Liangsheng Yin 5ea96ac7cc Reduce one step decode for draft model. (#11561)

2025-10-14 23:52:04 +08:00

..

attention

Reduce one step decode for draft model. (#11561)

2025-10-14 23:52:04 +08:00

moe

Add fused_moe_triton config: triton_3_4_0/E=256,N=256,device_name=NVIDIA_B200.json (#11587)

2025-10-14 13:24:43 +08:00

quantization

Tiny cleanup fp4 gemm calls (#11537)

2025-10-13 14:48:22 -07:00

activation.py

[Auto Sync] Update activation.py, chunk_cache.py, utils.py (20250917) (#10538)

2025-09-16 17:06:43 -07:00

amx_utils.py

[CPU] refine CPU integration code (#7647)

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

communicator.py

Depreate global_server_args_dict (#11528)

2025-10-13 19:34:43 +08:00

dp_attention.py

[Feat] Support Torch Symm Mem AllReduce (#10571)

2025-10-05 13:55:19 -07:00

elementwise.py

[Auto Sync] Update elementwise.py (20250923) (#10823)

2025-09-23 13:50:22 -07:00

flashinfer_comm_fusion.py

Support trtllm_allreduce_fusion in flashinfer for cuda<12.8 (#9339)

2025-08-20 16:54:30 -07:00

layernorm.py

Piecewise CUDA Graph Support & Torch Compile Backend (#10062)

2025-10-12 11:55:57 +08:00

linear.py

fix: fp8 quantization failure of qwen 2.5 VL 7B model (#10112)

2025-09-27 05:05:23 +00:00

logits_processor.py

Depreate global_server_args_dict (#11528)

2025-10-13 19:34:43 +08:00

model_parallel.py

Move parsers under a single folder (#9912)

2025-09-02 18:25:04 -07:00

modelopt_utils.py

Enable native ModelOpt quantization support (1/3) (#7149)

2025-10-06 13:24:15 -07:00

multimodal.py

[VLM] Improving multimodal tensor hash kernel (#9008)

2025-08-14 13:45:55 -07:00

parameter.py

fix: fp8 quantization failure of qwen 2.5 VL 7B model (#10112)

2025-09-27 05:05:23 +00:00

pooler.py

Feat/support rerank (#6058)

2025-06-16 10:50:01 -07:00

radix_attention.py

Compilation Folder Reset (#11539)

2025-10-14 09:19:12 +08:00

rocm_linear_utils.py

Optimized deepseek-v3/r1 model performance on mxfp4 run (#10008)

2025-09-04 15:11:22 -07:00

rotary_embedding.py

fix: fix video input for qwen3-vl (#11442)

2025-10-13 09:30:43 -07:00

sampler.py

Minor: improve sampler & remove unused fields from model_config.py (#11531)

2025-10-13 11:04:44 -07:00

torchao_utils.py

Support loading of larger models with on-the-fly quantization (#3061)

2025-01-22 21:33:17 -08:00

utils.py

fix: fp8 quantization failure of qwen 2.5 VL 7B model (#10112)

2025-09-27 05:05:23 +00:00

vocab_parallel_embedding.py

Revert "Revert the changes on NCCL symmetric memory" (#10238)

2025-09-09 12:11:49 -07:00