This is a follow up to:
https://github.com/meta-llama/llama-stack/pull/1140
Signed-off-by: Ihar Hrachyshka <ihar.hrachyshka@gmail.com>
# What does this PR do?
[Provide a short summary of what this PR does and why. Link to relevant
issues if applicable.]
Avoid unnecessary GPU memory clean attempt when the GPU is not used for
training.
[//]: # (If resolving an issue, uncomment and update the line below)
[//]: # (Closes #[issue-number])
## Test Plan
With CPU:
```
INFO 2025-02-26 16:43:56,267 torchtune.utils._logging:121: Model checkpoint of size 6.43 GB saved to /Users/ihrachys/.llama/checkpoints/meta-llama/Llama-3.2-3B-Instruct-sft-0/consolidated.00.pth
INFO 2025-02-26 16:43:56,274 torchtune.utils._logging:132: Adapter checkpoint of size 0.00 GB saved to /Users/ihrachys/.llama/checkpoints/meta-llama/Llama-3.2-3B-Instruct-sft-0/adapter/adapter.pth
model_file_path /Users/ihrachys/.llama/checkpoints/meta-llama/Llama-3.2-3B-Instruct-sft-0
```
With CUDA:
```
INFO 2025-02-26 21:39:24,314 torchtune.utils._logging:121: Model checkpoint of size 6.43 GB saved to /home/ec2-user/.llama/checkpoints/meta-llama/Llama-3.2-3B-Instruct-sft-0/consolidated.00.pth
INFO 2025-02-26 21:39:24,333 torchtune.utils._logging:132: Adapter checkpoint of size 0.00 GB saved to /home/ec2-user/.llama/checkpoints/meta-llama/Llama-3.2-3B-Instruct-sft-0/adapter/adapter.pth
model_file_path /home/ec2-user/.llama/checkpoints/meta-llama/Llama-3.2-3B-Instruct-sft-0
```
[//]: # (## Documentation)
Signed-off-by: Ihar Hrachyshka <ihar.hrachyshka@gmail.com>
# Summary:
Our tests sometimes error out with
```
========================== 11 passed, 342 warnings in 58.86s ==========================
Error exporting span to SQLite: Cannot operate on a closed database.
Fatal Python error: _enter_buffered_busy: could not acquire lock for <_io.BufferedWriter name='<stdout>'> at interpreter shutdown, possibly due to daemon threads
Python runtime state: finalizing (tstate=0x000000012af04280)
Current thread 0x00000001fa29c240 (most recent call first):
<no Python frame>
```
Usually able to repro this by running 10 times.
The proposed fix is to use threadsafe var for creating sqlite connection
to ensure connection is only used by one thread. Not 100% if this is the
fix, but am not able to repro with this.
# Test Plan:
Run 10 times and saw no more errors
```
for i in {1..10}; do
echo "=== Starting Run $i ==="
LLAMA_STACK_CONFIG=fireworks pytest -s -v tests/client-sdk/agents/test_agents.py --safety-shield meta-llama/Llama-Guard-3-8B
if [[ $? -ne 0 ]]; then
echo "=== Run $i FAILED with exit code $? ==="
break
else
echo "=== Run $i PASSED ==="
fi
echo
done
```
Summary:
Lets the model decide which tool it needs to call to respond to a query.
Test Plan:
```
LLAMA_STACK_CONFIG=fireworks pytest -s -v tests/client-sdk/ --safety-shield meta-llama/Llama-Guard-3-8B
```
Also evaluated on a small benchmark with 20 questions from HotpotQA.
With this PR and some prompting, the performance is 77% recall compared
to 50% currently.
---
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed
with
[ReviewStack](https://reviewstack.dev/meta-llama/llama-stack/pull/1015).
* #1268
* #1239
* __->__ #1015
# What does this PR do?
This PR makes a couple of changes required to get the test
`tests/client-sdk/agents/test_agents.py::test_builtin_tool_web_search`
passing on the remote-vllm provider.
First, we adjust agent_instance to also pass in the description and
parameters of builtin tools. We need these parameters so we can pass the
tool's expected parameters into vLLM. The meta-reference implementations
may not have needed these for builtin tools, as they are able to take
advantage of the Llama-model specific support for certain builtin tools.
However, with vLLM, our server-side chat templates for tool calling
treat all tools the same and don't separate out Llama builtin vs custom
tools. So, we need to pass the full set of parameter definitions and
list of required parameters for builtin tools as well.
Next, we adjust the vllm streaming chat completion code to fix up some
edge cases where it was returning an extra ChatCompletionResponseEvent
with an empty ToolCall with empty string call_id, tool_name, and
arguments properties. This is a bug discovered after the above fix,
where after a successful tool invocation we were sending extra chunks
back to the client with these empty ToolCalls.
## Test Plan
With these changes, the following test that previously failed now
passes:
```
VLLM_URL="http://localhost:8000/v1" \
INFERENCE_MODEL="meta-llama/Llama-3.2-3B-Instruct" \
LLAMA_STACK_CONFIG=remote-vllm \
python -m pytest -v \
tests/client-sdk/agents/test_agents.py::test_builtin_tool_web_search \
--inference-model "meta-llama/Llama-3.2-3B-Instruct"
```
Additionally, I ran the remote-vllm client-sdk and provider inference
tests as below to ensure they all still passed with this change:
```
VLLM_URL="http://localhost:8000/v1" \
INFERENCE_MODEL="meta-llama/Llama-3.2-3B-Instruct" \
LLAMA_STACK_CONFIG=remote-vllm \
python -m pytest -v \
tests/client-sdk/inference/test_text_inference.py \
--inference-model "meta-llama/Llama-3.2-3B-Instruct"
```
```
VLLM_URL="http://localhost:8000/v1" \
python -m pytest -s -v \
llama_stack/providers/tests/inference/test_text_inference.py \
--providers "inference=vllm_remote"
```
[//]: # (## Documentation)
Signed-off-by: Ben Browning <bbrownin@redhat.com>
## context
Now, in llama stack, we only support inference / eval a finetuned
checkpoint with meta-reference as inference provider. This is
sub-optimal since meta-reference is pretty slow.
Our vision is that developer can inference / eval a finetuned checkpoint
produced by post training apis with all the inference providers on the
stack. To achieve this, we'd like to define an unified output checkpoint
format for post training providers. So that, all the inference provider
can respect that format for customized model inference.
By spotting check how
[ollama](https://github.com/ollama/ollama/blob/main/docs/import.md) and
[fireworks](https://docs.fireworks.ai/models/uploading-custom-models) do
inference on a customized model, we defined the output checkpoint format
as /adapter/adapter_config.json and /adapter/adapter_model.safetensors
(as we only support LoRA post training now, we begin from adapter only
checkpoint)
## test
we kick off a post training job and configured checkpoint format as
'huggingface'. Output files
we did a proof of concept with ollama to see if ollama can inference our
finetuned checkpoint
1. create Modelfile like
<img width="799" alt="Screenshot 2025-01-22 at 5 04 18 PM"
src="https://github.com/user-attachments/assets/7fca9ac3-a294-44f8-aab1-83852c600609"
/>
2. create a customized model with `ollama create llama_3_2_finetuned`
and run inference successfully
This is just a proof of concept with ollama cmd line. As next step, we'd
like to wrap loading / inference customized model logic in the inference
provider implementation.
# What does this PR do?
[Provide a short summary of what this PR does and why. Link to relevant
issues if applicable.]
to the llama-stack-client-swift repo - PR:
https://github.com/meta-llama/llama-stack-client-swift/pull/22
[//]: # (If resolving an issue, uncomment and update the line below)
[//]: # (Closes #[issue-number])
## Test Plan
[Describe the tests you ran to verify your changes with result
summaries. *Provide clear instructions so the plan can be easily
re-executed.*]
[//]: # (## Documentation)
# What does this PR do?
- Fixed type hinting and missing imports across multiple modules.
- Improved compatibility by using `TYPE_CHECKING` for conditional
imports.
- Updated `pyproject.toml` to enforce stricter linting.
Signed-off-by: Sébastien Han <seb@redhat.com>
Signed-off-by: Sébastien Han <seb@redhat.com>
# What does this PR do?
When there are issues with the tool call function, an exception is
raised but the error message is not informative. This adds a clearer
message to tell users to check their functions.
```
Traceback (most recent call last):
File "/Users/phayes/projects/llama-stack/llama-stack/llama_stack/distribution/server/server.py", line 208, in sse_generator
async for item in event_gen:
File "/Users/phayes/projects/llama-stack/llama-stack/llama_stack/providers/inline/agents/meta_reference/agents.py", line 165, in _create_agent_turn_streaming
async for event in agent.create_and_execute_turn(request):
File "/Users/phayes/projects/llama-stack/llama-stack/llama_stack/providers/inline/agents/meta_reference/agent_instance.py", line 197, in create_and_execute_turn
async for chunk in self.run(
File "/Users/phayes/projects/llama-stack/llama-stack/llama_stack/providers/inline/agents/meta_reference/agent_instance.py", line 389, in run
async for res in self._run(
File "/Users/phayes/projects/llama-stack/llama-stack/llama_stack/providers/inline/agents/meta_reference/agent_instance.py", line 811, in _run
content=tool_result.content,
AttributeError: 'NoneType' object has no attribute 'content'
```
## Test Plan
Ran the same script and exception is raised with clearer error message.
Signed-off-by: Yuan Tang <terrytangyuan@gmail.com>
# What does this PR do?
[Provide a short summary of what this PR does and why. Link to relevant
issues if applicable.]
[//]: # (If resolving an issue, uncomment and update the line below)
[//]: # (Closes #[issue-number])
## Test Plan
[Describe the tests you ran to verify your changes with result
summaries. *Provide clear instructions so the plan can be easily
re-executed.*]
[//]: # (## Documentation)
Summary:
kotlin SDK expects this format
Test Plan:
python prints the expected format
>>> str(datetime.now().astimezone())
'2025-02-24 22:02:58.729763-08:00'
Summary:
Allows tools to output metadata. This is useful for evaluating tool
outputs, e.g. RAG tool will output document IDs, which can be used to
score recall.
Will need to make a similar change on the client side to support
ClientTool outputting metadata.
Test Plan:
LLAMA_STACK_CONFIG=fireworks pytest -s -v
tests/client-sdk/agents/test_agents.py
# Problem
Our current Agent framework has discrepancies in definition on how we
handle server side and client side tools.
1. Server Tools: a single Turn is returned including `ToolExecutionStep`
in agenst
2. Client Tools: `create_agent_turn` is called in loop with client agent
lib yielding the agent chunk
ad6ffc63df/src/llama_stack_client/lib/agents/agent.py (L186-L211)
This makes it inconsistent to work with server & client tools. It also
complicates the logs to telemetry to get information about agents turn /
history for observability.
#### Principle
The same `turn_id` should be used to represent the steps required to
complete a user message including client tools.
## Solution
1. `AgentTurnResponseEventType.turn_awaiting_input` status to indicate
that the current turn is not completed, and awaiting tool input
2. `continue_agent_turn` endpoint to update agent turn with client's
tool response.
# What does this PR do?
- Skeleton API as example
## Test Plan
[Describe the tests you ran to verify your changes with result
summaries. *Provide clear instructions so the plan can be easily
re-executed.*]
- Just API update, no functionality change
```
llama stack run + client-sdk test
```
<img width="842" alt="image"
src="https://github.com/user-attachments/assets/7ac56b5f-f424-4632-9476-7e0f57555bc3"
/>
[//]: # (## Documentation)
See Issue #922
The change is slightly backwards incompatible but no callsite (in our
client codebases or stack-apps) every passes a depth-2
`List[List[InterleavedContentItem]]` (which is now disallowed.)
## Test Plan
```bash
$ cd llama_stack/providers/tests/inference
$ pytest -s -v -k fireworks test_embeddings.py \
--inference-model nomic-ai/nomic-embed-text-v1.5 --env EMBEDDING_DIMENSION=784
$ pytest -s -v -k together test_embeddings.py \
--inference-model togethercomputer/m2-bert-80M-8k-retrieval --env EMBEDDING_DIMENSION=784
$ pytest -s -v -k ollama test_embeddings.py \
--inference-model all-minilm:latest --env EMBEDDING_DIMENSION=784
```
Also ran `tests/client-sdk/inference/test_embeddings.py`
# What does this PR do?
- Fully deprecate eval/tasks
[//]: # (If resolving an issue, uncomment and update the line below)
Closes#1088
NOTE: this will be a breaking change. We have introduced the new API in
0.1.3 .
Notebook has been updated to use the new endpoints.
## Test Plan
```
pytest -v -s --nbval-lax ./docs/notebooks/Llama_Stack_Benchmark_Evals.ipynb
```
<img width="611" alt="image"
src="https://github.com/user-attachments/assets/79f6efe1-81ba-494e-bf36-1fc0c2b9bc6f"
/>
cc @SLR722 for awareness
[//]: # (## Documentation)
# What does this PR do?
You are now able to run a training cycle on CPU. This is useful for
debugging and testing purposes.
[//]: # (If resolving an issue, uncomment and update the line below)
[//]: # (Closes #[issue-number])
## Test Plan
On a Mac machine without CUDA devices:
```
17:00:24.417 [START] /v1/post-training/supervised-fine-tune
DEBUG 2025-02-18 12:00:24,419 torchtune.utils._logging:60: Setting manual seed to local seed 3268931494. Local seed is seed + rank = 3268931494 + 0
INFO 2025-02-18 12:00:24,463 torchtune.utils._logging:64: Identified model_type = Llama3_2. Ignoring output.weight in checkpoint in favor of the tok_embedding.weight tied weights.
INFO 2025-02-18 12:00:46,699 llama_stack.providers.inline.post_training.torchtune.recipes.lora_finetuning_single_device:182: Model is initialized with precision torch.bfloat16.
INFO 2025-02-18 12:00:46,784 llama_stack.providers.inline.post_training.torchtune.recipes.lora_finetuning_single_device:185: Tokenizer is initialized.
INFO 2025-02-18 12:00:46,786 llama_stack.providers.inline.post_training.torchtune.recipes.lora_finetuning_single_device:188: Optimizer is initialized.
INFO 2025-02-18 12:00:46,786 llama_stack.providers.inline.post_training.torchtune.recipes.lora_finetuning_single_device:192: Loss is initialized.
INFO 2025-02-18 12:00:48,997 llama_stack.providers.inline.post_training.torchtune.recipes.lora_finetuning_single_device:209: Dataset and Sampler are initialized.
INFO 2025-02-18 12:00:48,998 llama_stack.providers.inline.post_training.torchtune.recipes.lora_finetuning_single_device:227: Learning rate scheduler is initialized.
Writing logs to /Users/ihrachys/.llama/checkpoints/meta-llama/Llama-3.2-3B-Instruct-sft-0/log_1739898049.txt
1|1|Loss: 1.7414989471435547: 100% 1/1 [03:46<00:00, 226.21s/it]INFO 2025-02-18 12:04:35,227 llama_stack.providers.inline.post_training.torchtune.recipes.lora_finetuning_single_device:528: Starting checkpoint save...
INFO 2025-02-18 12:04:49,974 torchtune.utils._logging:121: Model checkpoint of size 6.43 GB saved to /Users/ihrachys/.llama/checkpoints/meta-llama/Llama-3.2-3B-Instruct-sft-0/consolidated.00.pth
INFO 2025-02-18 12:04:49,981 torchtune.utils._logging:132: Adapter checkpoint of size 0.00 GB saved to /Users/ihrachys/.llama/checkpoints/meta-llama/Llama-3.2-3B-Instruct-sft-0/adapter/adapter.pth
model_file_path /Users/ihrachys/.llama/checkpoints/meta-llama/Llama-3.2-3B-Instruct-sft-0
1|1|Loss: 1.7414989471435547: 100% 1/1 [04:01<00:00, 241.18s/it]
INFO: ::1:64990 - "POST /v1/post-training/supervised-fine-tune HTTP/1.1" 200 OK
17:04:50.364 [END] /v1/post-training/supervised-fine-tune [StatusCode.OK] (265947.01ms)
17:00:24.419 [DEBUG] Setting manual seed to local seed 3268931494. Local seed is seed + rank = 3268931494 + 0
17:00:24.463 [INFO] Identified model_type = Llama3_2. Ignoring output.weight in checkpoint in favor of the tok_embedding.weight tied weights.
17:00:46.700 [INFO] Model is initialized with precision torch.bfloat16.
17:00:46.784 [INFO] Tokenizer is initialized.
17:00:46.786 [INFO] Optimizer is initialized.
17:00:46.786 [INFO] Loss is initialized.
17:00:48.997 [INFO] Dataset and Sampler are initialized.
17:00:48.998 [INFO] Learning rate scheduler is initialized.
17:04:35.227 [INFO] Starting checkpoint save...
17:04:49.974 [INFO] Model checkpoint of size 6.43 GB saved to /Users/ihrachys/.llama/checkpoints/meta-llama/Llama-3.2-3B-Instruct-sft-0/consolidated.00.pth
17:04:49.981 [INFO] Adapter checkpoint of size 0.00 GB saved to /Users/ihrachys/.llama/checkpoints/meta-llama/Llama-3.2-3B-Instruct-sft-0/adapter/adapter.pth
```
[//]: # (## Documentation)
Signed-off-by: Ihar Hrachyshka <ihar.hrachyshka@gmail.com>
as title, to let scoring function llm_as_judge_405b_simpleqa output
aggregated_results.
We can leverage categorical_count to calculate the % of correctness as
eval benchmark metrics
- **refactor: simplify job status extraction a bit**
- **torchtune: save job status on schedule**
- **refactor: get rid of job_list in torchtune job management code**
# What does this PR do?
A failed job is now registered in API, and one can consult its status.
[//]: # (If resolving an issue, uncomment and update the line below)
[//]: # (Closes #[issue-number])
## Test Plan
```
$ llama-stack-client post_training status --job-uuid test-jobe244b5b0-5053-4892-a4d9-d8fc8b116e73
JobStatusResponse(checkpoints=[], job_uuid='test-jobe244b5b0-5053-4892-a4d9-d8fc8b116e73', status='failed', completed_at=None, resources_allocated=None, scheduled_at=datetime.datetime(2025, 2, 18, 9, 4, 34, 3252), started_at=datetime.datetime(2025, 2, 18, 9, 4, 34, 10688))
```
[//]: # (## Documentation)
---------
Signed-off-by: Ihar Hrachyshka <ihar.hrachyshka@gmail.com>
# What does this PR do?
1. This PR adds batch inserts into sqlite-vec as requested in
https://github.com/meta-llama/llama-stack/pull/1040
- Note: the inserts uses a uuid generated from the hash of the document
id and chunk content.
2. This PR also adds unit tests for sqlite-vec. In a follow up PR, I can
add similar tests to Faiss.
## Test Plan
1. Integration tests:
```python
INFERENCE_MODEL=llama3.2:3b-instruct-fp16 LLAMA_STACK_CONFIG=ollama pytest -s -v tests/client-sdk/vector_io/test_vector_io.py
...
PASSED
tests/client-sdk/vector_io/test_vector_io.py::test_vector_db_retrieve[all-MiniLM-L6-v2-sqlite_vec] PASSED
tests/client-sdk/vector_io/test_vector_io.py::test_vector_db_list PASSED
tests/client-sdk/vector_io/test_vector_io.py::test_vector_db_register[all-MiniLM-L6-v2-faiss] PASSED
tests/client-sdk/vector_io/test_vector_io.py::test_vector_db_register[all-MiniLM-L6-v2-sqlite_vec] PASSED
tests/client-sdk/vector_io/test_vector_io.py::test_vector_db_unregister[faiss] PASSED
tests/client-sdk/vector_io/test_vector_io.py::test_vector_db_unregister[sqlite_vec] PASSED
```
3. Unit tests:
```python
pytest llama_stack/providers/tests/vector_io/test_sqlite_vec.py -v -s --tb=short --disable-warnings --asyncio-mode=auto
...
llama_stack/providers/tests/vector_io/test_sqlite_vec.py::test_add_chunks PASSED
llama_stack/providers/tests/vector_io/test_sqlite_vec.py::test_query_chunks PASSED
llama_stack/providers/tests/vector_io/test_sqlite_vec.py::test_chunk_id_conflict PASSED
llama_stack/providers/tests/vector_io/test_sqlite_vec.py::test_register_vector_db PASSED
llama_stack/providers/tests/vector_io/test_sqlite_vec.py::test_unregister_vector_db PASSED
llama_stack/providers/tests/vector_io/test_sqlite_vec.py::test_generate_chunk_id PASSED
```
I also tested using the same example RAG script in
https://github.com/meta-llama/llama-stack/pull/1040 and received the
output.
---------
Signed-off-by: Francisco Javier Arceo <farceo@redhat.com>
There should be a choke-point for llama3.api imports -- this is the
prompt adapter. Creating a ChatFormat() object on demand is inexpensive.
The underlying Tokenizer is a singleton anyway.
# What does this PR do?
Before this change, `distro_codegen.py` would only work if the user
manually installed multiple provider-specific dependencies (see #1122).
Now, users can run `distro_codegen.py` without any provider-specific
dependencies because we avoid importing the entire provider
implementations just to get the config needed to build the provider
template.
Concretely, this mostly means moving the
MODEL_ALIASES (and related variants) definitions to a new models.py
class within the provider implementation for those providers that
require additional dependencies. It also meant moving a couple of
imports from top-level imports to inside `get_adapter_impl` for some
providers, which follows the pattern used by multiple existing
providers.
To ensure we don't regress and accidentally add new imports that cause
distro_codegen.py to fail, the stubbed-in pre-commit hook for
distro_codegen.py was uncommented and slightly tweaked to run via `uv
run python ...` to ensure it runs with only the project's default
dependencies and to run automatically instead of manually.
Lastly, this updates distro_codegen.py itself to keep track of paths it
might have changed and to only `git diff` those specific paths when
checking for changed files instead of doing a diff on the entire working
tree. The latter was overly broad and would require a user have no other
unstaged changes in their working tree, even if those unstaged changes
were unrelated to generated code. Now it only flags uncommitted changes
for paths distro_codegen.py actually writes to.
Our generated code was also out-of-date, presumably because of these
issues, so this commit also has some updates to the generated code
purely because it was out of sync, and the pre-commit hook now enforces
things to be updated.
(Closes#1122)
## Test Plan
I manually tested distro_codegen.py and the pre-commit hook to verify
those work as expected, flagging any uncommited changes and catching any
imports that attempt to pull in provider-specific dependencies.
However, I do not have valid api keys to the impacted provider
implementations, and am unable to easily run the inference tests against
each changed provider. There are no functional changes to the provider
implementations here, but I'd appreciate a second set of eyes on the
changed import statements and moving of MODEL_ALIASES type code to a
separate models.py to ensure I didn't make any obvious errors.
---------
Signed-off-by: Ben Browning <bbrownin@redhat.com>
Co-authored-by: Ashwin Bharambe <ashwin.bharambe@gmail.com>
llama-models should have extremely minimal cruft. Its sole purpose
should be didactic -- show the simplest implementation of the llama
models and document the prompt formats, etc.
This PR is the complement to
https://github.com/meta-llama/llama-models/pull/279
## Test Plan
Ensure all `llama` CLI `model` sub-commands work:
```bash
llama model list
llama model download --model-id ...
llama model prompt-format -m ...
```
Ran tests:
```bash
cd tests/client-sdk
LLAMA_STACK_CONFIG=fireworks pytest -s -v inference/
LLAMA_STACK_CONFIG=fireworks pytest -s -v vector_io/
LLAMA_STACK_CONFIG=fireworks pytest -s -v agents/
```
Create a fresh venv `uv venv && source .venv/bin/activate` and run
`llama stack build --template fireworks --image-type venv` followed by
`llama stack run together --image-type venv` <-- the server runs
Also checked that the OpenAPI generator can run and there is no change
in the generated files as a result.
```bash
cd docs/openapi_generator
sh run_openapi_generator.sh
```
# What does this PR do?
- Remove hardcoded configurations from pre-commit.
- Allow configuration to be set via pyproject.toml.
- Merge .ruff.toml settings into pyproject.toml.
- Ensure the linter and formatter use the defined configuration instead
of being overridden by pre-commit.
Signed-off-by: Sébastien Han <seb@redhat.com>
[//]: # (If resolving an issue, uncomment and update the line below)
[//]: # (Closes #[issue-number])
## Test Plan
[Describe the tests you ran to verify your changes with result
summaries. *Provide clear instructions so the plan can be easily
re-executed.*]
[//]: # (## Documentation)
Signed-off-by: Sébastien Han <seb@redhat.com>
# What does this PR do?
- Update `/eval-tasks` to `/benchmarks`
- ⚠️ Remove differentiation between `app` v.s. `benchmark` eval task
config. Now we only have `BenchmarkConfig`. The overloaded `benchmark`
is confusing and do not add any value. Backward compatibility is being
kept as the "type" is not being used anywhere.
[//]: # (If resolving an issue, uncomment and update the line below)
[//]: # (Closes #[issue-number])
## Test Plan
- This change is backward compatible
- Run notebook test with
```
pytest -v -s --nbval-lax ./docs/getting_started.ipynb
pytest -v -s --nbval-lax ./docs/notebooks/Llama_Stack_Benchmark_Evals.ipynb
```
<img width="846" alt="image"
src="https://github.com/user-attachments/assets/d2fc06a7-593a-444f-bc1f-10ab9b0c843d"
/>
[//]: # (## Documentation)
[//]: # (- [ ] Added a Changelog entry if the change is significant)
---------
Signed-off-by: Ihar Hrachyshka <ihar.hrachyshka@gmail.com>
Signed-off-by: Ben Browning <bbrownin@redhat.com>
Signed-off-by: Sébastien Han <seb@redhat.com>
Signed-off-by: reidliu <reid201711@gmail.com>
Co-authored-by: Ihar Hrachyshka <ihar.hrachyshka@gmail.com>
Co-authored-by: Ben Browning <ben324@gmail.com>
Co-authored-by: Sébastien Han <seb@redhat.com>
Co-authored-by: Reid <61492567+reidliu41@users.noreply.github.com>
Co-authored-by: reidliu <reid201711@gmail.com>
Co-authored-by: Yuan Tang <terrytangyuan@gmail.com>
# What does this PR do?
This changes all VectorIO providers classes to follow the pattern
`<ProviderName>VectorIOConfig` and `<ProviderName>VectorIOAdapter`. All
API endpoints for VectorIOs are currently consistent with `/vector-io`.
Note that API endpoint for VectorDB stay unchanged as `/vector-dbs`.
## Test Plan
I don't have a way to test all providers. This is a simple renaming so
things should work as expected.
---------
Signed-off-by: Yuan Tang <terrytangyuan@gmail.com>
# What does this PR do?
- Configured ruff linter to automatically fix import sorting issues.
- Set --exit-non-zero-on-fix to ensure non-zero exit code when fixes are
applied.
- Enabled the 'I' selection to focus on import-related linting rules.
- Ran the linter, and formatted all codebase imports accordingly.
- Removed the black dep from the "dev" group since we use ruff
Signed-off-by: Sébastien Han <seb@redhat.com>
[//]: # (If resolving an issue, uncomment and update the line below)
[//]: # (Closes #[issue-number])
## Test Plan
[Describe the tests you ran to verify your changes with result
summaries. *Provide clear instructions so the plan can be easily
re-executed.*]
[//]: # (## Documentation)
[//]: # (- [ ] Added a Changelog entry if the change is significant)
Signed-off-by: Sébastien Han <seb@redhat.com>
# What does this PR do?
This commit enhances the signal handling mechanism in the server by
improving the `handle_signal` (previously handle_sigint) function. It
now properly retrieves the signal name, ensuring clearer logging when a
termination signal is received. Additionally, it cancels all running
tasks and waits for their completion before stopping the event loop,
allowing for a more graceful shutdown. Support for handling
SIGTERM has also been added alongside SIGINT.
Before the changes, handle_sigint used asyncio.run(run_shutdown()).
However, asyncio.run() is meant to start a new event loop, and calling
it inside an existing one (like when running Uvicorn) raises an error.
The fix replaces asyncio.run(run_shutdown()) with an async function
scheduled on the existing loop using loop.create_task(shutdown()). This
ensures that the shutdown coroutine runs within the current event loop
instead of trying to create a new one.
Furthermore, this commit updates the project dependencies. `fastapi` and
`uvicorn` have been added to the development dependencies in
`pyproject.toml` and `uv.lock`, ensuring that the necessary packages are
available for development and execution.
Closes: https://github.com/meta-llama/llama-stack/issues/1043
Signed-off-by: Sébastien Han <seb@redhat.com>
[//]: # (If resolving an issue, uncomment and update the line below)
[//]: # (Closes #[issue-number])
## Test Plan
Run a server and send SIGINT:
```
INFERENCE_MODEL="meta-llama/Llama-3.2-3B-Instruct" python -m llama_stack.distribution.server.server --yaml-config ./llama_stack/templates/ollama/run.yaml
Using config file: llama_stack/templates/ollama/run.yaml
Run configuration:
apis:
- agents
- datasetio
- eval
- inference
- safety
- scoring
- telemetry
- tool_runtime
- vector_io
container_image: null
datasets: []
eval_tasks: []
image_name: ollama
metadata_store:
db_path: /Users/leseb/.llama/distributions/ollama/registry.db
namespace: null
type: sqlite
models:
- metadata: {}
model_id: meta-llama/Llama-3.2-3B-Instruct
model_type: !!python/object/apply:llama_stack.apis.models.models.ModelType
- llm
provider_id: ollama
provider_model_id: null
- metadata:
embedding_dimension: 384
model_id: all-MiniLM-L6-v2
model_type: !!python/object/apply:llama_stack.apis.models.models.ModelType
- embedding
provider_id: sentence-transformers
provider_model_id: null
providers:
agents:
- config:
persistence_store:
db_path: /Users/leseb/.llama/distributions/ollama/agents_store.db
namespace: null
type: sqlite
provider_id: meta-reference
provider_type: inline::meta-reference
datasetio:
- config: {}
provider_id: huggingface
provider_type: remote::huggingface
- config: {}
provider_id: localfs
provider_type: inline::localfs
eval:
- config: {}
provider_id: meta-reference
provider_type: inline::meta-reference
inference:
- config:
url: http://localhost:11434
provider_id: ollama
provider_type: remote::ollama
- config: {}
provider_id: sentence-transformers
provider_type: inline::sentence-transformers
safety:
- config: {}
provider_id: llama-guard
provider_type: inline::llama-guard
scoring:
- config: {}
provider_id: basic
provider_type: inline::basic
- config: {}
provider_id: llm-as-judge
provider_type: inline::llm-as-judge
- config:
openai_api_key: '********'
provider_id: braintrust
provider_type: inline::braintrust
telemetry:
- config:
service_name: llama-stack
sinks: console,sqlite
sqlite_db_path: /Users/leseb/.llama/distributions/ollama/trace_store.db
provider_id: meta-reference
provider_type: inline::meta-reference
tool_runtime:
- config:
api_key: '********'
max_results: 3
provider_id: brave-search
provider_type: remote::brave-search
- config:
api_key: '********'
max_results: 3
provider_id: tavily-search
provider_type: remote::tavily-search
- config: {}
provider_id: code-interpreter
provider_type: inline::code-interpreter
- config: {}
provider_id: rag-runtime
provider_type: inline::rag-runtime
vector_io:
- config:
kvstore:
db_path: /Users/leseb/.llama/distributions/ollama/faiss_store.db
namespace: null
type: sqlite
provider_id: faiss
provider_type: inline::faiss
scoring_fns: []
server:
port: 8321
tls_certfile: null
tls_keyfile: null
shields: []
tool_groups:
- args: null
mcp_endpoint: null
provider_id: tavily-search
toolgroup_id: builtin::websearch
- args: null
mcp_endpoint: null
provider_id: rag-runtime
toolgroup_id: builtin::rag
- args: null
mcp_endpoint: null
provider_id: code-interpreter
toolgroup_id: builtin::code_interpreter
vector_dbs: []
version: '2'
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:213: Resolved 31 providers
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: inner-inference => ollama
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: inner-inference => sentence-transformers
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: models => __routing_table__
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: inference => __autorouted__
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: inner-vector_io => faiss
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: inner-safety => llama-guard
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: shields => __routing_table__
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: safety => __autorouted__
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: vector_dbs => __routing_table__
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: vector_io => __autorouted__
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: inner-tool_runtime => brave-search
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: inner-tool_runtime => tavily-search
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: inner-tool_runtime => code-interpreter
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: inner-tool_runtime => rag-runtime
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: tool_groups => __routing_table__
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: tool_runtime => __autorouted__
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: agents => meta-reference
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: inner-datasetio => huggingface
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: inner-datasetio => localfs
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: datasets => __routing_table__
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: datasetio => __autorouted__
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: telemetry => meta-reference
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: inner-scoring => basic
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: inner-scoring => llm-as-judge
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: inner-scoring => braintrust
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: scoring_functions => __routing_table__
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: scoring => __autorouted__
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: inner-eval => meta-reference
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: eval_tasks => __routing_table__
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: eval => __autorouted__
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:215: inspect => __builtin__
INFO 2025-02-12 10:21:03,540 llama_stack.distribution.resolver:216:
INFO 2025-02-12 10:21:03,723 llama_stack.providers.remote.inference.ollama.ollama:148: checking connectivity to Ollama at `http://localhost:11434`...
INFO 2025-02-12 10:21:03,734 httpx:1740: HTTP Request: GET http://localhost:11434/api/ps "HTTP/1.1 200 OK"
INFO 2025-02-12 10:21:03,843 faiss.loader:148: Loading faiss.
INFO 2025-02-12 10:21:03,865 faiss.loader:150: Successfully loaded faiss.
INFO 2025-02-12 10:21:03,868 faiss:173: Failed to load GPU Faiss: name 'GpuIndexIVFFlat' is not defined. Will not load constructor refs for GPU indexes.
Warning: `bwrap` is not available. Code interpreter tool will not work correctly.
INFO 2025-02-12 10:21:04,315 datasets:54: PyTorch version 2.6.0 available.
INFO 2025-02-12 10:21:04,556 httpx:1740: HTTP Request: GET http://localhost:11434/api/ps "HTTP/1.1 200 OK"
INFO 2025-02-12 10:21:04,557 llama_stack.providers.utils.inference.embedding_mixin:42: Loading sentence transformer for all-MiniLM-L6-v2...
INFO 2025-02-12 10:21:07,202 sentence_transformers.SentenceTransformer:210: Use pytorch device_name: mps
INFO 2025-02-12 10:21:07,202 sentence_transformers.SentenceTransformer:218: Load pretrained SentenceTransformer: all-MiniLM-L6-v2
INFO 2025-02-12 10:21:09,500 llama_stack.distribution.stack:102: Models: all-MiniLM-L6-v2 served by sentence-transformers
INFO 2025-02-12 10:21:09,500 llama_stack.distribution.stack:102: Models: meta-llama/Llama-3.2-3B-Instruct served by ollama
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Scoring_fns: basic::equality served by basic
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Scoring_fns: basic::regex_parser_multiple_choice_answer served by basic
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Scoring_fns: basic::subset_of served by basic
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Scoring_fns: braintrust::answer-correctness served by braintrust
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Scoring_fns: braintrust::answer-relevancy served by braintrust
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Scoring_fns: braintrust::answer-similarity served by braintrust
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Scoring_fns: braintrust::context-entity-recall served by braintrust
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Scoring_fns: braintrust::context-precision served by braintrust
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Scoring_fns: braintrust::context-recall served by braintrust
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Scoring_fns: braintrust::context-relevancy served by braintrust
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Scoring_fns: braintrust::factuality served by braintrust
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Scoring_fns: braintrust::faithfulness served by braintrust
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Scoring_fns: llm-as-judge::405b-simpleqa served by llm-as-judge
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Scoring_fns: llm-as-judge::base served by llm-as-judge
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Tool_groups: builtin::code_interpreter served by code-interpreter
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Tool_groups: builtin::rag served by rag-runtime
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:102: Tool_groups: builtin::websearch served by tavily-search
INFO 2025-02-12 10:21:09,501 llama_stack.distribution.stack:106:
Serving API eval
POST /v1/eval/tasks/{task_id}/evaluations
DELETE /v1/eval/tasks/{task_id}/jobs/{job_id}
GET /v1/eval/tasks/{task_id}/jobs/{job_id}/result
GET /v1/eval/tasks/{task_id}/jobs/{job_id}
POST /v1/eval/tasks/{task_id}/jobs
Serving API agents
POST /v1/agents
POST /v1/agents/{agent_id}/session
POST /v1/agents/{agent_id}/session/{session_id}/turn
DELETE /v1/agents/{agent_id}
DELETE /v1/agents/{agent_id}/session/{session_id}
GET /v1/agents/{agent_id}/session/{session_id}
GET /v1/agents/{agent_id}/session/{session_id}/turn/{turn_id}/step/{step_id}
GET /v1/agents/{agent_id}/session/{session_id}/turn/{turn_id}
Serving API scoring_functions
GET /v1/scoring-functions/{scoring_fn_id}
GET /v1/scoring-functions
POST /v1/scoring-functions
Serving API safety
POST /v1/safety/run-shield
Serving API inspect
GET /v1/health
GET /v1/inspect/providers
GET /v1/inspect/routes
GET /v1/version
Serving API tool_runtime
POST /v1/tool-runtime/invoke
GET /v1/tool-runtime/list-tools
POST /v1/tool-runtime/rag-tool/insert
POST /v1/tool-runtime/rag-tool/query
Serving API datasetio
POST /v1/datasetio/rows
GET /v1/datasetio/rows
Serving API shields
GET /v1/shields/{identifier}
GET /v1/shields
POST /v1/shields
Serving API eval_tasks
GET /v1/eval-tasks/{eval_task_id}
GET /v1/eval-tasks
POST /v1/eval-tasks
Serving API models
GET /v1/models/{model_id}
GET /v1/models
POST /v1/models
DELETE /v1/models/{model_id}
Serving API datasets
GET /v1/datasets/{dataset_id}
GET /v1/datasets
POST /v1/datasets
DELETE /v1/datasets/{dataset_id}
Serving API vector_io
POST /v1/vector-io/insert
POST /v1/vector-io/query
Serving API inference
POST /v1/inference/chat-completion
POST /v1/inference/completion
POST /v1/inference/embeddings
Serving API tool_groups
GET /v1/tools/{tool_name}
GET /v1/toolgroups/{toolgroup_id}
GET /v1/toolgroups
GET /v1/tools
POST /v1/toolgroups
DELETE /v1/toolgroups/{toolgroup_id}
Serving API vector_dbs
GET /v1/vector-dbs/{vector_db_id}
GET /v1/vector-dbs
POST /v1/vector-dbs
DELETE /v1/vector-dbs/{vector_db_id}
Serving API scoring
POST /v1/scoring/score
POST /v1/scoring/score-batch
Serving API telemetry
GET /v1/telemetry/traces/{trace_id}/spans/{span_id}
GET /v1/telemetry/spans/{span_id}/tree
GET /v1/telemetry/traces/{trace_id}
POST /v1/telemetry/events
GET /v1/telemetry/spans
GET /v1/telemetry/traces
POST /v1/telemetry/spans/export
Listening on ['::', '0.0.0.0']:5001
INFO: Started server process [65372]
INFO: Waiting for application startup.
INFO: ASGI 'lifespan' protocol appears unsupported.
INFO: Application startup complete.
INFO: Uvicorn running on http://['::', '0.0.0.0']:5001 (Press CTRL+C to quit)
^CINFO: Shutting down
INFO: Finished server process [65372]
Received signal SIGINT (2). Exiting gracefully...
INFO 2025-02-12 10:21:11,215 __main__:151: Shutting down ModelsRoutingTable
INFO 2025-02-12 10:21:11,216 __main__:151: Shutting down InferenceRouter
INFO 2025-02-12 10:21:11,216 __main__:151: Shutting down ShieldsRoutingTable
INFO 2025-02-12 10:21:11,216 __main__:151: Shutting down SafetyRouter
INFO 2025-02-12 10:21:11,216 __main__:151: Shutting down VectorDBsRoutingTable
INFO 2025-02-12 10:21:11,216 __main__:151: Shutting down VectorIORouter
INFO 2025-02-12 10:21:11,216 __main__:151: Shutting down ToolGroupsRoutingTable
INFO 2025-02-12 10:21:11,216 __main__:151: Shutting down ToolRuntimeRouter
INFO 2025-02-12 10:21:11,216 __main__:151: Shutting down MetaReferenceAgentsImpl
INFO 2025-02-12 10:21:11,216 __main__:151: Shutting down DatasetsRoutingTable
INFO 2025-02-12 10:21:11,216 __main__:151: Shutting down DatasetIORouter
INFO 2025-02-12 10:21:11,216 __main__:151: Shutting down TelemetryAdapter
INFO 2025-02-12 10:21:11,216 __main__:151: Shutting down ScoringFunctionsRoutingTable
INFO 2025-02-12 10:21:11,216 __main__:151: Shutting down ScoringRouter
INFO 2025-02-12 10:21:11,216 __main__:151: Shutting down EvalTasksRoutingTable
INFO 2025-02-12 10:21:11,216 __main__:151: Shutting down EvalRouter
INFO 2025-02-12 10:21:11,216 __main__:151: Shutting down DistributionInspectImpl
```
[//]: # (## Documentation)
[//]: # (- [ ] Added a Changelog entry if the change is significant)
Signed-off-by: Sébastien Han <seb@redhat.com>
# What does this PR do?
This PR adds `sqlite_vec` as an additional inline vectordb.
Tested with `ollama` by adding the `vector_io` object in
`./llama_stack/templates/ollama/run.yaml` :
```yaml
vector_io:
- provider_id: sqlite_vec
provider_type: inline::sqlite_vec
config:
kvstore:
type: sqlite
namespace: null
db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/ollama}/sqlite_vec.db
db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/ollama}/sqlite_vec.db
```
I also updated the `./tests/client-sdk/vector_io/test_vector_io.py` test
file with:
```python
INLINE_VECTOR_DB_PROVIDERS = ["faiss", "sqlite_vec"]
```
And parameterized the relevant tests.
[//]: # (If resolving an issue, uncomment and update the line below)
# Closes
https://github.com/meta-llama/llama-stack/issues/1005
## Test Plan
I ran the tests with:
```bash
INFERENCE_MODEL=llama3.2:3b-instruct-fp16 LLAMA_STACK_CONFIG=ollama pytest -s -v tests/client-sdk/vector_io/test_vector_io.py
```
Which outputs:
```python
...
PASSED
tests/client-sdk/vector_io/test_vector_io.py::test_vector_db_retrieve[all-MiniLM-L6-v2-sqlite_vec] PASSED
tests/client-sdk/vector_io/test_vector_io.py::test_vector_db_list PASSED
tests/client-sdk/vector_io/test_vector_io.py::test_vector_db_register[all-MiniLM-L6-v2-faiss] PASSED
tests/client-sdk/vector_io/test_vector_io.py::test_vector_db_register[all-MiniLM-L6-v2-sqlite_vec] PASSED
tests/client-sdk/vector_io/test_vector_io.py::test_vector_db_unregister[faiss] PASSED
tests/client-sdk/vector_io/test_vector_io.py::test_vector_db_unregister[sqlite_vec] PASSED
```
In addition, I ran the `rag_with_vector_db.py`
[example](https://github.com/meta-llama/llama-stack-apps/blob/main/examples/agents/rag_with_vector_db.py)
using the script below with `uv run rag_example.py`.
<details>
<summary>CLICK TO SHOW SCRIPT 👋 </summary>
```python
#!/usr/bin/env python3
import os
import uuid
from termcolor import cprint
# Set environment variables
os.environ['INFERENCE_MODEL'] = 'llama3.2:3b-instruct-fp16'
os.environ['LLAMA_STACK_CONFIG'] = 'ollama'
# Import libraries after setting environment variables
from llama_stack.distribution.library_client import LlamaStackAsLibraryClient
from llama_stack_client.lib.agents.agent import Agent
from llama_stack_client.lib.agents.event_logger import EventLogger
from llama_stack_client.types.agent_create_params import AgentConfig
from llama_stack_client.types import Document
def main():
# Initialize the client
client = LlamaStackAsLibraryClient("ollama")
vector_db_id = f"test-vector-db-{uuid.uuid4().hex}"
_ = client.initialize()
model_id = 'llama3.2:3b-instruct-fp16'
# Define the list of document URLs and create Document objects
urls = [
"chat.rst",
"llama3.rst",
"memory_optimizations.rst",
"lora_finetune.rst",
]
documents = [
Document(
document_id=f"num-{i}",
content=f"https://raw.githubusercontent.com/pytorch/torchtune/main/docs/source/tutorials/{url}",
mime_type="text/plain",
metadata={},
)
for i, url in enumerate(urls)
]
# (Optional) Use the documents as needed with your client here
client.vector_dbs.register(
provider_id='sqlite_vec',
vector_db_id=vector_db_id,
embedding_model="all-MiniLM-L6-v2",
embedding_dimension=384,
)
client.tool_runtime.rag_tool.insert(
documents=documents,
vector_db_id=vector_db_id,
chunk_size_in_tokens=512,
)
# Create agent configuration
agent_config = AgentConfig(
model=model_id,
instructions="You are a helpful assistant",
enable_session_persistence=False,
toolgroups=[
{
"name": "builtin::rag",
"args": {
"vector_db_ids": [vector_db_id],
}
}
],
)
# Instantiate the Agent
agent = Agent(client, agent_config)
# List of user prompts
user_prompts = [
"What are the top 5 topics that were explained in the documentation? Only list succinct bullet points.",
"Was anything related to 'Llama3' discussed, if so what?",
"Tell me how to use LoRA",
"What about Quantization?",
]
# Create a session for the agent
session_id = agent.create_session("test-session")
# Process each prompt and display the output
for prompt in user_prompts:
cprint(f"User> {prompt}", "green")
response = agent.create_turn(
messages=[
{
"role": "user",
"content": prompt,
}
],
session_id=session_id,
)
# Log and print events from the response
for log in EventLogger().log(response):
log.print()
if __name__ == "__main__":
main()
```
</details>
Which outputs a large summary of RAG generation.
# Documentation
Will handle documentation updates in follow-up PR.
# (- [ ] Added a Changelog entry if the change is significant)
---------
Signed-off-by: Francisco Javier Arceo <farceo@redhat.com>
# What does this PR do?
The current default system prompt for llama3.2 tends to overindex on
tool calling and doesn't work well when the prompt does not require tool
calling.
This PR adds an option to override the default system prompt, and
organizes tool-related configs into a new config object.
- [ ] Addresses issue (#issue)
## Test Plan
LLAMA_STACK_CONFIG=together pytest
\-\-inference\-model=meta\-llama/Llama\-3\.3\-70B\-Instruct -s -v
tests/client-sdk/agents/test_agents.py::test_override_system_message_behavior
## Sources
Please link relevant resources if necessary.
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the
other checks if that's the case).
- [ ] Ran pre-commit to handle lint / formatting issues.
- [ ] Read the [contributor
guideline](https://github.com/meta-llama/llama-stack/blob/main/CONTRIBUTING.md),
Pull Request section?
- [ ] Updated relevant documentation.
- [ ] Wrote necessary unit or integration tests.
# What does this PR do?
The current default system prompt for llama3.2 tends to overindex on
tool calling and doesn't work well when the prompt does not require tool
calling.
This PR adds an option to override the default system prompt, and
organizes tool-related configs into a new config object.
- [ ] Addresses issue (#issue)
## Test Plan
python -m unittest
llama_stack.providers.tests.inference.test_prompt_adapter
## Sources
Please link relevant resources if necessary.
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the
other checks if that's the case).
- [ ] Ran pre-commit to handle lint / formatting issues.
- [ ] Read the [contributor
guideline](https://github.com/meta-llama/llama-stack/blob/main/CONTRIBUTING.md),
Pull Request section?
- [ ] Updated relevant documentation.
- [ ] Wrote necessary unit or integration tests.
---
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed
with
[ReviewStack](https://reviewstack.dev/meta-llama/llama-stack/pull/937).
* #938
* __->__ #937
# What does this PR do?
To work with the updated iOSCalendarAssistantWithLocalInf
[here](https://github.com/meta-llama/llama-stack-apps/compare/ios_local).
In short, provide a summary of what this PR does and why. Usually, the
relevant context should be present in a linked issue.
- [ ] Addresses issue (#issue)
## Test Plan
Please describe:
- tests you ran to verify your changes with result summaries.
- provide instructions so it can be reproduced.
## Sources
Please link relevant resources if necessary.
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the
other checks if that's the case).
- [ ] Ran pre-commit to handle lint / formatting issues.
- [ ] Read the [contributor
guideline](https://github.com/meta-llama/llama-stack/blob/main/CONTRIBUTING.md),
Pull Request section?
- [ ] Updated relevant documentation.
- [ ] Wrote necessary unit or integration tests.
Lint check in main branch is failing. This fixes the lint check after we
moved to ruff in https://github.com/meta-llama/llama-stack/pull/921. We
need to move to a `ruff.toml` file as well as fixing and ignoring some
additional checks.
Signed-off-by: Yuan Tang <terrytangyuan@gmail.com>
This commit adds support for XPU and CPU devices into meta-reference
stack for text models. On creation stack automatically identifies which
device to use checking available accelerate capabilities in the
following order: CUDA, then XPU, finally CPU. This behaviour can be
overwritten with the `DEVICE` environment variable. In this case
explicitly specified device will be used.
Tested with:
```
torchrun pytest llama_stack/providers/tests/inference/test_text_inference.py -k meta_reference
```
Results:
* Tested on: system with single CUDA device, system with single XPU
device and on pure CPU system
* Results: all test pass except `test_completion_logprobs`
* `test_completion_logprobs` fails in the same way as on a baseline,
i.e. unrelated with this change: `AssertionError: Unexpected top_k=3`
Requires: https://github.com/meta-llama/llama-models/pull/233
Signed-off-by: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com>
# What does this PR do?
Fixes a bug where agents were not working when both rag and
code-interpreter were added as tools.
## Test Plan
Added a new client_sdk test which tests for this scenario
```
LLAMA_STACK_CONFIG=together pytest -s -v tests/client-sdk -k 'test_rag_and_code_agent'
```
---------
Co-authored-by: Hardik Shah <hjshah@fb.com>
# What does this PR do?
Add response format for agents structured output.
- [ ] Using structured output for agents (interior_design app as an
example) (#issue)
https://github.com/meta-llama/llama-stack-apps/issues/122
## Test Plan
E2E test plan with llama-stack-apps interior_design
Please describe:
Test ran:
- provide instructions so it can be reproduced.
Start your distro:
llama stack run llama_stack/templates/fireworks/run.yaml --env
FIREWORKS_API_KEY=<API_KEY>
Run api test:
```PYTHONPATH=. python examples/interior_design_assistant/api.py localhost 5000 examples/interior_design_assistant/resources/documents/ examples/interior_design_assistant/resources/images/fireplaces```
## Sources
Results:
https://github.com/meta-llama/llama-stack-client-python/pull/72
## Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [ ] Ran pre-commit to handle lint / formatting issues.
- [x] Read the [contributor guideline](https://github.com/meta-llama/llama-stack/blob/main/CONTRIBUTING.md),
Pull Request section?
- [ ] Updated relevant documentation.
- [ ] Wrote necessary unit or integration tests.
# What does this PR do?
When you re-initialize the library client in a notebook, we were seeing
this error:
```
Getting traces for session_id=5c8d1969-0957-49d2-b852-32cbb8ef8caf
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
[<ipython-input-11-d74bb6cdd3ab>](https://localhost:8080/#) in <cell line: 0>()
7 agent_logs = []
8
----> 9 for span in client.telemetry.query_spans(
10 attribute_filters=[
11 {"key": "session_id", "op": "eq", "value": session_id},
10 frames
[/usr/local/lib/python3.11/dist-packages/llama_stack/providers/inline/telemetry/meta_reference/telemetry.py](https://localhost:8080/#) in query_traces(self, attribute_filters, limit, offset, order_by)
246 ) -> QueryTracesResponse:
247 return QueryTracesResponse(
--> 248 data=await self.trace_store.query_traces(
249 attribute_filters=attribute_filters,
250 limit=limit,
AttributeError: 'TelemetryAdapter' object has no attribute 'trace_store'
```
This is happening because the we were skipping some required steps for
the object state as part of the global _TRACE_PROVIDER check. This PR
moves the initialization of the object state out of the TRACE_PROVIDER
init.
