Commit graph

Select branches

Hide pull requests

main

Mono Color

• ddcdf47539
  Support Windows ARL release (#12183) Yuwen Hu 2024-10-11 18:30:52 +0800
• f983f1a8f4
  Add Qwen2-VL gpu example (#12135) Jinhe 2024-10-11 18:25:23 +0800
• 310f18c8af
  update NPU pipeline generate (#12182) Ruonan Wang 2024-10-11 17:39:20 +0800
• 1daab4531f
  Upgrade oneccl to 0.0.4 in serving-xpu image (#12185) Shaojun Liu 2024-10-11 16:54:50 +0800
• 724b2ae66d
  add npu-level0 pipeline.dll to ipex-llm (#12181) Shaojun Liu 2024-10-11 16:05:20 +0800
• 4d93bb81fe
  Initial support of NPU level0 Model (#12177) Ruonan Wang 2024-10-11 09:45:53 +0800
• ac44e98b7d
  Update Windows guide regarding LNL support (#12178) Yuwen Hu 2024-10-11 09:20:08 +0800
• 0ef7e1d101
  fix vllm docs (#12176) Guancheng Fu 2024-10-10 15:44:36 +0800
• 890662610b
  Fix auto importer for LNL release (#12175) Yuwen Hu 2024-10-10 15:17:43 +0800
• 535bee5381
  fix qwen2 vl again (#12174) Yishuo Wang 2024-10-10 13:50:01 +0800
• aef1f671bd
  Support LNL Windows release (#12169) Yuwen Hu 2024-10-09 17:41:10 +0800
• 78d253165d
  optimize qwen2 vl perf again (#12167) Yishuo Wang 2024-10-09 16:43:48 +0800
• 412cf8e20c
  [UPDATE] update mddocs/DockerGuides/vllm_docker_quickstart.md (#12166) Jun Wang 2024-10-09 11:19:32 +0800
• 3d044dbf53
  add llama3.2-vision Pytorch example (#12165) Zijie Li 2024-10-08 21:20:42 -0400
• e2ef9e938e
  Delete deprecated docs/readthedocs directory (#12164) Shaojun Liu 2024-10-08 14:48:02 +0800
• 644af2a76e
  add basic llama 3.2 vision support (#12163) Yishuo Wang 2024-10-08 10:46:48 +0800
• 9b75806d14
  Update Windows GPU quickstart regarding demo (#12124) Ch1y0q 2024-09-29 18:08:49 +0800
• 17c23cd759
  add llama3.2 GPU example (#12137) Ch1y0q 2024-09-29 14:41:54 +0800
• f71b38a994
  Update MiniCPM_V_26 GPU example with save & load (#12127) Yuwen Hu 2024-09-26 17:40:22 +0800
• 669ff1a97b
  fix sd1.5 (#12129) Yishuo Wang 2024-09-26 17:15:16 +0800
• a266528719
  optimize llama 3.2 rope (#12128) Yishuo Wang 2024-09-26 16:08:10 +0800
• 584c3489e7
  add basic support for llama3.2 (#12125) Yishuo Wang 2024-09-26 15:46:19 +0800
• 66f419f8b7
  fix qwen2 vl (#12126) Yishuo Wang 2024-09-26 15:44:02 +0800
• 2ea13d502f
  Add minicpm3 gpu example (#12114) Ch1y0q 2024-09-26 13:51:37 +0800
• 77af9bc5fa
  support passing None to low_bit in optimize_model (#12121) Yishuo Wang 2024-09-26 11:09:35 +0800
• 47e0b83cbf
  optimize sd 1.5 (#12119) Yishuo Wang 2024-09-25 15:45:13 +0800
• 2bedb17be7
  Add Qwen2.5 NPU Example (#12110) Jin, Qiao 2024-09-25 15:20:03 +0800
• 657889e3e4
  use english prompt by default (#12115) Shaojun Liu 2024-09-24 17:40:50 +0800
• 5d63aef60b
  optimize qwen2 vl again (#12109) Yishuo Wang 2024-09-23 13:22:01 +0800
• 03bd01c99c
  optimize npu qwen2 (#12107) Ruonan Wang 2024-09-20 04:46:16 -0700
• 02399021d6
  add npu load_low_bit api in all-in-one benchmark (#12103) Jinhe 2024-09-20 17:56:08 +0800
• 47a9597f24
  Add missing link for Qwen2.5 to CN-ZH readme (#12106) Yuwen Hu 2024-09-20 17:30:30 +0800
• 9239fd4f12
  add basic support and optimization for qwen2-vl (#12104) Yishuo Wang 2024-09-20 17:23:06 +0800
• 828fa01ad3
  [NPU] Add mixed_precision for Qwen2 7B (#12098) Yuwen Hu 2024-09-20 16:36:21 +0800
• 2269768e71
  add internvl2 example (#12102) Ch1y0q 2024-09-20 16:31:54 +0800
• ad1fe77fe6
  Add language switching (#12096) joan726 2024-09-20 16:05:20 +0800
• 09b8c80d9d
  update code for NPU qwen2 (#12094) Ruonan Wang 2024-09-20 00:58:32 -0700
• db7500bfd4
  Add Qwen2.5 GPU example (#12101) Jin, Qiao 2024-09-20 15:55:57 +0800
• b36359e2ab
  Fix xpu serving image oneccl (#12100) Guancheng Fu 2024-09-20 15:25:41 +0800
• 54b973c744
  fix ipex_llm import in transformers 4.45 (#12099) Yishuo Wang 2024-09-20 15:24:59 +0800
• a6cbc01911
  Use new oneccl for ipex-llm serving image (#12097) Guancheng Fu 2024-09-20 14:52:49 +0800
• 1295898830
  update vllm_online_benchmark script to support long input (#12095) Shaojun Liu 2024-09-20 14:18:30 +0800
• 9650bf616a
  add transpose_value_cache for NPU benchmark (#12092) Ch1y0q 2024-09-19 18:45:05 +0800
• f7fb3c896c
  Update lm_head optimization for Qwen2 7B (#12090) Yuwen Hu 2024-09-18 17:02:02 +0800
• ee33b93464
  Longbench: NV code to ipex-llm (#11662) Xu, Shuo 2024-09-18 15:55:14 +0800
• 40e463c66b
  Enable vllm load gptq model (#12083) Wang, Jian4 2024-09-18 14:41:00 +0800
• c2774e1a43
  Update oneccl to 0.0.3 in serving-xpu image (#12088) Xiangyu Tian 2024-09-18 14:29:17 +0800
• 081af41def
  [NPU] Optimize Qwen2 lm_head to use INT4 (#12072) Ruonan Wang 2024-09-14 00:26:46 -0700
• 18714ceac7
  Update README.md (#12084) joan726 2024-09-14 15:24:08 +0800
• b4b8c3e495
  add lowbit_path for generate.py, fix npu_model (#12077) Ch1y0q 2024-09-13 17:28:05 +0800
• d703e4f127
  Enable vllm multimodal minicpm-v-2-6 (#12074) Wang, Jian4 2024-09-13 13:28:35 +0800
• a767438546
  fix typo (#12076) Ruonan Wang 2024-09-12 20:44:42 -0700
• 3f0b24ae2b
  update cpp quickstart (#12075) Ruonan Wang 2024-09-12 20:35:32 -0700
• 9b4fee8b5b
  disable nightly release for finetune images (#12070) Shaojun Liu 2024-09-12 15:10:50 +0800
• beb876665d
  pin gradio version to fix connection error (#12069) Shaojun Liu 2024-09-12 14:36:09 +0800
• 48d9092b5a
  upgrade OneAPI version for cpp Windows (#12063) Ruonan Wang 2024-09-11 20:12:12 -0700
• e78e45ee01
  update NPU readme: run conhost as administrator (#12066) Jinhe 2024-09-11 17:54:04 +0800
• 4ca330da15
  Fix NPU load error message and add minicpm npu lowbit feat (#12064) Jinhe 2024-09-11 16:56:35 +0800
• 32e8362da7
  added minicpm cpu examples (#12027) Jinhe 2024-09-11 15:51:21 +0800
• a0c73c26d8
  clean NPU code (#12060) Ruonan Wang 2024-09-11 00:10:35 -0700
• c75f3dd874
  vllm no padding glm4 to avoid nan error (#12062) Wang, Jian4 2024-09-11 13:44:40 +0800
• 649390c464
  fix: textual and env variable adjustment (#12038) Chu,Youcheng 2024-09-11 13:38:01 +0800
• c94032f97e
  Try to fix llamaindex ut again (#12061) Yuwen Hu 2024-09-11 12:11:04 +0800
• 7e1e51d91a
  Update vllm setting (#12059) Shaojun Liu 2024-09-11 11:45:08 +0800
• 30a8680645
  Update for vllm one card padding (#12058) Wang, Jian4 2024-09-11 10:52:55 +0800
• c5fdfde1bd
  fix npu-model prompt (#12057) Zijie Li 2024-09-11 10:06:45 +0800
• 94dade9aca
  Fix UT of ipex_llm.llamaindex (#12055) Yuwen Hu 2024-09-11 09:58:43 +0800
• 52863dd567
  fix vllm_online_benchmark.py (#12056) Shaojun Liu 2024-09-11 09:45:30 +0800
• d8c044e79d
  optimize minicpm3 kv cache (#12052) Yishuo Wang 2024-09-10 16:51:21 +0800
• 5d3ab16a80
  Add vllm glm and baichuan padding (#12053) Wang, Jian4 2024-09-10 15:57:28 +0800
• 69c8d36f16
  Switching from vLLM v0.3.3 to vLLM 0.5.4 (#12042) Guancheng Fu 2024-09-10 15:37:43 +0800
• 73a4360f3f
  update lowbit path for baichuan2, qwen2, generate.py (#12051) Ch1y0q 2024-09-10 15:35:24 +0800
• dc4af02b2a
  Fix qwen2 1.5B NPU load error (#12049) Ruonan Wang 2024-09-09 23:41:18 -0700
• abc370728c
  optimize minicpm3 again (#12047) Yishuo Wang 2024-09-10 14:19:57 +0800
• f0061a9916
  remove local import os to fix Baichuan NPU load issue (#12044) Ch1y0q 2024-09-10 14:13:24 +0800
• 640998edea
  update inter_pp of qwen2 (#12041) Ruonan Wang 2024-09-09 19:34:17 -0700
• 048b4590aa
  add basic minicpm3 optimization (#12039) Yishuo Wang 2024-09-09 17:25:08 +0800
• 16c658e732
  LLM: add known issues to harness evaluation (#12036) Chu,Youcheng 2024-09-09 14:15:42 +0800
• 6cedb601e4
  remove some useless code (#12035) Yishuo Wang 2024-09-06 17:51:08 +0800
• d2e1b9aaff
  Add input padding during prefill for qwen2-7b (#12033) binbin Deng 2024-09-06 16:39:59 +0800
• f61b1785fb
  Small update to NPU example readme (#12034) Yuwen Hu 2024-09-06 15:54:23 +0800
• 0d04531ae0
  update NPU readme of Qwen2 (#12032) Ruonan Wang 2024-09-06 00:02:39 -0700
• 58555bd9de
  Optimize broadcast for npu llama (#12028) Yang Wang 2024-09-05 22:28:20 -0700
• e5581e6ded
  Select the Appropriate APT Repository Based on CPU Type (#12023) Shaojun Liu 2024-09-05 17:06:07 +0800
• 5b18bb3c4a
  Add recommend version for mtl npu (#12024) binbin Deng 2024-09-05 16:28:53 +0800
• 845e5dc89e
  Support lm_head of minicpm-2b on NPU (#12019) binbin Deng 2024-09-05 16:19:22 +0800
• 820f8a4554
  add --lowbit-path option for NPU llama example (#12020) Ch1y0q 2024-09-05 15:31:01 +0800
• 8803242f5c
  fix llama on cpu (#12018) Guoqiong Song 2024-09-04 19:17:54 -0700
• b3b2cd64b4
  Support lightweight-serving glm-4v-9b (#11994) Wang, Jian4 2024-09-05 09:25:08 +0800
• 75b19f8522
  revert actions/download-artifact version to 3 (#12017) Shaojun Liu 2024-09-04 22:39:07 +0800
• c6348a4666
  Update action.yml (#12016) Shaojun Liu 2024-09-04 22:12:24 +0800
• b1408a1f1c
  fix UT (#12005) Yishuo Wang 2024-09-04 18:02:49 +0800
• 77cb348220
  fix dependabot alerts (#12006) Shaojun Liu 2024-09-04 17:13:45 +0800
• 2b993ad479
  vllm update for glm-4 model automatic not_convert (#12003) Wang, Jian4 2024-09-04 13:50:32 +0800
• 9eaff5e47d
  add save & load support for NPU optimized model (#11999) Ruonan Wang 2024-09-03 05:53:22 -0700
• 6eb55653ba
  Performance mode strategy update for input_embeds input (#11997) Yuwen Hu 2024-09-03 17:46:16 +0800
• 164f47adbd
  MiniCPM-V-2 & MiniCPM-Llama3-V-2_5 example updates (#11988) Jinhe 2024-09-03 17:02:06 +0800
• 2e54f4402b
  Rename MiniCPM-V-2_6 CPU example (#11998) Jin, Qiao 2024-09-03 16:50:42 +0800
• 643458d8f0
  Update GraphRAG QuickStart (#11995) Yuwen Hu 2024-09-03 15:52:08 +0800
• 01099f08ee
  Revert prefill logic of qwen2-7b (#11992) binbin Deng 2024-09-03 14:45:01 +0800