Upload TrailRAG cross-encoder model for hotpotqa

CECorrelationEvaluator_validation_results.csv

@@ -1,67 +1,15 @@
 epoch,steps,Pearson_Correlation,Spearman_Correlation
-0,50,nan,nan
-0,-1,nan,nan
-1,50,nan,nan
-1,-1,nan,nan
-2,50,nan,nan
-2,-1,nan,nan
-0,50,nan,nan
-0,-1,nan,nan
-1,50,nan,nan
-1,-1,nan,nan
-2,50,nan,nan
-2,-1,nan,nan
-3,50,nan,nan
-3,-1,nan,nan
-4,50,nan,nan
-4,-1,nan,nan
-5,50,nan,nan
-5,-1,nan,nan
-6,50,nan,nan
-6,-1,nan,nan
-7,50,nan,nan
-7,-1,nan,nan
-0,50,0.8844942424785296,0.8857172465764791
-0,-1,0.8844942424785296,0.8857172465764791
-1,50,0.9336146534653291,0.8904310386001277
-1,-1,0.9336146534653291,0.8904310386001277
-2,50,0.9450708705265047,0.908718841785034
-2,-1,0.9450708705265047,0.908718841785034
-3,50,0.9478919301215523,0.9122450535593611
-3,-1,0.9478919301215523,0.9122450535593611
-4,50,0.9495609740799951,0.9147266921300965
-4,-1,0.9495609740799951,0.9147266921300965
-5,50,0.9509284655955195,0.914543270049704
-5,-1,0.9509284655955195,0.914543270049704
-0,50,0.8769600515545288,0.8871269174764043
-0,-1,0.8769600515545288,0.8871269174764043
-1,50,0.9245579489171054,0.8784258567954037
-1,-1,0.9245579489171054,0.8784258567954037
-2,50,0.9480165887969222,0.9301595056146273
-2,-1,0.9480165887969222,0.9301595056146273
-3,50,0.9509537570683666,0.9320302181621813
-3,-1,0.9509537570683666,0.9320302181621813
-4,50,0.9527921664806371,0.9309925694044452
-4,-1,0.9527921664806371,0.9309925694044452
-5,50,0.9526781526872711,0.9313792659915328
-5,-1,0.9526781526872711,0.9313792659915328
-6,50,0.954482858927449,0.9329198527954008
-6,-1,0.954482858927449,0.9329198527954008
-7,50,0.953667993155727,0.9318349325109867
-7,-1,0.953667993155727,0.9318349325109867
-0,50,0.9009366205573937,0.8841251669454295
-0,-1,0.9009366205573937,0.8841251669454295
-1,50,0.9377902395000197,0.8864781167682224
-1,-1,0.9377902395000197,0.8864781167682224
-2,50,0.9526408925989197,0.9106428337931949
-2,-1,0.9526408925989197,0.9106428337931949
-3,50,0.9601144587325283,0.920010369671858
-3,-1,0.9601144587325283,0.920010369671858
-4,50,0.9629702193217928,0.9225955082725437
-4,-1,0.9629702193217928,0.9225955082725437
-5,50,0.9614396022220375,0.9227764446780587
-5,-1,0.9614396022220375,0.9227764446780587
-6,50,0.9626509279130091,0.9245291205031555
-6,-1,0.9626509279130091,0.9245291205031555
-7,50,0.9617904585351846,0.9228447811745623
-7,-1,0.9617904585351846,0.9228447811745623
+0,-1,0.8888841799688185,0.8860855585221055
+1,-1,0.9433172575408377,0.899095726435546
+2,-1,0.9543818279076624,0.9182857047232711
+3,-1,0.9561063731084964,0.9276779309546022
+4,-1,0.9568838576527665,0.92585888012172
+5,-1,0.958280584064253,0.9264295627359576
+0,-1,0.8674377304858821,0.8760607745831603
+1,-1,0.9274483825359976,0.9092044306717953
+2,-1,0.9414549054805882,0.9326797719917711
+3,-1,0.9434662406344414,0.9349552144265474
+4,-1,0.9423602934547362,0.9294904476564604
+5,-1,0.9457376917330784,0.9352871141032617
+6,-1,0.9482051166603169,0.9371009844294911
+7,-1,0.9478484140539114,0.9367443852419516

README.md

@@ -21,17 +21,17 @@ model-index:
       type: hotpotqa
     metrics:
     - type: mse
-      value: 0.0502763435546878
+      value: 0.0557947916534922
     - type: mae
-      value: 0.1253658650726789
+      value: 0.1418474710541999
     - type: rmse
-      value: 0.224223869279539
+      value: 0.2362092116186248
     - type: r2_score
-      value: 0.6987198891908811
+      value: 0.6484965021143569
     - type: pearson_correlation
-      value: 0.8868365534337148
+      value: 0.8754595236036868
     - type: spearman_correlation
-      value: 0.8719195144396966
+      value: 0.8618191776300459
 ---
 
 # TrailRAG Cross-Encoder: HotpotQA Enhanced
@@ -53,20 +53,20 @@ This is a fine-tuned cross-encoder model specifically optimized for **Multi-hop
 
 | Metric | Value | Description |
 |--------|-------|-------------|
-| **MSE** | **0.050276** | Mean Squared Error (lower is better) |
-| **MAE** | **0.125366** | Mean Absolute Error (lower is better) |
-| **RMSE** | **0.224224** | Root Mean Squared Error (lower is better) |
-| **R² Score** | **0.698720** | Coefficient of determination (higher is better) |
-| **Pearson Correlation** | **0.886837** | Linear correlation (higher is better) |
-| **Spearman Correlation** | **0.871920** | Rank correlation (higher is better) |
+| **MSE** | **0.055795** | Mean Squared Error (lower is better) |
+| **MAE** | **0.141847** | Mean Absolute Error (lower is better) |
+| **RMSE** | **0.236209** | Root Mean Squared Error (lower is better) |
+| **R² Score** | **0.648497** | Coefficient of determination (higher is better) |
+| **Pearson Correlation** | **0.875460** | Linear correlation (higher is better) |
+| **Spearman Correlation** | **0.861819** | Rank correlation (higher is better) |
 
 ### Training Details
 
-- **Training Duration**: 33 minutes
+- **Training Duration**: 28 minutes
 - **Epochs**: 8
 - **Early Stopping**: No
-- **Best Correlation Score**: 0.922845
-- **Final MSE**: 0.050276
+- **Best Correlation Score**: 0.936744
+- **Final MSE**: 0.055795
 
 ### Training Configuration
 

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:926392c30561825b4939ed008ad74e2fe368726eee6738fbebbd676e45063094
+oid sha256:b49f199a57085d2a7b518756e550554d7cb8b1275313e52ba6f1abb26a4456d1
 size 90866412