End of training
Browse files
README.md
CHANGED
|
@@ -8,9 +8,22 @@ tags:
|
|
| 8 |
- generated_from_trainer
|
| 9 |
datasets:
|
| 10 |
- AshtonLKY/Whisper_ASR_ATC
|
|
|
|
|
|
|
| 11 |
model-index:
|
| 12 |
- name: Whisper_ASR_ATC
|
| 13 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 14 |
---
|
| 15 |
|
| 16 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -20,13 +33,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 20 |
|
| 21 |
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the AshtonLKY/augmented_audio dataset.
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
-
-
|
| 24 |
-
-
|
| 25 |
-
- eval_runtime: 7410.9891
|
| 26 |
-
- eval_samples_per_second: 1.813
|
| 27 |
-
- eval_steps_per_second: 0.227
|
| 28 |
-
- epoch: 0.89
|
| 29 |
-
- step: 3000
|
| 30 |
|
| 31 |
## Model description
|
| 32 |
|
|
@@ -52,9 +60,25 @@ The following hyperparameters were used during training:
|
|
| 52 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 53 |
- lr_scheduler_type: linear
|
| 54 |
- lr_scheduler_warmup_steps: 500
|
| 55 |
-
- training_steps:
|
| 56 |
- mixed_precision_training: Native AMP
|
| 57 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 58 |
### Framework versions
|
| 59 |
|
| 60 |
- Transformers 4.36.2
|
|
|
|
| 8 |
- generated_from_trainer
|
| 9 |
datasets:
|
| 10 |
- AshtonLKY/Whisper_ASR_ATC
|
| 11 |
+
metrics:
|
| 12 |
+
- wer
|
| 13 |
model-index:
|
| 14 |
- name: Whisper_ASR_ATC
|
| 15 |
+
results:
|
| 16 |
+
- task:
|
| 17 |
+
name: Automatic Speech Recognition
|
| 18 |
+
type: automatic-speech-recognition
|
| 19 |
+
dataset:
|
| 20 |
+
name: AshtonLKY/augmented_audio
|
| 21 |
+
type: AshtonLKY/Whisper_ASR_ATC
|
| 22 |
+
args: 'split: test'
|
| 23 |
+
metrics:
|
| 24 |
+
- name: Wer
|
| 25 |
+
type: wer
|
| 26 |
+
value: 10.259091588129461
|
| 27 |
---
|
| 28 |
|
| 29 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 33 |
|
| 34 |
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the AshtonLKY/augmented_audio dataset.
|
| 35 |
It achieves the following results on the evaluation set:
|
| 36 |
+
- Loss: 0.0171
|
| 37 |
+
- Wer: 10.2591
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 38 |
|
| 39 |
## Model description
|
| 40 |
|
|
|
|
| 60 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 61 |
- lr_scheduler_type: linear
|
| 62 |
- lr_scheduler_warmup_steps: 500
|
| 63 |
+
- training_steps: 10000
|
| 64 |
- mixed_precision_training: Native AMP
|
| 65 |
|
| 66 |
+
### Training results
|
| 67 |
+
|
| 68 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 69 |
+
|:-------------:|:-----:|:-----:|:---------------:|:-------:|
|
| 70 |
+
| 0.2282 | 0.3 | 1000 | 0.2253 | 49.5224 |
|
| 71 |
+
| 0.1461 | 0.6 | 2000 | 0.1456 | 42.3271 |
|
| 72 |
+
| 0.1052 | 0.89 | 3000 | 0.1061 | 10.8325 |
|
| 73 |
+
| 0.0698 | 1.19 | 4000 | 0.0708 | 13.8258 |
|
| 74 |
+
| 0.043 | 1.49 | 5000 | 0.0537 | 11.0072 |
|
| 75 |
+
| 0.0407 | 1.79 | 6000 | 0.0383 | 10.9401 |
|
| 76 |
+
| 0.019 | 2.08 | 7000 | 0.0349 | 15.2078 |
|
| 77 |
+
| 0.0323 | 2.38 | 8000 | 0.0268 | 11.4068 |
|
| 78 |
+
| 0.0164 | 2.68 | 9000 | 0.0236 | 12.3902 |
|
| 79 |
+
| 0.0153 | 2.98 | 10000 | 0.0171 | 10.2591 |
|
| 80 |
+
|
| 81 |
+
|
| 82 |
### Framework versions
|
| 83 |
|
| 84 |
- Transformers 4.36.2
|