### Running the Setup Script (Bash)

Source: https://github.com/pico-lm/pico-analyze/blob/main/README.md

Executes the setup.sh script, which handles environment checks, tool installation, and setting up the Poetry virtual environment for the project. This completes the installation process.

```bash
source setup.sh
```

--------------------------------

### Cloning the Pico Analyze Repository (Bash)

Source: https://github.com/pico-lm/pico-analyze/blob/main/README.md

Clones the pico-lm/pico-analyze repository from GitHub and changes the current directory to the newly cloned repository's root. This is the first step in the installation process.

```bash
git clone https://github.com/pico-lm/pico-analyze
cd pico-analyze
```

--------------------------------

### Example Analysis Configuration (YAML)

Source: https://github.com/pico-lm/pico-analyze/blob/main/README.md

A sample YAML configuration file (configs/my_analysis_config.yaml) demonstrating how to specify analysis parameters. It defines which checkpoints (steps), metrics (CKA), components (simple), layers, and monitoring settings (output directory, Weights & Biases) to use for an analysis run.

```yaml
analysis_name: "my_analysis"
steps:
  - 0
  - 1000
  - 5000
metrics:
  - metric_name: cka
    data_split: "val"
    target_checkpoint: 5000
    components:
      - component_name: simple
        data_type: "weights"
        layer_suffixes: "attention.o_proj"
        layers: [0, 1, 2]
monitoring:
  output_dir: "analysis_results"
  save_to_wandb: true
  wandb:
    entity: "pico-lm"
    project: "pico-analysis"
```

--------------------------------

### Configuring Environment Tokens (Bash)

Source: https://github.com/pico-lm/pico-analyze/blob/main/README.md

Example content for a .env file used to set environment variables for Hugging Face and Weights & Biases API keys. These tokens are required for accessing models and logging results.

```bash
export HF_TOKEN=your_huggingface_token
export WANDB_API_KEY=your_wandb_key
```

--------------------------------

### Running the Analysis Command (Bash)

Source: https://github.com/pico-lm/pico-analyze/blob/main/README.md

Command line instruction using Poetry to execute the analyze script. It takes the path to the analysis configuration file, the Hugging Face repository ID, and the branch/revision as arguments to start the analysis process.

```bash
poetry run analyze \
    --config_path configs/my_analysis_config.yaml \
    --repo_id pico-lm/pico-decoder-small \
    --branch pico-decoder-small-1
```

--------------------------------

### BibTeX Citation for Pico Analyze

Source: https://github.com/pico-lm/pico-analyze/blob/main/README.md

BibTeX entry for citing the Pico project in academic or professional work, as recommended by the authors.

```bibtex
@software{pico2025,
    author = {Diehl Martinez, Richard},
    title = {Pico: A Lightweight Framework for Studying Language Model Learning Dynamics},
    year = {2025},
    url = {https://github.com/pico-lm}
}
```

--------------------------------

### Registering a Custom Metric (Python)

Source: https://github.com/pico-lm/pico-analyze/blob/main/README.md

Demonstrates the basic structure for creating and registering a custom metric class in Python. The class should inherit from BaseMetric and be decorated with @register_metric("your_metric_name") to be recognized by Pico Analyze.

```python
# src/metrics/custom.py

@register_metric("my_custom_metric")
class MyCustomMetric(BaseMetric):
    ...
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.