phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-13 19:32:37 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
334 commits 57 branches 151 tags 128 MiB
Commit graph

334 commits

This branch
This branch
All branches
Author SHA1 Message Date
Xi Yan
202667f3db delete templates 2024-10-21 11:03:34 -07:00
Xi Yan
3ca822f4cd build templates 2024-10-21 11:02:32 -07:00
Xi Yan
ca2e7f52bd vllm 2024-10-21 11:00:50 -07:00
nehal-a2z
8ef3d3d239 Update event_logger.py (#275) 
spelling error
 2024-10-21 10:48:50 -07:00
raghotham
af52c22c5e Create .readthedocs.yaml 
Trying out readthedocs
 2024-10-21 10:46:47 -07:00
Yuan Tang
74e6356b51 Add vLLM inference provider for OpenAI compatible vLLM server (#178) 
This PR adds vLLM inference provider for OpenAI compatible vLLM server.
 2024-10-21 10:46:45 -07:00
Ashwin Bharambe
391dedd1c0 update ollama for llama-guard3 2024-10-21 10:46:40 -07:00
Ashwin Bharambe
89759a0ad3 Improve an important error message 2024-10-21 10:46:40 -07:00
Ashwin Bharambe
5863f65874 Make all methods async def again; add completion() for meta-reference (#270) 
PR #201 had made several changes while trying to fix issues with getting the stream=False branches of inference and agents API working. As part of this, it made a change which was slightly gratuitous. Namely, making chat_completion() and brethren "def" instead of "async def".

The rationale was that this allowed the user (within llama-stack) of this to use it as:

```
async for chunk in api.chat_completion(params)
```

However, it causes unnecessary confusion for several folks. Given that clients (e.g., llama-stack-apps) anyway use the SDK methods (which are completely isolated) this choice was not ideal. Let's revert back so the call now looks like:

```
async for chunk in await api.chat_completion(params)
```

Bonus: Added a completion() implementation for the meta-reference provider. Technically should have been another PR :)
 2024-10-21 10:46:40 -07:00
Ashwin Bharambe
92aca57bfa Small rename 2024-10-21 10:46:40 -07:00
Ashwin Bharambe
6f4537b4c4 Allow overridding checkpoint_dir via config 2024-10-21 10:46:40 -07:00
Ashwin Bharambe
a90ab5878b Add an option to not use elastic agents for meta-reference inference (#269) 2024-10-21 10:46:40 -07:00
Xi Yan
2f5c410c73 [bugfix] fix case for agent when memory bank registered without specifying provider_id (#264) 
* fix case where memory bank is registered without provider_id

* memory test

* agents unit test
 2024-10-21 10:46:40 -07:00
Xi Yan
29c8edb4f6 readme 2024-10-21 09:11:25 -07:00
Xi Yan
5ea36b0274 readme 2024-10-21 09:03:05 -07:00
Xi Yan
d4caab3c67 developer cookbook 2024-10-21 09:01:34 -07:00
Xi Yan
302fa5c4bb build/developer cookbook/new api provider 2024-10-21 09:01:22 -07:00
Xi Yan
f58441cc21 readme 2024-10-18 18:55:29 -07:00
Xi Yan
100b5fecd4 readme 2024-10-18 18:53:49 -07:00
Xi Yan
955743ba7a kill distribution/templates 2024-10-18 17:32:11 -07:00
Xi Yan
c830235936 rename 2024-10-18 17:28:26 -07:00
Xi Yan
cbb423a32f move distribution/templates to distributions/ 2024-10-18 17:21:50 -07:00
Xi Yan
b4aca0aeb6 move distribution folders 2024-10-18 17:05:41 -07:00
Xi Yan
fd90d2ae97 readme 2024-10-18 14:30:44 -07:00
Xi Yan
a3f748a875 readme for distributions 2024-10-18 14:21:44 -07:00
Xi Yan
dcac9e4874 update compose file 2024-10-18 11:12:27 -07:00
Xi Yan
542ffbee72 comment 2024-10-17 19:37:22 -07:00
Xi Yan
293d8f2895 docker compose ollama 2024-10-17 19:31:29 -07:00
Ashwin Bharambe
9fcf5d58e0 Allow overriding MODEL_IDS for inference test 2024-10-17 10:03:27 -07:00
Xi Yan
02be26098a getting started 2024-10-16 23:56:21 -07:00
Xi Yan
cf9e5b76b2
Update getting_started.md 2024-10-16 23:52:29 -07:00
Xi Yan
7cc47da8f2
Update getting_started.md 2024-10-16 23:50:31 -07:00
Xi Yan
d787d1e84f
config templates restructure, docs (#262) 
* wip

* config templates

* readmes
 2024-10-16 23:25:10 -07:00
Tam
a07dfffbbf
initial changes (#261) 
Update the parsing logic for comma-separated list and download function
 2024-10-16 23:15:59 -07:00
ATH
319a6b5f83
Update getting_started.md (#260) 2024-10-16 18:05:36 -07:00
Xi Yan
c4d5d6bb91
Docker compose scripts for remote adapters (#241) 
* tgi docker compose

* path

* wait for tgi server to start before starting server

* update provider-id

* move scripts to distribution/ folder

* add readme

* readme
 2024-10-15 16:32:53 -07:00
Matthieu FRONTON
770647dede
Fix broken rendering in Google Colab (#247) 2024-10-15 15:41:49 -07:00
Ashwin Bharambe
09b793c4d6 Fix fp8 implementation which had bit-rotten a bit 
I only tested with "on-the-fly" bf16 -> fp8 conversion, not the "load
from fp8" codepath.

YAML I tested with:

```
providers:
  - provider_id: quantized
    provider_type: meta-reference-quantized
    config:
      model: Llama3.1-8B-Instruct
      quantization:
        type: fp8
```
 2024-10-15 13:57:01 -07:00
Yuan Tang
80ada04f76
Remove request arg from chat completion response processing (#240) 
Signed-off-by: Yuan Tang <terrytangyuan@gmail.com>
 2024-10-15 13:03:17 -07:00
Xi Yan
209cd3d35e Bump version to 0.0.42 2024-10-14 11:13:04 -07:00
Yuan Tang
a2b87ed0cb
Switch to pre-commit/action (#239) 2024-10-11 11:09:11 -07:00
Yuan Tang
05282d1234
Enable pre-commit on main branch (#237) 2024-10-11 10:03:59 -07:00
Yuan Tang
2128e61da2
Fix incorrect completion() signature for Databricks provider (#236) 2024-10-11 08:47:57 -07:00
Dalton Flanagan
9fbe8852aa
Add Swift Package Index badge 2024-10-10 23:39:25 -04:00
Xi Yan
ca29980c6b fix agents context retriever 2024-10-10 20:17:29 -07:00
Ashwin Bharambe
1ff0476002 Split off meta-reference-quantized provider 2024-10-10 16:03:19 -07:00
Xi Yan
7ff5800dea generate openapi 2024-10-10 15:30:34 -07:00
Dalton Flanagan
a3e65d58a9
Add logo 2024-10-10 15:04:21 -04:00
Russell Bryant
eba9d1ea14
ci: Run pre-commit checks in CI (#176) 
Run the pre-commit checks in a github workflow to validate that a PR
or a direct push to the repo does not introduce new errors.
 2024-10-10 11:21:59 -07:00
Ashwin Bharambe
89d24a07f0 Bump version to 0.0.41 2024-10-10 10:27:03 -07:00