./outs

README.md

@@ -1,7 +1,7 @@
 ---
 library_name: transformers
 license: apache-2.0
-base_model: biodatlab/whisper-th-medium
+base_model: biodatlab/whisper-th-medium-combined
 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-medium](https://huggingface.co/biodatlab/whisper-th-medium) on an unknown dataset.
+This model is a fine-tuned version of [biodatlab/whisper-th-medium-combined](https://huggingface.co/biodatlab/whisper-th-medium-combined) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.1143
-- Cer: 4.7054
+- Loss: 0.0874
+- Cer: 5.5267
 
 ## Model description
 
@@ -40,18 +40,27 @@ The following hyperparameters were used during training:
 - train_batch_size: 16
 - eval_batch_size: 16
 - seed: 42
-- gradient_accumulation_steps: 4
-- total_train_batch_size: 64
+- gradient_accumulation_steps: 2.0
+- total_train_batch_size: 32.0
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
+- lr_scheduler_warmup_ratio: 0.1
 - num_epochs: 10.0
-- mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss | Cer    |
 |:-------------:|:------:|:----:|:---------------:|:------:|
-| 0.0095        | 5.9259 | 200  | 0.1143          | 4.7054 |
+| No log        | 0.9926 | 67   | 0.0899          | 4.9847 |
+| 0.2506        | 2.0    | 135  | 0.0949          | 4.8051 |
+| 0.085         | 2.9926 | 202  | 0.0822          | 4.3363 |
+| 0.085         | 4.0    | 270  | 0.0841          | 5.2042 |
+| 0.0456        | 4.9926 | 337  | 0.0896          | 9.6136 |
+| 0.0392        | 6.0    | 405  | 0.0875          | 4.7087 |
+| 0.0392        | 6.9926 | 472  | 0.0879          | 4.7320 |
+| 0.0296        | 8.0    | 540  | 0.0874          | 5.3804 |
+| 0.0276        | 8.9926 | 607  | 0.0860          | 5.2873 |
+| 0.0276        | 9.9259 | 670  | 0.0874          | 5.5267 |
 
 
 ### Framework versions

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "biodatlab/whisper-th-medium",
+  "_name_or_path": "biodatlab/whisper-th-medium-combined",
   "activation_dropout": 0.0,
   "activation_function": "gelu",
   "apply_spec_augment": true,
@@ -30,10 +30,10 @@
   "is_encoder_decoder": true,
   "mask_feature_length": 64,
   "mask_feature_min_masks": 0,
-  "mask_feature_prob": 1e-05,
+  "mask_feature_prob": 0.1,
   "mask_time_length": 10,
   "mask_time_min_masks": 2,
-  "mask_time_prob": 1e-05,
+  "mask_time_prob": 0.1,
   "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:b479c16a1bd5744691525dead15d42c9d7054c90ea2eac26808884d44634763f
+oid sha256:3f217f32202aec544ca36e34d10bf1729140e94d4984c12151c203a65251216d
 size 3055544304

runs/Oct07_11-43-27_babb89e2e784/events.out.tfevents.1728301412.babb89e2e784.2316.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1d2f89bc35e614fe06a484708037306fe7ccd9a1fff35d431d80ef394f74c0a
+size 10485

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:558331cf01557fc487451746ec6d64cda2b2ddfe1bf821572d098225c7a4bbee
+oid sha256:bc352b343ade27a40fed5ff8303139b0b86d7232406ab3b3c53ff2ccb38ea02d
 size 5304