Test text processing functionality

Test timer functionality.

Test that the agent correctly uses OpenAI tool calling format

Verify extract_key_information is removed from tool definitions

Verify extract_key_information is removed from tool definitions

Test tool calling capabilities

Test various perception tools.

Test that different top_k values return the correct number of results

Test that different top_k values return the correct number of results

Test and display TTFT tracking across iterations

Test virtual terminal functionality.

Test 1: Workflow capture functionality.

Test 4: Workflow replay mechanism.

Convert workflow to dictionary for serialization

Convert to dictionary for JSON serialization

Convert to dictionary

Convert to dictionary

Convert to dictionary

Convert event to user message format for the agent

Train the agent for one episode. Returns: Total reward, number of steps, victory status

使用 SFTTrainer 训练模型 Args: model: 配置了 LoRA 的模型 tokenizer: 分词器 dataset: 训练数据集 output_dir: 输出目录 batc

Train the model using TRL SFTTrainer. Hyperparameters are based on the OpenAI Cookbook gpt-oss-20b example, which provides good defa

Transcribe audio file using OpenAI Whisper (local). Note: Requires whisper package installed. Args: file_path: Path to audio

Trim audio file to specified time range using ffmpeg. Args: audio_path: Path to audio file start_time: Start time in sec

Safely decode bytes to string.

Update the embedding for a document. Args: doc_id: Document ID embedding: Embedding vector

Update a specific line in file. Args: file_path: File path line_number: Line number to update (1-ind

Update performance metrics for a workflow after execution. Args: workflow_id: ID of the workflow that was execut

Validate the configuration.

Validate the test case structure.

Validate required configuration Args: provider: Provider to validate (defaults to configured provider)

Validate that the conversation has at least the minimum required rounds.

Verify that indexing worked by performing test searches

View evaluation results.

View evaluation results

View details of a specific test case.

View loaded test cases

View trajectory file with formatted output

Generate visualizations from saved results Args: results_path: Path to JSON results file output_dir: Directory to save v

Create visualizations of ablation study results Args: results: List of test results

Wait for server to be ready

Watch an LLM agent play with reasoning.

Watch a trained RL agent play.

Called during generation to track tokens

Cleanup when agent is destroyed

Initialize the Orpheus TTS inference engine. Args: model_path: HuggingFace model path or local path

Initialize coding agent Args: api_key: API key for the provider model: Model identifier

Initialize coding agent Args: api_key: API key for Anthropic/OpenRouter model: Model identifier

Initialize the learning agent. Args: task: The task to be performed llm: Language model to use (brow

Initialize the workflow replayer. Args: headless: Whether to run browser in headless mode

Initialize the knowledge base. Args: storage_path: Directory path for storing workflow data

Initialize the trajectory summarizer. Args: llm_config: LLM configuration model_name: Model to use f

Initialize the Experience Agent. Args: conf: Agent configuration name: Agent name system

Initialize the knowledge base. Args: index_path: Path to store the index files model_name: Name of t

Initialize demo with configuration.

Initialize Elo rating system. Args: initial_rating: Starting rating for all models k_factor: Learnin

Initialize NumPy-based Elo system.

Initialize the ablation agent Args: tools_info: Information about available tools wiki: Wiki/system