./outs

README.md

@@ -1,7 +1,7 @@
 ---
 library_name: transformers
-license: apache-2.0
-base_model: biodatlab/whisper-th-small-combined
+license: mit
+base_model: biodatlab/distill-whisper-th-small
 tags:
 - generated_from_trainer
 model-index:
@@ -14,10 +14,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # outs
 
-This model is a fine-tuned version of [biodatlab/whisper-th-small-combined](https://huggingface.co/biodatlab/whisper-th-small-combined) on an unknown dataset.
+This model is a fine-tuned version of [biodatlab/distill-whisper-th-small](https://huggingface.co/biodatlab/distill-whisper-th-small) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.4404
-- Cer: 22.8718
+- Loss: 0.5794
+- Cer: 22.7354
 
 ## Model description
 
@@ -37,26 +37,26 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 8e-05
-- train_batch_size: 2
+- train_batch_size: 4
 - eval_batch_size: 2
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 3.0
+- num_epochs: 6.0
 - mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Cer     |
 |:-------------:|:-----:|:----:|:---------------:|:-------:|
-| 0.3249        | 1.0   | 1073 | 0.5601          | 28.5382 |
-| 0.0853        | 2.0   | 2146 | 0.4791          | 22.7421 |
-| 0.0465        | 3.0   | 3219 | 0.4404          | 22.8718 |
+| 0.3046        | 1.0   | 537  | 0.5714          | 37.0843 |
+| 0.1192        | 2.0   | 1074 | 0.5905          | 29.1700 |
+| 0.0471        | 3.0   | 1611 | 0.5794          | 22.7354 |
 
 
 ### Framework versions
 
 - Transformers 4.44.2
 - Pytorch 2.4.1+cu121
-- Datasets 3.0.0
+- Datasets 3.0.1
 - Tokenizers 0.19.1

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "biodatlab/whisper-th-small-combined",
+  "_name_or_path": "biodatlab/distill-whisper-th-small",
   "activation_dropout": 0.0,
   "activation_function": "gelu",
   "apply_spec_augment": true,
@@ -17,7 +17,7 @@
   "decoder_attention_heads": 12,
   "decoder_ffn_dim": 3072,
   "decoder_layerdrop": 0.0,
-  "decoder_layers": 12,
+  "decoder_layers": 4,
   "decoder_start_token_id": 50258,
   "dropout": 0.0,
   "encoder_attention_heads": 12,
@@ -30,10 +30,10 @@
   "is_encoder_decoder": true,
   "mask_feature_length": 64,
   "mask_feature_min_masks": 0,
-  "mask_feature_prob": 0.1,
+  "mask_feature_prob": 1e-05,
   "mask_time_length": 10,
   "mask_time_min_masks": 2,
-  "mask_time_prob": 0.1,
+  "mask_time_prob": 1e-05,
   "max_length": 448,
   "max_source_positions": 1500,
   "max_target_positions": 448,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6cc774d0a74e4b398164d78e6d0c29cfedb6c3d7df50994411b871ff134e420f
-size 966995080
+oid sha256:cce52bbee5aaeec31aa652ec967274a497b0c3b5b4c3e60dc9d7632c97cafc0f
+size 664564920

runs/Oct08_22-05-54_ab5162ac8a56/events.out.tfevents.1728425155.ab5162ac8a56.827.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:148a8f0df00a5e6ce1a93542e02b0f75ba0d101f4e7c44684b971c667cc8c62b
+size 10372

tokenizer_config.json

@@ -12985,5 +12985,6 @@
   "processor_class": "WhisperProcessor",
   "return_attention_mask": false,
   "tokenizer_class": "WhisperTokenizer",
-  "unk_token": "<|endoftext|>"
+  "unk_token": "<|endoftext|>",
+  "use_fast": true
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:03872426129390240982662b0f722d805ebaf58e36adb4321d7f42234c9a1ae7
+oid sha256:88ddf1d9023ed0d0a44a60aab42edf6a7bb32fd900ce380a63a0d73798e2e5f5
 size 5304