Model save

README.md

@@ -3,16 +3,9 @@ license: apache-2.0
 base_model: openai/whisper-tiny
 tags:
 - generated_from_trainer
-- fa-asr-leaderboard
-- mozilla-foundation/common_voice_3_0
-metrics:
-- wer
 model-index:
 - name: whisper-tiny-fa
   results: []
-language:
-- fa
-pipeline_tag: automatic-speech-recognition
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -20,10 +13,15 @@ should probably proofread and complete it, then remove this comment. -->
 
 # whisper-tiny-fa
 
-This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the mozilla-foundation/common_voice_3_0 fa dataset.
+This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.9080
-- Wer: 137.7028
+- eval_loss: 0.9118
+- eval_wer: 98.7467
+- eval_runtime: 617.4423
+- eval_samples_per_second: 3.406
+- eval_steps_per_second: 0.107
+- epoch: 29.08
+- step: 4100
 
 ## Model description
 
@@ -42,32 +40,21 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 1e-05
+- learning_rate: 1e-06
 - train_batch_size: 16
-- eval_batch_size: 8
+- eval_batch_size: 32
 - seed: 42
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 32
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 500
-- training_steps: 4000
-
-### Training results
-
-| Training Loss | Epoch | Step | Validation Loss | Wer      |
-|:-------------:|:-----:|:----:|:---------------:|:--------:|
-| 0.9264        | 1.77  | 500  | 0.9697          | 119.7783 |
-| 0.5599        | 3.55  | 1000 | 0.8059          | 104.1866 |
-| 0.3672        | 5.32  | 1500 | 0.7827          | 131.1082 |
-| 0.2368        | 7.09  | 2000 | 0.8048          | 135.4507 |
-| 0.2093        | 8.87  | 2500 | 0.8282          | 157.7871 |
-| 0.1309        | 10.64 | 3000 | 0.8653          | 143.5641 |
-| 0.1043        | 12.41 | 3500 | 0.8969          | 133.2448 |
-| 0.0814        | 14.18 | 4000 | 0.9080          | 137.7028 |
-
+- training_steps: 5000
+- mixed_precision_training: Native AMP
 
 ### Framework versions
 
 - Transformers 4.37.0.dev0
 - Pytorch 2.1.0+cu121
 - Datasets 2.16.0
-- Tokenizers 0.15.0
+- Tokenizers 0.15.0

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bc8a5460fbcff4221f3d64b1d5da7fdd7bb278dbc901c8051c546cd984831556
+oid sha256:ea1d4543c5084bb686a0aa776a644b62f1172625dfb032e20cd9a044c4af41db
 size 151061672

runs/Dec30_10-21-02_f385658a7ea5/events.out.tfevents.1703931663.f385658a7ea5.4248.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8ab373598f8f9ca1362cde2c1e0e87c39fbe61f7c40a3fd22c25b918a1e314d3
-size 43002
+oid sha256:d07f1e1c29935fa32b24177b318c92bf7005d26f74e6c65d1d0f164757db74f5
+size 44262