Enhance serve.py with fine-tuning job management, including job creation, status tracking, and training process in a separate thread. Update serve_test.py to include a test for fine-tuning functionality. Modify .gitignore to exclude model files. This update improves model training capabilities and API integration.
Enhance serve.py to handle additional content types by converting dictionary text and joining list items. Update train.py to replace FastLanguageModel with FastModel and LiteLLMModel, streamline model loading, and adjust dataset preparation logic. Modify config.yaml to change max_samples for testing and add provider information for model configuration.
Add serve_test.py for testing chat completion functionality with the OpenAI client. Update serve.py to use FastModel for improved performance and adjust input handling for optional image processing. Include debugging output for better error tracking.
Update requirements.txt to include sse-starlette dependency, enhance serve.py with additional imports for FastLanguageModel and FastVisionModel, and refactor train.py for improved organization and memory tracking during model training.
Add serve.py for model deployment and API integration, update requirements.txt for smolagents with vllm support, and enhance .gitignore to exclude memory snapshot files. Additionally, implement testing configuration in config.yaml and modify train.py for memory tracking and model saving in VLLM format.
Refactor SFTTrainer configuration in train.py to remove data_collator from the SFT config, preventing duplication and enhancing clarity in trainer setup.
Refactor train.py to utilize a comprehensive configuration structure from config.yaml, enhancing model loading, dataset handling, and trainer setup. This update centralizes parameters for model, PEFT, dataset, and training settings, improving maintainability and flexibility.
Add hydra integration and configuration support in train.py, allowing dynamic model loading and training control. Update requirements.txt to include hydra-core dependency and introduce config.yaml for model parameters and training settings.
Refactor model loading in train.py to use a default model name parameter, enhancing flexibility. Adjust configuration for max sequence length and dtype for improved clarity and consistency.
Add final model files including adapter configuration, model weights, tokenizer settings, and README documentation for SmolLM2-135M. This commit establishes the foundational structure for model deployment and usage.
Update trainer configuration in train.py to align evaluation strategy with save strategy. Set eval_steps to match save_steps for consistent evaluation frequency.
Refactor trainer configuration in train.py for improved clarity. Clean up comments and ensure consistent formatting in evaluation strategy and model selection parameters.
Refactor train.py to improve code readability and organization. Adjust logging setup for clarity, streamline dependency installation commands, and enhance dataset splitting and formatting processes. Ensure consistent formatting in log messages and code structure.
Enhance training script for SmolLM2-135M model by adding logging functionality, improving error handling, and implementing dataset validation split. Refactor model loading and dataset preparation processes for better clarity and robustness. Update trainer configuration to include evaluation strategy and logging of final metrics.
Add training script for SmolLM2-135M model using Unsloth. Includes model loading, dataset preparation, and training configuration. Provides detailed instructions for setup and execution.
Refactor imports in main_v2.py to include SmartSearchTool, enhance code organization in model_factory.py by removing unnecessary whitespace, and improve formatting in test_questions.py for better readability and consistency.
Update unsloth_SmolLM2 notebook to refine model training parameters and output formatting. Adjust dropout rates for better performance, enhance layer configurations, and increase training steps for improved stability. Revise output messages for clarity and update execution counts to reflect changes in training duration.
Refactor unsloth_SmolLM2 notebook to improve model training parameters and output handling. Update dropout settings and layer configurations for enhanced performance. Adjust execution counts for consistency and streamline output messages. Remove redundant display data outputs to clean up notebook structure.
Enhance project structure by adding new files and updating dependencies. Introduce model_factory.py for streamlined model creation using environment variables. Update .gitignore to exclude additional cache and model files. Modify main_v2.py to utilize ModelFactory for model instantiation and adjust API key loading. Expand requirements.txt with new dependencies for improved functionality. Add new notebooks for fine-tuning and model training, enhancing project documentation and usability.
Refactor main_v2.py to update task formatting for dual answer requests, enhancing response structure. Implement error handling for JSON parsing in agent results, ensuring robust output. Add unit tests in test_questions.py to validate succinct answer accuracy against expected values. Remove unused extract_final_answer utility from utils.py, streamlining the codebase.
Update SmartSearchTool to utilize Wikipedia API for content retrieval, enhancing search functionality. Refactor methods for cleaning and formatting Wikipedia content, and update requirements.txt to include new dependencies such as beautifulsoup4 and markdown.
Refactor SmartSearchTool in tool.py to improve logging and variable naming for web search results. Update return statements to ensure consistency in terminology, enhancing clarity in the output messages.
Update requirements.txt to add new dependencies for enhanced functionality, including kagglehub, langchain, and llama-index packages. Refactor SmartSearchTool in tool.py to replace Wikipedia search with a webpage visiting approach, improving content retrieval from web search results. Update tool description and logging for better clarity and usability.
Add main function to tool.py for direct execution of SmartSearchTool. Implement logging and command-line argument handling for search queries, enhancing usability and testing capabilities.
Refactor main_v2.py to remove unused tools and agents, replacing them with the new SmartSearchTool for improved search functionality. Update prompt template loading to use the modified YAML file. Clean up imports and enhance overall code organization for better maintainability.
Enhance main_v2.py by adding ToolCallingAgent for improved web search capabilities and updating prompt template loading method. Comment out unused agent initializations and adjust task handling to streamline agent functionality. Introduce new YAML-based prompt template for CodeAgent to enhance task execution clarity.
Refactor app.py to comment out global question_counter and related logic, simplifying the agent's question handling. This change prepares for potential future enhancements while maintaining current functionality.
Update main_v2.py to format task instructions in GAIA-style, ensuring responses are concise and follow specific guidelines. Modify the task parameter to utilize the new formatted instructions for improved clarity and response accuracy.
Enhance agent functionality in main_v2.py by adding WikipediaSearchTool and updating DuckDuckGoSearchTool and VisitWebpageTool parameters. Modify agent initialization to accommodate new tools and increase max results and output length. Update requirements.txt to include Wikipedia-API dependency. Refactor imports for better organization across agent modules.
Refactor agent structure by modularizing agent implementations into separate directories for web, data analysis, and media agents. Remove legacy code from agents.py, prompts.py, and tools.py, enhancing maintainability. Update main_v2.py to reflect new import paths and agent initialization. Add new tools for enhanced functionality, including web searching and data extraction. Update requirements.txt to include necessary dependencies for new tools.
Refactor app.py to import main from main_v2.py, enhancing modularity. Introduce main_v2.py with a new agent implementation, including OpenTelemetry integration and YAML-based prompt templates. Update requirements.txt to reflect the latest smolagents version. Add tasks.json for VSCode to streamline development workflows.
Integrate telemetry support in main.py by registering the SmolagentsInstrumentor for enhanced monitoring. Update the main function to simplify question enhancement instructions and improve clarity. Modify extract_final_answer utility to prioritize 'final_answer_text' in results. Update requirements.txt to include telemetry dependencies for smolagents.
Enhance agent capabilities by integrating YAML-based prompt templates for web, data analysis, and media agents in agents.py. Update main.py to initialize agents with these templates, improving task handling and response accuracy. Introduce utility functions for extracting final answers and managing prompts, streamlining the overall agent workflow.
Add new agent functionalities by creating agents for web browsing, data analysis, and media handling in agents.py. Introduce various tools in tools.py for web searching, webpage browsing, image analysis, PDF reading, CSV parsing, and date extraction, enhancing the overall capabilities of the agent system.
Add DuckDuckGoSearchTool and VisitWebpageTool to agent initialization in main.py. Update tools configuration to enhance agent functionality and enable visualization of tools used during execution.
Add sleep functionality in process_step to introduce a delay after processing each action step. Enhance logging to include step details and sleep notifications for better traceability during execution.
Implement question counter in BasicAgent and modify response logic in app.py. Update main function in main.py to streamline task execution and improve logging. Prepare for potential rate limiting with a placeholder for wait time.
Add .env.example for API configuration and update .gitignore to exclude .env files. Refactor app.py to implement BasicAgent class, replacing AgentRunner, and improve error handling and logging. Remove unused configuration and graph modules to streamline the codebase.
Add configuration, graph, runner, and tools modules to enhance agent functionality. Introduce a Configuration class for managing parameters, implement an AgentRunner for executing the agent graph, and create tools for general search and mathematical calculations. Update test_agent.py to reflect new import paths and improve overall code organization.
Refactor app.py and update import paths in test_agent.py to improve code organization. Introduce new files for agent configuration, graph definition, and tools, enhancing the overall structure and functionality of the agent system.
Enhance AgentRunner and graph functionality by adding answer extraction logic and improving logging throughout the processing flow. Update the handling of interrupts and state management to ensure clarity in debug output. Refactor the should_continue function in graph.py to better manage completion states and improve user interaction.
Enhance AgentRunner and graph functionality by introducing memory management and improved state handling. Update __call__ method to support both question input and resuming from interrupts, while adding new memory-related fields to track context, actions, and success/error counts. Refactor step callback logic for better user interaction and state management.
Refactor graph.py and test_agent.py by removing unused imports to streamline code and improve readability. This includes the removal of uuid, requests, and unnecessary components from langchain_core.
