phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-17 23:59:47 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 2
716 commits 57 branches 151 tags 128 MiB
Commit graph

3 commits

Author SHA1 Message Date
Dinesh Yeduguru
ee3f0c6b55 fix failing memory tests 2024-12-12 11:43:06 -08:00
Dinesh Yeduguru
d362d2d740 implement embedding generation in supported inference providers (#589) 
This PR adds the ability to generate embeddings in all supported
inference providers.

```
pytest -v -s llama_stack/providers/tests/inference/test_embeddings.py -k "bedrock" --inference-model="amazon.titan-embed-text-v2:0"  --env EMBEDDING_DIMENSION=1024

 pytest -v -s -k "vllm"  --inferrence-model="intfloat/e5-mistral-7b-instruct"  llama_stack/providers/tests/inference/test_embeddings.py --env EMBEDDING_DIMENSION=4096  --env VLLM_URL="http://localhost:9798/v1"

pytest -v -s --inference-model="nomic-ai/nomic-embed-text-v1.5"  llama_stack/providers/tests/inference/test_embeddings.py  -k "fireworks"  --env FIREWORKS_API_KEY=<API_KEY>--env EMBEDDING_DIMENSION=128

pytest -v -s --inference-model="togethercomputer/m2-bert-80M-2k-retrieval"  llama_stack/providers/tests/inference/test_embeddings.py  -k "together"  --env TOGETHER_API_KEY=<API_KEY>--env EMBEDDING_DIMENSION=768

pytest -v -s -k "ollama"  --inference-model="all-minilm:v8"  llama_stack/providers/tests/inference/test_embeddings.py --env EMBEDDING_DIMENSION=384

 torchrun $CONDA_PREFIX/bin/pytest -v -s -k "meta_reference" --inference-model="sentence-transformers/all-MiniLM-L6-v2"  llama_stack/providers/tests/inference/test_embeddings.py --env EMBEDDING_DIMENSION=384

```
 2024-12-12 11:25:09 -08:00
Ashwin Bharambe
ffedb81c11
Significantly simpler and malleable test setup (#360) 
* Significantly simpler and malleable test setup

* convert memory tests

* refactor fixtures and add support for composable fixtures

* Fix memory to use the newer fixture organization

* Get agents tests working

* Safety tests work

* yet another refactor to make this more general

now it accepts --inference-model, --safety-model options also

* get multiple providers working for meta-reference (for inference + safety)

* Add README.md

---------

Co-authored-by: Ashwin Bharambe <ashwin@meta.com>
 2024-11-04 17:36:43 -08:00