Commit Graph

  • d334a6f9f4 Merge pull request #528 from wangzhaode/feat/add_mnn_support jingyaogong 2025-11-10 22:46:15 +0800
  • f3441b0078
    Merge pull request #528 from wangzhaode/feat/add_mnn_support jingyaogong 2025-11-10 22:46:15 +0800
  • 7088cc886f [feat] add MNN support to README. yanxing 2025-11-10 21:56:27 +0800
  • 5959396096 [feat] add MNN support to README. yanxing 2025-11-10 21:56:27 +0800
  • 237744d58a Merge remote-tracking branch 'upstream/master' into feat/as_a_tools yuyu5333 2025-11-10 11:08:05 +0000
  • 095a656cff support ddp yuyu5333 2025-11-10 10:00:44 +0000
  • 768f0708e3 add default model para yuyu5333 2025-11-10 07:24:38 +0000
  • 9b27e8b716 update web auto port yuyu5333 2025-11-10 14:55:23 +0800
  • d7cc69d4de [fix] model-name jingyaogong 2025-11-07 19:38:20 +0800
  • bf60bde8fb [fix] model-name jingyaogong 2025-11-07 19:38:20 +0800
  • 48a81d99bb update logfile && process yuyu5333 2025-11-07 09:34:19 +0000
  • 5bef3f08ab add: rl-dpo yuyu5333 2025-11-07 05:13:13 +0000
  • e9d2ccbb2d fix training logs show yuyu5333 2025-11-07 03:11:22 +0000
  • 9bfb58ad1e Init Web UI yuyu5333 2025-11-06 22:34:22 +0800
  • 7de4b9ecd6 fix: Loading LoRA parameters which saved from multi-card training yuyu5333 2025-11-06 19:03:18 +0800
  • 23d341df97 [fix] harmonize template jingyaogong 2025-11-06 13:14:08 +0800
  • 81e869fc3e [fix] harmonize template jingyaogong 2025-11-06 13:14:08 +0800
  • 86d5cc70ec [feat] clear cache jingyaogong 2025-11-06 13:12:28 +0800
  • 509d8dacf1 [feat] clear cache jingyaogong 2025-11-06 13:12:28 +0800
  • d4f2339c1b [feat] update model install method Mr_Dwj 2025-11-05 11:17:07 +0800
  • f5079ce090
    Update eval_llm.py LearnMan 2025-11-03 20:47:25 +0800
  • 289ca01679 [fix] harmonize template jingyaogong 2025-11-02 23:18:11 +0800
  • 8a0b04ed82 [fix] harmonize template jingyaogong 2025-11-02 23:18:11 +0800
  • 7c3d660eec [feat] update import jingyaogong 2025-10-31 23:45:55 +0800
  • 0323815729 [feat] update import jingyaogong 2025-10-31 23:45:55 +0800
  • fdc6c5db5d [feat] update readme jingyaogong 2025-10-30 23:39:25 +0800
  • 8d71754e05 [feat] update readme jingyaogong 2025-10-30 23:39:25 +0800
  • f10df43275 [feat] update readme jingyaogong 2025-10-30 23:30:12 +0800
  • d8ac558ce2 [feat] update readme jingyaogong 2025-10-30 23:30:12 +0800
  • 1869dc8315 [feat] update readme jingyaogong 2025-10-30 23:27:15 +0800
  • e4807a5214 [feat] update readme jingyaogong 2025-10-30 23:27:15 +0800
  • e544126ab5 [feat] update datasets jingyaogong 2025-10-30 11:08:13 +0800
  • de23e1ea39 [feat] update datasets jingyaogong 2025-10-30 11:08:13 +0800
  • e840d8b83e [feat] update args jingyaogong 2025-10-30 10:48:31 +0800
  • 08ce3da228 [feat] update args jingyaogong 2025-10-30 10:48:31 +0800
  • ccab16a9b9 [feat] add args jingyaogong 2025-10-30 10:05:12 +0800
  • bf123b585d [feat] add args jingyaogong 2025-10-30 10:05:12 +0800
  • d3f9161cec [feat] update readme jingyaogong 2025-10-29 12:13:25 +0800
  • 800fed4639 [feat] update readme jingyaogong 2025-10-29 12:13:25 +0800
  • bfa2b4bf0f
    Merge pull request #1 from jingyaogong/master WenQiang 2025-10-29 10:51:24 +0800
  • 230b29b94a [fix] model device jingyaogong 2025-10-29 10:36:28 +0800
  • 1713c24114 [fix] model device jingyaogong 2025-10-29 10:36:28 +0800
  • e29f695998 [feat] update trainer jingyaogong 2025-10-29 00:52:37 +0800
  • acd5925193 [feat] update trainer jingyaogong 2025-10-29 00:52:37 +0800
  • cd576b58e8 [feat] update trainer jingyaogong 2025-10-28 23:30:10 +0800
  • 8f7e07b8ef [feat] update trainer jingyaogong 2025-10-28 23:30:10 +0800
  • f48568f62e [feat] update readme jingyaogong 2025-10-27 16:31:40 +0800
  • eb96113cd4 [feat] update readme jingyaogong 2025-10-27 16:31:40 +0800
  • 2718fe1960
    Update train_grpo.py Litar 2025-10-26 22:22:23 +0800
  • 3052baff06
    Update train_grpo.py Litar 2025-10-26 22:20:58 +0800
  • 16434d79c2
    Create train_grpo.md Litar 2025-10-26 22:14:55 +0800
  • 1eccfec438 [feat] update html jingyaogong 2025-10-26 19:00:24 +0800
  • db16c62412 [feat] update readme jingyaogong 2025-10-26 18:52:50 +0800
  • cc81925ecc [feat] update readme jingyaogong 2025-10-26 18:52:50 +0800
  • d74d8451c7 [feat] update eval-llm jingyaogong 2025-10-26 18:52:01 +0800
  • 35bed8936d [feat] update eval-llm jingyaogong 2025-10-26 18:52:01 +0800
  • 43726a7c34 [feat] pause-training jingyaogong 2025-10-26 18:49:52 +0800
  • e8484874f5 [feat] pause-training jingyaogong 2025-10-26 18:49:52 +0800
  • 717d465689
    Create RESOURCES.md for MiniMind project VinodHatti_ 2025-10-24 20:18:47 +0530
  • d8c848de7f
    Merge fd6448ce61 into 6efba3249a Zhenyu Wang 2025-10-24 01:19:19 +0800
  • 6c78df4c64 [feat] update readme jingyaogong 2025-10-24 01:18:33 +0800
  • 6efba3249a [feat] update readme jingyaogong 2025-10-24 01:18:33 +0800
  • bded87a804 [feat] update readme jingyaogong 2025-10-24 00:45:11 +0800
  • ea2abb5fb3 [feat] update readme jingyaogong 2025-10-24 00:45:11 +0800
  • c3ca123216 [feat] update requirements jingyaogong 2025-10-23 23:54:55 +0800
  • c4e9789d7e [feat] update requirements jingyaogong 2025-10-23 23:54:55 +0800
  • dcf04e9136 [feat] update requirements jingyaogong 2025-10-23 23:50:21 +0800
  • 4276f8a72f [feat] update requirements jingyaogong 2025-10-23 23:50:21 +0800
  • c99f12746e [feat] update readme jingyaogong 2025-10-23 23:23:52 +0800
  • d762926f48 [feat] update readme jingyaogong 2025-10-23 23:23:52 +0800
  • 5e72e0f4d5 [feat] update readme jingyaogong 2025-10-23 20:23:46 +0800
  • 09be7dba11 [feat] update readme jingyaogong 2025-10-23 20:23:46 +0800
  • 36b6ada80b [feat] repetition-penalty jingyaogong 2025-10-23 20:23:25 +0800
  • fa6df82ff8 [feat] repetition-penalty jingyaogong 2025-10-23 20:23:25 +0800
  • 8284d44039 [feat] convert2llama jingyaogong 2025-10-23 20:22:42 +0800
  • 28cc44579a [feat] convert2llama jingyaogong 2025-10-23 20:22:42 +0800
  • 8d1778a186 [feat] shuffle data jingyaogong 2025-10-23 20:13:28 +0800
  • a82526da11 [feat] shuffle data jingyaogong 2025-10-23 20:13:28 +0800
  • 3658bdb20d [fix] loss-issues-430 jingyaogong 2025-10-23 19:08:42 +0800
  • 805744e60a [fix] loss-issues-430 jingyaogong 2025-10-23 19:08:42 +0800
  • 00c24568dc [fix] restore jingyaogong 2025-10-23 19:00:06 +0800
  • 4014e62cdf [fix] restore jingyaogong 2025-10-23 19:00:06 +0800
  • fa935a496e [fix] issue-431 jingyaogong 2025-10-23 18:56:30 +0800
  • 557bcc018d [fix] issue-431 jingyaogong 2025-10-23 18:56:30 +0800
  • 1e4f464553 [fix] sampler-ddp jingyaogong 2025-10-23 15:03:19 +0800
  • 463044e92a [fix] sampler-ddp jingyaogong 2025-10-23 15:03:19 +0800
  • 41cacac703 Stabilize full SFT Wen 2025-10-22 23:52:57 -0700
  • faec18117e [fix] graph-oom & ddp-pos_cis jingyaogong 2025-10-23 14:22:13 +0800
  • fa7dff8291 [fix] graph-oom & ddp-pos_cis jingyaogong 2025-10-23 14:22:13 +0800
  • a6e1ceff9f [fix] git track jingyaogong 2025-10-21 23:09:00 +0800
  • ce693f8e7f [fix] git track jingyaogong 2025-10-21 23:09:00 +0800
  • 3d1de90df0 [feat] update readme jingyaogong 2025-10-21 23:02:16 +0800
  • d8bf813eac [feat] update readme jingyaogong 2025-10-21 23:02:16 +0800
  • f44ee7a1b0 [feat] update docs v2 jingyaogong 2025-10-21 22:24:30 +0800
  • c050af959a [feat] update website jingyaogong 2025-10-21 22:22:49 +0800
  • 249fb5b7ef [feat] minimind-2510 jingyaogong 2025-10-21 21:19:47 +0800
  • 7d877db79b [feat] minimind-2510 jingyaogong 2025-10-21 21:19:47 +0800
  • c7c8a00564
    Merge 79e36e17e8 into 3812a16d20 Howard-Wang 2025-10-18 16:37:46 +0200
  • 79e36e17e8 Fix: Remove attn_mask when using is_causal=True in Flash Attention wanghao 2025-10-18 21:02:46 +0800
  • c4a1b4ada9
    Merge d9ad86500f into 3812a16d20 Daisy 2025-10-18 14:01:16 +0530