End of training

README.md

@@ -5,9 +5,9 @@ license: apache-2.0
 library_name: peft
 tags:
 - generated_from_trainer
-base_model: openai/whisper-small
 datasets:
 - mozilla-foundation/common_voice_16_1
+base_model: openai/whisper-small
 model-index:
 - name: Whisper small Taiwanese - LoRA
   results: []
@@ -20,7 +20,7 @@ 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 Common Voice 16.1 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.9338
+- Loss: 0.9244
 
 ## Model description
 
@@ -48,22 +48,21 @@ 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_ratio: 0.1
-- num_epochs: 5
+- num_epochs: 2
 - mixed_precision_training: Native AMP
 
 ### Training results
 
-| Training Loss | Epoch | Step  | Validation Loss |
-|:-------------:|:-----:|:-----:|:---------------:|
-| 1.0115        | 1.0   | 2500  | 1.0421          |
-| 0.7221        | 3.0   | 7500  | 0.9296          |
-| 0.4991        | 5.0   | 12500 | 0.9338          |
+| Training Loss | Epoch | Step | Validation Loss |
+|:-------------:|:-----:|:----:|:---------------:|
+| 0.9794        | 1.0   | 2500 | 1.0233          |
+| 0.8197        | 2.0   | 5000 | 0.9244          |
 
 
 ### Framework versions
 
 - PEFT 0.10.0
-- Transformers 4.40.2
-- Pytorch 2.3.0+cu121
-- Datasets 2.19.1
-- Tokenizers 0.19.1
+- Transformers 4.38.2
+- Pytorch 2.2.0+cu121
+- Datasets 2.18.0
+- Tokenizers 0.15.2

runs/Aug10_17-40-39_4ad312458ad9/events.out.tfevents.1723311640.4ad312458ad9.26.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c73866722e4631569930c47043098a6b206ca581a5fbc74b3ee7a087bc291c50
-size 47895
+oid sha256:426440407fed612863268749318df5a4e56de8957ad9d4e745c377555c54f428
+size 48520