second epoch

README.md

@@ -1,7 +1,7 @@
 ---
 library_name: transformers
 license: mit
-base_model: microsoft/speecht5_tts
+base_model: Lingua-connect/speecht5_sw_bible
 tags:
 - generated_from_trainer
 model-index:
@@ -14,9 +14,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # speecht5_sw_bible
 
-This model is a fine-tuned version of [microsoft/speecht5_tts](https://huggingface.co/microsoft/speecht5_tts) on an unknown dataset.
+This model is a fine-tuned version of [Lingua-connect/speecht5_sw_bible](https://huggingface.co/Lingua-connect/speecht5_sw_bible) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.4821
+- Loss: 0.4446
 
 ## Model description
 
@@ -36,11 +36,9 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 1e-05
-- train_batch_size: 4
+- train_batch_size: 16
 - eval_batch_size: 2
 - seed: 42
-- gradient_accumulation_steps: 8
-- total_train_batch_size: 32
 - optimizer: Use OptimizerNames.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
@@ -51,12 +49,12 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 0.5366        | 1.0   | 649  | 0.4821          |
+| 0.467         | 1.0   | 1298 | 0.4446          |
 
 
 ### Framework versions
 
-- Transformers 4.50.2
+- Transformers 4.49.0
 - Pytorch 2.6.0+cu124
 - Datasets 3.5.0
 - Tokenizers 0.21.1

generation_config.json

@@ -5,5 +5,5 @@
   "eos_token_id": 2,
   "max_length": 1876,
   "pad_token_id": 1,
-  "transformers_version": "4.50.2"
+  "transformers_version": "4.49.0"
 }