Faraz ff9b561817 Fix TRTLLM MLA Cuda KV Blocks Causing accuracy drop (#9675)

2025-08-29 17:16:10 -07:00

..

attention

Fix TRTLLM MLA Cuda KV Blocks Causing accuracy drop (#9675)

2025-08-29 17:16:10 -07:00

moe

[Feature] Support NPUGraph for DeepSeek on Ascend NPU (#9355)

2025-08-28 16:06:24 -07:00

quantization

[ModelOpt] Fix Weight Loading for DSR1-FP4 Quantization (#9712)

2025-08-29 17:13:52 -07:00

activation.py

[AMD] Add silu_and_mul, gelu_and_mul, gelu_tanh_and_mul, and gelu_quick kernels for AMD GPUs (#7135)

2025-07-24 23:44:28 -07:00

amx_utils.py

[CPU] refine CPU integration code (#7647)

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

communicator.py

move is_sm90_supported/is_sm100_supported to python/sglang/srt/utils.py (#9679)

2025-08-27 03:34:29 -07:00

dp_attention.py

[Bugfix] fix kv buffer register & dp attention & deepepmoe (#9327)

2025-08-19 10:09:48 -07:00

elementwise.py

Update grok.py and tiktoken tokenizer (#9532)

2025-08-23 05:40:18 -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

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

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

linear.py

[ModelOpt] Fix Weight Loading for DSR1-FP4 Quantization (#9712)

2025-08-29 17:13:52 -07:00

logits_processor.py

feat: add original logprobs to response (#8375)

2025-08-29 11:43:57 -07:00

multimodal.py

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

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

parameter.py

[CPU][Qwen3 MoE] Enable fused_topk CPU fusion and enhance FP8 TP padding (#7838)

2025-07-09 02:04:21 -07:00

pooler.py

Feat/support rerank (#6058)

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

radix_attention.py

Update grok.py and tiktoken tokenizer (#9532)

2025-08-23 05:40:18 -07:00

rotary_embedding.py

torch.compile() mrope (#9487)

2025-08-21 23:01:08 -07:00

sampler.py

feat: add original logprobs to response (#8375)

2025-08-29 11:43:57 -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

move is_sm90_supported/is_sm100_supported to python/sglang/srt/utils.py (#9679)

2025-08-27 03:34:29 -07:00

vocab_parallel_embedding.py

Replace torch.jit.script with torch.compile in get_masked_input_and_mask to fix benchmark underreporting (#8733)

2025-08-03 19:02:51 -07:00