Commit Graph

42 Commits (e653ca0a74b7ff135022e8929fca57586ced08ac)

Author SHA1 Message Date
Meng Zhang cfbcff64ec
Update README.md 2023-04-07 05:34:10 +08:00
Meng Zhang ef483564fe
feat: integrate caddy, re-org paths (#49)
* integrate caddy

* address comments
2023-04-06 17:02:10 +08:00
Meng Zhang 93e4f5b3ca refactor: default docker-compose w/ python backend 2023-04-06 00:50:24 +08:00
Meng Zhang e36ddbac6b
refactor: move scripts to tabby/ (#48) 2023-04-06 00:44:10 +08:00
Meng Zhang c06710f021
Update tabby.sh 2023-04-05 23:29:09 +08:00
Meng Zhang c46142385b fix: properly set /data ownership 2023-04-05 23:13:34 +08:00
Meng Zhang bc25593f93 fix: create LOGS_DIR if not exits 2023-04-05 23:04:42 +08:00
Meng Zhang 7f5189210a fix: don't set web concurrency by default 2023-04-05 22:20:07 +08:00
Meng Zhang db77d7f267
feat: support single container (#46)
* docs: update readme

* fix: do not exclude peft

* Free disk space before docker building

* fix: fix docker-compose

* fix: dockercompose user to 1000

* fix dockerfile

* fix: cachedir ownership
2023-04-05 20:19:43 +08:00
Meng Zhang 499a2adab9 set default theme to dark 2023-04-04 20:02:57 +08:00
Meng Zhang 658a1f1c24 refactor: move vector.toml to tabby/config
fix
2023-04-04 20:02:56 +08:00
Meng Zhang 79585cc2a4
feat: improve events system (#40)
* feat: improve events system

* docs: add Events.md for Event sub system.

* Link vector.toml
2023-04-04 13:22:16 +08:00
Meng Zhang 962f4b53b5 feat: set WEB_CONCURRENCY to number of CPU cores 2023-04-04 12:04:52 +08:00
Meng Zhang 1c61ef3944
feat: integrate projects / dataset information in admin. (#38)
* feat: add projects page in admin

* feat: integrate update_dataset job

* feat: display dataset info in projects
2023-04-03 13:04:04 +08:00
Meng Zhang 6e0ab3af49 fix: model replica not working for skypilot 2023-04-02 21:15:42 +08:00
Meng Zhang 22fbaefbd4
feat: add deployment script on lambda cloud with skypilot (#37)
* feat: add deployment script on lambda cloud with skypilot

* docs: adjust API documentation level

* fix: move docker-compose install before the docker-compose pull

* Fix documentation

* Update replica updating
2023-04-02 20:32:49 +08:00
Meng Zhang 0a66a9d498 feat: support MODEL_REPLICA to control number of model instances in triton 2023-04-02 13:19:09 +08:00
Meng Zhang 2cda4fd07b feat: add skypilot.yml for skypilot (https://github.com/skypilot-org/skypilot) deployment 2023-04-02 13:09:49 +08:00
Meng Zhang 477ef83319 Update startup_periord to 1200s as a HF model download might take time 2023-03-29 16:45:46 +08:00
Meng Zhang 20801bbe8c
Cleanup environment variable (#30)
* Remove EVENTS_LOG_DIR

* Rename supervisord.sh -> tabby.sh
2023-03-29 16:33:00 +08:00
Meng Zhang bf7d149a27
Add supervisord to support a single docker run deployment (#29)
* Add suppervisord in dockerfile

* Create supervisord

* Update README.md

* Update README.md
2023-03-29 12:57:03 +08:00
Meng Zhang d966f05abd
Add Completion Events & Acceptance Rate in metrics panel. (#26)
* Add duckdb

* Add basic Metrics w/duckdb
2023-03-28 20:12:03 +08:00
Meng Zhang 2f8714e6fe
Add vector logging for tabby-server events. (#25)
* Switch to dagu for init job

* Add processed logging
2023-03-28 16:32:35 +08:00
Meng Zhang c990ba843f Extract environment variable 2023-03-27 13:42:06 +08:00
Meng Zhang d5d58fbbec
Improve documentations. (#20)
* Improve help message of model preload

* Update development/scripts/triton.sh

* Improve documents

* Update deployment.md

* Update deployment.md
2023-03-27 11:46:18 +08:00
Meng Zhang 9d92821cf5
Add gptj converter (#19)
* Rename deployment-next to development

* Add GPTJ converter
2023-03-27 11:12:52 +08:00
Meng Zhang 92ddbe8705
Update README.md 2023-03-27 01:12:37 +08:00
Meng Zhang 9739683fba
Switch default deploy model (#18) 2023-03-27 01:10:15 +08:00
Meng Zhang da40363057
Update README.md 2023-03-27 00:57:47 +08:00
Meng Zhang 1c3ec20f93
Prepare public release with a minimal deployment setup (#16)
* Move deployment to deployment-next

* Add deployment setup

* Update deployment-next

* Remove vector label

* update README.md
2023-03-26 22:44:15 +08:00
Meng Zhang e6bf16711f
Fix misc mistakes (#13)
* Fix misc

* Fix docker-compose
2023-03-25 21:37:38 +08:00
Meng Zhang b622bd6762
use TabbyML/NeoX-70M for minimal e2e deployment (#10)
* use TabbyML/NeoX-70M for minimal e2e deployment

* Use python3 of triton image
2023-03-25 17:39:40 +08:00
Meng Zhang 8144e4f83a
Add tabby.tools.repository.updater for easier git repository synchronization. (#9)
* Move dags to tabby.tasks

* Add repository syncer

* Follow redirect for curl
2023-03-25 14:44:46 +08:00
Meng Zhang 8cf533016a
Move python code under tabby/ (#8)
* Add tabby config file

* Rename train.yaml to trainer.yaml

* Change server to relative import

* Move source files into tabby

* Rename conf
2023-03-25 12:20:29 +08:00
Meng Zhang 1038bb39a1
Add dagu for data processing job orchestration (#7)
* Install dagu

* Move dagu install to first stage

* Fix metrics

* Add DAGs for create dataset from code repository
2023-03-25 00:05:47 +08:00
Meng Zhang 0f5a959269 Remove tokenizer in testdata/ 2023-03-24 09:53:48 +08:00
Meng Zhang a0b438da06
Add python transformer backend for tabby (mainly used for local dev / test in non-cuda environment) (#6)
* Add python backend

* Split docker-compose.triton.yml

* update makefile
2023-03-23 14:14:33 +08:00
Meng Zhang df149fad61
Reduce repo size (#5)
* Remove binary files from git

* Add Makefile

* update
2023-03-23 13:10:59 +08:00
Meng Zhang 5c5a0ad8c3 Create config dir for config files in deployment 2023-03-22 23:58:55 +08:00
Meng Zhang dddbcdc116
Create README.md 2023-03-22 23:57:26 +08:00
Meng Zhang 3409a096af Fix docker-compose.yml 2023-03-22 23:47:35 +08:00
Meng Zhang bcce00794e
Add admin panel (w/ streamlit) and logging (w/ vectordev) (#4) 2023-03-22 23:18:12 +08:00