LeVy4/whisper-small-v3

README.md

@@ -4,6 +4,8 @@ license: apache-2.0
 base_model: openai/whisper-small
 tags:
 - generated_from_trainer
+metrics:
+- wer
 model-index:
 - name: whisper-small-vi
   results: []
@@ -16,13 +18,8 @@ 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 an unknown dataset.
 It achieves the following results on the evaluation set:
-- eval_loss: 0.2205
-- eval_wer: 8.8652
-- eval_runtime: 20.9686
-- eval_samples_per_second: 2.48
-- eval_steps_per_second: 0.334
-- epoch: 38.1053
-- step: 724
+- Loss: 0.0855
+- Wer: 14.0000
 
 ## Model description
 
@@ -45,15 +42,31 @@ The following hyperparameters were used during training:
 - train_batch_size: 16
 - eval_batch_size: 8
 - seed: 42
-- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- optimizer: Use 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: 200
 - training_steps: 1000
 - mixed_precision_training: Native AMP
 
+### Training results
+
+| Training Loss | Epoch   | Step | Validation Loss | Wer     |
+|:-------------:|:-------:|:----:|:---------------:|:-------:|
+| 1.6255        | 5.5556  | 100  | 1.6181          | 67.0    |
+| 0.0811        | 11.1111 | 200  | 0.1741          | 24.0    |
+| 0.0031        | 16.6667 | 300  | 0.0962          | 17.0    |
+| 0.0007        | 22.2222 | 400  | 0.0863          | 15.0    |
+| 0.0004        | 27.7778 | 500  | 0.0860          | 15.0    |
+| 0.0002        | 33.3333 | 600  | 0.0855          | 14.0000 |
+| 0.0002        | 38.8889 | 700  | 0.0851          | 14.0000 |
+| 0.0002        | 44.4444 | 800  | 0.0854          | 14.0000 |
+| 0.0002        | 50.0    | 900  | 0.0855          | 14.0000 |
+| 0.0002        | 55.5556 | 1000 | 0.0855          | 14.0000 |
+
+
 ### Framework versions
 
-- Transformers 4.45.2
-- Pytorch 2.4.1+cu121
-- Datasets 3.0.1
+- Transformers 4.46.1
+- Pytorch 2.5.0+cu121
+- Datasets 3.1.0
 - Tokenizers 0.20.1

generation_config.json

@@ -250,5 +250,5 @@
     "transcribe": 50359,
     "translate": 50358
   },
-  "transformers_version": "4.45.2"
+  "transformers_version": "4.46.1"
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:719c5e9593555466fa21ba0b2375b6c2cc657ebcfa58767e59ebd431294db984
+oid sha256:3a6b62e3bb6d5512028519051149f6f750864f8d24de838bdda7db8c3b88851b
 size 966995080

runs/Nov03_04-30-43_f914acea8591/events.out.tfevents.1730608300.f914acea8591.270.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8d5546cc2f23fb009678159de00651d00db56cf3ef96df2a8ffee097b7c7ad66
-size 31001
+oid sha256:f8ebc79d8c0b95bef1997e6a2aba661c5e53ca70d69bd5d8106abb53c02b7a48
+size 31355

runs/Nov03_04-30-43_f914acea8591/events.out.tfevents.1730612898.f914acea8591.270.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc2fae5e7845348baae10140ec27c85ecc9823d72b273345ed75b26cdd9c70b6
+size 406