Commits · 65d20f7fb81a0901dd2169eb5003922dafcd8402 · OpenDAS / vllm_cscc

30 Aug, 2024 1 commit
- fix error in test_long_context.py::test_batched_rope_kernel · 65d20f7f
  wangmin6 authored Aug 30, 2024
  
  65d20f7f
23 Aug, 2024 1 commit
- [Bugfix] adding chunking mechanism to fused_moe to handle large inputs · 4ac087d9
  zhuwenwen authored Aug 23, 2024
  
  4ac087d9
22 Aug, 2024 2 commits
- use opt layernorm_kernels · 4440e8c0
  zhuwenwen authored Aug 22, 2024
  
  4440e8c0
- use ori layernorm_kernels · bf278a88
  zhuwenwen authored Aug 22, 2024
  
  bf278a88
21 Aug, 2024 4 commits
- Update refactoring operation · 668ec4ef
  zhuwenwen authored Aug 21, 2024
  
  668ec4ef
- update VLLM_USE_OPT_OP to use opt kernels · 8a325c18
  zhuwenwen authored Aug 21, 2024
  
  8a325c18
- Add VLLM_USE_PA_PRINT_PARAM flag to print pa size · ef037256
  zhuwenwen authored Aug 21, 2024
  
  ef037256
- Update refactoring operation · bd93e661
  zhuwenwen authored Aug 21, 2024
  
  bd93e661
20 Aug, 2024 2 commits
- Refactoring the optimized kernel · 4405f82c
  zhuwenwen authored Aug 20, 2024
  
  4405f82c
- fix fa and triton tests · 3a6764a4
  zhuwenwen authored Aug 20, 2024
  
  3a6764a4
19 Aug, 2024 2 commits
- update log of seq_len · c7c03f73
  zhuwenwen authored Aug 19, 2024
  
  c7c03f73
- Add fa pad conditions and automatic switching strategy · fa973559
  zhuwenwen authored Aug 19, 2024
  
  fa973559
17 Aug, 2024 1 commit
- update fp16 model layout conversion conditions · a528f350
  zhuwenwen authored Aug 17, 2024
  
  a528f350
16 Aug, 2024 1 commit
- 添加要使用rocbla所需要的模块导入 · 1057935a
  gaoqiong authored Aug 16, 2024
  
  1057935a
15 Aug, 2024 4 commits
- 修改awq workspace 申请 · 70267267
  gaoqiong authored Aug 15, 2024
  
  70267267
- 修改awq workspace 申请 · 8376cc41
  gaoqiong authored Aug 15, 2024
  
  8376cc41
- update triton fa config for head_dim>128 · 3b2b3046
  zhuwenwen authored Aug 15, 2024
  
  3b2b3046
- 修改awq 申请sharememshapce方式 · 2d250236
  gaoqiong authored Aug 15, 2024
  
  2d250236
14 Aug, 2024 1 commit
- update the configuration of triton fa · 3e2c63a7
  zhuwenwen authored Aug 14, 2024
  
  3e2c63a7
13 Aug, 2024 1 commit
- update prefix_prefill params · f39feef5
  zhuwenwen authored Aug 13, 2024
  
  f39feef5
10 Aug, 2024 2 commits
- Revert feat:optimize act_and_mul_kernel · 880b2e41
  zhuwenwen authored Aug 10, 2024
  
  880b2e41
- add benchmarks to vllm and add env of VLLM_USE_FLASH_ATTN_AUTO · efb2f75f
  zhuwenwen authored Aug 10, 2024
  
  efb2f75f
09 Aug, 2024 1 commit
- 增添lmslim gptq支持 · dcaabcf7
  gaoqiong authored Aug 09, 2024
  
  dcaabcf7
08 Aug, 2024 3 commits
- fix awq index of qwen · 04633c47
  zhuwenwen authored Aug 08, 2024
  
  04633c47
- Fix multiprocessing shutdown errors · 708d897e
  zhuwenwen authored Aug 08, 2024
  
  708d897e
- The timeout for each iteration in the engine is set to 120 seconds by default · 3d087876
  zhuwenwen authored Aug 08, 2024
  
  3d087876
06 Aug, 2024 2 commits
- update version and add gptq_kernels info · c7e7ca7b
  zhuwenwen authored Aug 06, 2024
  
  c7e7ca7b
- update version and skip install gptq_kernels · 58fb0c33
  zhuwenwen authored Aug 06, 2024
  
  58fb0c33
04 Aug, 2024 3 commits
- add qwen awq support · 982c1545
  gaoqiong authored Aug 05, 2024
  
  982c1545
- add qwen2 awq support · ec98d390
  gaoqiong authored Aug 04, 2024
  
  ec98d390
- add llama model awq support · 5f5ddc3d
  gaoqiong authored Aug 04, 2024
  
  5f5ddc3d
01 Aug, 2024 1 commit
- support fa pad · ffbef65c
  zhuwenwen authored Aug 01, 2024
  
  ffbef65c
25 Jul, 2024 1 commit
- update prefix_prefill params · d69fb7b2
  zhuwenwen authored Jul 25, 2024
  
  d69fb7b2
24 Jul, 2024 2 commits
- add gemm pad and fa pad for 7b model · 0b5e4e11
  zhuwenwen authored Jul 24, 2024
  
  0b5e4e11
- add a list that supports nn model architectures · 0e7df36b
  zhuwenwen authored Jul 24, 2024
  
  0e7df36b
22 Jul, 2024 2 commits
- fix index error of baichuan2 · 795ce518
  zhuwenwen authored Jul 22, 2024
  
  795ce518
- fix gptq performance degradation when batch size>4 issue · f423ad60
  huangwb authored Jul 22, 2024
  
  f423ad60
20 Jul, 2024 3 commits
- use two method to add bias · 4caf1539
  zhuwenwen authored Jul 20, 2024
  
  4caf1539
- support nn layout · 1e0cb1f4
  zhuwenwen authored Jul 20, 2024
  
  1e0cb1f4
- 去掉调试信息 · 9653385f
  gaoqiong authored Jul 20, 2024
  
  9653385f