aicloud
/
chatglm-6b
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Commit Graph

  • f831824845 Add test for modeling_chatglm main duzx16 2023-04-19 18:05:28 +0800
  • 35ca52301f Fix input embeds duzx16 2023-04-18 20:46:39 +0800
  • 0829959f96 Update slack link duzx16 2023-04-17 16:05:26 +0800
  • 4de8efebc8 Change mask positions to batch duzx16 2023-04-14 15:54:43 +0800
  • 3a99d7951d Always add gmask in token ids duzx16 2023-04-14 15:54:32 +0800
  • 53f019758b Fix bug duzx16 2023-04-14 15:13:54 +0800
  • eb55ff050e Add empty_init option duzx16 2023-04-13 20:33:07 +0800
  • 969290547e Update README duzx16 2023-04-13 15:43:34 +0800
  • aa51e62ddc Fix eos token in tokenizer duzx16 2023-04-11 13:24:10 +0800
  • cde457b39f Fix attention score on mps duzx16 2023-04-09 16:10:23 +0800
  • acd41f7731 Update dependency duzx16 2023-04-08 21:45:46 +0800
  • 6650ae3a53 Merge branch 'main' of https://huggingface.co/THUDM/chatglm-6b duzx16 2023-04-08 12:04:14 +0800
  • 7e69b85627 Fix tokenizer config saving duzx16 2023-04-08 12:04:08 +0800
  • 61eee50c9f Fix LogitsProcessor using slim checkpoint (#29) Zhengxiao Du 2023-04-08 02:54:27 +0000
  • 9324de70a9 Use gmask in first place duzx16 2023-04-06 23:25:10 +0800
  • d467effe91 Update slim checkpoint (#28) Zhengxiao Du 2023-04-06 14:47:57 +0000
  • 06a22a39fa Merge branch 'slim' of https://huggingface.co/THUDM/chatglm-6b into slim duzx16 2023-04-06 22:45:49 +0800
  • 36b7f2d0ad Add gmask token id duzx16 2023-04-06 22:43:35 +0800
  • 6461061e82 Update slim checkpoint Zhengxiao Du 2023-04-06 14:25:30 +0000
  • 63ce1bac4a Update code for slim duzx16 2023-04-06 22:19:30 +0800
  • 72985e820c Drop icetk dependency duzx16 2023-04-06 19:18:52 +0800
  • 551a50efec fix typo in use_gmask (#21) Zhengxiao Du 2023-04-05 11:11:33 +0000
  • 23ad39b571 Fix decode method for torch tensor duzx16 2023-04-05 18:26:09 +0800
  • fdb7a601d8 Support single integer or empty list as input to decode (#7) Zhengxiao Du 2023-04-04 09:48:07 +0000
  • f82b180d8d Fix position ids expand duzx16 2023-04-03 14:14:20 +0800
  • fb23542cfe Fix generate duzx16 2023-04-02 14:51:17 +0800
  • 08bc85104d Fix attention mask for prefix prompt duzx16 2023-04-02 02:25:03 +0800
  • 4b7ffbf070 No padding for chat function duzx16 2023-04-02 02:03:05 +0800
  • 373fd6b9d4 Fix attention_mask and position_ids duzx16 2023-04-02 01:58:45 +0800
  • e22cddf212 Fix encode method duzx16 2023-04-02 01:04:40 +0800
  • e1494f222d Fix batch input duzx16 2023-04-01 22:38:53 +0800
  • cc96a2271a Implement batch generation duzx16 2023-04-01 19:41:28 +0800
  • 11c270c26c Fix position id for training duzx16 2023-03-31 20:18:10 +0800
  • 9c7416d834 fix GLM6BBlock name typo (#20) Zhengxiao Du 2023-03-31 07:34:06 +0000
  • 2e1be30ac4 Add support for loading quantized model duzx16 2023-03-31 10:48:38 +0800
  • c949d03152 Use dynamic dtype for prompts duzx16 2023-03-31 01:13:32 +0800
  • 0cfae21ef8 Fix backward for quantization duzx16 2023-03-30 21:49:06 +0800
  • aea6cefcf5 Implement gradient checkpointing duzx16 2023-03-30 19:42:01 +0800
  • 0564795e6e Fix bugs duzx16 2023-03-30 17:35:58 +0800
  • 2200e2bc52 Add pad_token_id in config.json Fix position_ids in ChatGLMModel Add batch position_ids duzx16 2023-03-29 21:52:46 +0800
  • db2249979c Change padding side duzx16 2023-03-29 21:25:40 +0800
  • 5c64357295 Set ignore_index for CrossEntropyLoss duzx16 2023-03-29 21:19:38 +0800
  • 8127ab6abf Support batch training duzx16 2023-03-29 21:15:30 +0800
  • fbda1206cb Merge branch 'main' into dev_pt duzx16 2023-03-29 20:37:39 +0800
  • 812f43f9ff Add p-tuning v2 duzx16 2023-03-29 20:22:57 +0800
  • 096f3de6b4 Fix context length in get_position_ids duzx16 2023-03-28 17:37:46 +0800
  • 4a9b711e61 Close CPU fusion on Mac duzx16 2023-03-23 22:43:06 +0800
  • d2bbc82a2c Fix Chinese punctuation duzx16 2023-03-22 14:37:21 +0800
  • 2449bdc9d8 Add English duzx16 2023-03-21 23:27:46 +0800
  • 1b54948bb2 Fix typo in tokenization_chatglm.py songxxzp 2023-03-19 22:52:12 +0800
  • 2460dc2430 Remove hardcode bos_token_id duzx16 2023-03-19 14:56:15 +0800
  • 42095d42ff Add support for streaming output duzx16 2023-03-19 14:31:26 +0800
  • 220f772e9a Fix overflow in FP16 duzx16 2023-03-16 09:26:05 +0800
  • f9f74fda55 Set is_parallelizable to False duzx16 2023-03-16 00:30:43 +0800
  • 8492687842 Remove image tokens when decoding duzx16 2023-03-16 00:24:42 +0800
  • 0d8b08d1bb Revert "Update quantization" duzx16 2023-03-15 21:31:05 +0800
  • 0101afdfa7 Update quantization duzx16 2023-03-15 21:18:24 +0800
  • c3dece3f01 Add logit processor for NaN or Inf scores duzx16 2023-03-15 18:14:34 +0800
  • 9d1509a1ad Fix default history argument duzx16 2023-03-14 18:38:49 +0800
  • bcb053bda6 Merge branch 'main' of https://huggingface.co/THUDM/chatglm-6b duzx16 2023-03-14 14:49:47 +0800
  • d4832e8142 Add support for float32 duzx16 2023-03-14 14:49:14 +0800
  • a034f2a1ed Update requirements duzx16 2023-03-14 12:19:54 +0800
  • 3d0afeeed7 Update requirements duzx16 2023-03-14 12:15:19 +0800
  • 534e8b7045 Update README.md Zeng Aohan 2023-03-14 03:18:04 +0000
  • 6331ebe869 Update README.md Zeng Aohan 2023-03-14 03:17:32 +0000
  • 4cfd1ea181 Update README.md Zeng Aohan 2023-03-13 18:44:47 +0000
  • ac23f2a8f2 Update requirements duzx16 2023-03-14 02:21:04 +0800
  • 27d477b51f Update README duzx16 2023-03-14 02:15:35 +0800
  • 0f573d111a Update README duzx16 2023-03-14 02:11:53 +0800
  • cd8041ea53 Fix past_key_values duzx16 2023-03-14 02:08:43 +0800
  • 65bb3f00a7 add ice_text.model duzx16 2023-03-14 01:50:11 +0800
  • e8b5e82fbe delete ice_text.model duzx16 2023-03-14 01:49:46 +0800
  • 9671e13e0e Merge branch 'main' of https://huggingface.co/THUDM/chatglm-6b duzx16 2023-03-14 01:48:56 +0800
  • c4575e73d0 Update tokenizer duzx16 2023-03-14 01:32:34 +0800
  • 0772a2f0c8 Update README.md Zeng Aohan 2023-03-13 17:19:13 +0000
  • 7424d1fead Update README.md Zeng Aohan 2023-03-13 17:18:06 +0000
  • 452e0aa476 Update README.md Zeng Aohan 2023-03-13 17:17:31 +0000
  • eb1daff875 Update README.md Zeng Aohan 2023-03-13 17:16:28 +0000
  • a77bce8299 Update README.md Zeng Aohan 2023-03-13 17:15:46 +0000
  • fc937144cc Add model file Sengxian 2023-03-14 00:29:56 +0800
  • d11c6aaed8 Add chatglm-6b Sengxian 2023-03-14 00:21:01 +0800