Spaces:
Sleeping
Sleeping
Commit History
-
Fix: Correct attribute references for error detection/correction metrics
RGB Evaluation- Changed error_detection_count to errors_detected (actual attribute name) - Changed error_correction_count to errors_corrected (actual attribute name) - Fixed both app.py (main) and RGBMetrics/app.py (mirror) - Resolves 'EvaluationResult' object has no attribute 'error_detection_count' error005cdbf -
fix: Information Integration evaluation - handle multiple answer variants with pipe-separated format
RGB Evaluation5253a83 -
feat: Show all 9 LLM models in app dropdown, add comprehensive code review and metric analysis documentation
RGB Evaluationb1ccc5d
-
feat: Add noise level percentage to Noise Robustness cards
RGB Evaluation- Extract noise level (0%, 20%, 40%, 60%, 80%) from task name - Display as metric in each card for noise robustness - Applied to both current results and past results sections - Improves clarity of which noise level each result represents - Extracted from task format: noise_robustness_XX%ca20603 -
fix: Convert past results section to grid layout
RGB Evaluation- Replace past results table with grid display - Each task shows in separate grid with all models - Maintains same grid structure as current results - Applies to: Noise Robustness, Negative Rejection, Information Integration, Counterfactual Robustness - Improved readability and consistency90b0944 -
fix: Remove duplicate results display section
RGB Evaluation- Removed duplicate evaluation results heading - Keep only grid layout display for 4 RAG abilities - Each task shows separate grid with all models - Tasks: Noise Robustness, Negative Rejection, Information Integration, Counterfactual Robustness0cfbd9f -
feat: Add Moonshot Kimi K2 Instruct model to DEFAULT_MODELS
RGB Evaluation- Added moonshotai/kimi-k2-instruct to primary evaluation models - Now evaluating with 5 models total - Kimi K2 provides multi-lingual capabilities and strong reasoning - Maintains all existing models in list1b8d19d -
fix: Add HF Spaces YAML configuration to README.md
RGB Evaluation- Add title: RGB RAG Evaluation Dashboard - Add emoji: - Add color scheme: blue to green - Set sdk: streamlit - Set app_file: app.py - Fixes HF configuration errorca5ddcb -
Merge main: accept local changes with new grid layout and model updates
RGB Evaluationbdba13c -
feat: Add separate grid layout for 4 RAG abilities in Streamlit UI
RGB Evaluation- Display Noise Robustness results in separate grid - Display Negative Rejection results in separate grid - Display Information Integration results in separate grid - Display Counterfactual Robustness results in separate grid - All models shown side-by-side for easy comparison - Task-specific metrics displayed for each ability - Updated gemma2-9b-it as replacement for decommissioned mixtral - Maintained detailed visualization tabs for analyticsaf25c62
-
Fix results file pattern matching - search for results_*.json instead of evaluation_*.json
RGB Deployment Bota757c21 -
Update all LLMs to new model list with 4 default models
RGB Deployment Bot3d27cc5 -
Fix debug function - replace st.debug() with st.write() in expander
RGB Deployment Bote7842fe -
Add debug info to past results viewer to diagnose missing results
RGB Deployment Bot8796cb2 -
Replace decommissioned llama-3.1-70b-versatile with llama2-70b-4096
RGB Deployment Bot800ecaf -
Reduce RPM limit to 25 (safe margin below 30) and increase request interval to 2.5s
RGB Deployment Bot4649c35 -
Implement RPM rate limiting (30 requests/minute) with sliding window
RGB Deployment Bot68627ba -
Add deepseek-r1-distill-llama-70b free chat model
RGB Deployment Botb8b9c59 -
Add persistent results storage with past results viewer
RGB Deployment Bot2397c87 -
Replace decommissioned mixtral-8x7b-32768 with llama-3.1-70b-versatile
RGB Deployment Botdd2d90a -
Fix continuous page refresh during background evaluation
RGB Deployment Bot0deb4ed -
Add 4th model: meta-llama/llama-4-maverick-17b-128e-instruct
RGB Deployment Botd68feb2 -
Replace deprecated gemma2-9b-it with mixtral-8x7b-32768
RGB Deployment Bot8268ab9 -
Add background evaluation and downloadable reports (CSV, JSON, PDF)
RGB Deployment Bot81c6a30 -
Add dataset files for HF Space
RGB Deployment Bot07844bd -
Add YAML metadata to README
RGB Deployment Bot921aa55 -
Deploy RGB Metrics dashboard - 2026-01-04 22:05:37
RGB Deployment Bot3f89944 -
initial commit
Verified signaturec1063cf