End of training

README.md

@@ -18,9 +18,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [Mat17892/t5small_enfr_opus](https://huggingface.co/Mat17892/t5small_enfr_opus) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.4839
-- Bleu: 6.4
-- Gen Len: 17.5697
+- Loss: 1.3217
+- Bleu: 7.3648
+- Gen Len: 17.5145
 
 ## Model description
 
@@ -45,17 +45,21 @@ The following hyperparameters were used during training:
 - seed: 42
 - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: linear
-- num_epochs: 4
+- num_epochs: 8
 - mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch | Step  | Validation Loss | Bleu   | Gen Len |
 |:-------------:|:-----:|:-----:|:---------------:|:------:|:-------:|
-| 1.7415        | 1.0   | 6355  | 1.5201          | 6.1792 | 17.5724 |
-| 1.6892        | 2.0   | 12710 | 1.4981          | 6.2914 | 17.572  |
-| 1.6825        | 3.0   | 19065 | 1.4873          | 6.3889 | 17.5701 |
-| 1.6803        | 4.0   | 25420 | 1.4839          | 6.4    | 17.5697 |
+| 1.542         | 1.0   | 6355  | 1.3460          | 7.1462 | 17.5281 |
+| 1.5196        | 2.0   | 12710 | 1.3419          | 7.2241 | 17.5219 |
+| 1.5005        | 3.0   | 19065 | 1.3346          | 7.3005 | 17.5137 |
+| 1.5083        | 4.0   | 25420 | 1.3298          | 7.3326 | 17.5174 |
+| 1.4999        | 5.0   | 31775 | 1.3269          | 7.3337 | 17.5214 |
+| 1.4906        | 6.0   | 38130 | 1.3237          | 7.3359 | 17.5185 |
+| 1.4944        | 7.0   | 44485 | 1.3223          | 7.3664 | 17.5161 |
+| 1.5018        | 8.0   | 50840 | 1.3217          | 7.3648 | 17.5145 |
 
 
 ### Framework versions

runs/Dec05_16-28-30_9efd5b24bb80/events.out.tfevents.1733416112.9efd5b24bb80.427.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3a192fad03ea43ff9834b3c183393e63d22e0f7a38ce277ad654b098e94b4a18
-size 30300
+oid sha256:8f8e743dafaf3f5f8a706e40261b1ffd9d3842dc7374c2bd92df931b949d78c4
+size 31037