End of training

README.md

@@ -19,11 +19,13 @@ model-index:
     dataset:
       name: ASR Wolof Dataset
       type: IndabaxSenegal/asr-wolof-dataset
+      config: wo_sn
+      split: test
       args: 'config: wo, split: test'
     metrics:
     - name: Wer
       type: wer
-      value: 52.14391822141652
+      value: 43.507061617297836
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -33,8 +35,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the ASR Wolof Dataset dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.9443
-- Wer: 52.1439
+- Loss: 0.9866
+- Wer: 43.5071
 
 ## Model description
 
@@ -54,28 +56,26 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 1e-05
-- train_batch_size: 48
-- eval_batch_size: 16
+- train_batch_size: 16
+- eval_batch_size: 8
 - 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
-- lr_scheduler_warmup_steps: 500
-- num_epochs: 4
+- num_epochs: 3.0
 - mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Wer     |
 |:-------------:|:-----:|:----:|:---------------:|:-------:|
-| 0.2474        | 1.0   | 967  | 0.8705          | 51.7197 |
-| 0.1774        | 2.0   | 1934 | 0.8798          | 52.3592 |
-| 0.1011        | 3.0   | 2901 | 0.9154          | 51.8179 |
-| 0.0541        | 4.0   | 3868 | 0.9443          | 52.1439 |
+| 1.1308        | 1.0   | 142  | 0.9861          | 43.7820 |
+| 0.6317        | 2.0   | 284  | 0.9646          | 43.4946 |
+| 0.4205        | 3.0   | 426  | 0.9866          | 43.5071 |
 
 
 ### Framework versions
 
 - Transformers 4.46.3
-- Pytorch 2.1.0+cu118
+- Pytorch 2.4.0
 - Datasets 3.1.0
-- Tokenizers 0.20.3
+- Tokenizers 0.20.0

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6447b0eb5e44d7962db882a2316781d6ea5a1b3a7d7b45ba0b7f0dc458063e8a
+oid sha256:53289b0744bcee57815395931d3cabb227e4c6732cb3a21c2324f4b1b3405d0f
 size 966995080