llama-stack/llama_stack/distribution
Ashwin Bharambe 8de8eb03c8
Update the "InterleavedTextMedia" type (#635)
## What does this PR do?

This is a long-pending change and particularly important to get done
now.

Specifically:
- we cannot "localize" (aka download) any URLs from media attachments
anywhere near our modeling code. it must be done within llama-stack.
- `PIL.Image` is infesting all our APIs via `ImageMedia ->
InterleavedTextMedia` and that cannot be right at all. Anything in the
API surface must be "naturally serializable". We need a standard `{
type: "image", image_url: "<...>" }` which is more extensible
- `UserMessage`, `SystemMessage`, etc. are moved completely to
llama-stack from the llama-models repository.

See https://github.com/meta-llama/llama-models/pull/244 for the
corresponding PR in llama-models.

## Test Plan

```bash
cd llama_stack/providers/tests

pytest -s -v -k "fireworks or ollama or together" inference/test_vision_inference.py
pytest -s -v -k "(fireworks or ollama or together) and llama_3b" inference/test_text_inference.py
pytest -s -v -k chroma memory/test_memory.py \
  --env EMBEDDING_DIMENSION=384 --env CHROMA_DB_PATH=/tmp/foobar

pytest -s -v -k fireworks agents/test_agents.py  \
   --safety-shield=meta-llama/Llama-Guard-3-8B \
   --inference-model=meta-llama/Llama-3.1-8B-Instruct
```

Updated the client sdk (see PR ...), installed the SDK in the same
environment and then ran the SDK tests:

```bash
cd tests/client-sdk
LLAMA_STACK_CONFIG=together pytest -s -v agents/test_agents.py
LLAMA_STACK_CONFIG=ollama pytest -s -v memory/test_memory.py

# this one needed a bit of hacking in the run.yaml to ensure I could register the vision model correctly
INFERENCE_MODEL=llama3.2-vision:latest LLAMA_STACK_CONFIG=ollama pytest -s -v inference/test_inference.py
```
2024-12-17 11:18:31 -08:00
..
routers Update the "InterleavedTextMedia" type (#635) 2024-12-17 11:18:31 -08:00
server fix telemetry import (#585) 2024-12-09 11:18:53 -08:00
store Update the "InterleavedTextMedia" type (#635) 2024-12-17 11:18:31 -08:00
tests Fix bedrock inference impl 2024-12-16 14:22:34 -08:00
ui Add eval/scoring/datasetio API providers to distribution templates & UI developer guide (#564) 2024-12-05 16:29:32 -08:00
utils Ensure model_local_dir does not mangle "C:\" on Windows 2024-11-24 14:18:59 -08:00
__init__.py API Updates (#73) 2024-09-17 19:51:35 -07:00
build.py Miscellaneous fixes around telemetry, library client and run yaml autogen 2024-12-08 20:40:22 -08:00
build_conda_env.sh fix prompt guard (#177) 2024-10-03 11:07:53 -07:00
build_container.sh Make run yaml optional so dockers can start with just --env (#492) 2024-11-20 13:11:40 -08:00
build_venv.sh Miscellaneous fixes around telemetry, library client and run yaml autogen 2024-12-08 20:40:22 -08:00
client.py use API version in "remote" stack client 2024-11-19 15:59:47 -08:00
common.sh API Updates (#73) 2024-09-17 19:51:35 -07:00
configure.py use logging instead of prints (#499) 2024-11-21 11:32:53 -08:00
configure_container.sh docker: Check for selinux before using --security-opt (#167) 2024-10-02 10:37:41 -07:00
datatypes.py Miscellaneous fixes around telemetry, library client and run yaml autogen 2024-12-08 20:40:22 -08:00
distribution.py Kill "remote" providers and fix testing with a remote stack properly (#435) 2024-11-12 21:51:29 -08:00
inspect.py Remove "routing_table" and "routing_key" concepts for the user (#201) 2024-10-10 10:24:13 -07:00
library_client.py Update the "InterleavedTextMedia" type (#635) 2024-12-17 11:18:31 -08:00
request_headers.py fixes tests & move braintrust api_keys to request headers (#535) 2024-11-26 13:11:21 -08:00
resolver.py [1/n] torchtune <> llama-stack integration skeleton (#540) 2024-12-13 11:05:35 -08:00
stack.py Update the "InterleavedTextMedia" type (#635) 2024-12-17 11:18:31 -08:00
start_conda_env.sh Move to use argparse, fix issues with multiple --env cmdline options 2024-11-18 16:31:59 -08:00
start_container.sh Move to use argparse, fix issues with multiple --env cmdline options 2024-11-18 16:31:59 -08:00