Model save
Browse files- README.md +16 -30
- configuration_xvector.py +1 -1
README.md
CHANGED
|
@@ -1,39 +1,25 @@
|
|
| 1 |
---
|
| 2 |
library_name: transformers
|
| 3 |
tags:
|
| 4 |
-
- audio-classification
|
| 5 |
- generated_from_trainer
|
| 6 |
datasets:
|
| 7 |
- voxceleb
|
| 8 |
metrics:
|
| 9 |
- accuracy
|
| 10 |
model-index:
|
| 11 |
-
- name:
|
| 12 |
-
results:
|
| 13 |
-
- task:
|
| 14 |
-
name: Audio Classification
|
| 15 |
-
type: audio-classification
|
| 16 |
-
dataset:
|
| 17 |
-
name: confit/voxceleb
|
| 18 |
-
type: voxceleb
|
| 19 |
-
config: verification
|
| 20 |
-
split: train
|
| 21 |
-
args: verification
|
| 22 |
-
metrics:
|
| 23 |
-
- name: Accuracy
|
| 24 |
-
type: accuracy
|
| 25 |
-
value: 0.9410023545240498
|
| 26 |
---
|
| 27 |
|
| 28 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 29 |
should probably proofread and complete it, then remove this comment. -->
|
| 30 |
|
| 31 |
-
#
|
| 32 |
|
| 33 |
-
This model is a fine-tuned version of [](https://huggingface.co/) on the
|
| 34 |
It achieves the following results on the evaluation set:
|
| 35 |
-
- Loss: 0.
|
| 36 |
-
- Accuracy: 0.
|
| 37 |
|
| 38 |
## Model description
|
| 39 |
|
|
@@ -66,16 +52,16 @@ The following hyperparameters were used during training:
|
|
| 66 |
|
| 67 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 68 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
| 69 |
-
| 4.
|
| 70 |
-
| 3.
|
| 71 |
-
| 2.
|
| 72 |
-
| 1.
|
| 73 |
-
| 1.
|
| 74 |
-
| 1.
|
| 75 |
-
| 1.
|
| 76 |
-
| 0.
|
| 77 |
-
| 0.
|
| 78 |
-
| 0.
|
| 79 |
|
| 80 |
|
| 81 |
### Framework versions
|
|
|
|
| 1 |
---
|
| 2 |
library_name: transformers
|
| 3 |
tags:
|
|
|
|
| 4 |
- generated_from_trainer
|
| 5 |
datasets:
|
| 6 |
- voxceleb
|
| 7 |
metrics:
|
| 8 |
- accuracy
|
| 9 |
model-index:
|
| 10 |
+
- name: xvector-voxceleb1
|
| 11 |
+
results: []
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
---
|
| 13 |
|
| 14 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 15 |
should probably proofread and complete it, then remove this comment. -->
|
| 16 |
|
| 17 |
+
# xvector-voxceleb1
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [](https://huggingface.co/) on the voxceleb dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.2981
|
| 22 |
+
- Accuracy: 0.9405
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
|
|
| 52 |
|
| 53 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 54 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
| 55 |
+
| 4.6869 | 1.0 | 523 | 4.1199 | 0.1960 |
|
| 56 |
+
| 3.2423 | 2.0 | 1046 | 2.2824 | 0.5047 |
|
| 57 |
+
| 2.4164 | 3.0 | 1569 | 1.4862 | 0.6816 |
|
| 58 |
+
| 1.8625 | 4.0 | 2092 | 0.9794 | 0.7917 |
|
| 59 |
+
| 1.5637 | 5.0 | 2615 | 0.7048 | 0.8490 |
|
| 60 |
+
| 1.265 | 6.0 | 3138 | 0.5389 | 0.8862 |
|
| 61 |
+
| 1.0888 | 7.0 | 3661 | 0.4364 | 0.9101 |
|
| 62 |
+
| 0.9296 | 8.0 | 4184 | 0.3617 | 0.9265 |
|
| 63 |
+
| 0.8066 | 9.0 | 4707 | 0.3207 | 0.9353 |
|
| 64 |
+
| 0.7675 | 10.0 | 5230 | 0.2981 | 0.9405 |
|
| 65 |
|
| 66 |
|
| 67 |
### Framework versions
|
configuration_xvector.py
CHANGED
|
@@ -158,7 +158,7 @@ class XVectorConfig(PretrainedConfig):
|
|
| 158 |
# Decoder configuration
|
| 159 |
self.emb_sizes = emb_sizes
|
| 160 |
self.pool_mode = pool_mode
|
| 161 |
-
self.angular = True if objective in ['additive_angular_margin'] else False
|
| 162 |
self.attention_channels = attention_channels
|
| 163 |
self.decoder_config = {
|
| 164 |
"feat_in": filters[-1],
|
|
|
|
| 158 |
# Decoder configuration
|
| 159 |
self.emb_sizes = emb_sizes
|
| 160 |
self.pool_mode = pool_mode
|
| 161 |
+
self.angular = True if objective in ['additive_angular_margin', 'additive_margin'] else False
|
| 162 |
self.attention_channels = attention_channels
|
| 163 |
self.decoder_config = {
|
| 164 |
"feat_in": filters[-1],
|