Model save

README.md

@@ -18,14 +18,14 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [google-bert/bert-base-cased](https://huggingface.co/google-bert/bert-base-cased) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.9110
-- F1 Macro: 0.8953
-- F1: 0.9272
-- F1 Neg: 0.8633
-- Acc: 0.905
-- Prec: 0.9326
-- Recall: 0.9219
-- Mcc: 0.7906
+- Loss: 0.3886
+- F1 Macro: 0.8984
+- F1: 0.9288
+- F1 Neg: 0.8679
+- Acc: 0.9075
+- Prec: 0.9379
+- Recall: 0.92
+- Mcc: 0.7970
 
 ## Model description
 
@@ -51,7 +51,7 @@ The following hyperparameters were used during training:
 - distributed_type: multi-GPU
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 15
+- num_epochs: 3
 - mixed_precision_training: Native AMP
 
 ### Training results