Commit Graph

  • 050b745a6e Merge pull request #643 from Azure-Tang/support-fp8 Azure 2025-02-25 16:22:12 +08:00
  • 36fbeee341 Update doc Azure 2025-02-25 08:21:18 +00:00
  • f639fbc19e feat: basic api key support ceerrep 2025-02-25 14:11:39 +08:00
  • 4dc5518e4d update fp8 kernel tutorial Azure 2025-02-24 15:37:01 +00:00
  • 7b2a6690ab Merge pull request #608 from makllama/fix_musa_ext Atream 2025-02-24 23:12:54 +08:00
  • 6f9ea689a9 Merge pull request #645 from makllama/torch2.2 Atream 2025-02-24 23:12:33 +08:00
  • f88c05a6f1 Ensure backward compatibility with Torch 2.2 Xiaodong Ye 2025-02-24 21:55:30 +08:00
  • 07eb712a73 Left out lazymio 2025-02-24 21:51:14 +08:00
  • 91062a834f Default values lazymio 2025-02-24 21:38:01 +08:00
  • 76487c4dcb Revert repetition_penalty as it is not in API spec lazymio 2025-02-24 21:30:03 +08:00
  • 05ad288453 Also /chat/completions lazymio 2025-02-24 21:08:36 +08:00
  • bf36547f98 Also allow repetition_penalty lazymio 2025-02-24 21:07:35 +08:00
  • 8704c09192 Allow temperature and top_p from requests lazymio 2025-02-24 21:01:33 +08:00
  • ca7366d2db Merge remote-tracking branch 'upstream/develop-0.2.2' into support-fp8 Azure 2025-02-24 11:58:10 +00:00
  • 581a524f65 Add data loader to read special weights for fp8; Add special weight process script Azure 2025-02-24 11:16:23 +00:00
  • e9b1216a9a Merge branch 'main' into feat-absorb-for-long-prefill Atream 2025-02-24 09:44:17 +00:00
  • 4b5991e77e Merge pull request #638 from kvcache-ai/feat-moonlight Atream 2025-02-24 17:32:05 +08:00
  • f327695079 fix KExpertsMarlin on GPU with out CUDA Graph Atream 2025-02-24 09:30:54 +00:00
  • cea07d1998 Feat: Clear cache during weight loading to prevent OOM on GPUs with <=8GB VRAM Yuhao Tsui 2025-02-24 10:09:42 +08:00
  • 706e69f4fc Fix missing macro definition for KTRANSFORMERS_USE_CUDA and <chrono> includes on MSVC akemimadoka 2025-02-24 01:37:50 +08:00
  • eb039b723d Merge pull request #621 from kvcache-ai/feat-moonlight Atream 2025-02-23 22:39:08 +08:00
  • f5f6c6b95d update yaml Atream 2025-02-23 14:33:58 +00:00
  • e8e02e5ccc support Moonlight Atream 2025-02-23 14:21:18 +00:00
  • 95d937c51d tmp DDong Jianwei 2025-02-23 18:51:42 +08:00
  • 006e8c6abc remove causal mask Atream 2025-02-23 07:40:47 +00:00
  • cdb6f896bb Merge pull request #612 from kvcache-ai/fix-bf16-load Atream 2025-02-23 15:37:23 +08:00
  • 036ae25a89 fix bf16 load, TODO: refactor cpu dequant Atream 2025-02-23 15:37:09 +08:00
  • 18b1d18367 musa: support bf16 Xiaodong Ye 2025-02-23 10:19:19 +08:00
  • 7b7c6a657d Add fp8 linear kernel;\n Add empty cache to fit in 16G VRAM; By 'wkGCaSS - 知乎 https://zhuanlan.zhihu.com/p/25491611225' Azure 2025-02-22 13:05:08 +00:00
  • 94ab2de3b9 Merge pull request #523 from miaooo0000OOOO/main Atream 2025-02-22 17:38:18 +08:00
  • 72d09f3f6e Merge pull request #597 from kvcache-ai/feat-more-context Atream 2025-02-22 17:17:09 +08:00
  • f7f1059873 fix merge bug, this branch also padding Marlin Atream 2025-02-22 09:00:09 +00:00
  • 4c568e04bb deploy: e90896314c Atream 2025-02-22 08:46:01 +00:00
  • e90896314c Merge pull request #577 from JiamingMai/dev Atream 2025-02-22 16:45:41 +08:00
  • 954796123c Merge pull request #582 from twobob/patch-1 Atream 2025-02-22 16:44:48 +08:00
  • 024009675e Merge branch 'main' into feat-more-context Atream 2025-02-22 06:17:39 +00:00
  • 5ec33d046d optimize gguf dequant, save mem, support Q2_K use marlin for lm_head, lm_head only calc last token for prefill extend context window to 19K for DeepSeek-V3/R1 within 24GB VRAM Atream 2025-02-22 06:13:01 +00:00
  • 5ed441a0f5 Update README.md _ 2025-02-21 14:15:50 +00:00
  • 45faddf668 fix the link addresses JiamingMai 2025-02-21 17:53:20 +08:00
  • 7e1fe256c8 optimize GPU Atream 2025-02-21 05:06:57 +00:00
  • 25c5bddd08 Merge pull request #506 from makllama/musa Azure 2025-02-20 22:50:31 +08:00
  • 1dd84b4a5b Merge pull request #550 from kvcache-ai/docker_dev ZiWei Yuan 2025-02-20 22:29:56 +08:00
  • ed74f91a25 ⚡ docker-iamge workflow liam 2025-02-20 22:27:51 +08:00
  • a6aee1c54a Merge branch 'main' of https://github.com/kvcache-ai/ktransformers into main liam 2025-02-20 22:00:55 +08:00
  • 64a00651b1 Merge pull request #232 from HandSonic/patch-1 ZiWei Yuan 2025-02-20 22:00:16 +08:00
  • 97e1dc97f6 ⚡ fix .so bug liam 2025-02-20 21:24:46 +08:00
  • 574b13390b ⚡ fin docker-image liam 2025-02-20 12:23:11 +08:00
  • b05cc5685d fix: fix SSE formatting xv44586 2025-02-20 15:01:35 +08:00
  • 0cdc446a19 默认并行编译 miaooo0000OOOO 2025-02-20 11:09:16 +08:00
  • 2207f6cd14 feat: Support Moore Threads GPU Xiaodong Ye 2025-02-19 18:26:55 +08:00
  • b7db4d3ad3 deploy: b4fb633991 Atream 2025-02-19 08:47:03 +00:00
  • b4fb633991 Merge pull request #501 from Zhoneym/main Atream 2025-02-19 16:46:48 +08:00
  • c92a300bd8 Necessary tips for Node.js related issues Zhoneym 2025-02-19 16:37:18 +08:00
  • c8d11c431f deploy: c9eb1f76f6 Atream 2025-02-19 05:47:02 +00:00
  • c9eb1f76f6 Merge pull request #384 from ubergarm/add-tutorial-notes Atream 2025-02-19 13:46:49 +08:00
  • 89f8218a2a Merge pull request #487 from kvcache-ai/clean_pr Atream 2025-02-19 12:46:48 +08:00
  • a529518346 clean PR code and disable flashinfer Atream 2025-02-19 04:42:47 +00:00
  • cf4da5fd47 Merge pull request #382 from ceerRep/server-prefix-cache Atream 2025-02-19 12:33:36 +08:00
  • be801a3645 deploy: ea75849d03 Azure-Tang 2025-02-19 03:19:30 +00:00
  • ea75849d03 Merge pull request #481 from Zhoneym/main Azure 2025-02-19 11:19:11 +08:00
  • c72291832d Update install.md Zhoneym 2025-02-19 10:39:43 +08:00
  • 3015e9dfa7 Modify and add any incorrect or missing content in the install.md file. Zhoneym 2025-02-19 10:01:16 +08:00
  • 584c7d5639 fix: object type for non-streaming response ceerrep 2025-02-18 23:49:51 +08:00
  • 6d45871de8 fix: workaround return dummy usage ceerrep 2025-02-18 22:39:49 +08:00
  • 09f5c5e77e Merge pull request #476 from kvcache-ai/V0.2.1.post1 v0.2.1.post1 ZiWei Yuan 2025-02-18 21:48:30 +08:00
  • 75d7d423b0 🔖 release v0.2.1.post1 liam 2025-02-18 20:45:48 +08:00
  • a435e1bbcd Merge pull request #406 from hoshinohikari/main Azure 2025-02-18 20:09:59 +08:00
  • db61375f1a Merge remote-tracking branch 'kvai/main' into server-prefix-cache ceerrep 2025-02-18 17:51:22 +08:00
  • 19fd24f461 Merge pull request #412 from haojiubujian1985/delete-unused-link Azure 2025-02-18 14:58:10 +08:00
  • 1c1769a579 Merge pull request #413 from kvcache-ai/fix_precision_MLA Atream 2025-02-18 14:44:50 +08:00
  • 592e13d453 ⚡ add mmlu_pro test liam 2025-02-18 14:43:38 +08:00
  • 1054ee3ae4 Merge pull request #450 from kvcache-ai/update-WeChatGroup Atream 2025-02-18 14:34:07 +08:00
  • 261d5d393f Add files via upload Atream 2025-02-18 14:28:29 +08:00
  • 0c1fcac2ea deploy: 2671fc7f4b Azure-Tang 2025-02-18 06:23:58 +00:00
  • 2671fc7f4b Merge pull request #445 from dedfaf/main Azure 2025-02-18 14:23:29 +08:00
  • 2259ff6749 deploy: ccf0f1910f UnicornChan 2025-02-18 05:59:00 +00:00
  • ccf0f1910f Merge pull request #442 from Zhoneym/patch-1 UnicornChan 2025-02-18 13:58:44 +08:00
  • 07a0555016 ⚡ fix device and add test liam 2025-02-18 12:52:17 +08:00
  • 73d072f609 Merge branch 'fix_precision_MLA' of https://github.com/kvcache-ai/ktransformers into server-prefix-cache ceerrep 2025-02-18 11:44:28 +08:00
  • 2ffb43f974 fix: adapt prefix cache in forward_linux_flashinfer ceerrep 2025-02-18 11:40:15 +08:00
  • f029588b61 fix server warmup Xie Weiyu 2025-02-18 11:39:45 +08:00
  • 8b221f532a Fix `GLIBCXX_3.4.30' not found in Anaconda Zhoneym 2025-02-18 11:32:42 +08:00
  • c70b6f4d5b fix: use 'cuda:0' by default if torch_device is 'cuda' ceerrep 2025-02-18 11:15:17 +08:00
  • 3e5c5a2f4a Fix typo in DeepseekR1_V3_tutorial.md Line 174 dedfaf 2025-02-18 10:57:12 +08:00
  • 9f1da18630 update install.sh Xie Weiyu 2025-02-17 20:50:19 +08:00
  • c176e516b5 server mix mla Xie Weiyu 2025-02-17 20:40:28 +08:00
  • ee24eb8dc3 fix: fix server for triton kernel ceerrep 2025-02-17 18:08:45 +08:00
  • bb1cadfff3 Merge branch 'fix_precision_MLA' of https://github.com/kvcache-ai/ktransformers into server-prefix-cache ceerrep 2025-02-17 18:08:04 +08:00
  • 038bc30888 fix precision bug imported by position_ids in 0.2.0 Atream 2025-02-17 09:23:14 +00:00
  • 49c6e2fc04 Update README_ZH.md xubo 2025-02-17 16:56:44 +08:00
  • 6d9182795c Fix cmake error caused by lack of environment variables in Windows environment hoshinohikari 2025-02-17 15:56:56 +08:00
  • cd9f7f8f34 fix: server: drop <think> tag in chat template ceerrep 2025-02-17 14:14:10 +08:00
  • bb39eeb005 Add notes to DeepSeek-R1 tutorial documentation John W. Leimgruber III 2025-02-16 13:40:27 -05:00
  • ca2090d89b feat: use model name in openai endpoint ceerrep 2025-02-17 00:11:54 +08:00
  • 5ac266085e fix: use flash_attn for faster prefill ceerrep 2025-02-17 00:11:24 +08:00
  • bb0ccc7b1a feat: add prefix cache for server ceerrep 2025-02-17 00:10:55 +08:00
  • 41a0b63068 deploy: 24de607be4 qiyuxinlin 2025-02-16 07:06:13 +00:00
  • 2f26a7efd6 deploy: c515cc49a5 qiyuxinlin 2025-02-16 07:05:55 +00:00
  • c515cc49a5 Merge pull request #355 from kvcache-ai/qiyuxinlin-patch-1 wang jiahao 2025-02-16 15:05:43 +08:00
  • 1a03698c94 deploy: 24de607be4 qiyuxinlin 2025-02-16 07:05:32 +00:00