Create README.md

README.md

@@ -0,0 +1,38 @@
+Whisper Small Tr - Saadin</br></br>
+This model is a fine-tuned version of openai/whisper-small on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:</br></br>
+
+Loss: 0.2572</br>
+Wer: 23.18</br>
+Model description</br>
+More information needed</br></br>
+
+Intended uses & limitations</br>
+More information needed</br></br>
+
+Training and evaluation data</br>
+More information needed</br></br>
+
+Training procedure</br>
+Training hyperparameters</br>
+The following hyperparameters were used during training:</br></br>
+
+learning_rate: 1e-05</br>
+train_batch_size: 16</br>
+eval_batch_size: 16</br>
+seed: 42</br>
+optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08</br>
+lr_scheduler_type: linear</br>
+lr_scheduler_warmup_steps: 500</br>
+training_steps: 1000</br>
+mixed_precision_training: Native AMP</br></br>
+
+Training results</br>
+Training Loss --Epoch---Validation Loss--Wer </br>
+0.1743----------1000-----0.2572------- -23.18</br></br>
+
+
+Framework versions</br>
+Transformers 4.25.0.dev0</br>
+Pytorch 1.12.1</br>
+Datasets 2.5.3.dev0</br>
+Tokenizers 0.12.1</br>