iHore/vLLM
8
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity
Files
cddce79fdae2bb843e8988d7f0cd145e918b2d1f
vLLM/vllm/model_executor/models
T
History
Lucas WilkinsonandBenjamin Chislett 29255cfc3b [Spec-Decode] Support piecewise cudagraphs for Eagle head (#25109) 
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Co-authored-by: Benjamin Chislett <chislett.ben@gmail.com>
2025-10-10 01:20:31 -04:00
..
__init__.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
adapters.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
aimv2.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
apertus.py
[Models] Improve iteration over layers (#26425)
2025-10-08 20:48:33 +00:00
arcee.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
arctic.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
aria.py
[Model] Apply shared experts overlap optimization to all models with shared experts (#26145)
2025-10-09 11:31:04 -04:00
aya_vision.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
baichuan.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
bailing_moe.py
[Model] Apply shared experts overlap optimization to all models with shared experts (#26145)
2025-10-09 11:31:04 -04:00
bamba.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
bert_with_rope.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
bert.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
blip2.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
blip.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
bloom.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
chameleon.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
chatglm.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
clip.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
cohere2_vision.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
commandr.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
config.py
[Hybrid]: Decouple Kernel Block Size from KV Page Size (#24486)
2025-10-08 23:43:39 -07:00
dbrx.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
deepseek_eagle.py
[Bugfix] Allow skipping MoE in NVFP4 (fix for MTP) (#25987)
2025-10-06 16:16:30 -04:00
deepseek_mtp.py
[Spec-Decode] Support piecewise cudagraphs for Eagle head (#25109)
2025-10-10 01:20:31 -04:00
deepseek_v2.py
[Model] Apply shared experts overlap optimization to all models with shared experts (#26145)
2025-10-09 11:31:04 -04:00
deepseek_vl2.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
deepseek.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
dots1.py
[Model] Apply shared experts overlap optimization to all models with shared experts (#26145)
2025-10-09 11:31:04 -04:00
dots_ocr.py
[Models][Qwen] Replace pad with cat for better performance (#26486)
2025-10-09 14:51:26 +00:00
ernie45_moe.py
[Model] Apply shared experts overlap optimization to all models with shared experts (#26145)
2025-10-09 11:31:04 -04:00
ernie45_vl_moe.py
[Model] Apply shared experts overlap optimization to all models with shared experts (#26145)
2025-10-09 11:31:04 -04:00
ernie45_vl.py
[Models][Qwen] Replace pad with cat for better performance (#26486)
2025-10-09 14:51:26 +00:00
ernie45.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
ernie_mtp.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
exaone4.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
exaone.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
fairseq2_llama.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
falcon_h1.py
[Models] Improve iteration over layers (#26425)
2025-10-08 20:48:33 +00:00
falcon.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
fuyu.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
gemma2.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
gemma3_mm.py
Remove all references to yapf as it's no longer used (#26251)
2025-10-05 09:18:11 -07:00
gemma3.py
[Model] Gemma3: Fix GGUF loading and quantization (#26189)
2025-10-09 07:00:53 +00:00
gemma3n_mm.py
[Misc] Clean up unnecessary E501 ignore (#26274)
2025-10-06 07:29:18 +00:00
gemma3n.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
gemma.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
glm4_1v.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
glm4_moe_mtp.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
glm4_moe.py
[Model] Apply shared experts overlap optimization to all models with shared experts (#26145)
2025-10-09 11:31:04 -04:00
glm4.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
glm4v.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
glm.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
gpt2.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
gpt_bigcode.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
gpt_j.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
gpt_neox.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
gpt_oss.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
granite_speech.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
granite.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
granitemoe.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
granitemoehybrid.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
granitemoeshared.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
gritlm.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
grok1.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
h2ovl.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
hunyuan_v1.py
[Model] Apply shared experts overlap optimization to all models with shared experts (#26145)
2025-10-09 11:31:04 -04:00
hyperclovax_vision.py
Fix per file ruff ignores related to line length (#26262)
2025-10-06 05:12:40 +00:00
idefics2_vision_model.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
idefics3.py
Remove all references to yapf as it's no longer used (#26251)
2025-10-05 09:18:11 -07:00
interfaces_base.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
interfaces.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
intern_vit.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
internlm2_ve.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
internlm2.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
interns1_vit.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
interns1.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
internvl.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
jais.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
jamba.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
jina_vl.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
keye_vl1_5.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
keye.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
kimi_vl.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
lfm2_moe.py
[Models] Improve iteration over layers (#26425)
2025-10-08 20:48:33 +00:00
lfm2.py
[Model] Lfm2Moe (#26344)
2025-10-07 16:03:05 +00:00
llama4_eagle.py
Fix per file ruff ignores related to line length (#26262)
2025-10-06 05:12:40 +00:00
llama4.py
[Model] Apply shared experts overlap optimization to all models with shared experts (#26145)
2025-10-09 11:31:04 -04:00
llama_eagle3.py
[Spec-Decode] Support piecewise cudagraphs for Eagle head (#25109)
2025-10-10 01:20:31 -04:00
llama_eagle.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
llama.py
Support llama3 eagle3 head with llama4 verifier (#25961)
2025-10-06 13:56:08 -04:00
llava_next_video.py
[Model] Use merge_by_field_config for MM models (Llava family) (#26280)
2025-10-06 09:45:26 +00:00
llava_next.py
[Model] Use merge_by_field_config for MM models (Llava family) (#26280)
2025-10-06 09:45:26 +00:00
llava_onevision.py
[Model] Use merge_by_field_config for MM models (Llava family) (#26280)
2025-10-06 09:45:26 +00:00
llava.py
[Model] Use merge_by_field_config for MM models (Llava family) (#26280)
2025-10-06 09:45:26 +00:00
longcat_flash_mtp.py
Fix per file ruff ignores related to line length (#26262)
2025-10-06 05:12:40 +00:00
longcat_flash.py
[Models] Improve iteration over layers (#26425)
2025-10-08 20:48:33 +00:00
mamba2.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
mamba.py
[Models] Improve iteration over layers (#26425)
2025-10-08 20:48:33 +00:00
medusa.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
midashenglm.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
mimo_mtp.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
mimo.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
minicpm3.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
minicpm_eagle.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
minicpm.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
minicpmo.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
minicpmv.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
minimax_text_01.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
minimax_vl_01.py
[Model] Use merge_by_field_config for MM models (Llava family) (#26280)
2025-10-06 09:45:26 +00:00
mistral3.py
[Model] Use merge_by_field_config for MM models (Llava family) (#26280)
2025-10-06 09:45:26 +00:00
mixtral.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
mllama4.py
Support llama3 eagle3 head with llama4 verifier (#25961)
2025-10-06 13:56:08 -04:00
mlp_speculator.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
modernbert.py
[Model] Add support for ModernBertForTokenClassification (#26340)
2025-10-07 14:29:19 +00:00
module_mapping.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
molmo.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
moonvit.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
mpt.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
nano_nemotron_vl.py
[Model] Allow passing custom number of max tiles to Nano 2 VL (#26403)
2025-10-08 11:19:39 +00:00
nemotron_h.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
nemotron_nas.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
nemotron_vl.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
nemotron.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
nvlm_d.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
olmo2.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
olmo.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
olmoe.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
opt.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
orion.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
ovis2_5.py
[Model] Use merge_by_field_config for MM models (Ovis family) (#26308)
2025-10-07 12:54:22 +00:00
ovis.py
[Model] Use merge_by_field_config for MM models (Ovis family) (#26308)
2025-10-07 12:54:22 +00:00
paligemma.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
persimmon.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
phi3.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
phi3v.py
Remove all references to yapf as it's no longer used (#26251)
2025-10-05 09:18:11 -07:00
phi4_multimodal.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
phi4mm_audio.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
phi4mm_utils.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
phi4mm.py
Fix per file ruff ignores related to line length (#26262)
2025-10-06 05:12:40 +00:00
phi.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
phimoe.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
pixtral.py
Refactor MistralTokenizer (#26358)
2025-10-09 22:48:58 +00:00
plamo2.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
qwen2_5_omni_thinker.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
qwen2_5_vl.py
[Model] EVS support for nano_nemotron_vl (#26269)
2025-10-07 00:23:37 +08:00
qwen2_audio.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
qwen2_moe.py
[Model] Apply shared experts overlap optimization to all models with shared experts (#26145)
2025-10-09 11:31:04 -04:00
qwen2_rm.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
qwen2_vl.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
qwen2.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
qwen3_moe.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
qwen3_next_mtp.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
qwen3_next.py
[Model] Apply shared experts overlap optimization to all models with shared experts (#26145)
2025-10-09 11:31:04 -04:00
qwen3_vl_moe.py
[Models] Improve iteration over layers (#26425)
2025-10-08 20:48:33 +00:00
qwen3_vl.py
[Models][Qwen] Replace pad with cat for better performance (#26486)
2025-10-09 14:51:26 +00:00
qwen3.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
qwen_vl.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
qwen.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
radio.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
registry.py
[Misc] add usedforsecurity=False in md5 hash call (#26357)
2025-10-08 10:18:32 +00:00
roberta.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
rvl.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
seed_oss.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
siglip2navit.py
[Models][Qwen] Replace pad with cat for better performance (#26486)
2025-10-09 14:51:26 +00:00
siglip.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
skyworkr1v.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
smolvlm.py
Remove all references to yapf as it's no longer used (#26251)
2025-10-05 09:18:11 -07:00
solar.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
stablelm.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
starcoder2.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
step3_text.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
step3_vl.py
[Model] Define merge_by_field_config MM interface (R-T) (#26260)
2025-10-07 16:10:55 +08:00
swin.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
tarsier.py
[Model] Define merge_by_field_config MM interface (R-T) (#26260)
2025-10-07 16:10:55 +08:00
telechat2.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
teleflm.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
terratorch.py
[Model] Define merge_by_field_config MM interface (R-T) (#26260)
2025-10-07 16:10:55 +08:00
transformers_moe.py
[Bugfix] Fix gemma3 with transformers backend (#23178)
2025-10-06 18:42:32 +00:00
transformers_pooling.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
transformers.py
Enable RMSNorm substitution for Transformers backend (#26353)
2025-10-09 07:28:51 +00:00
ultravox.py
[Model] Define merge_by_field_config MM interface (U-Z) (#26261)
2025-10-07 06:45:49 +00:00
utils.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
vision.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00
voxtral.py
Refactor MistralTokenizer (#26358)
2025-10-09 22:48:58 +00:00
whisper.py
[Model] Define merge_by_field_config MM interface (U-Z) (#26261)
2025-10-07 06:45:49 +00:00
zamba2.py
Convert formatting to use ruff instead of yapf + isort (#26247)
2025-10-05 07:06:22 -07:00