Ke Bao 799c4bb502 Fuse MLA set kv cache kernel (#5748)

2025-04-26 18:42:22 -07:00

..

baichuan.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

bert.py

Feat/support encoder model (like bert) (#4887)

2025-04-17 01:50:48 -07:00

chatglm.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

clip.py

refactor: multimodal data (#4754)

2025-03-31 09:57:51 -07:00

commandr.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

dbrx.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

deepseek_janus_pro.py

refactor: multimodal data (#4754)

2025-03-31 09:57:51 -07:00

deepseek_nextn.py

Reorder loop in shared expert weight loading (#5719)

2025-04-25 17:27:42 -07:00

deepseek_v2.py

Fuse MLA set kv cache kernel (#5748)

2025-04-26 18:42:22 -07:00

deepseek_vl2.py

[Fix]: support deepseek-vl2-tiny model (#5552)

2025-04-26 17:52:53 +08:00

deepseek.py

[Fix]: support deepseek-vl2-tiny model (#5552)

2025-04-26 17:52:53 +08:00

exaone.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

gemma2_reward.py

[QUANT] Add GPTQModel Dynamic Quantization + lm_head Quantization (#3790)

2025-03-05 01:11:00 -08:00

gemma2.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

gemma3_causal.py

fix: gemma 3 not use softcap (#5622)

2025-04-22 01:16:08 -07:00

gemma3_mm.py

refactor: multimodal data (#4754)

2025-03-31 09:57:51 -07:00

gemma.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

gpt2.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

gpt_bigcode.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

granite.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

grok.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

internlm2_reward.py

[QUANT] Add GPTQModel Dynamic Quantization + lm_head Quantization (#3790)

2025-03-05 01:11:00 -08:00

internlm2.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

llama4.py

Optimize attention in llama4 (#5127)

2025-04-11 00:32:41 -07:00

llama_classification.py

[QUANT] Add GPTQModel Dynamic Quantization + lm_head Quantization (#3790)

2025-03-05 01:11:00 -08:00

llama_eagle3.py

Fix EAGLE3 for llama3.3 70b (#4716)

2025-03-24 17:31:19 -07:00

llama_eagle.py

[Feature] Support EAGLE 3 (#4247)

2025-03-18 07:35:23 -07:00

llama_embedding.py

[QUANT] Add GPTQModel Dynamic Quantization + lm_head Quantization (#3790)

2025-03-05 01:11:00 -08:00

llama_reward.py

[QUANT] Add GPTQModel Dynamic Quantization + lm_head Quantization (#3790)

2025-03-05 01:11:00 -08:00

llama.py

[Model Support] unsloth/Phi-4-mini bnb model (#4982)

2025-04-16 19:58:20 -07:00

llava.py

refactor: multimodal data (#4754)

2025-03-31 09:57:51 -07:00

llavavid.py

refactor: multimodal data (#4754)

2025-03-31 09:57:51 -07:00

minicpm3.py

[misc] remove is_cuda_available (#5319)

2025-04-20 18:16:51 -07:00

minicpm.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

minicpmo.py

vlm: enable radix cache for qwen-vl models (#5349)

2025-04-23 20:35:05 -07:00

minicpmv.py

refactor: multimodal data (#4754)

2025-03-31 09:57:51 -07:00

mistral.py

docs: fix module docstrings and copyright headers (#2077)

2024-11-22 22:16:53 +08:00

mixtral_quant.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

mixtral.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

mllama4.py

vlm: enable radix cache for qwen-vl models (#5349)

2025-04-23 20:35:05 -07:00

mllama.py

Support BNB quantization for llama/mllama (#5038)

2025-04-15 18:00:31 -07:00

olmo2.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

olmo.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

olmoe.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

phi3_small.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

qwen2_5_vl.py

vlm: enable radix cache for qwen-vl models (#5349)

2025-04-23 20:35:05 -07:00

qwen2_classification.py

[Model] Support Qwen2ForSequenceClassification (#4609)

2025-03-24 19:13:44 -07:00

qwen2_eagle.py

[QUANT] Add GPTQModel Dynamic Quantization + lm_head Quantization (#3790)

2025-03-05 01:11:00 -08:00

qwen2_moe.py

[Model] Adding Qwen3 and Qwen3MoE (#4693)

2025-04-18 09:51:29 -07:00

qwen2_rm.py

[QUANT] Add GPTQModel Dynamic Quantization + lm_head Quantization (#3790)

2025-03-05 01:11:00 -08:00

qwen2_vl.py

vlm: enable radix cache for qwen-vl models (#5349)

2025-04-23 20:35:05 -07:00

qwen2.py

[Model] Adding Qwen3 and Qwen3MoE (#4693)

2025-04-18 09:51:29 -07:00

qwen3_moe.py

[Model] Adding Qwen3 and Qwen3MoE (#4693)

2025-04-18 09:51:29 -07:00

qwen3.py

[Model] Adding Qwen3 and Qwen3MoE (#4693)

2025-04-18 09:51:29 -07:00

qwen.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

registry.py

Update model_loader deps and qqq quantization deps (#2220) (#2318)

2024-12-02 23:22:13 +08:00

stablelm.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

torch_native_llama.py

[QUANT] Add GPTQModel Dynamic Quantization + lm_head Quantization (#3790)

2025-03-05 01:11:00 -08:00

xverse_moe.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

xverse.py

Fix loading KV quantization scale; Enable modelopt kv cache (#4686)

2025-04-08 09:11:35 -07:00

yivl.py

[QUANT] Add GPTQModel Dynamic Quantization + lm_head Quantization (#3790)

2025-03-05 01:11:00 -08:00