Training in progress, step 50

README.md

@@ -1,34 +1,103 @@
 ---
-base_model: openai/whisper-small
-datasets:
-- artifacts/sv-SE_mozilla-foundation_common_voice_17_0
 library_name: transformers
 license: apache-2.0
+base_model: openai/whisper-small
+tags:
+- generated_from_trainer
+datasets:
+- common_voice_17_0
+metrics:
+- wer
 model-index:
-- name: Finetuned openai/whisper-small on Swedish
+- name: whisper-small-swedish
   results:
   - task:
+      name: Automatic Speech Recognition
       type: automatic-speech-recognition
-      name: Speech-to-Text
     dataset:
-      name: Common Voice (Swedish)
-      type: common_voice
+      name: common_voice_17_0
+      type: common_voice_17_0
+      config: sv-SE
+      split: test
+      args: sv-SE
     metrics:
-    - type: wer
-      value: 20.147
+    - name: Wer
+      type: wer
+      value: 20.14691254112786
 ---
 
-# Finetuned openai/whisper-small on 12954 Swedish training audio samples from artifacts/sv-SE_mozilla-foundation_common_voice_17_0.
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# whisper-small-swedish
+
+This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the common_voice_17_0 dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.3111
+- Model Preparation Time: 0.0044
+- Wer: 20.1469
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 1e-05
+- train_batch_size: 258
+- eval_batch_size: 64
+- seed: 42
+- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: linear
+- lr_scheduler_warmup_steps: 50
+- training_steps: 1250
+- mixed_precision_training: Native AMP
+
+### Training results
 
-This model was created from the Mozilla.ai Blueprint:
-[speech-to-text-finetune](https://github.com/mozilla-ai/speech-to-text-finetune).
+| Training Loss | Epoch   | Step | Validation Loss | Model Preparation Time | Wer     |
+|:-------------:|:-------:|:----:|:---------------:|:----------------------:|:-------:|
+| 0.6188        | 0.9804  | 50   | 0.3424          | 0.0044                 | 23.1107 |
+| 0.2417        | 1.9608  | 100  | 0.3089          | 0.0044                 | 21.0881 |
+| 0.1456        | 2.9412  | 150  | 0.3038          | 0.0044                 | 20.7412 |
+| 0.0928        | 3.9216  | 200  | 0.3111          | 0.0044                 | 20.1469 |
+| 0.0554        | 4.9020  | 250  | 0.3231          | 0.0044                 | 20.3739 |
+| 0.035         | 5.8824  | 300  | 0.3367          | 0.0044                 | 20.8432 |
+| 0.0215        | 6.8627  | 350  | 0.3619          | 0.0044                 | 20.8458 |
+| 0.0143        | 7.8431  | 400  | 0.3768          | 0.0044                 | 20.9299 |
+| 0.0101        | 8.8235  | 450  | 0.3880          | 0.0044                 | 20.8509 |
+| 0.0084        | 9.8039  | 500  | 0.3960          | 0.0044                 | 20.8993 |
+| 0.0072        | 10.7843 | 550  | 0.3999          | 0.0044                 | 21.0014 |
+| 0.0059        | 11.7647 | 600  | 0.4069          | 0.0044                 | 20.8942 |
+| 0.0053        | 12.7451 | 650  | 0.4130          | 0.0044                 | 20.9656 |
+| 0.0047        | 13.7255 | 700  | 0.4177          | 0.0044                 | 20.9963 |
+| 0.0043        | 14.7059 | 750  | 0.4208          | 0.0044                 | 20.9478 |
+| 0.004         | 15.6863 | 800  | 0.4241          | 0.0044                 | 21.0371 |
+| 0.0037        | 16.6667 | 850  | 0.4265          | 0.0044                 | 21.0600 |
+| 0.0035        | 17.6471 | 900  | 0.4298          | 0.0044                 | 21.1034 |
+| 0.0034        | 18.6275 | 950  | 0.4317          | 0.0044                 | 21.0983 |
+| 0.0032        | 19.6078 | 1000 | 0.4334          | 0.0044                 | 21.1416 |
+| 0.0031        | 20.5882 | 1050 | 0.4351          | 0.0044                 | 21.1518 |
+| 0.003         | 21.5686 | 1100 | 0.4361          | 0.0044                 | 21.1748 |
+| 0.0029        | 22.5490 | 1150 | 0.4368          | 0.0044                 | 21.1620 |
+| 0.0029        | 23.5294 | 1200 | 0.4374          | 0.0044                 | 21.1799 |
+| 0.0029        | 24.5098 | 1250 | 0.4377          | 0.0044                 | 21.1722 |
 
-## Evaluation results on 5259 audio samples of Swedish:
 
-### Baseline model (before finetuning) on Swedish
-- Word Error Rate: 28.413
-- Loss: 1.066
+### Framework versions
 
-### Finetuned model (after finetuning) on Swedish
-- Word Error Rate: 20.147
-- Loss: 0.311
+- Transformers 4.49.0
+- Pytorch 2.6.0+cu124
+- Datasets 3.3.2
+- Tokenizers 0.21.0

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e85d224783b788d49f3bd261eb5521577b73bab5a9170dc0cf163e0f4fd60ee0
+oid sha256:c2959e53dcd67c9c0abaed3f0f0f9390a55f7f2c233612d728189ef24997f760
 size 966995080

runs/Mar06_11-07-10_gpu-pod/events.out.tfevents.1741252962.gpu-pod.61799.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0d658132bb792f32212cdd2d426f84ae3b7464c0a96b4093a8bb3d243f5e1447
-size 7982
+oid sha256:3d5aa607770acde88a40acacc461cfa192053e248f856692ebbbe7c1a020dc55
+size 8359

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:921a700027d7bdc4f4cdf9b4840a82854023950a8dcd2c5a54c600ece5137ea0
+oid sha256:60e7f45ccee89aa482141b136aa8d426b55c7f4b91ed263c47d2be378a7df601
 size 5560