Model save

README.md

@@ -1,11 +1,11 @@
 ---
 library_name: transformers
-language:
-- zh
 license: apache-2.0
 base_model: openai/whisper-small
 tags:
 - generated_from_trainer
+metrics:
+- wer
 model-index:
 - name: Whisper-squeezeformer-v2
   results: []
@@ -16,10 +16,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # Whisper-squeezeformer-v2
 
-This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Voice_Data_Collection_second_edition dataset.
+This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3045
-- Cer: 14.6384
+- Loss: 2.4412
+- Wer: 111.4063
 
 ## Model description
 
@@ -39,29 +39,29 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 1e-05
-- train_batch_size: 16
+- train_batch_size: 20
 - eval_batch_size: 8
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- lr_scheduler_warmup_steps: 10
-- training_steps: 100
+- lr_scheduler_warmup_steps: 2000
+- training_steps: 20000
 - mixed_precision_training: Native AMP
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss | Cer     |
-|:-------------:|:-----:|:----:|:---------------:|:-------:|
-| 0.4443        | 0.4   | 10   | 0.3512          | 16.0494 |
-| 0.2747        | 0.8   | 20   | 0.3033          | 15.5203 |
-| 0.188         | 1.2   | 30   | 0.2901          | 15.6966 |
-| 0.1052        | 1.6   | 40   | 0.2948          | 14.2857 |
-| 0.1038        | 2.0   | 50   | 0.2973          | 13.5802 |
-| 0.0394        | 2.4   | 60   | 0.2951          | 14.2857 |
-| 0.0306        | 2.8   | 70   | 0.2969          | 13.7566 |
-| 0.0241        | 3.2   | 80   | 0.3003          | 14.2857 |
-| 0.0164        | 3.6   | 90   | 0.3028          | 14.2857 |
-| 0.013         | 4.0   | 100  | 0.3045          | 14.6384 |
+| Training Loss | Epoch | Step  | Validation Loss | Wer      |
+|:-------------:|:-----:|:-----:|:---------------:|:--------:|
+| 2.8658        | 0.8   | 2000  | 2.8024          | 136.7925 |
+| 2.7568        | 1.6   | 4000  | 2.6706          | 138.5043 |
+| 2.6111        | 2.4   | 6000  | 2.5906          | 124.6691 |
+| 2.51          | 3.2   | 8000  | 2.5565          | 121.7742 |
+| 2.4191        | 4.0   | 10000 | 2.4921          | 115.4728 |
+| 2.3168        | 4.8   | 12000 | 2.4751          | 112.5951 |
+| 2.248         | 5.6   | 14000 | 2.4565          | 112.4620 |
+| 2.1859        | 6.4   | 16000 | 2.4505          | 113.1866 |
+| 2.1358        | 7.2   | 18000 | 2.4463          | 110.7977 |
+| 2.092         | 8.0   | 20000 | 2.4412          | 111.4063 |
 
 
 ### Framework versions

generation_config.json

@@ -150,7 +150,7 @@
     "<|yo|>": 50325,
     "<|zh|>": 50260
   },
-  "language": "chinese",
+  "language": "english",
   "max_initial_timestamp_index": 50,
   "max_length": 448,
   "no_timestamps_token_id": 50363,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4871e4f7ed76f80e28b5bf75a8f066ad4fc181a2bce464cf315c821e51a4c878
+oid sha256:9dcf8896e8d8169bdcb050ecf6a2920421ecfc7a9ac50a7b331be5092ce4a5ac
 size 1128344152

runs/Nov15_06-59-33_DESKTOP-Q3M97DH/events.out.tfevents.1731625184.DESKTOP-Q3M97DH.21508.1

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3e13860b321d11adfe5bdae90ed058349d2b648b4e3d25abe26fa3908d234c40
-size 11171
+oid sha256:30e2ba5d1d0fee8091d8235ce49318533d685f4e3a450dff7370519d2a1b5029
+size 11531