update model card README.md

README.md

@@ -1,6 +1,4 @@
 ---
-license: mit
-base_model: microsoft/speecht5_tts
 tags:
 - generated_from_trainer
 datasets:
@@ -15,9 +13,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # speecht5_finetuned_voxpopuli_sw
 
-This model is a fine-tuned version of [microsoft/speecht5_tts](https://huggingface.co/microsoft/speecht5_tts) on the common_voice_11_0 dataset.
-It achieves the following results on the evaluation set:
-- Loss: 0.5298
+This model was trained from scratch on the common_voice_11_0 dataset.
 
 ## Model description
 
@@ -37,25 +33,15 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 1e-05
-- train_batch_size: 4
-- eval_batch_size: 2
+- train_batch_size: 8
+- eval_batch_size: 4
 - seed: 42
-- gradient_accumulation_steps: 8
-- total_train_batch_size: 32
+- gradient_accumulation_steps: 16
+- total_train_batch_size: 128
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 500
-- training_steps: 4000
-
-### Training results
-
-| Training Loss | Epoch | Step | Validation Loss |
-|:-------------:|:-----:|:----:|:---------------:|
-| 0.6196        | 1.65  | 1000 | 0.5635          |
-| 0.5928        | 3.29  | 2000 | 0.5452          |
-| 0.5803        | 4.94  | 3000 | 0.5339          |
-| 0.574         | 6.59  | 4000 | 0.5298          |
-
+- training_steps: 19000
 
 ### Framework versions