Hugging Face's logo

Akchunks
/
speaker-segmentation-fine-tuned-hindi

Transformers
TensorBoard
Safetensors
Hindi
pyannet
speaker-diarization
speaker-segmentation
Generated from Trainer
Model card Files Metrics Community
Akchunks commited on
Commit
8cce44e
·
verified ·
1 Parent(s): 036b78e

End of training

Browse files
Files changed (2) hide show
  1. README.md +29 -24
  2. model.safetensors +1 -1
README.md CHANGED
@@ -11,23 +11,23 @@ tags:
11
  datasets:
12
  - Akchunks/synthetic-speaker-diarization-dataset-hindi-short
13
  model-index:
14
- - name: speaker-segmentation-fine-tuned-hindi-v2
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
- # speaker-segmentation-fine-tuned-hindi-v2
22
 
23
  This model is a fine-tuned version of [pyannote/speaker-diarization-3.1](https://huggingface.co/pyannote/speaker-diarization-3.1) on the Akchunks/synthetic-speaker-diarization-dataset-hindi-short dataset.
24
  It achieves the following results on the evaluation set:
25
- - Loss: 0.3801
26
- - Model Preparation Time: 0.0043
27
- - Der: 0.1089
28
- - False Alarm: 0.0383
29
- - Missed Detection: 0.0234
30
- - Confusion: 0.0472
31
 
32
  ## Model description
33
 
@@ -52,27 +52,32 @@ The following hyperparameters were used during training:
52
  - seed: 42
53
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
54
  - lr_scheduler_type: cosine
55
- - num_epochs: 15
56
 
57
  ### Training results
58
 
59
  | Training Loss | Epoch | Step | Validation Loss | Model Preparation Time | Der | False Alarm | Missed Detection | Confusion |
60
  |:-------------:|:-----:|:----:|:---------------:|:----------------------:|:------:|:-----------:|:----------------:|:---------:|
61
- | No log | 1.0 | 24 | 0.4602 | 0.0043 | 0.1444 | 0.0350 | 0.0256 | 0.0838 |
62
- | 0.5195 | 2.0 | 48 | 0.3554 | 0.0043 | 0.1305 | 0.0323 | 0.0243 | 0.0739 |
63
- | 0.3063 | 3.0 | 72 | 0.3681 | 0.0043 | 0.1243 | 0.0405 | 0.0252 | 0.0586 |
64
- | 0.2147 | 4.0 | 96 | 0.3841 | 0.0043 | 0.1262 | 0.0370 | 0.0253 | 0.0639 |
65
- | 0.1941 | 5.0 | 120 | 0.3978 | 0.0043 | 0.1259 | 0.0349 | 0.0245 | 0.0666 |
66
- | 0.1596 | 6.0 | 144 | 0.3656 | 0.0043 | 0.1145 | 0.0393 | 0.0232 | 0.0520 |
67
- | 0.1435 | 7.0 | 168 | 0.3477 | 0.0043 | 0.1114 | 0.0359 | 0.0248 | 0.0507 |
68
- | 0.1172 | 8.0 | 192 | 0.4034 | 0.0043 | 0.1264 | 0.0393 | 0.0229 | 0.0643 |
69
- | 0.1253 | 9.0 | 216 | 0.3766 | 0.0043 | 0.1169 | 0.0380 | 0.0240 | 0.0548 |
70
- | 0.0963 | 10.0 | 240 | 0.3694 | 0.0043 | 0.1114 | 0.0399 | 0.0231 | 0.0483 |
71
- | 0.1055 | 11.0 | 264 | 0.3746 | 0.0043 | 0.1101 | 0.0379 | 0.0238 | 0.0484 |
72
- | 0.0924 | 12.0 | 288 | 0.3782 | 0.0043 | 0.1090 | 0.0379 | 0.0235 | 0.0476 |
73
- | 0.0975 | 13.0 | 312 | 0.3780 | 0.0043 | 0.1082 | 0.0379 | 0.0236 | 0.0468 |
74
- | 0.082 | 14.0 | 336 | 0.3798 | 0.0043 | 0.1087 | 0.0382 | 0.0234 | 0.0471 |
75
- | 0.0935 | 15.0 | 360 | 0.3801 | 0.0043 | 0.1089 | 0.0383 | 0.0234 | 0.0472 |
 
 
 
 
 
76
 
77
 
78
  ### Framework versions
 
11
  datasets:
12
  - Akchunks/synthetic-speaker-diarization-dataset-hindi-short
13
  model-index:
14
+ - name: speaker-segmentation-fine-tuned-hindi-v3
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
+ # speaker-segmentation-fine-tuned-hindi-v3
22
 
23
  This model is a fine-tuned version of [pyannote/speaker-diarization-3.1](https://huggingface.co/pyannote/speaker-diarization-3.1) on the Akchunks/synthetic-speaker-diarization-dataset-hindi-short dataset.
24
  It achieves the following results on the evaluation set:
25
+ - Loss: 0.3447
26
+ - Model Preparation Time: 0.007
27
+ - Der: 0.0985
28
+ - False Alarm: 0.0375
29
+ - Missed Detection: 0.0235
30
+ - Confusion: 0.0375
31
 
32
  ## Model description
33
 
 
52
  - seed: 42
53
  - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
54
  - lr_scheduler_type: cosine
55
+ - num_epochs: 20
56
 
57
  ### Training results
58
 
59
  | Training Loss | Epoch | Step | Validation Loss | Model Preparation Time | Der | False Alarm | Missed Detection | Confusion |
60
  |:-------------:|:-----:|:----:|:---------------:|:----------------------:|:------:|:-----------:|:----------------:|:---------:|
61
+ | No log | 1.0 | 24 | 0.4600 | 0.007 | 0.1443 | 0.0349 | 0.0256 | 0.0837 |
62
+ | 0.5196 | 2.0 | 48 | 0.3562 | 0.007 | 0.1304 | 0.0325 | 0.0242 | 0.0737 |
63
+ | 0.306 | 3.0 | 72 | 0.3732 | 0.007 | 0.1251 | 0.0402 | 0.0253 | 0.0596 |
64
+ | 0.2116 | 4.0 | 96 | 0.3712 | 0.007 | 0.1265 | 0.0408 | 0.0242 | 0.0615 |
65
+ | 0.1944 | 5.0 | 120 | 0.3846 | 0.007 | 0.1223 | 0.0337 | 0.0260 | 0.0627 |
66
+ | 0.1538 | 6.0 | 144 | 0.3544 | 0.007 | 0.1191 | 0.0375 | 0.0228 | 0.0587 |
67
+ | 0.1417 | 7.0 | 168 | 0.4045 | 0.007 | 0.1213 | 0.0358 | 0.0241 | 0.0614 |
68
+ | 0.1122 | 8.0 | 192 | 0.4213 | 0.007 | 0.1267 | 0.0438 | 0.0228 | 0.0601 |
69
+ | 0.1053 | 9.0 | 216 | 0.4171 | 0.007 | 0.1178 | 0.0368 | 0.0255 | 0.0555 |
70
+ | 0.0897 | 10.0 | 240 | 0.3561 | 0.007 | 0.1142 | 0.0409 | 0.0228 | 0.0505 |
71
+ | 0.1043 | 11.0 | 264 | 0.3738 | 0.007 | 0.1122 | 0.0380 | 0.0248 | 0.0495 |
72
+ | 0.0825 | 12.0 | 288 | 0.3383 | 0.007 | 0.1025 | 0.0377 | 0.0237 | 0.0411 |
73
+ | 0.0894 | 13.0 | 312 | 0.3328 | 0.007 | 0.0995 | 0.0388 | 0.0237 | 0.0370 |
74
+ | 0.0699 | 14.0 | 336 | 0.3272 | 0.007 | 0.0988 | 0.0376 | 0.0237 | 0.0375 |
75
+ | 0.0785 | 15.0 | 360 | 0.3374 | 0.007 | 0.0991 | 0.0378 | 0.0235 | 0.0378 |
76
+ | 0.0759 | 16.0 | 384 | 0.3414 | 0.007 | 0.0978 | 0.0383 | 0.0233 | 0.0362 |
77
+ | 0.0653 | 17.0 | 408 | 0.3417 | 0.007 | 0.0973 | 0.0375 | 0.0234 | 0.0364 |
78
+ | 0.0726 | 18.0 | 432 | 0.3439 | 0.007 | 0.0981 | 0.0374 | 0.0236 | 0.0370 |
79
+ | 0.0684 | 19.0 | 456 | 0.3445 | 0.007 | 0.0984 | 0.0374 | 0.0235 | 0.0375 |
80
+ | 0.0731 | 20.0 | 480 | 0.3447 | 0.007 | 0.0985 | 0.0375 | 0.0235 | 0.0375 |
81
 
82
 
83
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c098a3b7b7577e336ba8cb9fa63a96e5f95997cdbf75d3c6a7120892c20350a3
3
  size 5899124
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:edb82510cc88d375c6a03a919b01c27c602cd7bf4eab2f140da89d3acd4e8b1d
3
  size 5899124