Commit Graph

  • 0586df7b4f
    Merge 3577858ef7 into 6d9c84f527 m3ndax 2023-10-04 19:08:37 +0100
  • 6d9c84f527
    Goodnight! main James Ravenscroft 2023-09-30 09:16:58 +0100
  • 3577858ef7 removed unused semicolon mendax0110 2023-09-10 03:58:36 +0200
  • 0bc30c3923
    Merge branch 'ravenscroftj:main' into main m3ndax 2023-09-10 03:43:09 +0200
  • eaeb52fcb0
    Merge pull request #67 from ravenscroftj/fix/mac-memory-usage v0.2.1 James Ravenscroft 2023-08-28 11:55:55 +0100
  • 83cb7c042f Temporary fix for stablecode and starcoder on mac fix/mac-memory-usage James Ravenscroft 2023-08-28 09:50:40 +0200
  • 8fd357e0a5
    Merge pull request #65 from ravenscroftj/ravenscroftj-patch-1 James Ravenscroft 2023-08-26 17:11:21 +0100
  • 30c437700a
    Update README.md ravenscroftj-patch-1 James Ravenscroft 2023-08-26 17:11:12 +0100
  • 86f07745bb
    Merge pull request #64 from ravenscroftj/fix/docker-release v0.2.0 James Ravenscroft 2023-08-26 17:02:05 +0100
  • 7bb93b6f4e fix docker build for tags fix/docker-release James Ravenscroft 2023-08-26 17:01:35 +0100
  • 2b27760a7f
    Merge pull request #55 from ravenscroftj/feature/gpu_layers James Ravenscroft 2023-08-26 16:35:52 +0100
  • a00de2a332 recomment the cuda preprocessor check feature/gpu_layers James Ravenscroft 2023-08-26 16:21:42 +0100
  • 215a69b5af update clblast code in gpt-j model James Ravenscroft 2023-08-26 16:16:01 +0100
  • 91639b8fc0 disable clblast docker images James Ravenscroft 2023-08-26 16:12:50 +0100
  • 0b408510f4 add gpu offload for gpt-j models (codegen) James Ravenscroft 2023-08-26 16:11:16 +0100
  • 604183380d tidy up prints in stablecoder and starcoder James Ravenscroft 2023-08-26 16:04:41 +0100
  • 88683abe50 update run script to incorporate GPU layers James Ravenscroft 2023-08-26 16:03:16 +0100
  • 326e76c9bb Merge branch 'main' into feature/gpu_layers James Ravenscroft 2023-08-26 15:59:13 +0100
  • 23c0a3d19e Merge branch 'feature/gpu_layers' of github.com:ravenscroftj/turbopilot into feature/gpu_layers James Ravenscroft 2023-08-26 15:34:28 +0100
  • 31bb33c731 use latest upstream ggml instead of mine James Ravenscroft 2023-08-26 15:22:14 +0100
  • d4989b543c
    Merge pull request #62 from ravenscroftj/feature/blasdocker James Ravenscroft 2023-08-26 15:32:13 +0100
  • e9dc6a304a use latest upstream ggml instead of mine James Ravenscroft 2023-08-26 15:22:14 +0100
  • 97a0377cd6 remove llama.cpp submodule James Ravenscroft 2023-08-26 15:21:01 +0100
  • 6d26c9b064 Merge branch 'feature/gpu_layers' of github.com:ravenscroftj/turbopilot into feature/gpu_layers James Ravenscroft 2023-08-26 15:20:25 +0100
  • 63b554793d tidy cmakelist James Ravenscroft 2023-08-26 15:20:02 +0100
  • 0cf7a9c341 remove llama James Ravenscroft 2023-08-26 15:19:51 +0100
  • 356a83c5fd remove crow submodule James Ravenscroft 2023-08-26 15:19:17 +0100
  • a5517b0fcd use ggerganov ggml instead of mine James Ravenscroft 2023-08-26 15:19:04 +0100
  • 8fa70e1518 update for gpu build James Ravenscroft 2023-08-21 20:40:17 +0100
  • b79ab46b50 add gpu offload for gptneox James Ravenscroft 2023-08-21 20:03:25 +0100
  • 4a47251822 update for gpu build James Ravenscroft 2023-08-21 20:40:17 +0100
  • b2b4a1480f increase scratch on starcoder James Ravenscroft 2023-08-21 20:20:35 +0100
  • 5f7155a314 add gpu offload for gptneox James Ravenscroft 2023-08-21 20:03:25 +0100
  • 77cde95cb9 remove deprecated cuda dockerfiles feature/blasdocker James Ravenscroft 2023-08-26 14:17:31 +0100
  • bea7ebdb34 correct runtime libs for openblas and clblast James Ravenscroft 2023-08-26 14:16:39 +0100
  • 812bbea9d7 correct typo with clblast James Ravenscroft 2023-08-26 13:53:26 +0100
  • 08e8834390 add changes to dockerfile James Ravenscroft 2023-08-26 13:51:50 +0100
  • 25680e64d8 remove all the quotes James Ravenscroft 2023-08-26 13:43:33 +0100
  • dca25d8456 remove quotes James Ravenscroft 2023-08-26 13:42:11 +0100
  • 1f6f84a783 add quotes to args James Ravenscroft 2023-08-26 13:39:02 +0100
  • 0183b30502 always use ubuntu 22.04 James Ravenscroft 2023-08-26 13:32:01 +0100
  • b465eae818 break out vars James Ravenscroft 2023-08-26 13:28:08 +0100
  • e8adff5339 try again James Ravenscroft 2023-08-26 13:22:09 +0100
  • f12dacaa15 read the readme properly James Ravenscroft 2023-08-26 13:16:34 +0100
  • 0b0b914f92 add commas? James Ravenscroft 2023-08-26 13:12:27 +0100
  • b21dd0799d fix basenames James Ravenscroft 2023-08-26 13:10:23 +0100
  • c73c196364 build nvidia with default dockerfile James Ravenscroft 2023-08-26 13:08:19 +0100
  • 30834e3121 remove brew update to prevent python breaking build James Ravenscroft 2023-08-26 12:58:02 +0100
  • 39c3182a3a try to fix build args James Ravenscroft 2023-08-26 12:57:00 +0100
  • 2abdcabf02 use lists for build args James Ravenscroft 2023-08-26 12:45:23 +0100
  • 6877542ad8 blas docker build James Ravenscroft 2023-08-26 12:43:52 +0100
  • 5b561f7b7e
    Merge pull request #61 from c01o/patch-1 James Ravenscroft 2023-08-26 12:37:21 +0100
  • e85492d8ba
    Fix download link on MODELS.md c01o 2023-08-26 19:16:15 +0900
  • f840ea0b73
    Merge pull request #60 from nvtienanh/update-dockerfile-default James Ravenscroft 2023-08-26 09:43:24 +0100
  • 3e37c4bb7c Update cmake command Anh Nguyen 2023-08-26 14:27:27 +0700
  • 20b1460bd8 Using GGML_STATIC Anh Nguyen 2023-08-26 14:21:18 +0700
  • 2308b9ae21 Change from alpine to ubuntu in dockerfile.default Anh Nguyen 2023-08-26 13:21:04 +0700
  • c4e57e0aab
    Merge pull request #58 from ravenscroftj/feature/model-lock James Ravenscroft 2023-08-25 06:57:05 +0100
  • dc81abbc52 Merge branch 'main' into feature/model-lock feature/model-lock James Ravenscroft 2023-08-24 14:58:44 +0100
  • ae2d505a2f use std mutex instead of boost mutex James Ravenscroft 2023-08-24 14:55:23 +0100
  • 143155dac3 boost James Ravenscroft 2023-08-24 14:28:01 +0100
  • 596c835939 boost James Ravenscroft 2023-08-24 14:26:40 +0100
  • 6b0a25cb71
    Merge pull request #59 from ravenscroftj/feature/batch-flag James Ravenscroft 2023-08-24 14:13:40 +0100
  • 227501188c try to set boost librarydir James Ravenscroft 2023-08-24 13:57:22 +0100
  • f69a8f65d4 fix build? James Ravenscroft 2023-08-24 13:49:14 +0100
  • 22f2993db4 try using stage lib dir for boost root James Ravenscroft 2023-08-24 13:28:58 +0100
  • e8beac34e7 more attempts to build with boost threads James Ravenscroft 2023-08-24 13:17:51 +0100
  • ccf425f019 update deps for boost James Ravenscroft 2023-08-24 13:05:10 +0100
  • cceee41f79 add boost threads James Ravenscroft 2023-08-24 13:03:12 +0100
  • 113544400a try adding build boost dirs explicitely James Ravenscroft 2023-08-24 12:04:45 +0100
  • f0627cd567 add boost libraries to cmake James Ravenscroft 2023-08-24 11:50:21 +0100
  • 2d617b458e expose batch size flag to cli feature/batch-flag James Ravenscroft 2023-08-24 11:40:19 +0100
  • 0c1fc1a04e implement locking of model per request to prevent crashing when multiple requests reeived James Ravenscroft 2023-08-24 11:30:44 +0100
  • e0adf0519b
    Merge pull request #57 from aperullo/model-docs-fix James Ravenscroft 2023-08-24 10:18:36 +0100
  • ef1402a1a8 Fix errant command in model docs aperullo 2023-08-23 17:59:43 -0400
  • c164deb042 Merge branch 'feature/gpu_layers' of github.com:ravenscroftj/turbopilot into feature/gpu_layers James Ravenscroft 2023-08-23 17:22:15 +0100
  • b3d8d996a0 update for gpu build James Ravenscroft 2023-08-21 20:40:17 +0100
  • 1e14d91bc3 increase scratch on starcoder James Ravenscroft 2023-08-21 20:20:35 +0100
  • cf7c5285e6 add gpu offload for gptneox James Ravenscroft 2023-08-21 20:03:25 +0100
  • 11f385066a
    Merge pull request #54 from ravenscroftj/feature/debug-timing-logs James Ravenscroft 2023-08-23 17:17:20 +0100
  • 6a72e32dab add debug output to starcoder feature/debug-timing-logs James Ravenscroft 2023-08-23 15:53:38 +0000
  • bad53ad190 add debug logs to codegen James Ravenscroft 2023-08-23 14:57:31 +0000
  • cd9baacdd7 add debug logs to stablecode James Ravenscroft 2023-08-23 14:56:36 +0000
  • 6d90e5d870
    Merge pull request #53 from ravenscroftj/feature/expose-temp-top_p James Ravenscroft 2023-08-23 10:42:10 +0100
  • 88c5ede768 fix cast problem due to incorrect float notation feature/expose-temp-top_p James Ravenscroft 2023-08-23 08:12:53 +0000
  • 1e0d2d3434
    Merge pull request #52 from ravenscroftj/feature/expose-temp-top_p James Ravenscroft 2023-08-23 09:03:28 +0100
  • 757cc64a59 add temp and top p as cli args James Ravenscroft 2023-08-23 06:27:37 +0100
  • 5f5e9f90be update for gpu build James Ravenscroft 2023-08-21 20:40:17 +0100
  • 68760434b2 Merge branch 'fix/starcoder_segfault' into feature/gpu_layers James Ravenscroft 2023-08-21 20:28:43 +0100
  • 364168dda7 increase scratch on starcoder James Ravenscroft 2023-08-21 20:20:35 +0100
  • f818e2d09f add gpu offload for gptneox James Ravenscroft 2023-08-21 20:03:25 +0100
  • 8be7171573
    Merge pull request #48 from ravenscroftj/fix/docker-static-builds James Ravenscroft 2023-08-13 09:51:45 +0100
  • 2f0e2dee94 add check for GGML_STATIC and compile accordingly James Ravenscroft 2023-08-13 08:34:07 +0100
  • df954e45bf add early prelit build feature/replit James Ravenscroft 2023-08-10 14:45:23 +0100
  • 6fb323bb11 Merge remote-tracking branch 'origin/main' into feature/replit James Ravenscroft 2023-08-10 10:09:22 +0100
  • d568c3d703 implement replit model James Ravenscroft 2023-08-10 10:09:16 +0100
  • c0cde10046
    Merge pull request #44 from ravenscroftj/feature/hf-code James Ravenscroft 2023-08-10 10:08:20 +0100
  • 12a3fe9367 Merge remote-tracking branch 'origin/main' into feature/hf-code James Ravenscroft 2023-08-10 09:35:22 +0100
  • de68ab022c
    Merge pull request #43 from ravenscroftj/feature/gptneox James Ravenscroft 2023-08-10 09:34:54 +0100
  • f7f1991e2c add huggingface request handler and refactor old req handler James Ravenscroft 2023-08-10 09:26:54 +0100