Spaces:
Sleeping
Sleeping
Commit History
-
Fix PyABSA v2.4.2 API
2e2ec5c -
Fix PyABSA v2.4.2 API - use ATEPC
d0d45bd -
Fix PyABSA v2.4.2 API - use ATEPCCheckpointManager
07164fa -
Fixing models for PyABSA
b2f2eab -
ATPC fixes
95fbfc5 -
Adding git and git-lfs to the HF spaces
5b0f561 -
Updating models
99cc145
-
Fixes1
a8443d7 -
Fixes
70de5e9 -
Fixes for model calling
f70514c -
Lifting rate limits
a3b4a6e -
added dotenv
eb5b662 -
Fixed dockerignore
317c94f -
Fixed dockerfile
a9680a1 -
Adding dockerfile
a4da994 -
Adding app.py
7175d0a -
Adding Mongo+Redis concept
4b62d23 -
Updating FastAPI calls
dd42f74
-
Integrated PyABSA again
2b4639d -
Changed translation model
ac8959c -
Fx1
9b1ad5f -
Adding fix
b17ff40 -
Adding few fixes
db53222 -
Changing streamlit warning in Backend optional;
ad40a54 -
Fixing things again
41146f6 -
Fixing docker file
c65ea29 -
Adding cancellation button:
102e87a
-
ADD: FastAPI server for HF Spaces API endpoints
Added api_server.py - FastAPI backend with REST endpoints /process-reviews endpoint for frontend integration /health endpoint for status checking Updated Dockerfile to run FastAPI server Added FastAPI, uvicorn, pydantic to requirements CORS enabled for frontend connections Frontend can now connect: POST /process-reviews Status: API architecture complete, ready for testingc08a089 -
FIX: Complete DataManager session management
Added missing load_session() method Added save_session() method for session persistence Enhanced get_history_list() with error handling Backward compatibility for different data formats Robust directory handling (creates if missing) Status: Session management fully functional with JSON storagebaecac6 -
FIX: Add spaCy model for PyABSA compatibility
Fixed PyABSA dependency (uncommented in requirements) Added spaCy English model to requirements Updated Dockerfile to install en_core_web_sm PyABSA multilingual checkpoint already working! Status: Final dependency fix for complete deployment6a45d88 -
FIX: Docker permission issue in Dockerfile
Fixed file permission error in HF Spaces deployment Changed order: set ownership before switching to user PyABSA dependencies successfully installed (106.7s) Status: Ready for deployment testingcf02b75 -
ADD: API architecture files for split deployment
Added app_api.py - HF Spaces API entry point Added backend_api.py - Dedicated ML API service Added frontend_light.py - Streamlit Cloud frontend Added requirements-backend.txt - ML dependencies Added requirements-frontend.txt - Lightweight UI only Architecture: HF Spaces (PyABSA backend) + Streamlit Cloud (frontend) Next: Test HF Spaces deployment and configure API endpoints1b4a2d1 -
ROLLBACK: Restore PyABSA approach for high accuracy
Restored PyABSA implementation in data_processor.py Updated requirements for ML dependencies (torch, transformers, pyabsa) Fixed requirements-docker.txt for HF Spaces deployment Updated secrets template for PyABSA approach Created comprehensive PyABSA deployment guide Rationale: Higher accuracy needed than HF API transformers Strategy: HF Spaces backend + Streamlit Cloud frontend Next: Deploy backend to HF Spaces with PyABSA modelsf71c767 -
MAJOR: Switch to HF Inference API + Streamlit Cloud approach
BREAKING CHANGE: Complete architecture overhaul for better reliability **API Integration:** - Replaced PyABSA with HF Inference API (cardiffnlp/twitter-roberta-base-sentiment-latest) - Replaced local M2M100 with HF Translation API - Added smart retry logic with exponential backoff - Graceful fallbacks when API unavailable **Optimized Dependencies:** - Removed torch (~1GB), transformers (~500MB), pyabsa - Added requests for API calls - Result: ~1.5GB smaller, much faster builds **Performance Improvements:** - Enhanced aspects: 10 categories with better keywords - Progress logging for user feedback - Robust error handling with fallbacks - Production-ready timeout and retry logic **Streamlit Cloud Ready:** - Created .streamlit/config.toml for production settings - Added secrets template for HF_TOKEN - Comprehensive deployment guide - Optimized for file uploads and better UX **Benefits:** - Faster builds (3-5 min vs 10-15 min) - Lower memory (500MB vs 2-3GB) - Better reliability …ee7ef03
-
Fix PyABSA checkpoint loading with fallback options
- Fixed 'multilingual' checkpoint download failure - Added fallback checkpoint options: english, chinese, default - Tries multiple checkpoints until one succeeds - Better error handling and logging for each attempt - Should resolve 'expected str, bytes or os.PathLike object, not NoneType' error The app will now try different PyABSA checkpoints instead of failing on 'multilingual'.ebe1bc5 -
Major upgrade: Python 3.11 + Clean ABSA implementation
1. Upgraded to Python 3.11: - Updated Dockerfile from python:3.9-slim to python:3.11-slim - Better compatibility with PyABSA library - Improved performance and modern language features 2. Cleaned ABSA implementation: - Removed TextBlob fallback sentiment analysis - Simplified ABSAProcessor to use only PyABSA - Returns empty results when ABSA unavailable instead of fallback - Removed textblob dependency from requirements 3. Benefits: - Cleaner, more focused codebase - Better PyABSA compatibility with Python 3.11 - Faster builds with fewer dependencies - More predictable behavior This should resolve the type annotation compatibility issues.2eed23e -
Critical fixes for Docker deployment
1. Fixed cache directory permissions: - Added cache directory setup in Dockerfile with proper permissions - Added HF_HOME and TRANSFORMERS_CACHE environment variables - Added cache directory initialization in app startup 2. Fixed type annotation compatibility: - Removed @st.cache_resource decorators from class methods - Fixed Python 3.9 compatibility issues with union types - Updated TranslationService and ABSAProcessor model loading These fixes should resolve the permission denied and type annotation errors.1798510 -
Fix Dockerfile: Remove software-properties-common package
- Removed software-properties-common which is not available in python:3.9-slim - Kept essential packages: build-essential, curl, git - This should resolve the Docker build failure28acaaf -
Deploy enhanced sentiment analysis dashboard with advanced NLP features
Parthnuwal7- Added app_enhanced.py with comprehensive ABSA capabilities - Included network analysis, Sankey diagrams, timeline charts - Optimized Docker configuration for HF Spaces deployment - Added multilingual support and intent classification - Enhanced visualizations with KPI dashboard and export features66b9109 -
initial commit
Verified signature07910ac -
Duplicate from streamlit/streamlit-template-space
Verified signatureCo-authored-by: Franck Abgrall <FranckAbgrall@users.noreply.huggingface.co>41a3533
