	Update loader_utils.py	2025-05-12 21:06:30 -07:00
mapper.py	DeepSeek R1 Qwen	2025-05-30 01:38:53 -07:00
mistral.py	Prelim release	2025-03-01 00:13:11 -08:00
qwen2.py	Fix exec, eval (#743)	2024-07-07 09:33:01 -07:00
qwen3.py	Qwen 3	2025-05-02 03:09:44 -07:00
qwen3_moe.py	add comments and use modified function	2025-04-02 06:49:06 +00:00
rl.py	Bug fixes (#2651)	2025-05-29 09:59:29 -07:00
rl_replacements.py	Latest TRL, GRPO + Bug fixes (#2645)	2025-05-28 06:15:12 -07:00
vision.py	Fix Whisper, ModernBERT (#2565)	2025-05-17 05:11:50 -07:00