End of training

README.md

@@ -1,11 +1,14 @@
 ---
 library_name: transformers
-license: apache-2.0
-base_model: facebook/wav2vec2-base
+license: bsd-3-clause
+base_model: MIT/ast-finetuned-audioset-10-10-0.4593
 tags:
 - generated_from_trainer
 metrics:
 - accuracy
+- precision
+- recall
+- f1
 model-index:
 - name: model_dialect
   results: []
@@ -16,10 +19,13 @@ should probably proofread and complete it, then remove this comment. -->
 
 # model_dialect
 
-This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
+This model is a fine-tuned version of [MIT/ast-finetuned-audioset-10-10-0.4593](https://huggingface.co/MIT/ast-finetuned-audioset-10-10-0.4593) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.8299
-- Accuracy: 0.7182
+- Loss: 0.6309
+- Accuracy: 0.7529
+- Precision: 0.7608
+- Recall: 0.7661
+- F1: 0.7623
 
 ## Model description
 
@@ -38,42 +44,33 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 4e-05
-- train_batch_size: 32
-- eval_batch_size: 32
+- learning_rate: 5e-05
+- train_batch_size: 8
+- eval_batch_size: 8
 - seed: 42
-- gradient_accumulation_steps: 4
-- total_train_batch_size: 128
-- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- lr_scheduler_warmup_ratio: 0.1
-- num_epochs: 16
+- num_epochs: 10
 
 ### Training results
 
-| Training Loss | Epoch   | Step | Validation Loss | Accuracy |
-|:-------------:|:-------:|:----:|:---------------:|:--------:|
-| 6.4248        | 0.9455  | 13   | 1.5911          | 0.2748   |
-| 6.3186        | 1.9636  | 27   | 1.4754          | 0.4249   |
-| 5.5404        | 2.9818  | 41   | 1.2967          | 0.4850   |
-| 5.136         | 4.0     | 55   | 1.1871          | 0.5289   |
-| 4.8487        | 4.9455  | 68   | 1.1144          | 0.5612   |
-| 4.337         | 5.9636  | 82   | 1.0689          | 0.5820   |
-| 4.1023        | 6.9818  | 96   | 0.9931          | 0.6628   |
-| 3.762         | 8.0     | 110  | 0.9842          | 0.6305   |
-| 3.4974        | 8.9455  | 123  | 0.9379          | 0.6420   |
-| 3.4265        | 9.9636  | 137  | 0.9304          | 0.6559   |
-| 3.111         | 10.9818 | 151  | 0.8909          | 0.6767   |
-| 2.9471        | 12.0    | 165  | 0.8744          | 0.6859   |
-| 2.8061        | 12.9455 | 178  | 0.8533          | 0.7021   |
-| 2.7032        | 13.9636 | 192  | 0.8541          | 0.7067   |
-| 2.5328        | 14.9818 | 206  | 0.8299          | 0.7136   |
-| 2.5328        | 15.1273 | 208  | 0.8299          | 0.7182   |
+| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1     |
+|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|
+| 1.3499        | 1.0   | 217  | 1.4002          | 0.4065   | 0.2570    | 0.3730 | 0.2609 |
+| 1.2635        | 2.0   | 434  | 1.0922          | 0.5242   | 0.6868    | 0.4970 | 0.4952 |
+| 1.0379        | 3.0   | 651  | 1.0788          | 0.5335   | 0.7047    | 0.5137 | 0.5053 |
+| 0.9536        | 4.0   | 868  | 0.8706          | 0.6513   | 0.6804    | 0.6631 | 0.6668 |
+| 0.9321        | 5.0   | 1085 | 0.9052          | 0.6397   | 0.6693    | 0.6474 | 0.6307 |
+| 0.9395        | 6.0   | 1302 | 0.8028          | 0.6767   | 0.7494    | 0.6692 | 0.6845 |
+| 0.752         | 7.0   | 1519 | 0.7386          | 0.7344   | 0.7650    | 0.7386 | 0.7441 |
+| 0.7984        | 8.0   | 1736 | 0.7100          | 0.7206   | 0.7370    | 0.7383 | 0.7323 |
+| 0.5841        | 9.0   | 1953 | 0.6509          | 0.7413   | 0.7525    | 0.7494 | 0.7503 |
+| 0.6678        | 10.0  | 2170 | 0.6309          | 0.7529   | 0.7608    | 0.7661 | 0.7623 |
 
 
 ### Framework versions
 
-- Transformers 4.46.0
+- Transformers 4.45.1
 - Pytorch 2.4.0
 - Datasets 3.0.1
 - Tokenizers 0.20.0