trained on CommonVoice 13

README.md

@@ -4,16 +4,29 @@ language:
 license: apache-2.0
 base_model: openai/whisper-tiny
 tags:
+- audio
 - automatic-speech-recognition
 - generated_from_trainer
+datasets:
+- audio
 metrics:
 - wer
 model-index:
 - name: Whisper Tiny Uzbek
-  results: []
-datasets:
-- mozilla-foundation/common_voice_13_0
-pipeline_tag: automatic-speech-recognition
+  results:
+  - task:
+      name: Automatic Speech Recognition
+      type: automatic-speech-recognition
+    dataset:
+      name: mozilla-foundation/common_voice_13_0
+      type: audio
+      config: uz
+      split: test
+      args: uz
+    metrics:
+    - name: Wer
+      type: wer
+      value: 36.79056163528213
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -23,9 +36,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the mozilla-foundation/common_voice_13_0 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3775
-- Wer Ortho: 56.3536
-- Wer: 45.8937
+- Loss: 0.2981
+- Wer Ortho: 47.7812
+- Wer: 36.7906
 
 ## Model description
 
@@ -49,21 +62,20 @@ The following hyperparameters were used during training:
 - eval_batch_size: 16
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
-- lr_scheduler_type: constant_with_warmup
+- lr_scheduler_type: cosine_with_restarts
 - lr_scheduler_warmup_steps: 50
-- num_epochs: 1
+- num_epochs: 5
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss | Wer Ortho | Wer     |
-|:-------------:|:-----:|:----:|:---------------:|:---------:|:-------:|
-| 0.6542        | 0.13  | 500  | 0.6243          | 76.5585   | 67.7862 |
-| 0.5377        | 0.27  | 1000 | 0.5227          | 68.8556   | 60.2594 |
-| 0.4573        | 0.4   | 1500 | 0.4727          | 66.7551   | 56.0715 |
-| 0.4353        | 0.53  | 2000 | 0.4380          | 62.1211   | 52.5453 |
-| 0.3907        | 0.66  | 2500 | 0.4159          | 61.1252   | 50.8035 |
-| 0.4122        | 0.8   | 3000 | 0.3897          | 58.2628   | 47.8918 |
-| 0.3698        | 0.93  | 3500 | 0.3775          | 56.3536   | 45.8937 |
+| Training Loss | Epoch | Step  | Validation Loss | Wer Ortho | Wer     |
+|:-------------:|:-----:|:-----:|:---------------:|:---------:|:-------:|
+| 0.2929        | 0.8   | 3000  | 0.3281          | 50.8851   | 40.4395 |
+| 0.2194        | 1.59  | 6000  | 0.3110          | 49.2325   | 37.9320 |
+| 0.177         | 2.39  | 9000  | 0.3003          | 47.8700   | 36.8366 |
+| 0.1574        | 3.18  | 12000 | 0.2997          | 48.2291   | 37.0491 |
+| 0.1524        | 3.98  | 15000 | 0.2958          | 47.2395   | 36.4400 |
+| 0.1455        | 4.77  | 18000 | 0.2981          | 47.7812   | 36.7906 |
 
 
 ### Framework versions
@@ -71,4 +83,4 @@ The following hyperparameters were used during training:
 - Transformers 4.35.1
 - Pytorch 2.1.0
 - Datasets 2.14.6
-- Tokenizers 0.14.1
+- Tokenizers 0.14.1

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3b2adc980edbd221cd30f2cb40caac62549c99e3eb736d422292c3c5a9328d0b
+oid sha256:32770cd750470731caac857e7f1ef4b4733716940bda6164c8c33766d0c4b987
 size 151061672

runs/Nov26_00-28-37_L67DDV9G7R/events.out.tfevents.1700987323.L67DDV9G7R.63068.3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0ed3d6d150e26d9b441182211bd8c3895ee9b53e8ccc8c4673f859c75ce0ce5
+size 127271

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:445459a2541e1ef8a2633a1a0e8fb12d8697e4b16fdf2b2d5e219d2f10599ecd
+oid sha256:d6475deeca5ff756eef35b1bf90f1a9620fcb5bb2663a5548a9310de3e805131
 size 4728