p04798526
/
LLaMA-Factory-310P3
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Commit Graph

  • 36932ddb55
         Merge branch 'main' into feature/Support-Qwenvl marko1616 2024-07-18 23:40:19 +0800
  • 779aae83d2 follow #4878 fix #4684 hiyouga 2024-07-18 22:06:12 +0800
  • 2516763d69
         Merge pull request #4878 from ly863/main hoshi-hiyouga 2024-07-18 22:03:41 +0800
  • 545e64afe6 Update metric.py 01WarpDrive 2024-07-18 16:10:15 +0800
  • 1e7b396ff2 仅仅训练最后一轮对话 Shiyu Zhang 2024-07-18 15:30:25 +0800
  • beec77a089 fix metrics #4786 hiyouga 2024-07-17 00:47:00 +0800
  • d774b94f12 support batch_eval_metrics, fix #4826 hiyouga 2024-07-17 00:33:00 +0800
  • 0543306b1c
         Merge pull request #5 from ZJLab-DataHub-Security/qianhao qianhao 2024-07-16 18:25:02 +0800
  • bda302fbfb tiny fix hiyouga 2024-07-15 23:09:50 +0800
  • abdc2fa1f1
         Merge branch 'hiyouga:main' into main Zhangchi Feng 2024-07-15 23:09:02 +0800
  • f2aaebdbde
         Merge pull request #4822 from codemayq/test-ci hoshi-hiyouga 2024-07-15 23:07:55 +0800
  • 10289eab15
         Update test_template.py hoshi-hiyouga 2024-07-15 23:04:39 +0800
  • da990f76b8
         Update test_template.py hoshi-hiyouga 2024-07-15 23:00:27 +0800
  • 38bc411d42
         Merge pull request #4821 from codemayq/feature-eval-split hoshi-hiyouga 2024-07-15 22:59:44 +0800
  • 91ba083f37
         Update llama3_lora_eval.yaml hoshi-hiyouga 2024-07-15 22:55:12 +0800
  • 33420bab81
         Update test_template.py hoshi-hiyouga 2024-07-15 22:55:05 +0800
  • 52a4256ad9
         Update test_template.py hoshi-hiyouga 2024-07-15 22:52:25 +0800
  • fd8cc49008 fix #4820 hiyouga 2024-07-15 22:32:07 +0800
  • b0aa321a4a update wechat hiyouga 2024-07-15 22:02:52 +0800
  • 89c28fb65f fix compute_loss for cpt qianhao0713 2024-07-15 20:30:09 +0800
  • 70bd600d8c add cpt test launch shell qianhao0713 2024-07-15 19:39:25 +0800
  • 34f70cec65 add dp&sp hybrid for cpt qianhao0713 2024-07-15 19:08:29 +0800
  • ca44c8dde8 solve the predict problem of llava-next-video and the multi-gpu finetuning problem of idefics2 BUAADreamer 2024-07-15 17:27:37 +0800
  • 92554c2d42
         Merge pull request #4 from ZJLab-DataHub-Security/qianhao_dev qianhao 2024-07-15 17:15:46 +0800
  • d5563d3030
         Merge branch 'hiyouga:main' into main Zhangchi Feng 2024-07-15 16:22:59 +0800
  • 9b360dde49
         Merge pull request #3 from ZJLab-DataHub-Security/qianhao luckyqsz 2024-07-15 16:10:30 +0800
  • 4a4ea30960 fix bug qianhao0713 2024-07-15 16:08:37 +0800
  • d31c8f764c fix 70b launch shell qianhao0713 2024-07-15 15:49:58 +0800
  • d5e513528e rename variables qianhao0713 2024-07-15 15:39:24 +0800
  • 32c3afdfa1 add IN_GITHUB_ACTIONS codingma 2024-07-15 10:28:07 +0800
  • 645211dc01 1. change the task name format 2. delete split param in data_args.py codingma 2024-07-15 09:55:33 +0800
  • 99ab7a8c1c allow computing rouge in training hiyouga 2024-07-15 01:16:26 +0800
  • 29ebcd75d5 fix up hiyouga 2024-07-15 01:04:56 +0800
  • 15b399a82f
         Merge pull request #4691 from codemayq/feature-suppot-eval-dataset hoshi-hiyouga 2024-07-15 01:00:34 +0800
  • cba673f491
         Update data_args.py hoshi-hiyouga 2024-07-15 00:56:03 +0800
  • df52fb05b1
         Update preprocess.py hoshi-hiyouga 2024-07-15 00:55:36 +0800
  • 84e4047f8a
         Update parser.py hoshi-hiyouga 2024-07-15 00:55:21 +0800
  • 97a0e291c7
         Update data_utils.py hoshi-hiyouga 2024-07-15 00:54:34 +0800
  • a5b809516e
         Update loader.py hoshi-hiyouga 2024-07-15 00:50:06 +0800
  • a4ae3ab4ab update test template hiyouga 2024-07-15 00:49:34 +0800
  • 3d39d74003
         Update parser.py hoshi-hiyouga 2024-07-14 23:04:34 +0800
  • 9d64507bd5
         Update README.md hoshi-hiyouga 2024-07-14 21:27:04 +0800
  • f1d8d29bc3 add gemma test hiyouga 2024-07-14 18:01:45 +0800
  • 173921419d fix test hiyouga 2024-07-14 15:44:30 +0800
  • 88a20ba797 fix #4699 hiyouga 2024-07-14 15:34:22 +0800
  • d3c01552e0 tiny fix hiyouga 2024-07-14 10:56:45 +0800
  • 2f6af73da2 fix gemma2 attention hiyouga 2024-07-13 23:33:45 +0800
  • 7b19e99ed7 update workflows hiyouga 2024-07-13 22:31:15 +0800
  • 5da54deb50
         Merge pull request #4781 from hzhaoy/fix-dockerfile-cuda hoshi-hiyouga 2024-07-13 22:25:32 +0800
  • 6b48308ef9 fix #4792 hiyouga 2024-07-13 22:07:58 +0800
  • 32699a82a6
         Merge pull request #4804 from codemayq/fix-examples hoshi-hiyouga 2024-07-13 20:49:13 +0800
  • f618b80fa2
         Update llava1_5.yaml hoshi-hiyouga 2024-07-13 20:30:06 +0800
  • 982a1cdd24 1. fix output_dir in llama3_lora_pretrain.yaml 2. add llava1_5.yaml for inference codingma 2024-07-13 13:16:22 +0800
  • 37d097a297 fix bug qianhao0713 2024-07-12 16:29:59 +0800
  • 83a307ce5f
         Update tuner.py lgqnb 2024-07-12 15:06:59 +0800
  • 2f8f583f13
         Update callbacks.py lgqnb 2024-07-12 15:06:24 +0800
  • 567347aa86
         Update finetuning_args.py lgqnb 2024-07-12 15:05:24 +0800
  • 294a855f90
         Update workflow.py lgqnb 2024-07-12 15:04:28 +0800
  • 061f9684a0
         Update trainer.py lgqnb 2024-07-12 15:03:54 +0800
  • 5544a757e6 fix bug qianhao0713 2024-07-12 14:34:44 +0800
  • b7007ece37 fix loss normalizer for sp&dp hybrid qianhao0713 2024-07-12 10:59:53 +0800
  • 8bab99c582 tiny fix hzhaoy 2024-07-12 00:28:44 +0800
  • 642c6d666f fix #4780 hzhaoy 2024-07-12 00:25:48 +0800
  • a8bf1abf0f fix #4779 hzhaoy 2024-07-12 00:15:15 +0800
  • c039a82c94 fix bug when sequence_parallel_size=-1 qianhao0713 2024-07-11 21:23:19 +0800
  • 67040f149c update wechat_npu.jpg codemayq 2024-07-11 20:03:39 +0800
  • fccd0c3b26 fix sp bug in hybrid sp&dp mode qianhao0713 2024-07-11 17:23:00 +0800
  • b4c8d2325b fix sp bug in hybrid sp&dp mode qianhao0713 2024-07-11 17:23:00 +0800
  • 2d7c8d4400
         Merge branch 'hiyouga:main' into main khazzz1c 2024-07-11 08:49:50 +0800
  • ba4e7686dc add sp_group_size params for launch shell qianhao 2024-07-10 09:39:48 +0000
  • 751c6a15e2 hybrid dp & sp for sft qianhao0713 2024-07-10 17:10:58 +0800
  • 555194e150
         Merge pull request #4700 from marko1616/patch-1 hoshi-hiyouga 2024-07-10 13:51:50 +0800
  • 40c3b88b68
         Merge pull request #4746 from yzoaim/fix hoshi-hiyouga 2024-07-10 13:32:49 +0800
  • 39cd89ce17
         Update callbacks.py hoshi-hiyouga 2024-07-10 13:32:20 +0800
  • cff89a2e89 fix src/llamafactory/train/callbacks.py -.- 2024-07-10 12:05:51 +0800
  • 51942acee8 fix #4731 hiyouga 2024-07-10 11:32:36 +0800
  • fb0c400116 fix ppo trainer hiyouga 2024-07-10 11:05:45 +0800
  • 2f09520c0d fix #4742 hiyouga 2024-07-09 23:24:24 +0800
  • 62367a4ab2
         Merge pull request #2 from ZJLab-DataHub-Security/bxf qianhao 2024-07-09 20:33:42 +0800
  • 49bd709da2 remove useless line qianhao0713 2024-07-09 20:17:24 +0800
  • 7544719497 merge easycontext into llamafactory qianhao0713 2024-07-09 20:03:11 +0800
  • 3a0f5b4fe7
         Merge pull request #1 from ZJLab-DataHub-Security/qianhao qianhao 2024-07-09 16:56:36 +0800
  • ab6bfb4874 merge easycontext qianhao 2024-07-09 15:00:52 +0800
  • 86b1594823 Update wechat.jpg hiyouga 2024-07-09 09:25:11 +0800
  • 8ce6d2b5e3
         Update README.md (Bill) Yuchen Lin 2024-07-08 18:02:11 -0700
  • 4e320b506f added ppo v2 from TRL khazic 2024-07-08 11:08:29 +0800
  • 633140a3a0
         Merge branch 'hiyouga:main' into main khazzz1c 2024-07-08 09:37:22 +0800
  • 563a27dab7
         Merge pull request #4706 from T-Atlas/main hoshi-hiyouga 2024-07-07 15:50:38 +0800
  • f84b007ebb
         Update packages.py hoshi-hiyouga 2024-07-07 15:48:29 +0800
  • 322663bf90 chore: Update vllm_engine.py to support vllm version >= 0.5.1 Lian Junhong 2024-07-07 15:08:12 +0800
  • a15782cb9f fix #4705 hiyouga 2024-07-07 13:10:06 +0800
  • b932bc0b8e Phi3v dataset processor fix. marko1616 2024-07-06 20:45:21 +0800
  • e0562521bb
         Update utils.py marko1616 2024-07-06 20:40:13 +0800
  • 53b1002fb7 add codegeex4, internlm2.5 hiyouga 2024-07-06 16:16:47 +0800
  • c9bb0757ec update pissa example hiyouga 2024-07-06 15:47:32 +0800
  • 1cb9f175f2 optimize the definition of the evaluation task. codingma 2024-07-05 16:17:38 +0800
  • 76f3bbcfc0 1. add custom eval dataset support 2. merge load dataset and split dataset function codingma 2024-07-05 15:52:10 +0800
  • 89c3066600
         Merge branch 'hiyouga:main' into main khazzz1c 2024-07-05 15:38:42 +0800
  • d18952ef32 support ollama modelfile export codingma 2024-07-05 09:55:09 +0800
  • 9f33f1edf5 fix processors hiyouga 2024-07-05 08:33:22 +0800