Update from HF.

.github/workflows/quality.yml

@@ -16,20 +16,15 @@ jobs:
           python-version: "3.12"
 
       # Setup venv
-      # TODO: revisit when https://github.com/astral-sh/uv/issues/1526 is addressed.
       - name: Setup venv + uv
         run: |
           pip install --upgrade uv
           uv venv
 
       - name: Install dependencies
-        run: uv pip install "smolagents[test] @ ."
-      - run: uv run ruff check tests src # linter
-      - run: uv run ruff format --check tests src # formatter
+        run: uv pip install "smolagents[quality] @ ."
 
-      # Run type checking at least on smolagents root file to check all modules
-      # that can be lazy-loaded actually exist.
-      # - run: uv run mypy src/smolagents/__init__.py --follow-imports=silent --show-traceback
-
-      # Run mypy on full package
-      # - run: uv run mypy src
+      # Equivalent of "make quality" but step by step
+      - run: uv run ruff check examples src tests utils # linter
+      - run: uv run ruff format --check examples src tests utils # formatter
+      - run: uv run python utils/check_tests_in_ci.py

.github/workflows/tests.yml

@@ -20,9 +20,7 @@ jobs:
         with:
           python-version: ${{ matrix.python-version }}
 
-
       # Setup venv
-      # TODO: revisit when https://github.com/astral-sh/uv/issues/1526 is addressed.
       - name: Setup venv + uv
         run: |
           pip install --upgrade uv
@@ -33,33 +31,59 @@ jobs:
         run: |
           uv pip install "smolagents[test] @ ."
 
+      # Run all tests separately for individual feedback
+      # Use 'if success() || failure()' so that all tests are run even if one failed
+      # See https://stackoverflow.com/a/62112985
       - name: Agent tests
         run: |
-          uv run pytest -sv ./tests/test_agents.py
+          uv run pytest ./tests/test_agents.py
+        if: ${{ success() || failure() }}
+
       - name: Default tools tests
         run: |
-          uv run pytest -sv ./tests/test_default_tools.py
+          uv run pytest ./tests/test_default_tools.py
+        if: ${{ success() || failure() }}
+
+      # - name: Docs tests # Disabled for now (slow test + requires API keys)
+      #   run: |
+      #     uv run pytest ./tests/test_all_docs.py
+
       - name: Final answer tests
         run: |
-          uv run pytest -sv ./tests/test_final_answer.py
+          uv run pytest ./tests/test_final_answer.py
+        if: ${{ success() || failure() }}
+
       - name: Models tests
         run: |
-          uv run pytest -sv ./tests/test_models.py
+          uv run pytest ./tests/test_models.py
+        if: ${{ success() || failure() }}
+
       - name: Monitoring tests
         run: |
-          uv run pytest -sv ./tests/test_monitoring.py
+          uv run pytest ./tests/test_monitoring.py
+        if: ${{ success() || failure() }}
+
       - name: Python interpreter tests
         run: |
-          uv run pytest -sv ./tests/test_python_interpreter.py
+          uv run pytest ./tests/test_python_interpreter.py
+        if: ${{ success() || failure() }}
+
       - name: Search tests
         run: |
-          uv run pytest -sv ./tests/test_search.py
+          uv run pytest ./tests/test_search.py
+        if: ${{ success() || failure() }}
+
       - name: Tools tests
         run: |
-          uv run pytest -sv ./tests/test_tools.py
+          uv run pytest ./tests/test_tools.py
+        if: ${{ success() || failure() }}
+
       - name: Types tests
         run: |
-          uv run pytest -sv ./tests/test_types.py
+          uv run pytest ./tests/test_types.py
+        if: ${{ success() || failure() }}
+
       - name: Utils tests
         run: |
-          uv run pytest -sv ./tests/test_utils.py
+          uv run pytest ./tests/test_utils.py
+        if: ${{ success() || failure() }}

CONTRIBUTING.md

@@ -91,8 +91,8 @@ happy to make the changes or help you make a contribution if you're interested!
 
 ## I want to become a maintainer of the project. How do I get there?
 
-Smolagents is a project led and managed by Hugging Face as an initial fork of Transformers Agents. We are  more than 
-happy to have motivated individuals from other organizations join us as maintainers with the goal of helping Smolagents
+smolagents is a project led and managed by Hugging Face. We are more than
+happy to have motivated individuals from other organizations join us as maintainers with the goal of helping smolagents
 make a dent in the world of Agents.
 
 If you are such an individual (or organization), please reach out to us and let's collaborate.

Makefile

@@ -1,53 +1,18 @@
 .PHONY: quality style test docs utils
 
-check_dirs := .
+check_dirs := examples src tests utils
 
-# Check that source code meets quality standards
-
-extra_quality_checks:
-	python utils/check_copies.py
-	python utils/check_dummies.py
-	python utils/check_repo.py
-	doc-builder style smolagents docs/source --max_len 119
-
-# this target runs checks on all files
+# Check code quality of the source code
 quality:
 	ruff check $(check_dirs)
 	ruff format --check $(check_dirs)
-	doc-builder style smolagents docs/source --max_len 119 --check_only
+	python utils/check_tests_in_ci.py
 
-# Format source code automatically and check is there are any problems left that need manual fixing
+# Format source code automatically
 style:
 	ruff check $(check_dirs) --fix
 	ruff format $(check_dirs)
-	doc-builder style smolagents docs/source --max_len 119
 
-# Run tests for the library
-test_big_modeling:
-	python -m pytest -s -v ./tests/test_big_modeling.py ./tests/test_modeling_utils.py $(if $(IS_GITHUB_CI),--report-log "$(PYTORCH_VERSION)_big_modeling.log",)
-
-test_core:
-	python -m pytest -s -v ./tests/ --ignore=./tests/test_examples.py $(if $(IS_GITHUB_CI),--report-log "$(PYTORCH_VERSION)_core.log",)
-
-test_cli:
-	python -m pytest -s -v ./tests/test_cli.py $(if $(IS_GITHUB_CI),--report-log "$(PYTORCH_VERSION)_cli.log",)
-
-
-# Since the new version of pytest will *change* how things are collected, we need `deepspeed` to 
-# run after test_core and test_cli
+# Run smolagents tests
 test:
-	$(MAKE) test_core
-	$(MAKE) test_cli
-	$(MAKE) test_big_modeling
-	$(MAKE) test_deepspeed
-	$(MAKE) test_fsdp
-
-test_examples:
-	python -m pytest -s -v ./tests/test_examples.py $(if $(IS_GITHUB_CI),--report-log "$(PYTORCH_VERSION)_examples.log",)
-
-# Same as test but used to install only the base dependencies
-test_prod:
-	$(MAKE) test_core
-
-test_rest:
-	python -m pytest -s -v ./tests/test_examples.py::FeatureExamplesTests $(if $(IS_GITHUB_CI),--report-log "$(PYTORCH_VERSION)_rest.log",)
+	pytest ./tests/

README.md

@@ -24,7 +24,10 @@ limitations under the License.
 </p>
 
 <h3 align="center">
-  <p>🤗 smolagents - a smol library to build great agents!</p>
+  <div style="display:flex;flex-direction:row;">
+    <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/smolagents/mascot.png" alt="Hugging Face mascot as James Bond" width=100px>
+    <p>smolagents - a smol library to build great agents!</p>
+  </div>
 </h3>
 
 `smolagents` is a library that enables you to run powerful agents in a few lines of code. It offers:
@@ -66,8 +69,8 @@ In our `CodeAgent`,  the LLM engine writes its actions in code. This approach is
 and [reaches higher performance on difficult benchmarks](https://huggingface.co/papers/2411.01747). Head to [our high-level intro to agents](https://huggingface.co/docs/smolagents/conceptual_guides/intro_agents) to learn more on that.
 
 Especially, since code execution can be a security concern (arbitrary code execution!), we provide options at runtime:
-  - a secure python interpreter to run code more safely in your environment
-  - a sandboxed environment using [E2B](https://e2b.dev/).
+  - a secure python interpreter to run code more safely in your environment (more secure than raw code execution but still risky)
+  - a sandboxed environment using [E2B](https://e2b.dev/) (removes the risk to your own system).
 
 ## How smol is it really?
 
@@ -88,6 +91,36 @@ We've created [`CodeAgent`](https://huggingface.co/docs/smolagents/reference/age
 
 This comparison shows that open source models can now take on the best closed models!
 
+## Contributing
+
+To contribute, follow our [contribution guide](https://github.com/huggingface/smolagents/blob/main/CONTRIBUTING.md).
+
+At any moment, feel welcome to open an issue, citing your exact error traces and package versions if it's a bug.
+It's often even better to open a PR with your proposed fixes/changes!
+
+To install dev dependencies, run:
+```
+pip install -e ".[dev]"
+```
+
+When making changes to the codebase, please check that it follows the repo's code quality requirements by running:
+To check code quality of the source code:
+```
+make quality
+```
+
+If the checks fail, you can run the formatter with:
+```
+make style
+```
+
+And commit the changes.
+
+To run tests locally, run this command:
+```bash
+pytest .
+```
+
 ## Citing smolagents
 
 If you use `smolagents` in your publication, please cite it by using the following BibTeX entry.

docs/source/en/_toctree.yml

@@ -8,6 +8,8 @@
   sections:
   - local: tutorials/building_good_agents
     title: ✨ Building good agents
+  - local: tutorials/inspect_runs
+    title: 📊 Inspect your agent runs using telemetry
   - local: tutorials/tools
     title: 🛠️ Tools - in-depth guide
   - local: tutorials/secure_code_execution

docs/source/en/examples/multiagents.md

@@ -64,7 +64,7 @@ model_id = "Qwen/Qwen2.5-Coder-32B-Instruct"
 
 ## 🔍 Create a web search tool
 
-For web browsing, we can already use our pre-existing [`DuckDuckGoSearchTool`](https://github.com/huggingface/smolagents/blob/main/src/smolagents/default_tools/search.py) tool to provide a Google search equivalent.
+For web browsing, we can already use our pre-existing [`DuckDuckGoSearchTool`](https://github.com/huggingface/smolagents/blob/main/src/smolagents/default_tools.py#L151-L176) tool to provide a Google search equivalent.
 
 But then we will also need to be able to peak into the page found by the `DuckDuckGoSearchTool`.
 To do so, we could import the library's built-in `VisitWebpageTool`, but we will build it again to see how it's done.
@@ -196,4 +196,4 @@ Seems like we'll need some sizeable powerplants if the [scaling hypothesis](http
 
 Our agents managed to efficiently collaborate towards solving the task! ✅
 
-💡 You can easily extend this orchestration to more agents: one does the code execution, one the web search, one handles file loadings...
+💡 You can easily extend this orchestration to more agents: one does the code execution, one the web search, one handles file loadings...

docs/source/en/guided_tour.md

@@ -89,8 +89,9 @@ from smolagents import CodeAgent, LiteLLMModel
 
 model = LiteLLMModel(
     model_id="ollama_chat/llama3.2", # This model is a bit weak for agentic behaviours though
-    api_base="http://localhost:11434", # replace with remote open-ai compatible server if necessary
+    api_base="http://localhost:11434", # replace with 127.0.0.1:11434 or remote open-ai compatible server if necessary
     api_key="YOUR_API_KEY" # replace with API key if necessary
+    num_ctx=8192 # ollama default is 2048 which will fail horribly. 8192 works for easy tasks, more is better. Check https://huggingface.co/spaces/NyxKrage/LLM-Model-VRAM-Calculator to calculate how much VRAM this will need for the selected model.
 )
 
 agent = CodeAgent(tools=[], model=model, add_base_tools=True)
@@ -336,7 +337,7 @@ from smolagents import (
 )
 
 # Import tool from Hub
-image_generation_tool = load_tool("m-ric/text-to-image")
+image_generation_tool = load_tool("m-ric/text-to-image", trust_remote_code=True)
 
 model = HfApiModel(model_id)
 

docs/source/en/index.md

@@ -15,6 +15,10 @@ rendered properly in your Markdown viewer.
 
 # `smolagents`
 
+<div class="flex justify-center">
+    <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/smolagents/license_to_call.png" width=100%/>
+</div>
+
 This library is the simplest framework out there to build powerful agents! By the way, wtf are "agents"? We provide our definition [in this page](conceptual_guides/intro_agents), where you'll also find tips for when to use them or not (spoilers: you'll often be better off without agents).
 
 This library offers:

docs/source/en/reference/agents.md

@@ -136,8 +136,22 @@ messages = [
   {"role": "user", "content": "No need to help, take it easy."},
 ]
 
-model = LiteLLMModel("anthropic/claude-3-5-sonnet-latest", temperature=0.2)
-print(model(messages, max_tokens=10))
+model = LiteLLMModel("anthropic/claude-3-5-sonnet-latest", temperature=0.2, max_tokens=10)
+print(model(messages))
 ```
 
 [[autodoc]] LiteLLMModel
+
+### OpenAiServerModel
+
+This class lets you call any OpenAIServer compatible model.
+Here's how you can set it (you can customise the `api_base` url to point to another server):
+```py
+from smolagents import OpenAIServerModel
+
+model = OpenAIServerModel(
+    model_id="gpt-4o",
+    api_base="https://api.openai.com/v1",
+    api_key=os.environ["OPENAI_API_KEY"],
+)
+```

docs/source/en/reference/tools.md

@@ -57,6 +57,10 @@ contains the API docs for the underlying classes.
 
 [[autodoc]] VisitWebpageTool
 
+### UserInputTool
+
+[[autodoc]] UserInputTool
+
 ## ToolCollection
 
 [[autodoc]] ToolCollection

docs/source/en/tutorials/building_good_agents.md

@@ -19,7 +19,7 @@ rendered properly in your Markdown viewer.
 
 There's a world of difference between building an agent that works and one that doesn't.
 How can we build agents that fall into the latter category?
-In this guide, we're going to see best practices for building agents.
+In this guide, we're going to talk about best practices for building agents.
 
 > [!TIP]
 > If you're new to building agents, make sure to first read the [intro to agents](../conceptual_guides/intro_agents) and the [guided tour of smolagents](../guided_tour).
@@ -67,7 +67,7 @@ def get_weather_report_at_coordinates(coordinates, date_time):
     # Dummy function, returns a list of [temperature in °C, risk of rain on a scale 0-1, wave height in m]
     return [28.0, 0.35, 0.85]
 
-def get_coordinates_from_location(location):
+def convert_location_to_coordinates(location):
     # Returns dummy coordinates
     return [3.3, -42.0]