gnh1201/Open-Llama
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Commit Graph

  • 87f75f2dfe Fix huggingface download error main Namhyeon Go 2024-08-18 06:46:44 +0000
  • 9b8fe37cd1 Update requirements.txt Namhyeon Go 2024-08-18 06:21:38 +0000
  • eef8ae1477 Update Dockerfile Namhyeon Go 2024-06-15 18:59:00 +0000
  • e6df81ae9d Add docker-compose.yml Namhyeon Go 2024-06-15 18:32:23 +0000
  • bbb1d210aa Add Dockerfile Namhyeon Go 2024-06-15 18:31:53 +0000
  • 26a0ea81c0 Update requirements.txt Namhyeon Go 2024-06-15 17:12:47 +0000
  • a24271d48e Update data/download_the_pile.sh Namhyeon Go 2024-06-14 09:08:55 +0000
  • 5635f7d08d Update data/download_wudao.sh Namhyeon Go 2024-06-14 09:08:03 +0000
  • 0157b6938d update readme LiangSong 2023-05-17 22:45:04 +0700
  • 95973b5de1 update header LiangSong 2023-05-17 22:21:46 +0700
  • d269affb42 update readme LiangSong 2023-05-17 21:17:49 +0700
  • 6988c69884
         Merge pull request #64 from eltociear/patch-1 s-JoL 2023-05-16 22:24:13 +0700
  • 7bacd6cb93
         Update README.md Ikko Eltociear Ashimine 2023-05-17 00:11:35 +0900
  • 77b1c552c3 add discord invite link LiangSong 2023-05-15 23:00:25 +0700
  • 82c845a8ce update readme LiangSong 2023-05-15 00:21:13 +0800
  • 1ce8c18d83 add logo LiangSong 2023-05-14 10:52:43 +0800
  • 52e8df9a8d update readme LiangSong 2023-05-14 10:48:49 +0800
  • a07d9b0ac8 update readme LiangSong 2023-05-14 01:06:03 +0800
  • bf2cac0a45 update config LiangSong 2023-05-14 01:00:50 +0800
  • e18ead00cc update server LiangSong 2023-05-12 15:07:46 +0800
  • 7231d53ca4 update readme add new model LiangSong 2023-05-12 11:32:42 +0800
  • ceb1fd067b update vocab_size LiangSong 2023-05-11 14:15:12 +0800
  • 73dafa7ad6 add rounding vocab_size LiangSong 2023-05-10 17:49:52 +0800
  • 26f7421f05 add star history to readme LiangSong 2023-05-10 15:52:55 +0800
  • 72a6f81b61 update readme LiangSong 2023-05-09 18:47:29 +0800
  • 7d505ea303 update readme LiangSong 2023-05-09 17:03:13 +0800
  • 59b79af9d7 add comment LiangSong 2023-05-09 16:53:05 +0800
  • f6ac834ef9 update default config LiangSong 2023-05-09 15:16:50 +0800
  • 21fdd25b94 update Citation LiangSong 2023-05-09 15:12:49 +0800
  • 30ab306c56 update readme LiangSong 2023-05-09 15:06:47 +0800
  • 32583a41a7 update wudao download and preprocess LiangSong 2023-05-09 14:47:59 +0800
  • 7dc90c2558 fix typo LiangSong 2023-05-09 10:46:11 +0800
  • 6814fdb59e support gradient ckpt for peft LiangSong 2023-05-08 23:40:03 +0800
  • 3ba0c77053 update optimizer for lora LiangSong 2023-05-08 22:56:37 +0800
  • 58586112c1 fix table LiangSong 2023-05-08 22:30:02 +0800
  • 16811d0efe update readme LiangSong 2023-05-08 22:29:24 +0800
  • 92caa94490 support peft LiangSong 2023-05-08 22:26:39 +0800
  • 7da40f1c83 fix typo LiangSong 2023-05-08 19:00:06 +0800
  • 2df3e622e9 update readme LiangSong 2023-05-08 18:59:01 +0800
  • ec2b4d6ee7 fix split by shard bug LiangSong 2023-05-08 14:03:05 +0800
  • 4a1e7bb44b Optimized the structure of configs, added support for deepspeed stage3, reduced memory usage by using Auto class to load models, and added support for training 65B models. LiangSong 2023-05-06 23:37:17 +0800
  • 5b1f6a4861 fix epoch bug LiangSong 2023-05-06 09:45:37 +0800
  • f893a0f5b8 update dataset LiangSong 2023-05-05 19:23:16 +0800
  • 758af69c73 update science instruct-tuning datasets LiangSong 2023-05-05 19:00:37 +0800
  • d24b4cce54 update preprocess format LiangSong 2023-05-05 18:20:59 +0800
  • 85caa97a6a add xP3 dataset and belle_2M LiangSong 2023-05-05 17:05:41 +0800
  • 00cbdbbf26 fix typo LiangSong 2023-05-04 22:55:40 +0800
  • 693e3970d9 update readme LiangSong 2023-05-04 22:54:10 +0800
  • fbb7997607 fix typo LiangSong 2023-05-04 22:32:15 +0800
  • 98ffab3a97 update readme and add half to server LiangSong 2023-05-04 22:28:36 +0800
  • 5c876121cb update gradio, fix code format bug LiangSong 2023-05-04 18:18:52 +0800
  • a1acc90988 fix train_tokenizer bug LiangSong 2023-05-04 16:00:56 +0800
  • 51686b5fb8 add split dataset by shard option to accelerate data loading LiangSong 2023-05-04 09:20:23 +0800
  • f0d41f937b update instruct_config and set all random seed to 42 LiangSong 2023-05-04 08:45:21 +0800
  • dba2e2d680 update ShareGPT_90K preprocess LiangSong 2023-05-04 08:34:38 +0800
  • 154456c976 set dataset shuffle seed to 42 LiangSong 2023-05-04 00:31:12 +0800
  • c2184c6dd1 support multiple epochs LiangSong 2023-05-03 00:02:01 +0800
  • f05e929aad update config LiangSong 2023-05-02 21:42:55 +0800
  • 0466673f76 support load model from accelerate ckpt LiangSong 2023-04-29 20:40:42 +0800
  • 52cd09f664 update readme LiangSong 2023-04-29 20:30:24 +0800
  • fc21a75d1e add continue training LiangSong 2023-04-29 20:28:39 +0800
  • 28b11a5bed update requirements LiangSong 2023-04-29 13:39:03 +0800
  • 8b439dec4a update flops LiangSong 2023-04-29 12:31:11 +0800
  • a2816bd23d update readme LiangSong 2023-04-29 12:06:55 +0800
  • 4c5e50e4aa update readme LiangSong 2023-04-29 11:41:28 +0800
  • c8037746c3 update readme LiangSong 2023-04-28 22:45:45 +0800
  • 0ff8b2353f
         Merge pull request #30 from s-JoL/dev s-JoL 2023-04-28 19:54:52 +0800
  • 724265b435 update readme dev LiangSong 2023-04-28 19:54:14 +0800
  • 0fd7dbd636
         Merge pull request #29 from s-JoL/dev s-JoL 2023-04-28 19:50:29 +0800
  • 8c85535db3 update readme LiangSong 2023-04-28 19:49:51 +0800
  • 676dcfd995 add hardward configuration to readme LiangSong 2023-04-28 17:29:11 +0800
  • f3c664bde3
         Merge pull request #25 from s-JoL/dev v2 s-JoL 2023-04-28 15:11:02 +0800
  • c890bce69c update readme LiangSong 2023-04-28 15:10:41 +0800
  • 9baebfd49c Merge branch 'main' into dev LiangSong 2023-04-28 15:08:25 +0800
  • 2fd13ff075 fix typo LiangSong 2023-04-28 15:05:33 +0800
  • 0fdca8b949 update readme LiangSong 2023-04-28 15:01:01 +0800
  • 49118aad42 update header config and add padding to concat_multiple_sequence LiangSong 2023-04-27 23:42:11 +0800
  • db6cdb51d0 unified pre-training and instrcution-tuning both use train_lm and dataset LiangSong 2023-04-27 19:42:06 +0800
  • 97aff0e051 use split_dataset_by_node instead accelerate.prepare to accelerate data loading by 50% LiangSong 2023-04-27 00:04:11 +0800
  • 0377b43628 update tokenizer to LlamaTokenizer LiangSong 2023-04-26 18:53:30 +0800
  • f41f5558ec update header LiangSong 2023-04-24 23:19:07 +0800
  • f8f4cde228 using huggingface datasets to accelerate training, using open-llama to pretrain LiangSong 2023-04-24 19:13:53 +0800
  • 92af968637
         Update README.md v1.0 s-JoL 2023-04-23 16:26:58 +0800
  • cf852bc459
         Update README.md s-JoL 2023-04-23 16:26:21 +0800
  • ad3d943a7d update readme add ckpt from hf LiangSong 2023-04-16 23:50:36 +0800
  • b21441b14b disable concat docs LiangSong 2023-04-15 19:35:24 +0800
  • 3f62a23ee2 update format LiangSong 2023-04-12 22:16:15 +0800
  • a4aa109dd3 add trainer and utils LiangSong 2023-04-12 17:59:05 +0800
  • ae0691c509 update utils LiangSong 2023-04-12 17:15:40 +0800
  • da1c927016 update speed test LiangSong 2023-04-12 17:15:07 +0800
  • 0ee9612f40 add speed test LiangSong 2023-04-11 21:59:18 +0800
  • 4cb94d2687
         Update README.md s-JoL 2023-04-11 15:59:02 +0800
  • d2632467ec
         Update README_en.md s-JoL 2023-04-11 15:53:43 +0800
  • ce8bc5249f
         Update README_en.md s-JoL 2023-04-11 15:53:05 +0800
  • be2f0960c7
         Update README.md s-JoL 2023-04-11 15:51:43 +0800
  • 5f7a4a69d3
         Merge pull request #3 from Bayes-Song/dev S 2023-04-09 22:49:40 +0800
  • f9e7a3376a update readme LiangSong 2023-04-09 22:48:56 +0800
  • ce06d9feab
         Merge pull request #2 from Bayes-Song/dev S 2023-04-08 00:05:02 +0800
  • 00cda9e265 update readme_en LiangSong 2023-04-08 00:04:11 +0800
  • 56f71e24df
         Merge pull request #1 from Bayes-Song/dev S 2023-04-07 23:21:06 +0800