Joash2024/code-review-assistant

Update huggingface deploy changes

Update README with complete content

Update README.md

Update README metadata to trigger build

Trigger Hugging Face Spaces build

Fix history and metrics updates with proper state management

Add refresh functionality and auto-updates for history and metrics

Update interface to handle history and metrics updates

Add history and metrics persistence with file storage

Improve model initialization with better state management

Fix tokenizer initialization and improve error handling

Merge branch 'main' of https://huggingface.co/spaces/Joash2024/code-review-assistant

Add proper ZeroGPU configuration with spaces decorator

Update README.md

Update app with CPU-first approach

Update requirements for better CPU/GPU support

Improve model initialization with CPU-first approach

Merge branch 'main' of https://huggingface.co/spaces/Joash2024/code-review-assistant

Configure app for ZeroGPU dynamic GPU allocation

Update README.md

Simplify model loading to use CPU by default

Update Space metadata for proper ZeroGPU configuration

Simplify implementation and remove Space SDK dependencies

Update requirements for ZeroGPU and space configuration

Add ZeroGPU decorator and space configuration

Fix Gradio launch parameters and improve error handling

Improve model loading with fallback options and memory settings

Update requirements for improved model loading and error handling

Merge branch 'main' of https://huggingface.co/spaces/Joash2024/code-review-assistant

Add robust error handling and improve model loading

Update README.md

Simplify app.py and update configuration for ZeroGPU

Add ZeroGPU support and update dependencies

Update app to properly utilize ZeroGPU with GPU optimizations

Merge branch 'main' of https://huggingface.co/spaces/Joash2024/code-review-assistant

Add history and metrics tabs to Gradio interface

Update README.md

Convert to Gradio app for simpler deployment

Remove 4-bit quantization and use regular model loading

Fix offline mode and improve model loading

Optimize code reviewer with memory management and improved parsing

Switch to smaller model and reduce memory limits

Add comprehensive memory optimizations for model and Docker

Optimize memory usage with 4-bit quantization and Docker settings

Add detailed logging and improve error handling in model manager

Fix model_manager to use actual model inference instead of mock response

Remove hardcoded token and require it from environment variables

Resolve merge conflicts and update configurations

Update Dockerfile with proper user permissions

Fix Dockerfile permissions and user management