Commit Graph

  • c13b9d87c9
    fix(router): fix truncation (#190) OlivierDehaene 2023-04-17 16:51:53 +0200
  • cea9ae48c8 fix(router): fix truncation OlivierDehaene 2023-04-17 16:50:51 +0200
  • 7a1ba58557
    fix(docker): fix docker image dependencies (#187) OlivierDehaene 2023-04-17 00:26:47 +0200
  • 69a66b0669 revert build OlivierDehaene 2023-04-16 23:13:45 +0200
  • 396ce5f111 add requirements.txt OlivierDehaene 2023-04-16 22:54:18 +0200
  • a37e6edd5c add logs OlivierDehaene 2023-04-16 17:34:57 +0200
  • c23cc3e2f7 force push OlivierDehaene 2023-04-15 22:03:03 +0200
  • ded92301a2 fix(docker): fix docker image dependencies OlivierDehaene 2023-04-15 21:37:31 +0200
  • 7caea42573 feat(launcher): parse all shard logs feat/parse_logs OlivierDehaene 2023-04-15 21:25:02 +0200
  • 379c5c4da2
    fix(docker): revert dockerfile changes (#186) OlivierDehaene 2023-04-14 19:30:30 +0200
  • 4096ca4eed fix(docker): revert dockerfile changes OlivierDehaene 2023-04-14 19:27:57 +0200
  • f9047562d0
    fix(docker): fix image (#185) OlivierDehaene 2023-04-14 18:58:38 +0200
  • 1aecb60570 final OlivierDehaene 2023-04-14 18:58:06 +0200
  • 2078a9a406 fix OlivierDehaene 2023-04-14 18:26:40 +0200
  • 1ebea22857 fix(docker): fix image OlivierDehaene 2023-04-14 18:25:38 +0200
  • 1bb394631d
    fix(docker): fix docker image (#184) OlivierDehaene 2023-04-14 17:31:13 +0200
  • 2611e4826b use large-runner OlivierDehaene 2023-04-14 16:16:54 +0200
  • 10fd8e700c fix(docker): fix docker image OlivierDehaene 2023-04-14 15:59:14 +0200
  • 01c0e368e5
    fix(ci): fix cosign error (#183) OlivierDehaene 2023-04-14 12:35:26 +0200
  • 61ba530cf6 fix(ci): fix cosign error OlivierDehaene 2023-04-14 12:34:53 +0200
  • 53ee09c0b0
    fea(dockerfile): better layer caching (#159) OlivierDehaene 2023-04-14 10:12:21 +0200
  • e9dbfd3e76 runs-on ubuntu latest again OlivierDehaene 2023-04-14 10:10:07 +0200
  • d5929e90eb change cache-from OlivierDehaene 2023-04-14 09:57:29 +0200
  • aca0e81a9e setup buildx OlivierDehaene 2023-04-13 22:21:11 +0200
  • 88962ac73a remove setup buildx OlivierDehaene 2023-04-13 22:18:53 +0200
  • f3438d43a1 cache to azure instead OlivierDehaene 2023-04-13 18:18:05 +0200
  • c7803c41e8 fix test OlivierDehaene 2023-04-13 16:52:12 +0200
  • 42196a1af8 Merge remote-tracking branch 'origin/main' into feat/better_docker_image OlivierDehaene 2023-04-13 16:48:46 +0200
  • f1ddbf5c72 use mamba OlivierDehaene 2023-04-13 16:48:21 +0200
  • 12e5633c4d
    fix(ci): fix ci permissions (#181) OlivierDehaene 2023-04-13 16:32:37 +0200
  • dc3e7e14c7 fix(ci): fix ci permissions OlivierDehaene 2023-04-13 16:32:21 +0200
  • 4cfef0441f same syntax OlivierDehaene 2023-04-13 16:31:46 +0200
  • c1e2ea3b78
    feat(ci): faster scanning (#180) OlivierDehaene 2023-04-13 16:23:47 +0200
  • a0f223f9f6 feat(ci): faster scanning OlivierDehaene 2023-04-13 16:23:29 +0200
  • 0ffd0af94a update OlivierDehaene 2023-04-13 16:21:14 +0200
  • ccc8c7997e use larger machines OlivierDehaene 2023-04-12 17:20:54 +0200
  • 32e8c06a1a fix build OlivierDehaene 2023-04-11 17:54:36 +0200
  • a89d745d02 use private registry for caching OlivierDehaene 2023-04-11 17:52:14 +0200
  • 158d803383 run on self hosted runners OlivierDehaene 2023-04-11 16:39:15 +0200
  • a265dde4e0 fix install OlivierDehaene 2023-04-11 11:46:34 +0200
  • b5fec41033 better makefiles OlivierDehaene 2023-04-11 11:41:48 +0200
  • 4d8972fc9a merge some layers OlivierDehaene 2023-04-11 11:26:55 +0200
  • f8655e4683 add compression OlivierDehaene 2023-04-07 18:19:47 +0200
  • 384acff1a8 fix OlivierDehaene 2023-04-07 14:17:26 +0200
  • 49b66db53e rework OlivierDehaene 2023-04-07 14:15:35 +0200
  • ffa385031c fea(dockerfile): better layer caching OlivierDehaene 2023-04-07 14:11:44 +0200
  • 13f1cd024b
    feat(ci): use large runners (#179) OlivierDehaene 2023-04-13 16:11:48 +0200
  • 1d983ce9ac runs all build on large OlivierDehaene 2023-04-13 16:05:48 +0200
  • ad830b9440 fix label OlivierDehaene 2023-04-13 15:57:00 +0200
  • ddc1f1a1a5 feat(ci): use large runners OlivierDehaene 2023-04-13 15:54:49 +0200
  • 9683c37bd3
    feat(ci): add Trivy and scan docker image (#178) OlivierDehaene 2023-04-13 15:43:17 +0200
  • 5f1d7a4520 fix OlivierDehaene 2023-04-13 15:37:15 +0200
  • 6091e27c99 fix OlivierDehaene 2023-04-13 15:32:38 +0200
  • 1816b09a41 fix OlivierDehaene 2023-04-13 15:29:48 +0200
  • 824f18fdc5 add trivy OlivierDehaene 2023-04-13 15:26:47 +0200
  • a11772bfb9 feat(ci): add Trivy and scan docker image OlivierDehaene 2023-04-13 15:17:26 +0200
  • 643a39d556
    feat(ci): add image signing with cosign (#175) OlivierDehaene 2023-04-13 15:26:34 +0200
  • 430525306a login to access image cache OlivierDehaene 2023-04-13 15:18:33 +0200
  • 61e6e880d5 activate cosign OlivierDehaene 2023-04-13 14:51:18 +0200
  • 979f58c061 test OlivierDehaene 2023-04-13 13:04:00 +0200
  • 2e5c75c3bc test OlivierDehaene 2023-04-13 13:03:40 +0200
  • 44ddbb7277 feat(ci): add image signing with cosign OlivierDehaene 2023-04-13 13:00:28 +0200
  • 64347b05ff
    fix(ci): fix CVE in github-slug-action (#174) OlivierDehaene 2023-04-13 12:43:05 +0200
  • 0dd6fe488a fix(ci): fix CVE in github-slug-action OlivierDehaene 2023-04-13 12:18:10 +0200
  • e3a63b6fbc
    fix(launcher): revert change on shard errors (#173) OlivierDehaene 2023-04-13 11:07:11 +0200
  • d2ce06cd1e fix(launcher): revert change on shard errors OlivierDehaene 2023-04-13 10:06:49 +0200
  • 880a76eed5
    feat(server): support sharded santacoder (#167) OlivierDehaene 2023-04-12 17:18:08 +0200
  • 5fa8ae041c
    feat(server): optimize decode for sane tokenizers (#170) OlivierDehaene 2023-04-12 12:03:10 +0200
  • b163aef8ed fmt OlivierDehaene 2023-04-12 11:31:55 +0200
  • 2aa5004482 feat(server): optimize decode for sane tokenizers OlivierDehaene 2023-04-12 11:24:02 +0200
  • 6f0f1d70f6
    v0.5.0 (#168) v0.5.0 OlivierDehaene 2023-04-11 20:32:18 +0200
  • f4e6de1c3f v0.5.0 OlivierDehaene 2023-04-11 19:17:49 +0200
  • 7c281908cf fix load_weights OlivierDehaene 2023-04-11 20:10:33 +0200
  • 622daeb0c8 working model OlivierDehaene 2023-04-11 20:00:12 +0200
  • 9541c8f146 wip OlivierDehaene 2023-04-06 16:13:32 +0200
  • 2378529c15 wip OlivierDehaene 2023-04-06 15:03:13 +0200
  • e8a3ec36c3 wip OlivierDehaene 2023-04-05 16:29:34 +0200
  • f26dfd0dc1
    feat(server): support OPT models (#55) OlivierDehaene 2023-04-11 19:16:41 +0200
  • 5632fc5bad fix galactica OlivierDehaene 2023-04-11 18:59:13 +0200
  • aafec48ff3 add to readme OlivierDehaene 2023-04-11 18:52:07 +0200
  • 9b06248395 fix OlivierDehaene 2023-04-11 18:37:00 +0200
  • 34931a2111 patch safetensors loading OlivierDehaene 2023-02-28 15:56:35 +0100
  • ef51a1e0b7 rebase OlivierDehaene 2023-02-28 11:48:15 +0100
  • 438883cb10 feat(server): support OPT models OlivierDehaene 2023-02-03 15:54:39 +0100
  • 299217c95c
    feat(server): add flash attention llama (#144) OlivierDehaene 2023-04-11 16:38:22 +0200
  • d7548aef9b add llama to readme OlivierDehaene 2023-04-11 16:08:06 +0200
  • c2beaa279e revert build OlivierDehaene 2023-04-11 11:08:20 +0200
  • a1a6b5cc20 Merge remote-tracking branch 'origin/main' into feat/flash_llama OlivierDehaene 2023-04-09 20:24:16 +0200
  • 9987960062
    feat(router): make router input validation optional (#164) OlivierDehaene 2023-04-09 20:22:27 +0200
  • 7451196a78 fmt OlivierDehaene 2023-04-09 20:18:37 +0200
  • 7dec65a244
    fix(router): use buckets for metrics histograms (#163) OlivierDehaene 2023-04-09 20:13:28 +0200
  • 5cddc055e6
    fix(rust-client): use join_all instead of select_all to hopefully fix nccl issues (#162) OlivierDehaene 2023-04-09 20:07:02 +0200
  • e63a21eb4d
    feat(launcher): allow disabling hf_transfer (#161) OlivierDehaene 2023-04-09 20:00:05 +0200
  • 1883d8ecde
    feat(docker): improve flash_attention caching (#160) OlivierDehaene 2023-04-09 19:59:16 +0200
  • f19c2c3bf9 feat(router): make router input validation optional OlivierDehaene 2023-04-09 19:53:55 +0200
  • 98cfc9e70c fix(router): use buckets for metrics histograms OlivierDehaene 2023-04-09 19:41:24 +0200
  • 3e1a281f3f fix(rust-client): use join_all instead of select_all OlivierDehaene 2023-04-09 19:36:24 +0200
  • e26d28cef9 feat(launcher): allow disabling hf_transfer OlivierDehaene 2023-04-09 19:34:02 +0200
  • 98094f4d24 feat(docker): improve flash_attention caching OlivierDehaene 2023-04-09 19:29:11 +0200
  • 3795c19dcb minimum duration to 0.1 ms OlivierDehaene 2023-04-09 19:01:53 +0200