End of training

README.md

@@ -14,7 +14,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model was trained from scratch on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 3.0511
+- Loss: 3.7876
 
 ## Model description
 
@@ -33,55 +33,24 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 4e-05
-- train_batch_size: 12
-- eval_batch_size: 12
+- learning_rate: 0.0005
+- train_batch_size: 8
+- eval_batch_size: 8
 - seed: 42
 - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: cosine
-- num_epochs: 3
+- num_epochs: 2
 
 ### Training results
 
-| Training Loss | Epoch  | Step | Validation Loss |
-|:-------------:|:------:|:----:|:---------------:|
-| 3.341         | 0.08   | 100  | 3.2488          |
-| 3.1192        | 0.16   | 200  | 3.1795          |
-| 3.0514        | 0.24   | 300  | 3.1372          |
-| 3.0135        | 0.32   | 400  | 3.1121          |
-| 2.9543        | 0.4    | 500  | 3.0921          |
-| 2.9889        | 0.48   | 600  | 3.0757          |
-| 2.8721        | 0.56   | 700  | 3.0619          |
-| 2.8771        | 0.64   | 800  | 3.0505          |
-| 2.8444        | 0.72   | 900  | 3.0402          |
-| 2.7972        | 0.8    | 1000 | 3.0302          |
-| 2.7936        | 0.88   | 1100 | 3.0244          |
-| 2.8006        | 0.96   | 1200 | 3.0151          |
-| 2.6007        | 1.04   | 1300 | 3.0296          |
-| 2.4391        | 1.12   | 1400 | 3.0369          |
-| 2.4741        | 1.2    | 1500 | 3.0313          |
-| 2.4998        | 1.28   | 1600 | 3.0313          |
-| 2.4442        | 1.3600 | 1700 | 3.0305          |
-| 2.5129        | 1.44   | 1800 | 3.0245          |
-| 2.4257        | 1.52   | 1900 | 3.0234          |
-| 2.4279        | 1.6    | 2000 | 3.0208          |
-| 2.4327        | 1.6800 | 2100 | 3.0196          |
-| 2.4375        | 1.76   | 2200 | 3.0167          |
-| 2.406         | 1.8400 | 2300 | 3.0164          |
-| 2.3819        | 1.92   | 2400 | 3.0140          |
-| 2.4157        | 2.0    | 2500 | 3.0126          |
-| 2.2551        | 2.08   | 2600 | 3.0426          |
-| 2.2469        | 2.16   | 2700 | 3.0473          |
-| 2.185         | 2.24   | 2800 | 3.0506          |
-| 2.2112        | 2.32   | 2900 | 3.0524          |
-| 2.2294        | 2.4    | 3000 | 3.0501          |
-| 2.2252        | 2.48   | 3100 | 3.0495          |
-| 2.2259        | 2.56   | 3200 | 3.0514          |
-| 2.2325        | 2.64   | 3300 | 3.0507          |
-| 2.2441        | 2.7200 | 3400 | 3.0508          |
-| 2.1913        | 2.8    | 3500 | 3.0512          |
-| 2.2303        | 2.88   | 3600 | 3.0511          |
-| 2.2099        | 2.96   | 3700 | 3.0511          |
+| Training Loss | Epoch | Step | Validation Loss |
+|:-------------:|:-----:|:----:|:---------------:|
+| 2.9597        | 0.32  | 200  | 3.7846          |
+| 2.9834        | 0.64  | 400  | 3.7458          |
+| 2.8111        | 0.96  | 600  | 3.6531          |
+| 1.6601        | 1.28  | 800  | 3.8347          |
+| 1.439         | 1.6   | 1000 | 3.8058          |
+| 1.3692        | 1.92  | 1200 | 3.7876          |
 
 
 ### Framework versions