Model save
Browse files- README.md +31 -3
- model.safetensors +1 -1
README.md
CHANGED
|
@@ -5,9 +5,24 @@ tags:
|
|
| 5 |
- generated_from_trainer
|
| 6 |
datasets:
|
| 7 |
- common_voice_11_0
|
|
|
|
|
|
|
| 8 |
model-index:
|
| 9 |
- name: wav2vec2-large-mms-1b-turkish-colab
|
| 10 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
---
|
| 12 |
|
| 13 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -16,6 +31,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
# wav2vec2-large-mms-1b-turkish-colab
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on the common_voice_11_0 dataset.
|
|
|
|
|
|
|
|
|
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
@@ -40,17 +58,27 @@ The following hyperparameters were used during training:
|
|
| 40 |
- seed: 42
|
| 41 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 42 |
- lr_scheduler_type: linear
|
| 43 |
-
- lr_scheduler_warmup_steps:
|
| 44 |
- num_epochs: 1
|
| 45 |
- mixed_precision_training: Native AMP
|
| 46 |
|
| 47 |
### Training results
|
| 48 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 49 |
|
| 50 |
|
| 51 |
### Framework versions
|
| 52 |
|
| 53 |
- Transformers 4.37.0.dev0
|
| 54 |
- Pytorch 2.1.0+cu121
|
| 55 |
-
- Datasets 2.
|
| 56 |
- Tokenizers 0.15.0
|
|
|
|
| 5 |
- generated_from_trainer
|
| 6 |
datasets:
|
| 7 |
- common_voice_11_0
|
| 8 |
+
metrics:
|
| 9 |
+
- wer
|
| 10 |
model-index:
|
| 11 |
- name: wav2vec2-large-mms-1b-turkish-colab
|
| 12 |
+
results:
|
| 13 |
+
- task:
|
| 14 |
+
name: Automatic Speech Recognition
|
| 15 |
+
type: automatic-speech-recognition
|
| 16 |
+
dataset:
|
| 17 |
+
name: common_voice_11_0
|
| 18 |
+
type: common_voice_11_0
|
| 19 |
+
config: tr
|
| 20 |
+
split: test[:10]
|
| 21 |
+
args: tr
|
| 22 |
+
metrics:
|
| 23 |
+
- name: Wer
|
| 24 |
+
type: wer
|
| 25 |
+
value: 0.42857142857142855
|
| 26 |
---
|
| 27 |
|
| 28 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 31 |
# wav2vec2-large-mms-1b-turkish-colab
|
| 32 |
|
| 33 |
This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on the common_voice_11_0 dataset.
|
| 34 |
+
It achieves the following results on the evaluation set:
|
| 35 |
+
- Loss: 0.4348
|
| 36 |
+
- Wer: 0.4286
|
| 37 |
|
| 38 |
## Model description
|
| 39 |
|
|
|
|
| 58 |
- seed: 42
|
| 59 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 60 |
- lr_scheduler_type: linear
|
| 61 |
+
- lr_scheduler_warmup_steps: 1
|
| 62 |
- num_epochs: 1
|
| 63 |
- mixed_precision_training: Native AMP
|
| 64 |
|
| 65 |
### Training results
|
| 66 |
|
| 67 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 68 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
| 69 |
+
| 1.3297 | 0.12 | 100 | 0.5375 | 0.4571 |
|
| 70 |
+
| 0.2726 | 0.25 | 200 | 0.5256 | 0.4714 |
|
| 71 |
+
| 0.265 | 0.37 | 300 | 0.4696 | 0.4571 |
|
| 72 |
+
| 0.263 | 0.49 | 400 | 0.4405 | 0.4286 |
|
| 73 |
+
| 0.2574 | 0.61 | 500 | 0.4363 | 0.4143 |
|
| 74 |
+
| 0.2517 | 0.74 | 600 | 0.4592 | 0.4286 |
|
| 75 |
+
| 0.2454 | 0.86 | 700 | 0.4445 | 0.4143 |
|
| 76 |
+
| 0.2425 | 0.98 | 800 | 0.4348 | 0.4286 |
|
| 77 |
|
| 78 |
|
| 79 |
### Framework versions
|
| 80 |
|
| 81 |
- Transformers 4.37.0.dev0
|
| 82 |
- Pytorch 2.1.0+cu121
|
| 83 |
+
- Datasets 2.16.1
|
| 84 |
- Tokenizers 0.15.0
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3858942164
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:87f764e576b8ec5fe54b44b6adb9d2d843f6eeab3e79344b9c70c08a31033fd5
|
| 3 |
size 3858942164
|