Commit Graph

  • ac7e35eb41 Use new inference model config / API (#2208) AbdBarho 2023-03-25 20:45:23 +01:00
  • 45b6e8c7f7 Update Ukrainian translation (#2214) Nazarii 2023-03-25 20:41:12 +02:00
  • 49c6dc01cd Fix horizontal scrolling on mobile (#2211) AbdBarho 2023-03-25 18:16:38 +01:00
  • a5793a548d fixes text client to work with new debug login workflow (#2212) Yannic Kilcher 2023-03-25 18:11:04 +01:00
  • 6abfd91c2d Get available auth providers from inference server (#2207) AbdBarho 2023-03-25 17:48:04 +01:00
  • 8a461b9d7b Instruction Dataset: Retrieval-based grounded model generated Q-A pairs (BART version) (#2170) Michael Wan 2023-03-25 21:58:13 +08:00
  • b5669483a2 Russian Translation Updated + Stuff (#2197) almostEvil___ 2023-03-25 16:12:36 +03:00
  • 3b55e3cf6d feature : Alpaca dataset (#2205) theblackcat102 2023-03-25 18:13:26 +08:00
  • a38db4fa15 Post llama merge fixes (#2188) Andreas Köpf 2023-03-25 11:11:37 +01:00
  • c8c6dcd0fd Update CODEOWNERS for website (#2200) AbdBarho 2023-03-25 10:44:41 +01:00
  • bcd2fb3a7d Revert "fix: ghcr.io build for mulitplatform. includes Apple silicon" (#2199) Andreas Köpf 2023-03-25 00:20:49 +01:00
  • fa4c42d89f fix: ghcr.io build for mulitplatform. includes Apple silicon (#2151) James Melvin Ebenezer 2023-03-25 04:26:46 +05:30
  • 2a141e96c4 Add recent changes to eval_model/manual/sampling_report.py (#2191) Andreas Köpf 2023-03-24 23:47:25 +01:00
  • 0331e69780 Introduce model configs to abstract pairings of models and hardware (#2194) Yannic Kilcher 2023-03-24 22:28:11 +01:00
  • 5847dc2391 Evaluate sampling report using RM (#2190) v0.0.1-beta57 Shahul ES 2023-03-24 22:30:14 +05:30
  • d631e93941 rework layout (#2172) Nguyen Minh Duc 2023-03-24 22:22:21 +07:00
  • 8276076ccb Login to inference through website (#2176) AbdBarho 2023-03-24 16:21:21 +01:00
  • fc35a63ecb Update website's dockerfile (#2189) AbdBarho 2023-03-24 16:20:50 +01:00
  • f701540bb4 Fix output of RankingDataCollator when max_length is None (#2185) Andreas Köpf 2023-03-24 09:56:59 +01:00
  • 42d9a5f058 Add llama training support (#2055) Andreas Köpf 2023-03-23 22:24:16 +01:00
  • f8a73a8dc7 Add GPTNeoXRewardModel (#2182) Andreas Köpf 2023-03-23 21:23:19 +01:00
  • f6fc5c624e Fix inconsistent sources of max token length (#2178) Yannic Kilcher 2023-03-23 18:28:24 +01:00
  • b89838c825 Add @dvruette, @andreaskoepf & @yk as codeowners of /model/ (#2186) Andreas Köpf 2023-03-23 18:16:07 +01:00
  • 5cac336af9 Update pythia 6b RM training config (#2187) Andreas Köpf 2023-03-23 18:06:30 +01:00
  • 3500182698 Chat config UI (#2115) Nguyen Minh Duc 2023-03-23 15:04:36 +07:00
  • 7eb95aecd0 Fix: Correctly sending error to waiting client on message abort (#2169) Yannic Kilcher 2023-03-22 21:41:48 +01:00
  • 30cfe1d1e4 Added param configs from eval to pass to the frontend (#2163) Yannic Kilcher 2023-03-22 15:55:58 +01:00
  • 7e3768fc06 Diabling stop sequences in inference, since we are using eos_tokens in most models (#2168) Yannic Kilcher 2023-03-22 13:23:42 +01:00
  • 15c477fda7 Add first RM training run params (#2167) Andreas Köpf 2023-03-22 12:01:43 +01:00
  • 0497ccd125 Only allow prompter answers to complete parent message (#2164) Yannic Kilcher 2023-03-22 09:45:50 +01:00
  • 51077741b8 check per_digit_tokens exists (#2166) Huanglk 2023-03-22 16:37:12 +08:00
  • 99b42bc7cf Derive RewardModel from PreTrainedModel (#2158) Andreas Köpf 2023-03-22 01:58:04 +01:00
  • a911e0cd6b Added hardware info dependencies to oasst-shared to make api tests pass (#2162) Yannic Kilcher 2023-03-21 22:54:51 +01:00
  • 6165f2e1cb Changed worker config into worker info and unified model input and total lengths (#2160) Yannic Kilcher 2023-03-21 22:39:55 +01:00
  • bcaa7a066d building separate llama image for inference (#2161) Yannic Kilcher 2023-03-21 21:32:40 +01:00
  • 4d34caa2f8 Fix stacking of pos/neg logits for RM compute_metrics (#2138) Andreas Köpf 2023-03-21 21:32:08 +01:00
  • e0d7883f68 Updated team page to fit visually more the the rest of the pages, added a button to the call-to-action section which links to the team page. (#2126) Theodor Peifer 2023-03-21 17:05:32 +01:00
  • b9068ca389 Fixes changed access token model for website. Does not handle refresh yet (#2157) Yannic Kilcher 2023-03-21 17:05:04 +01:00
  • e2f2184b66 Fixes auth TokenPair change bug (also fix for null logprob from inference server) (#2155) Yannic Kilcher 2023-03-21 16:37:38 +01:00
  • f1e41d1bca Robustification: Specific message type for protocol upgrade, everything else will re-connect (#2152) Yannic Kilcher 2023-03-21 16:16:41 +01:00
  • dae74b603b Add refresh tokens to inference server auth (#2129) Oliver Stanley 2023-03-21 14:14:15 +00:00
  • dafc373975 Remove invalid badge from README (#2149) Oliver Stanley 2023-03-21 14:13:05 +00:00
  • b045d247fc Added various people to codeowners for backend and inference (#2150) Yannic Kilcher 2023-03-21 15:12:31 +01:00
  • 6bf071cb1a Added a timeout for inference messages (#2147) Yannic Kilcher 2023-03-21 12:09:22 +01:00
  • 410036df2e Cancelling pending chat messages of user on submission of a new message (#2141) Yannic Kilcher 2023-03-21 11:56:07 +01:00
  • 5f492e41fd Implemented ability to press ctrl+enter (or cmd+enter) to Review/Subm… (#1237) Austin Cooper 2023-03-21 02:49:13 -07:00
  • 998d13cd59 hotfix for max_new_tokens parameter being None (#2146) Yannic Kilcher 2023-03-21 09:52:54 +01:00
  • 792fd18dc2 Llama inference (#2144) Yannic Kilcher 2023-03-21 09:45:51 +01:00
  • cb44f4da9e Added ENABLE_CHAT to chat id page and docker-compose (#2137) Yannic Kilcher 2023-03-21 09:28:31 +01:00
  • 87119c73dd [#2077] Possbility to revert message deletion (#2079) Shadowner 2023-03-21 09:27:37 +01:00
  • f6b7360ab1 Add multi-turn conversation dataset format (#2113) Vechtomov 2023-03-21 11:26:26 +03:00
  • f83a104fe6 Display message tree state admin (#2133) Robert Lu Zheng 2023-03-21 01:29:27 -05:00
  • c351ada4a3 added HF, W&B, and Redmond to sponsors (#2142) Yannic Kilcher 2023-03-21 07:01:02 +01:00
  • 4474123f9f Simplify celery worker use (#2131) Andreas Köpf 2023-03-20 21:59:00 +01:00
  • 721eba408e hotfix for max_new_tokens parameter being None max-new-tokens-fix Yannic Kilcher 2023-03-20 18:12:43 +01:00
  • 237a1a64c6 added timeout setting to inference messages Yannic Kilcher 2023-03-20 17:52:31 +01:00
  • b2e9415cf3 Cancelling pending chat messages of user on submission of a new message Yannic Kilcher 2023-03-20 17:24:10 +01:00
  • 739099d10b Update team.json (#2139) Ken Tsui 2023-03-20 23:52:04 +08:00
  • bffdc51a8f Add me to team.json (#2132) John Tapsell 2023-03-20 02:32:50 -07:00
  • 7c5685ff40 fix team description (#2135) Nguyen Minh Duc 2023-03-20 15:12:51 +07:00
  • 4c3a0a4626 Evaluation datasets for Reward Model (#2112) Shahul ES 2023-03-20 12:39:04 +05:30
  • 3e5a8e1176 Add Joak to team (#2134) notmd 2023-03-20 13:28:14 +07:00
  • b1f37f9298 Boilerplate OAuth website <=> inference (#2127) AbdBarho 2023-03-20 07:10:12 +01:00
  • 42c8c3dd33 enable chat via environment variable instead of flag (#2128) Yannic Kilcher 2023-03-20 04:54:11 +01:00
  • 4a4e7db000 updated text client sampling parameters Yannic Kilcher 2023-03-19 23:19:54 +01:00
  • 307db951b6 some improvements to hf server hf-server-improvements Yannic Kilcher 2023-03-19 23:05:43 +01:00
  • 657c3f8c36 feature: Using Celery for APIs and periodic tasks (#1754) Andreas Köpf 2023-03-19 22:30:39 +01:00
  • 94e9bb325d Basic-hf-server (#2130) Yannic Kilcher 2023-03-19 22:02:58 +01:00
  • bd0acd6d11 Add tlcv2.0_oa (#1965) Gareth Davidson 2023-03-19 20:14:50 +00:00
  • 1ea571f197 Add tlcv_2_0_oa Wannaphong Phatthiyaphaibun 2023-03-05 22:44:30 +07:00
  • 7048b9e9ce Merge branch 'main' into 954_celery_worker_1 Andreas Köpf 2023-03-19 20:55:44 +01:00
  • 86ded7ae52 Merge branch 'main' into add-tlcv2.0_oa Gareth Davidson 2023-03-19 19:29:10 +00:00
  • 0cef54a107 Fix GitHub auth flow on inference server (#2121) Oliver Stanley 2023-03-19 19:13:11 +00:00
  • 0e0fdb12ec Add ~Zy~ to team.json (#2125) Zyphre 2023-03-19 14:08:26 -04:00
  • f858bd2150 Merge branch 'main' of github.com:LAION-AI/Open-Assistant Yannic Kilcher 2023-03-19 16:28:59 +01:00
  • ebf50cb5a8 updated dockerfile Yannic Kilcher 2023-03-19 16:28:55 +01:00
  • bd8e5f1eba Only import flash_attn if actually needed John Tapsell 2023-03-17 17:55:52 +00:00
  • 3223518422 Add missing dependencies John Tapsell 2023-03-17 09:14:13 +00:00
  • 72c3fe6a7d Add support for minimath John Tapsell 2023-03-17 08:45:33 +00:00
  • f327fffdfb Make sampling_report.py executable John Tapsell 2023-03-16 05:49:42 +00:00
  • 5568837f33 You in teams and more (#2123) almostEvil___ 2023-03-19 18:22:19 +03:00
  • f1252ce531 calling hf server from worker Yannic Kilcher 2023-03-19 15:51:25 +01:00
  • 78a0ba63b3 Merge branch 'main' of github.com:LAION-AI/Open-Assistant Yannic Kilcher 2023-03-19 15:39:43 +01:00
  • dd0657ab7a added hf server Yannic Kilcher 2023-03-19 15:39:31 +01:00
  • a4ee6c2902 Reward Model Training (#2071) Dimitri 2023-03-19 14:15:05 +01:00
  • 9704172646 ignoring pytest cache globally Yannic Kilcher 2023-03-19 14:15:01 +01:00
  • e4289f25b1 Update team info (#2120) almostEvil___ 2023-03-19 16:05:19 +03:00
  • 3bad5f7944 Merge branch 'main' of github.com:LAION-AI/Open-Assistant Yannic Kilcher 2023-03-19 13:33:39 +01:00
  • 28d09fc15c enable chat via flag Yannic Kilcher 2023-03-19 13:33:22 +01:00
  • 8044461ec2 Minor update on team.json (#2118) James Melvin Ebenezer 2023-03-19 16:12:54 +05:30
  • 1091e59cd1 Update team.json (#2116) James Melvin Ebenezer 2023-03-19 16:05:02 +05:30
  • a31d0f7cb4 Added James Mete (JMete) to team page. (#2117) James Mete 2023-03-19 13:18:14 +03:00
  • 7d24c4f5da Add Team page (#2103) AbdBarho 2023-03-19 09:24:57 +01:00
  • d74963282b added sentencepiece to inference worker Yannic Kilcher 2023-03-19 02:15:10 +01:00
  • 4b6c6ddd87 added llama tokenizer Yannic Kilcher 2023-03-19 01:42:23 +01:00
  • 8eff31fbf7 updated main scripts Yannic Kilcher 2023-03-19 01:23:25 +01:00
  • 79f012e854 installing git into worker Yannic Kilcher 2023-03-19 01:13:24 +01:00
  • 397dff2be9 added newest transformers Yannic Kilcher 2023-03-19 01:04:22 +01:00
  • 75fac1a07c added hf token to deploy Yannic Kilcher 2023-03-19 00:59:45 +01:00
  • aa772c2274 added typical_p Yannic Kilcher 2023-03-19 00:34:20 +01:00