bkh6722
/

bach-arb

@@ -3,18 +3,18 @@ license: apache-2.0
 tags:
 - generated_from_trainer
 model-index:
-  name: xlsr-vlbg
 ---
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 should probably proofread and complete it, then remove this comment. -->
-# xlsr-vlbg
 This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.0589
-- Wer: 0.4221
 ## Model description
@@ -42,29 +42,29 @@ The following hyperparameters were used during training:
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 500
-- num_epochs: 60
 - mixed_precision_training: Native AMP
 ### Training results
-| Training Loss | Epoch | Step | Validation Loss | Wer    |
-|:-------------:|:-----:|:----:|:---------------:|:------:|
-| 10.2457       | 3.56  | 100  | 3.8979          | 1.0    |
-| 3.3871        | 7.14  | 200  | 2.9923          | 1.0    |
-| 2.9516        | 10.7  | 300  | 2.9019          | 1.0    |
-| 2.4874        | 14.28 | 400  | 1.7228          | 0.9452 |
-| 1.1546        | 17.84 | 500  | 1.0064          | 0.6758 |
-| 0.5481        | 21.42 | 600  | 0.9380          | 0.5828 |
-| 0.3193        | 24.98 | 700  | 0.9767          | 0.5367 |
-| 0.2157        | 28.56 | 800  | 0.9501          | 0.4780 |
-| 0.168         | 32.14 | 900  | 0.9843          | 0.4711 |
-| 0.1106        | 35.7  | 1000 | 1.0452          | 0.4623 |
-| 0.0861        | 39.28 | 1100 | 1.0691          | 0.4515 |
-| 0.0733        | 42.84 | 1200 | 1.0813          | 0.4241 |
-| 0.0565        | 46.42 | 1300 | 1.0311          | 0.4486 |
-| 0.0443        | 49.98 | 1400 | 1.0479          | 0.4329 |
-| 0.0352        | 53.56 | 1500 | 1.0901          | 0.4221 |
-| 0.0277        | 57.14 | 1600 | 1.0589          | 0.4221 |
 ### Framework versions

 tags:
 - generated_from_trainer
 model-index:
+  name: bach-arb
 ---
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 should probably proofread and complete it, then remove this comment. -->
+# bach-arb
 This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on an unknown dataset.
 It achieves the following results on the evaluation set:
+- Loss: 2.6864
+- Wer: 0.7986
 ## Model description
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 500
+- num_epochs: 120
 - mixed_precision_training: Native AMP
 ### Training results
+| Training Loss | Epoch  | Step | Validation Loss | Wer    |
+|:-------------:|:------:|:----:|:---------------:|:------:|
+| 14.2075       | 7.14   | 100  | 3.9465          | 1.0    |
+| 3.4162        | 14.28  | 200  | 3.0725          | 1.0    |
+| 2.9811        | 21.41  | 300  | 2.8973          | 1.0    |
+| 2.5037        | 28.55  | 400  | 2.1000          | 1.0192 |
+| 1.2743        | 35.69  | 500  | 1.6965          | 0.9448 |
+| 0.6392        | 42.83  | 600  | 1.8077          | 0.8657 |
+| 0.3871        | 49.97  | 700  | 2.1009          | 0.8585 |
+| 0.2691        | 57.14  | 800  | 2.0475          | 0.8201 |
+| 0.2032        | 64.28  | 900  | 2.1936          | 0.8537 |
+| 0.1437        | 71.41  | 1000 | 2.1932          | 0.8058 |
+| 0.1144        | 78.55  | 1100 | 2.3580          | 0.8201 |
+| 0.0938        | 85.69  | 1200 | 2.4276          | 0.7938 |
+| 0.0672        | 92.83  | 1300 | 2.7036          | 0.8297 |
+| 0.055         | 99.97  | 1400 | 2.6035          | 0.7674 |
+| 0.0467        | 107.14 | 1500 | 2.7449          | 0.7746 |
+| 0.0372        | 114.28 | 1600 | 2.6864          | 0.7986 |
 ### Framework versions