End of training
Browse files- README.md +82 -0
- adapter_model.safetensors +1 -1
README.md
ADDED
|
@@ -0,0 +1,82 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
language:
|
| 4 |
+
- fr
|
| 5 |
+
license: apache-2.0
|
| 6 |
+
base_model: openai/whisper-small
|
| 7 |
+
tags:
|
| 8 |
+
- base_model:adapter:openai/whisper-small
|
| 9 |
+
- lora
|
| 10 |
+
- transformers
|
| 11 |
+
metrics:
|
| 12 |
+
- wer
|
| 13 |
+
model-index:
|
| 14 |
+
- name: 52Hz Small Fr - IMT Atlantique X 52 Hertz
|
| 15 |
+
results: []
|
| 16 |
+
---
|
| 17 |
+
|
| 18 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 19 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 20 |
+
|
| 21 |
+
# 52Hz Small Fr - IMT Atlantique X 52 Hertz
|
| 22 |
+
|
| 23 |
+
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Premier dataset organisé de 52 Hertz dataset.
|
| 24 |
+
It achieves the following results on the evaluation set:
|
| 25 |
+
- Loss: 0.4837
|
| 26 |
+
- Wer: 38.6555
|
| 27 |
+
|
| 28 |
+
## Model description
|
| 29 |
+
|
| 30 |
+
More information needed
|
| 31 |
+
|
| 32 |
+
## Intended uses & limitations
|
| 33 |
+
|
| 34 |
+
More information needed
|
| 35 |
+
|
| 36 |
+
## Training and evaluation data
|
| 37 |
+
|
| 38 |
+
More information needed
|
| 39 |
+
|
| 40 |
+
## Training procedure
|
| 41 |
+
|
| 42 |
+
### Training hyperparameters
|
| 43 |
+
|
| 44 |
+
The following hyperparameters were used during training:
|
| 45 |
+
- learning_rate: 0.0001
|
| 46 |
+
- train_batch_size: 16
|
| 47 |
+
- eval_batch_size: 8
|
| 48 |
+
- seed: 42
|
| 49 |
+
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 50 |
+
- lr_scheduler_type: cosine
|
| 51 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 52 |
+
- num_epochs: 15
|
| 53 |
+
- mixed_precision_training: Native AMP
|
| 54 |
+
|
| 55 |
+
### Training results
|
| 56 |
+
|
| 57 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 58 |
+
|:-------------:|:-----:|:----:|:---------------:|:-------:|
|
| 59 |
+
| 1.2749 | 1.0 | 23 | 1.1067 | 61.7647 |
|
| 60 |
+
| 0.6381 | 2.0 | 46 | 0.5765 | 43.5574 |
|
| 61 |
+
| 0.3656 | 3.0 | 69 | 0.4715 | 40.4762 |
|
| 62 |
+
| 0.2315 | 4.0 | 92 | 0.4486 | 39.6359 |
|
| 63 |
+
| 0.1314 | 5.0 | 115 | 0.4368 | 38.9356 |
|
| 64 |
+
| 0.0851 | 6.0 | 138 | 0.4391 | 40.8964 |
|
| 65 |
+
| 0.0555 | 7.0 | 161 | 0.4716 | 41.1765 |
|
| 66 |
+
| 0.0378 | 8.0 | 184 | 0.4584 | 39.7759 |
|
| 67 |
+
| 0.0256 | 9.0 | 207 | 0.4799 | 41.1765 |
|
| 68 |
+
| 0.0183 | 10.0 | 230 | 0.4760 | 41.4566 |
|
| 69 |
+
| 0.0105 | 11.0 | 253 | 0.4782 | 38.7955 |
|
| 70 |
+
| 0.0101 | 12.0 | 276 | 0.4826 | 38.9356 |
|
| 71 |
+
| 0.0083 | 13.0 | 299 | 0.4835 | 38.6555 |
|
| 72 |
+
| 0.0094 | 14.0 | 322 | 0.4836 | 38.6555 |
|
| 73 |
+
| 0.0078 | 15.0 | 345 | 0.4837 | 38.6555 |
|
| 74 |
+
|
| 75 |
+
|
| 76 |
+
### Framework versions
|
| 77 |
+
|
| 78 |
+
- PEFT 0.18.1
|
| 79 |
+
- Transformers 4.57.3
|
| 80 |
+
- Pytorch 2.9.1+cu130
|
| 81 |
+
- Datasets 4.4.2
|
| 82 |
+
- Tokenizers 0.22.2
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 51957816
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b5d48a44a2b9861a6a345a79c353811f19d41a48c8c6498c83baf0c7ea1c7c8b
|
| 3 |
size 51957816
|