Commit Graph

37 Commits

Author SHA1 Message Date
Aaron Pham
0d83cefcb6 fix(mixtral): setup hack atm to load weights from pt specifically instead of safetensors (#776)
fix(mixtral): setup hack atm to load weights from pt specifically
instead of safetensors

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
2023-12-13 18:18:51 -05:00
Aaron
ce6efc2a9e chore(style): cleanup bytes
Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
2023-11-28 01:27:27 -05:00
Aaron Pham
52a44b1bfa chore: cleanup loader (#729)
Signed-off-by: Aaron Pham <29749331+aarnphm@users.noreply.github.com>
2023-11-22 21:51:51 -05:00
Aaron Pham
8bb2742a9a chore(types): append additional types change (#719)
* chore(types): append additional types change

Signed-off-by: Aaron Pham <29749331+aarnphm@users.noreply.github.com>

* chore: add arguments for parsing dir

Signed-off-by: Aaron Pham <29749331+aarnphm@users.noreply.github.com>

---------

Signed-off-by: Aaron Pham <29749331+aarnphm@users.noreply.github.com>
2023-11-21 22:38:20 -05:00
Aaron Pham
77bd6f090a chore(logger): fix warnings and streamline style (#717)
Sorry but there are too much wasted spacing in `_llm.py`, and I'm unhappy and not productive anytime I look or want to do anything with it

---------

Signed-off-by: Aaron Pham <29749331+aarnphm@users.noreply.github.com>
Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2023-11-21 18:54:51 -05:00
Aaron Pham
816c1ee80e feat(engine): CTranslate2 (#698)
* chore: update instruction for dependencies

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* feat(experimental): CTranslate2

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

---------

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
2023-11-19 10:25:08 -05:00
Aaron Pham
206521e02d feat(ctranslate): initial infrastructure support (#694)
* perf: compact and improve speed and agility

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* --wip--

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* chore: cleanup infrastructure

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* chore: update styles notes and autogen mypy configuration

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

---------

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
2023-11-19 01:48:33 -05:00
Aaron Pham
09cc84a56c chore(loading): include verbose warning about trust_remote_code (#674)
Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
2023-11-16 20:09:50 -05:00
Aaron Pham
a58d947bc8 perf: improve build logics and cleanup speed (#657)
Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
2023-11-15 00:18:31 -05:00
Aaron Pham
126e6c9d63 fix(ruff): correct consistency between isort and formatter (#624)
Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
2023-11-12 21:12:50 -05:00
Aaron Pham
7e1fb35a71 chore(llm): expose quantise and lazy load heavy imports (#617)
* chore(llm): expose quantise and lazy load heavy imports

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* chore: move transformers to TYPE_CHECKING block

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

---------

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
2023-11-12 14:55:37 -05:00
Aaron Pham
c41828f68f feat(client): support authentication token and shim implementation (#605)
* chore: synch generate_iterator to be the same as server

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* --wip--

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* wip

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* feat: cleanup shim implementation

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* ci: auto fixes from pre-commit.ci

For more information, see https://pre-commit.ci

* chore: fix pre-commit

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* chore: update changelog

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* chore: update check with tuple

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

---------

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2023-11-10 17:44:31 -05:00
Aaron Pham
fa2038f4e2 fix: loading correct local models (#599)
* fix(model): loading local correctly

Signed-off-by: Aaron Pham <29749331+aarnphm@users.noreply.github.com>

* chore: update repr and correct bentomodel processor

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* ci: auto fixes from pre-commit.ci

For more information, see https://pre-commit.ci

* chore: cleanup transformers implementation

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* fix: ruff to ignore I001 on all stubs

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

---------

Signed-off-by: Aaron Pham <29749331+aarnphm@users.noreply.github.com>
Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2023-11-10 02:36:12 -05:00
Aaron Pham
ac377fe490 infra: using ruff formatter (#594)
Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
2023-11-09 12:44:05 -05:00
Aaron Pham
b8a2e8cf91 refactor(cli): cleanup API (#592)
* chore: remove unused imports

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* refactor(cli): update to only need model_id

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* feat: `openllm start model-id`

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* chore: add changelog

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* chore: update changelog notice

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* chore: update correct config and running tools

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* chore: update backward compat options and treat JSON outputs
corespondingly

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

---------

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
2023-11-09 11:40:17 -05:00
Aaron Pham
47107727b3 feat(vllm): squeezellm (#588)
* feat(vllm): squeezellm

Signed-off-by: Aaron Pham <29749331+aarnphm@users.noreply.github.com>

* fix: correct import_model with awq and gatekeep squeezellm for PyTorch

Signed-off-by: Aaron Pham <29749331+aarnphm@users.noreply.github.com>

---------

Signed-off-by: Aaron Pham <29749331+aarnphm@users.noreply.github.com>
2023-11-08 07:21:27 -05:00
Aaron Pham
ff8b6377c8 fix(awq): correct awq detection for support (#586)
* fix(awq): correct detection for awq

Signed-off-by: Aaron Pham <29749331+aarnphm@users.noreply.github.com>

* chore: update base docker to work

Signed-off-by: Aaron Pham <29749331+aarnphm@users.noreply.github.com>

* chore: disable awq on pytorch for now

Signed-off-by: Aaron Pham <29749331+aarnphm@users.noreply.github.com>

* ci: auto fixes from pre-commit.ci

For more information, see https://pre-commit.ci

---------

Signed-off-by: Aaron Pham <29749331+aarnphm@users.noreply.github.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2023-11-08 06:57:11 -05:00
Aaron Pham
97d7c38fea refactor: cleanup typing to expose correct API (#576)
Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
2023-11-08 01:24:03 -05:00
Aaron Pham
e2029c934b perf: unify LLM interface (#518)
Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
Signed-off-by: paperspace <29749331+aarnphm@users.noreply.github.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2023-11-06 20:39:43 -05:00
Aaron Pham
35e6945e86 fix(serialisation): vLLM safetensors support (#324)
* fix(serilisation): vllm support for safetensors

Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>

* chore: running tools

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* chore: generalize one shot generation

Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>

* chore: add changelog [skip ci]

Signed-off-by: paperspace <29749331+aarnphm@users.noreply.github.com>

---------

Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>
Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
Signed-off-by: paperspace <29749331+aarnphm@users.noreply.github.com>
2023-09-12 17:44:01 -04:00
aarnphm-ec2-dev
8530a067ea chore(serialisation): dump quantization_config.json to conform with
optimum load

Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>
2023-09-07 16:50:50 +00:00
Aaron Pham
956b3a53bc fix(gptq): use upstream integration (#297)
* wip

Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>

* feat: GPTQ transformers integration

Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>

* fix: only load if variable is available and add changelog

Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>

* chore: remove boilerplate check

Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>

---------

Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>
2023-09-04 14:05:50 -04:00
aarnphm-ec2-dev
7d893e6cd2 chore: ignore new lines split [skip ci]
Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>
2023-09-01 17:00:49 +00:00
Aaron Pham
608de0b667 fix(serving): vllm distributed size (#285)
* chore(weights): ignore gguf pattern for non GGML backend

Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>

* chore: correct fix num_gpus to be divisble by 2

This depends on the attention_heads from given models

Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>

---------

Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>
2023-09-01 12:37:10 -04:00
Aaron Pham
b7af7765d4 fix(yapf): align weird new lines break [generated] [skip ci] (#284)
fix(yapf): align weird new lines break

Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>
2023-09-01 05:34:22 -04:00
Aaron Pham
3e45530abd refactor(breaking): unify LLM API (#283)
Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>
Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
2023-09-01 05:15:19 -04:00
Aaron
b545ad2ad1 style: google
Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
2023-08-30 13:52:35 -04:00
Aaron Pham
c9cef1d773 fix: persistent styling between ruff and yapf (#279) 2023-08-30 11:37:41 -04:00
aarnphm-ec2-dev
806a663e4a chore(style): add one blank line
to conform with Google style

Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>
2023-08-26 11:36:57 +00:00
Aaron Pham
46c8904806 cron(style): run formatter [generated] [skip ci] (#257) 2023-08-25 06:38:59 -04:00
Aaron
787ce1b3b6 chore(style): synchronized style across packages [skip ci]
Signed-off-by: Aaron <29749331+aarnphm@users.noreply.github.com>
2023-08-23 08:46:22 -04:00
aarnphm-ec2-dev
eddbc06374 chore(style): reduce line length and truncate compression
Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>
2023-08-22 17:02:00 +00:00
aarnphm-ec2-dev
1488fbb167 chore(style): enable yapf to match with style guidelines
Signed-off-by: aarnphm-ec2-dev <29749331+aarnphm@users.noreply.github.com>
2023-08-22 14:03:06 +00:00
Aaron Pham
3ffb25a872 refactor: packages (#249) 2023-08-22 08:55:46 -04:00
Aaron Pham
9e205b4963 feat: token streaming and SSE support (#240) 2023-08-20 07:32:49 -04:00
Aaron Pham
4140d160b8 feat(embedding): Adding generic endpoint (#227) 2023-08-17 15:17:00 -04:00
Aaron Pham
cd872ef631 refactor: monorepo (#203) 2023-08-15 02:11:14 -04:00