Commit graph

Select branches

Hide pull requests

main

Mono Color

• 6f3441ba4c
  fix glm4-9b overflow (#12455) Yishuo Wang 2024-11-27 17:39:13 +0800
• 281c9b0bb9
  [NPU] Add L0 support for NPU C++ (#12454) Ruonan Wang 2024-11-27 01:04:13 -0800
• ce6fcaa9ba
  update transformers version in example of glm4 (#12453) Chu,Youcheng 2024-11-27 15:02:25 +0800
• effb9bb41c
  Small update to LangChain examples readme (#12452) Yuwen Hu 2024-11-27 14:02:25 +0800
• acd77d9e87
  Remove env variable BIGDL_LLM_XMX_DISABLED in documentation (#12445) Chu,Youcheng 2024-11-27 11:16:36 +0800
• f8c2bb2943
  [NPU] optimize qwen2 prefill performance for C++ (#12451) Ruonan Wang 2024-11-26 18:46:18 -0800
• 8331875f34
  Fix (#12390) Guancheng Fu 2024-11-27 10:41:58 +0800
• cb7b08948b
  update vllm-docker-quick-start for vllm0.6.2 (#12392) Jun Wang 2024-11-27 08:47:03 +0800
• 7b40f9b372
  [NPU] Support GW for NPU C++ (#12450) Ruonan Wang 2024-11-26 01:46:40 -0800
• c2efa264d9
  Update LangChain examples to use upstream (#12388) Jin, Qiao 2024-11-26 16:43:15 +0800
• 24b46b2b19
  [NPU] further fix of qwen2 int8 pipeline & C++ (#12449) Ruonan Wang 2024-11-26 00:39:39 -0800
• 303b104c10
  Fix abnormal output for Qwen2-7B when sym_int8 (#12446) Yuwen Hu 2024-11-26 15:53:04 +0800
• 71e1f11aa6
  update serving image runtime (#12433) Pepijn de Vos 2024-11-26 07:55:30 +0100
• 52c17fe104
  Optimize first token of C++ NPU by adding npu_dpu_groups (#12443) Ruonan Wang 2024-11-25 19:41:32 -0800
• 66bd7abae4
  add sdxl and lora-lcm optimization (#12444) Jinhe 2024-11-26 11:38:09 +0800
• 0e23bd779f
  Add support of llama3.2 for NPU C++ (#12442) Ruonan Wang 2024-11-25 17:26:55 -0800
• cdd41f5e4c
  optimize sdxl again (#12441) Yishuo Wang 2024-11-25 17:46:46 +0800
• b9abb8a285
  Support qwen2.5 3B for NPU & update related examples (#12438) Ruonan Wang 2024-11-25 00:38:31 -0800
• b633fbf26c
  add chinese prompt troubleshooting for npu cpp examples (#12437) Jinhe 2024-11-25 15:28:47 +0800
• 8164aed802
  small change (#12439) Yishuo Wang 2024-11-25 14:35:49 +0800
• be132c4209
  fix and optimize sd (#12436) Yishuo Wang 2024-11-25 14:09:48 +0800
• f41405368a
  Support minicpm for NPU C++ (#12434) Ruonan Wang 2024-11-24 18:42:02 -0800
• 0819fad34e
  support Llama2-7B / Llama3-8B for NPU C++ (#12431) Ruonan Wang 2024-11-22 02:47:19 -0800
• 4ffa6c752c
  New convert support for C++ NPU (#12430) Ruonan Wang 2024-11-21 22:28:30 -0800
• c089b6c10d
  Update english prompt to 34k (#12429) Shaojun Liu 2024-11-22 11:20:35 +0800
• e61ae88c5b
  Upgrade denpendency for xpu_lnl and xpu_arl option (#12424) Yuwen Hu 2024-11-21 18:37:15 +0800
• 2935e97610
  small fix of cpp readme(#12425) Ruonan Wang 2024-11-21 02:21:34 -0800
• 8fdc36c140
  Optimize with new batch kernel when batch_size=1 on LNL (#12419) Yuwen Hu 2024-11-21 16:21:35 +0800
• 7e0a840f74
  add optimization to openjourney (#12423) Jinhe 2024-11-21 15:23:51 +0800
• 145e8b480f
  update batch kernel condition (#12421) Yishuo Wang 2024-11-21 10:12:46 +0800
• 7288c759ce
  Initial NPU C++ Example (#12417) Ruonan Wang 2024-11-20 18:09:26 -0800
• d2a37b6ab2
  add Stable diffusion examples (#12418) Jinhe 2024-11-20 17:18:36 +0800
• 54c62feb74
  [NPU] dump prefill IR for further C++ solution (#12402) Ruonan Wang 2024-11-19 23:20:05 -0800
• 1bfcbc0640
  Add multimodal benchmark (#12415) Wang, Jian4 2024-11-20 14:21:13 +0800
• ff3f7cb25f
  Fix speech_paraformer issue with unexpected changes (#12416) SONG Ge 2024-11-18 23:01:20 -0800
• a9cb70a71c
  Add install_windows_gpu.zh-CN.md and install_linux_gpu.zh-CN.md (#12409) joan726 2024-11-19 14:39:53 +0800
• d6057f6dd2
  Update benchmark_vllm_throughput.py (#12414) Guancheng Fu 2024-11-19 10:41:43 +0800
• a69395f31f
  Support performance mode of GLM4 model (#12401) Yuwen Hu 2024-11-18 18:46:52 +0800
• d2c821d458
  Add missing arguments in pipeline parallel generate method (#12142) Song Fuchang 2024-11-18 13:50:18 +0800
• 3d5fbf2069
  update batch kernel condition (#12408) Yishuo Wang 2024-11-15 13:47:05 +0800
• 6c5e8fc70c
  fix again (#12407) Ruonan Wang 2024-11-15 11:57:58 +0800
• fcc0fa7316
  fix workflow again (#12406) Ruonan Wang 2024-11-15 11:01:35 +0800
• d1cde7fac4
  Tiny doc fix (#12405) Yuwen Hu 2024-11-15 10:28:38 +0800
• 548dec5185
  fix npu pipeline workflow (#12404) Ruonan Wang 2024-11-15 10:01:33 +0800
• d4d949443f
  [NPU] change attention_mask to fp16 (#12400) binbin Deng 2024-11-14 17:20:29 +0800
• 7e50ff113c
  Add padding_token=eos_token for GPU trl QLora example (#12398) Qiyuan Gong 2024-11-14 10:51:30 +0800
• d2cbcb060c
  Add initial support for modeling_xlm encoder on NPU (#12393) SONG Ge 2024-11-14 10:50:27 +0800
• 6726b198fd
  Update readme & doc for the vllm upgrade to v0.6.2 (#12399) Xu, Shuo 2024-11-14 10:28:15 +0800
• 59b01fa7d2
  small fix (#12397) Yina Chen 2024-11-14 04:03:36 +0200
• 00fce5c940
  use new q4_0 batch kernel (#12396) Yishuo Wang 2024-11-13 18:37:34 +0800
• d6d63d6b84
  [NPU] Qwen prefill attn_mask type hotfix (#12395) Yina Chen 2024-11-13 11:51:34 +0200
• 9220babaab
  qwen prefill attn_mask type fp16 (#12394) Yina Chen 2024-11-13 11:45:26 +0200
• 1158f91648
  Fix llava with multi-image inputs (#12384) Yuwen Hu 2024-11-13 09:27:50 +0800
• 27152476e1
  minor fix (#12389) Shaojun Liu 2024-11-12 22:36:43 +0800
• dd8964ba9c
  changed inference-cpp/Dockerfile (#12386) Xu, Shuo 2024-11-12 20:40:21 +0800
• 0ee54fc55f
  Upgrade to vllm 0.6.2 (#12338) Guancheng Fu 2024-11-12 20:35:34 +0800
• 4376fdee62
  Decouple the openwebui and the ollama. in inference-cpp-xpu dockerfile (#12382) Jun Wang 2024-11-12 20:15:23 +0800
• 6bf5a8c230
  [NPU] Update qwen2 compile config (#12383) Ruonan Wang 2024-11-12 16:59:44 +0800
• 7a97fbb779
  Support vpm and resampler module of minicpm-v on NPU (#12375) binbin Deng 2024-11-12 15:59:55 +0800
• 85c9279e6e
  Update llama-cpp docker usage (#12387) Wang, Jian4 2024-11-12 15:30:17 +0800
• c92d76b997
  Update oneccl-binding.patch (#12377) Shaojun Liu 2024-11-11 22:34:08 +0800
• e0918934c8
  Add fused_mlp to glm4v models (#12378) Yuwen Hu 2024-11-11 17:10:25 +0800
• dc34e8c51f
  optimize glm4v vision attention (#12369) Yishuo Wang 2024-11-08 17:01:57 +0800
• 2dfcc36825
  Fix trl version and padding in trl qlora example (#12368) Qiyuan Gong 2024-11-08 16:05:17 +0800
• fad15c8ca0
  Update fastchat demo script (#12367) Shaojun Liu 2024-11-08 15:42:17 +0800
• 51f7f87768
  fix ipex 2.3 bug (#12366) Yishuo Wang 2024-11-08 13:29:15 +0800
• b2e69a896c
  [NPU] Support Baichuan groupwise & gw code refactor (#12337) Yina Chen 2024-11-08 05:42:42 +0200
• 812d5cc32e
  [NPU L0] Support llama3.2 in L0 pipeline (#12361) binbin Deng 2024-11-08 10:01:23 +0800
• 7ef7696956
  update linux installation doc (#12365) Xin Qiu 2024-11-08 09:44:58 +0800
• 8fe294e01f
  Small fix to all-in-one benchmark (#12362) Yuwen Hu 2024-11-07 18:56:34 +0800
• 1a6cbc473f
  Add fused mlp optimizations to glm4 models (#12360) Yuwen Hu 2024-11-07 18:52:47 +0800
• 520af4e9b5
  Update install_linux_gpu.md (#12353) Xin Qiu 2024-11-07 16:08:01 +0800
• ad68c56573
  small improvement (#12359) Yishuo Wang 2024-11-07 15:57:41 +0800
• 71ea539351
  Add troubleshootings for ollama and llama.cpp (#12358) Jinhe 2024-11-07 15:49:20 +0800
• ce0c6ae423
  Update Readme for FastChat docker demo (#12354) Xu, Shuo 2024-11-07 15:22:42 +0800
• d880e534d2
  [NPU] acclib llama3.2 support groupwise (#12355) Yina Chen 2024-11-07 05:19:55 +0200
• 79f2877413
  add minicpm-v models to transformers_int4_npu_win api (#12352) Jinhe 2024-11-07 10:05:10 +0800
• a7b66683f1
  [NPU] Add Optimized Support for Llama3.2-1B/3B on NPU (#12339) SONG Ge 2024-11-06 19:21:40 +0800
• 872a74481a
  Small optimization to glm4 models (#12351) Yuwen Hu 2024-11-06 19:16:58 +0800
• c267355b35
  fix three NPU benchmark issues (#12350) Ruonan Wang 2024-11-06 19:01:01 +0800
• f24352aef9
  llama 3.1/3.2 support compresskv (#12347) Yina Chen 2024-11-06 11:33:43 +0200
• d984c0672a
  Add MiniCPM-V-2_6 to arc perf test (#12349) Jin, Qiao 2024-11-06 16:32:28 +0800
• e23ef7d088
  optimize glm4v's vision part (#12346) Yishuo Wang 2024-11-06 15:43:40 +0800
• c8b7265359
  Add basic glm4v support (#12345) Yishuo Wang 2024-11-06 13:50:10 +0800
• 69e3a56943
  [NPU] Hot fix of load_low_bit (#12344) binbin Deng 2024-11-06 10:07:00 +0800
• 899a30331a
  Replace gradio_web_server.patch to adjust webui (#12329) Xu, Shuo 2024-11-06 09:16:32 +0800
• 7240c283a3
  Add dummy model in iGPU perf (#12341) Jin, Qiao 2024-11-05 17:56:10 +0800
• 8e9a3a1158
  fix chatglm2 cpu ut (#12336) Zhao Changmin 2024-11-05 16:43:57 +0800
• d872639395
  [NPU] Llama3, Qwen2 1.5b, MiniCPM 1/2B groupwise support (#12327) Yina Chen 2024-11-05 09:51:31 +0200
• 82a61b5cf3
  Limit trl version in example (#12332) Jin, Qiao 2024-11-05 14:50:10 +0800
• 923d696854
  Small fix to LNL performance tests (#12333) Yuwen Hu 2024-11-05 13:24:58 +0800
• 45b0d371aa
  update benchmark readme (#12323) Zijie Li 2024-11-04 19:19:08 -0500
• e2adc974fd
  Small fix to LNL performance tests (#12331) Yuwen Hu 2024-11-04 19:22:41 +0800
• 522cdf8e9d
  Add initial support for LNL nightly performance tests (#12326) Yuwen Hu 2024-11-04 18:53:51 +0800
• 1b637e4477
  Add chatglm2&3 fuse mlp (#12328) Zhao Changmin 2024-11-04 18:04:41 +0800
• 94c4ce389f
  [NPU] Add env to disable compile opt (#12330) Yina Chen 2024-11-04 11:46:17 +0200
• e54af44ed6
  Add transformers_int4_npu_pipeline_win in all-in-one benchmark (#12325) Ch1y0q 2024-11-04 16:00:20 +0800
• 5ee6f97d6f
  [NPU L0] Add layernorm weight as const / input setting (#12322) binbin Deng 2024-11-04 15:46:24 +0800
• a01371f90b
  Doc: update harness readme (#12324) Chu,Youcheng 2024-11-04 14:58:54 +0800
• 4644cb640c
  Perf test further fix regarding trl version (#12321) Yuwen Hu 2024-11-04 11:01:25 +0800