Commit Graph

63 Commits

Author SHA1 Message Date
Lianmin Zheng
86d10d220f Update grok.py and tiktoken tokenizer (#9532) 2025-08-23 05:40:18 -07:00
Cheng Wan
c0fb25e949 DP Enhancement (#8280) 2025-07-24 21:36:21 -07:00
Lianmin Zheng
38af4f68a9 Fix grammar abort & Minor style fixes (#7204) 2025-06-14 22:49:41 -07:00
Ke Bao
799c4bb502 Fuse MLA set kv cache kernel (#5748) 2025-04-26 18:42:22 -07:00
Ke Bao
6b6e748775 Remove q concat in FA3 backend for DeepSeek decode (#5638) 2025-04-22 11:43:12 -07:00
woodx
3bface15e6 Feat/support encoder model (like bert) (#4887) 2025-04-17 01:50:48 -07:00
Yun Dai
2695ab0537 Fix loading KV quantization scale; Enable modelopt kv cache (#4686)
Co-authored-by: qingquansong <ustcsqq@gmail.com>
2025-04-08 09:11:35 -07:00
Chang Su
f04c80dc42 Add Llama4 support (#5092)
Co-authored-by: Cheng Wan <cwan39@gatech.edu>
Co-authored-by: fzyzcjy <ch271828n@outlook.com>
Co-authored-by: ispobock <ispobaoke@163.com>
2025-04-07 00:29:36 -07:00
Qubitium-ModelCloud
56a724eba3 [QUANT] Add GPTQModel Dynamic Quantization + lm_head Quantization (#3790)
Signed-off-by: ZX-ModelCloud <zx@modelcloud.ai>
Co-authored-by: ZX-ModelCloud <zx@modelcloud.ai>
2025-03-05 01:11:00 -08:00
Lianmin Zheng
dc1881326f Fix perf regression on small batch sizes (#3008) 2025-01-20 03:39:49 -08:00
bjmsong
0bb0f76311 Support FP8 E4M3 KV Cache (#2786)
Co-authored-by: root <bjmsong@126.com>
2025-01-12 21:17:11 -08:00
Lianmin Zheng
641b7d0ae0 [Minor] Improve code style (#2422) 2024-12-09 06:30:35 -08:00
Ke Bao
ec52464dde MLA prefill w/o weight absorption (#2349) 2024-12-05 01:50:28 +08:00
Xuehai Pan
62a4a339eb docs: fix module docstrings and copyright headers (#2077) 2024-11-22 22:16:53 +08:00
Liangsheng Yin
100f5b8bc9 Simplify flashinfer dispatch (#1552) 2024-10-01 00:28:42 -07:00
Lianmin Zheng
36d5acfca5 Rename InputMetadata -> ForwardBatch (#1543) 2024-09-30 02:41:11 -07:00
Lianmin Zheng
fec185ce0c Refactor attention backend (#1381) 2024-09-11 11:44:26 -07:00
Lianmin Zheng
46094e0c1b Deprecate --disable-flashinfer and introduce --attention-backend (#1380) 2024-09-10 17:11:16 -07:00
Lianmin Zheng
3a6e8b6d78 [Minor] move triton attention kernels into a separate folder (#1379) 2024-09-10 15:15:08 -07:00
Liangsheng Yin
69b3bb9ae1 Unify forward mode (#1360) 2024-09-09 13:49:29 -07:00
Ke Bao
2c615d120f [Feature] Support fp8 e5m2 kv cache with flashinfer (#1204)
Co-authored-by: Yineng Zhang <me@zhyncs.com>
2024-08-25 17:38:11 -07:00
Ying Sheng
93d4e354d8 [Fix] Window attention compatible with RadixAttention and chunked prefill (#1112) 2024-08-15 10:33:20 -07:00
Ying Sheng
14cb544d56 [Fix] fix flashinfer usage for window attention (#1107) 2024-08-15 00:53:24 -07:00
Ying Sheng
96a2093ef0 [Fix] Compatibility of window attention and cuda graph (#1090) 2024-08-14 10:37:01 -07:00
Ying Sheng
0909bb0d2f [Feat] Add window attention for gemma-2 (#1056) 2024-08-13 17:01:26 -07:00
Lianmin Zheng
fb1f28cbbb Clean up the comments and names under python/sglang/srt/layers (#1047) 2024-08-12 05:54:37 +00:00
Liangsheng Yin
87e8c090e9 Organize code (rename, movement) (#953) 2024-08-06 20:50:32 -07:00
Ke Bao
e1eae1fd15 Support MLA for DeepSeek-V2 with Triton - step 1 (#905) 2024-08-05 03:40:33 +10:00
Ying Sheng
e7487b08bc Adjust default mem fraction to avoid OOM (#823) 2024-07-30 01:58:31 -07:00
Liangsheng Yin
cdcbde5fc3 Code structure refactor (#807) 2024-07-29 23:04:48 -07:00
Yineng Zhang
dd7e8b9421 chore: add copyright for srt (#790) 2024-07-28 23:07:12 +10:00
Lianmin Zheng
752e643007 Allow disabling flashinfer sampling kernel (#778) 2024-07-27 20:18:56 -07:00
Mingyi
a523a3c13a Reduce hardcoded logic of kernel usage (#707) 2024-07-23 16:42:21 -07:00
Ying Sheng
cf99eab7d5 Fix flashinfer (#700) 2024-07-23 01:27:01 -07:00
Liangsheng Yin
69d19188fc Decouple kv (#679) 2024-07-20 14:16:45 -07:00
Liangsheng Yin
a9ef49c12c Detokenize incrementally when streaming (#653) 2024-07-18 17:57:40 -07:00
Liangsheng Yin
abd5385ac5 Move global_server_args_dict (#642) 2024-07-17 13:49:15 -07:00
Ying Sheng
5949b1ca0e Fix memory pool index error (#616) 2024-07-13 16:45:11 -07:00
Liangsheng Yin
10143e1a5f Memorypool chunked prefetch (#614) 2024-07-13 15:24:03 -07:00
Lianmin Zheng
396a69240f Cleanup attention backend: flashinfer and triton (#611) 2024-07-12 18:21:11 -07:00
Lianmin Zheng
af4e7910e7 Clean up the usage of flashinfer (#610) 2024-07-12 13:00:03 -07:00
Lianmin Zheng
519e20cfda Code clean up: Remove deprecated prefill move InputMetadata to infer_batch.py (#609) 2024-07-12 12:28:09 -07:00
Liangsheng Yin
f25b76c02a add LogitsMetadata (#604) 2024-07-08 17:46:55 -07:00
Mingyi
f4e885b7c3 Reduce number of workspaces (#601) 2024-07-07 19:35:22 -07:00
Ying Sheng
dc1b8bcfaa Format (#593) 2024-07-05 10:06:17 -07:00
Ying Sheng
5a57b8addd Add Gemma2 (#592) 2024-07-05 09:48:54 -07:00
Ying Sheng
2a754e57b0 2x performance improvement for large prefill & Fix workspace conflicts (#579) 2024-07-03 16:14:57 -07:00
Ying Sheng
9380f50ff9 Turn on flashinfer by default (#578) 2024-07-02 02:25:07 -07:00
Yueyang Pan
d9ac639202 Fix flashinfer version (#576) 2024-07-01 22:08:39 -07:00
Lianmin Zheng
b7e2f800ac Update flashinfer to 0.0.5 (#554) 2024-06-20 20:29:06 -07:00