update model card README.md

README.md

@@ -1,41 +1,38 @@
 ---
-language:
-- be
 license: apache-2.0
 tags:
-- whisper-event
 - generated_from_trainer
 datasets:
-- mozilla-foundation/common_voice_11_0
+- common_voice_11_0
 metrics:
 - wer
 model-index:
-- name: Whisper Small Belarusian
+- name: whisper-tiny-be-test
   results:
   - task:
       name: Automatic Speech Recognition
       type: automatic-speech-recognition
     dataset:
-      name: mozilla-foundation/common_voice_11_0 be
-      type: mozilla-foundation/common_voice_11_0
+      name: common_voice_11_0
+      type: common_voice_11_0
       config: be
       split: validation
       args: be
     metrics:
     - name: Wer
       type: wer
-      value: 75.27472527472527
+      value: 61.53846153846154
 ---
 
 <!-- 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. -->
 
-# Whisper Small Belarusian
+# whisper-tiny-be-test
 
-This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the mozilla-foundation/common_voice_11_0 be dataset.
+This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the common_voice_11_0 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.9745
-- Wer: 75.2747
+- Loss: 0.5759
+- Wer: 61.5385
 
 ## Model description
 
@@ -61,7 +58,7 @@ 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: 5
-- training_steps: 20
+- training_steps: 100
 - mixed_precision_training: Native AMP
 
 ### Training results
@@ -70,6 +67,14 @@ The following hyperparameters were used during training:
 |:-------------:|:-----:|:----:|:---------------:|:-------:|
 | 2.4473        | 0.5   | 10   | 1.3675          | 95.4212 |
 | 1.256         | 1.0   | 20   | 0.9745          | 75.2747 |
+| 0.9934        | 0.3   | 30   | 0.8114          | 72.1612 |
+| 0.9568        | 0.4   | 40   | 0.7814          | 72.7106 |
+| 0.6856        | 0.5   | 50   | 0.7517          | 76.9231 |
+| 0.7808        | 0.6   | 60   | 0.6514          | 63.5531 |
+| 0.6826        | 0.7   | 70   | 0.6197          | 60.4396 |
+| 0.7832        | 0.8   | 80   | 0.6129          | 65.9341 |
+| 0.6031        | 0.9   | 90   | 0.5877          | 61.3553 |
+| 0.6678        | 1.0   | 100  | 0.5759          | 61.5385 |
 
 
 ### Framework versions

train.log

@@ -31,3 +31,5 @@
 {'loss': 0.6031, 'learning_rate': 1.3684210526315791e-05, 'epoch': 0.9}
 {'eval_loss': 0.5876654982566833, 'eval_wer': 61.35531135531136, 'eval_runtime': 20.4075, 'eval_samples_per_second': 3.136, 'eval_steps_per_second': 0.098, 'epoch': 0.9}
 {'loss': 0.6678, 'learning_rate': 3.1578947368421056e-06, 'epoch': 1.0}
+{'eval_loss': 0.5758526921272278, 'eval_wer': 61.53846153846154, 'eval_runtime': 19.5593, 'eval_samples_per_second': 3.272, 'eval_steps_per_second': 0.102, 'epoch': 1.0}
+{'train_runtime': 782.3972, 'train_samples_per_second': 4.09, 'train_steps_per_second': 0.128, 'train_loss': 0.6153274965286255, 'epoch': 1.0}