tocmo0nlord/axolotl
1
0
Fork 0
Code Issues Pull Requests Actions 3 Packages Projects Releases Wiki Activity

Commit Graph

  • cb18856fc2 Merge pull request #85 from NanoCode012/fix/add-dataset-shard-readme NanoCode012 2023-05-27 23:52:50 +09:00
  • 8626b54aab Add dataset_shard_num and dataset_shard_idx NanoCode012 2023-05-27 23:51:17 +09:00
  • 87dffbc451 Merge pull request #75 from Thytu/refactor/rename-4b-to-gptq Wing Lian 2023-05-27 09:37:57 -04:00
  • 147241ca66 Merge branch 'main' into refactor/rename-4b-to-gptq Wing Lian 2023-05-27 09:37:52 -04:00
  • 7e974decb7 Merge pull request #76 from OpenAccess-AI-Collective/truthy-validation Wing Lian 2023-05-27 09:36:10 -04:00
  • 11fd39b1f5 Merge pull request #78 from OpenAccess-AI-Collective/falcoln-support Wing Lian 2023-05-27 09:35:56 -04:00
  • 157420df13 sane starter defaults and add lora Wing Lian 2023-05-27 09:33:14 -04:00
  • 679ffd7395 Merge pull request #77 from OpenAccess-AI-Collective/falcoln-support Wing Lian 2023-05-27 09:18:48 -04:00
  • d5f944ce2a add example for falcoln support Wing Lian 2023-05-27 09:16:43 -04:00
  • 4c906339f7 fix auto linear modules for lora w/o any set already Wing Lian 2023-05-27 08:49:43 -04:00
  • 4c500f5830 checking for False is not sufficent for NoneType/unset configs Wing Lian 2023-05-27 08:43:48 -04:00
  • 7cf07fc8b3 refactor(example): rename 4bit-lora-7b by gptq-lora-7b Thytu 2023-05-27 12:37:53 +00:00
  • dd0065773a refactor(param): rename load_4bit config param by gptq Thytu 2023-05-27 12:36:03 +00:00
  • ca1bb92337 Update src/axolotl/utils/models.py Wing Lian 2023-05-26 17:51:24 -04:00
  • 933e970cb5 Update src/axolotl/utils/models.py Wing Lian 2023-05-26 17:51:17 -04:00
  • c3d256271e fix wheel install glob Wing Lian 2023-05-26 10:37:02 -04:00
  • 46c5a44003 Merge pull request #69 from OpenAccess-AI-Collective/NanoCode012-quickstart-disable-xformers NanoCode012 2023-05-26 22:40:16 +09:00
  • ec3c0314bf Merge pull request #65 from NanoCode012/feat/target-linear NanoCode012 2023-05-26 22:39:38 +09:00
  • 79560934f9 Disable formers for QuickStart config NanoCode012 2023-05-26 22:23:38 +09:00
  • 353cebd838 Merge pull request #68 from OpenAccess-AI-Collective/NanoCode012-patch-1 NanoCode012 2023-05-26 22:20:31 +09:00
  • fe0e69f4f9 Fix recommendation condition NanoCode012 2023-05-26 22:19:50 +09:00
  • 1fc9b44e3d fix wheel blobs in dockerfile Wing Lian 2023-05-26 07:40:11 -04:00
  • 32e6fe9286 load the tokenizer seperately from the model Wing Lian 2023-05-26 07:29:35 -04:00
  • 919623793a Add cfg.lora_target_linear NanoCode012 2023-05-26 14:32:30 +09:00
  • bbfc333a01 Merge pull request #62 from OpenAccess-AI-Collective/qlora-fixes Wing Lian 2023-05-26 00:28:16 -04:00
  • a5bf838685 add logging and make sure model unloads to float16 Wing Lian 2023-05-26 00:09:55 -04:00
  • a4f12415a0 update readme and add typehints Wing Lian 2023-05-25 23:10:11 -04:00
  • 48f4c0571e fix validation for qlora merge Wing Lian 2023-05-25 23:02:03 -04:00
  • 1987e5cf56 qlora and 4bit check so we are able to merge and unload Wing Lian 2023-05-25 22:55:13 -04:00
  • e7e1a777bd fix bool args according to python fire docs Wing Lian 2023-05-25 22:45:41 -04:00
  • 7b5e762be2 fix merge conflict failure, black format Wing Lian 2023-05-25 22:40:27 -04:00
  • 3f6017db9e qlora merge and load requires that base model isn't loaded in 4 or 8 bit Wing Lian 2023-05-25 22:39:13 -04:00
  • 34c99f9812 fixes to make qlora actually work Wing Lian 2023-05-25 22:37:23 -04:00
  • 3815c054b6 Merge pull request #61 from NanoCode012/feat/update-readme NanoCode012 2023-05-26 11:27:31 +09:00
  • 85326bfbf3 Update quickstart config NanoCode012 2023-05-26 11:15:57 +09:00
  • e689069afd Add xformers error NanoCode012 2023-05-26 11:12:03 +09:00
  • d7d8bc739e Add strict yml NanoCode012 2023-05-26 11:10:59 +09:00
  • 60e32ff457 Fix shard config NanoCode012 2023-05-26 11:09:28 +09:00
  • 259262bf42 fix xentropy wheel name typo Wing Lian 2023-05-25 17:25:38 -04:00
  • 2e56203b50 another fix for shard and train split Wing Lian 2023-05-25 17:23:57 -04:00
  • be3d3963cd Merge pull request #58 from OpenAccess-AI-Collective/shards-fix Wing Lian 2023-05-25 16:32:31 -04:00
  • ac79360161 shard fix Wing Lian 2023-05-25 16:31:59 -04:00
  • b2fb61845e Merge pull request #54 from OpenAccess-AI-Collective/winglian-patch-1 Wing Lian 2023-05-25 12:45:19 -04:00
  • 71d600fc43 Merge branch 'main' into winglian-patch-1 Wing Lian 2023-05-25 12:45:13 -04:00
  • 4fd0c2d1b9 Merge pull request #57 from OpenAccess-AI-Collective/fixes-for-basic-samples Wing Lian 2023-05-25 12:43:22 -04:00
  • 943961fd10 missed ... Wing Lian 2023-05-25 12:42:56 -04:00
  • d2a6f79fd1 change auth token setting back Wing Lian 2023-05-25 12:41:17 -04:00
  • 98b1bce57e pr comments addressed Wing Lian 2023-05-25 12:24:52 -04:00
  • 004820209d Update src/axolotl/prompters.py Wing Lian 2023-05-25 12:21:02 -04:00
  • 8d6a28953f fix relative path in flash-attn build: Wing Lian 2023-05-25 12:18:28 -04:00
  • e396654319 fix tokenizer loading, got openllama 3b working Wing Lian 2023-05-25 12:15:12 -04:00
  • a5d739b66b fixes w/ example for super basic lora starter Wing Lian 2023-05-25 11:59:08 -04:00
  • 951facbb1f Merge pull request #56 from OpenAccess-AI-Collective/fix-build-flash-attn Wing Lian 2023-05-25 11:29:47 -04:00
  • f5fa3d131b fix cd within flash-attn Wing Lian 2023-05-25 11:29:15 -04:00
  • 7ec105041d Merge pull request #48 from NanoCode012/feat/update-readme NanoCode012 2023-05-25 23:49:58 +09:00
  • a9e502ef45 Update 4bit notes NanoCode012 2023-05-25 23:48:18 +09:00
  • 68f0c71424 Merge pull request #49 from NanoCode012/feat/gitignore NanoCode012 2023-05-25 23:42:49 +09:00
  • 52fb6d8a34 Update gitignore using standard Python template NanoCode012 2023-05-25 17:48:49 +09:00
  • f92245dbd6 Fix missing closing code block NanoCode012 2023-05-25 23:04:33 +09:00
  • e65c203e9e Add more detail on minimum GPU NanoCode012 2023-05-25 22:45:58 +09:00
  • 1377400c33 Add info on Runtime Error NanoCode012 2023-05-25 22:33:45 +09:00
  • 2c34f8d0c7 Update dataset type NanoCode012 2023-05-25 17:44:58 +09:00
  • 7bc28eb8a8 Add more data formats NanoCode012 2023-05-25 17:43:37 +09:00
  • 29273b5a5b Add other minor configs NanoCode012 2023-05-25 17:34:43 +09:00
  • 05c18340d6 Update scheduler configs NanoCode012 2023-05-25 17:32:03 +09:00
  • 5b712afbe4 Update bf16 options NanoCode012 2023-05-25 17:28:03 +09:00
  • 9083910036 Update lora config NanoCode012 2023-05-25 17:26:39 +09:00
  • 8552218491 Improve Inference instruction NanoCode012 2023-05-25 17:04:44 +09:00
  • de2a7335e6 Merge pull request #55 from OpenAccess-AI-Collective/missing-validation-file Wing Lian 2023-05-25 09:58:51 -04:00
  • 1d7da3b389 add missing file Wing Lian 2023-05-25 09:58:29 -04:00
  • e07bd8a441 add discord link to #axolotl-help channel Wing Lian 2023-05-25 09:45:45 -04:00
  • d092cdb19b Merge pull request #52 from OpenAccess-AI-Collective/bugfix-cfg-cfg Wing Lian 2023-05-25 09:35:24 -04:00
  • f523a0894c stray s Wing Lian 2023-05-25 09:23:56 -04:00
  • 676d7da661 cfg.cfg fix, also de-dupe lora module list Wing Lian 2023-05-25 09:18:57 -04:00
  • a617f1b65e Merge pull request #44 from OpenAccess-AI-Collective/qlora-add-modules-tuple Wing Lian 2023-05-24 23:46:40 -04:00
  • a8771b0aad fix tuple add to list Wing Lian 2023-05-24 23:46:04 -04:00
  • cf48ff7cac Merge pull request #41 from OpenAccess-AI-Collective/qlora-modules Wing Lian 2023-05-24 23:31:19 -04:00
  • 1cf21daf51 Update src/axolotl/utils/models.py Wing Lian 2023-05-24 23:31:12 -04:00
  • ffd1043607 attempt to find linear modules for qlora Wing Lian 2023-05-24 23:18:08 -04:00
  • 3369c4dcf8 Merge pull request #39 from OpenAccess-AI-Collective/dev Wing Lian 2023-05-24 23:03:22 -04:00
  • bc97f9c584 remove dev specific remark Wing Lian 2023-05-24 23:00:53 -04:00
  • ce34d64e8a apply black formatting Wing Lian 2023-05-24 22:59:33 -04:00
  • ce694e20a3 Merge branch 'main' of github.com:OpenAccess-AI-Collective/axolotl into dev Wing Lian 2023-05-24 22:59:09 -04:00
  • cebea372da Merge pull request #36 from OpenAccess-AI-Collective/qlora Wing Lian 2023-05-24 22:57:37 -04:00
  • 1f5d83ea72 remove un-needed code, add validation Wing Lian 2023-05-24 22:47:33 -04:00
  • 6e7d4d5344 Merge pull request #35 from NanoCode012/update-readme Wing Lian 2023-05-24 21:31:32 -04:00
  • 362821ce84 Add trust_remote_code config NanoCode012 2023-05-25 09:53:49 +09:00
  • 224d186ec9 Simplify docker instruction NanoCode012 2023-05-25 09:51:22 +09:00
  • 5417824b31 Add seq length NanoCode012 2023-05-25 09:50:43 +09:00
  • e1a91b0918 Remove redundant formats NanoCode012 2023-05-25 09:48:18 +09:00
  • 2a1b5728e6 Add line break NanoCode012 2023-05-25 09:37:18 +09:00
  • 702f2eee4b Fix inference command NanoCode012 2023-05-25 09:36:33 +09:00
  • 88bba24d9e Clean up data readme NanoCode012 2023-05-25 09:34:35 +09:00
  • ba9ac723f1 Update quickstart. Add common error and contribution section. NanoCode012 2023-05-25 09:31:34 +09:00
  • db73b94a58 Add image. Add quickstart. Simplify dataset. NanoCode012 2023-05-25 09:11:20 +09:00
  • 00dfe43b1d Add image NanoCode012 2023-05-25 09:02:03 +09:00
  • 9aab0b8cfe Update Docker instructions NanoCode012 2023-05-21 23:34:48 +09:00
  • 857a80b70e Format dataset types NanoCode012 2023-05-21 23:28:06 +09:00
  • cba0048067 Update typo NanoCode012 2023-05-21 23:27:08 +09:00
  • c22df8db9b Add all dataset types NanoCode012 2023-05-21 23:26:51 +09:00