Hugging Face's logo

objects76
/
full-rsup_eval-2.25sec-250513_1109

Transformers
Safetensors
pyannet
speaker-diarization
speaker-segmentation
Generated from Trainer
Model card Files
xet
 Community
objects76 commited on
Commit
88c06d2
·
verified ·
1 Parent(s): 67524e4

End of training

Browse files
Files changed (4) hide show
  1. README.md +130 -0
  2. config.json +23 -0
  3. model.safetensors +3 -0
  4. training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,130 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ license: mit
4
+ base_model: pyannote/segmentation-3.0
5
+ tags:
6
+ - speaker-diarization
7
+ - speaker-segmentation
8
+ - generated_from_trainer
9
+ datasets:
10
+ - objects76/rsup-eval-ja-522-250513
11
+ model-index:
12
+ - name: full-eval-ja-522-pivot
13
+ results: []
14
+ ---
15
+
16
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
17
+ should probably proofread and complete it, then remove this comment. -->
18
+
19
+ # full-eval-ja-522-pivot
20
+
21
+ This model is a fine-tuned version of [pyannote/segmentation-3.0](https://huggingface.co/pyannote/segmentation-3.0) on the objects76/rsup-eval-ja-522-250513 dataset.
22
+ It achieves the following results on the evaluation set:
23
+ - Loss: 0.7260
24
+ - Der: 0.2451
25
+ - False Alarm: 0.0501
26
+ - Missed Detection: 0.1370
27
+ - Confusion: 0.0579
28
+
29
+ ## Model description
30
+
31
+ More information needed
32
+
33
+ ## Intended uses & limitations
34
+
35
+ More information needed
36
+
37
+ ## Training and evaluation data
38
+
39
+ More information needed
40
+
41
+ ## Training procedure
42
+
43
+ ### Training hyperparameters
44
+
45
+ The following hyperparameters were used during training:
46
+ - learning_rate: 0.001
47
+ - train_batch_size: 2048
48
+ - eval_batch_size: 2048
49
+ - seed: 42
50
+ - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
51
+ - lr_scheduler_type: cosine
52
+ - num_epochs: 200
53
+
54
+ ### Training results
55
+
56
+ | Training Loss | Epoch | Step | Validation Loss | Der | False Alarm | Missed Detection | Confusion |
57
+ |:-------------:|:-----:|:----:|:---------------:|:------:|:-----------:|:----------------:|:---------:|
58
+ | No log | 1.0 | 1 | 1.3787 | 0.4863 | 0.0407 | 0.2177 | 0.2279 |
59
+ | No log | 2.0 | 2 | 1.3678 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
60
+ | No log | 3.0 | 3 | 1.3538 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
61
+ | No log | 4.0 | 4 | 1.3302 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
62
+ | No log | 5.0 | 5 | 1.2866 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
63
+ | No log | 6.0 | 6 | 1.2250 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
64
+ | No log | 7.0 | 7 | 1.1777 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
65
+ | No log | 8.0 | 8 | 1.1856 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
66
+ | No log | 9.0 | 9 | 1.2031 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
67
+ | No log | 10.0 | 10 | 1.1904 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
68
+ | No log | 11.0 | 11 | 1.1680 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
69
+ | No log | 12.0 | 12 | 1.1535 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
70
+ | No log | 13.0 | 13 | 1.1492 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
71
+ | No log | 14.0 | 14 | 1.1495 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
72
+ | No log | 15.0 | 15 | 1.1498 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
73
+ | No log | 16.0 | 16 | 1.1489 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
74
+ | No log | 17.0 | 17 | 1.1472 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
75
+ | No log | 18.0 | 18 | 1.1465 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
76
+ | No log | 19.0 | 19 | 1.1478 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
77
+ | No log | 20.0 | 20 | 1.1504 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
78
+ | No log | 21.0 | 21 | 1.1519 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
79
+ | No log | 22.0 | 22 | 1.1523 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
80
+ | No log | 23.0 | 23 | 1.1512 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
81
+ | No log | 24.0 | 24 | 1.1492 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
82
+ | 1.1857 | 25.0 | 25 | 1.1470 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
83
+ | 1.1857 | 26.0 | 26 | 1.1445 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
84
+ | 1.1857 | 27.0 | 27 | 1.1416 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
85
+ | 1.1857 | 28.0 | 28 | 1.1394 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
86
+ | 1.1857 | 29.0 | 29 | 1.1369 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
87
+ | 1.1857 | 30.0 | 30 | 1.1354 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
88
+ | 1.1857 | 31.0 | 31 | 1.1340 | 0.4526 | 0.0501 | 0.1370 | 0.2655 |
89
+ | 1.1857 | 32.0 | 32 | 1.1328 | 0.4503 | 0.0501 | 0.1370 | 0.2631 |
90
+ | 1.1857 | 33.0 | 33 | 1.0884 | 0.4307 | 0.0501 | 0.1370 | 0.2435 |
91
+ | 1.1857 | 34.0 | 34 | 1.0495 | 0.3814 | 0.0501 | 0.1370 | 0.1942 |
92
+ | 1.1857 | 35.0 | 35 | 1.0340 | 0.2475 | 0.0501 | 0.1370 | 0.0603 |
93
+ | 1.1857 | 36.0 | 36 | 1.0111 | 0.3258 | 0.0501 | 0.1370 | 0.1386 |
94
+ | 1.1857 | 37.0 | 37 | 0.9765 | 0.2803 | 0.0501 | 0.1370 | 0.0932 |
95
+ | 1.1857 | 38.0 | 38 | 0.9547 | 0.2443 | 0.0501 | 0.1370 | 0.0572 |
96
+ | 1.1857 | 39.0 | 39 | 0.9093 | 0.2428 | 0.0501 | 0.1370 | 0.0556 |
97
+ | 1.1857 | 40.0 | 40 | 0.8913 | 0.2475 | 0.0501 | 0.1370 | 0.0603 |
98
+ | 1.1857 | 41.0 | 41 | 0.8402 | 0.2412 | 0.0501 | 0.1370 | 0.0540 |
99
+ | 1.1857 | 42.0 | 42 | 0.8096 | 0.2373 | 0.0501 | 0.1370 | 0.0501 |
100
+ | 1.1857 | 43.0 | 43 | 0.7950 | 0.2404 | 0.0501 | 0.1370 | 0.0532 |
101
+ | 1.1857 | 44.0 | 44 | 0.7625 | 0.2428 | 0.0501 | 0.1370 | 0.0556 |
102
+ | 1.1857 | 45.0 | 45 | 0.7550 | 0.2420 | 0.0501 | 0.1370 | 0.0548 |
103
+ | 1.1857 | 46.0 | 46 | 0.7273 | 0.2381 | 0.0501 | 0.1370 | 0.0509 |
104
+ | 1.1857 | 47.0 | 47 | 0.7126 | 0.2396 | 0.0501 | 0.1370 | 0.0525 |
105
+ | 1.1857 | 48.0 | 48 | 0.7289 | 0.2459 | 0.0501 | 0.1370 | 0.0587 |
106
+ | 1.1857 | 49.0 | 49 | 0.6965 | 0.2381 | 0.0501 | 0.1370 | 0.0509 |
107
+ | 0.9285 | 50.0 | 50 | 0.6970 | 0.2420 | 0.0619 | 0.1323 | 0.0478 |
108
+ | 0.9285 | 51.0 | 51 | 0.7081 | 0.2388 | 0.0603 | 0.1308 | 0.0478 |
109
+ | 0.9285 | 52.0 | 52 | 0.6896 | 0.2373 | 0.0720 | 0.1261 | 0.0392 |
110
+ | 0.9285 | 53.0 | 53 | 0.6952 | 0.2412 | 0.0752 | 0.1261 | 0.0399 |
111
+ | 0.9285 | 54.0 | 54 | 0.7025 | 0.2365 | 0.0642 | 0.1284 | 0.0439 |
112
+ | 0.9285 | 55.0 | 55 | 0.6864 | 0.2341 | 0.0713 | 0.1253 | 0.0376 |
113
+ | 0.9285 | 56.0 | 56 | 0.6920 | 0.2357 | 0.0713 | 0.1261 | 0.0384 |
114
+ | 0.9285 | 57.0 | 57 | 0.7058 | 0.2396 | 0.0658 | 0.1284 | 0.0454 |
115
+ | 0.9285 | 58.0 | 58 | 0.6878 | 0.2334 | 0.0681 | 0.1284 | 0.0368 |
116
+ | 0.9285 | 59.0 | 59 | 0.7039 | 0.2388 | 0.0634 | 0.1308 | 0.0446 |
117
+ | 0.9285 | 60.0 | 60 | 0.7155 | 0.2365 | 0.0532 | 0.1339 | 0.0493 |
118
+ | 0.9285 | 61.0 | 61 | 0.7017 | 0.2373 | 0.0540 | 0.1355 | 0.0478 |
119
+ | 0.9285 | 62.0 | 62 | 0.7291 | 0.2475 | 0.0501 | 0.1370 | 0.0603 |
120
+ | 0.9285 | 63.0 | 63 | 0.7072 | 0.2381 | 0.0501 | 0.1370 | 0.0509 |
121
+ | 0.9285 | 64.0 | 64 | 0.7189 | 0.2357 | 0.0501 | 0.1370 | 0.0486 |
122
+ | 0.9285 | 65.0 | 65 | 0.7260 | 0.2451 | 0.0501 | 0.1370 | 0.0579 |
123
+
124
+
125
+ ### Framework versions
126
+
127
+ - Transformers 4.51.3
128
+ - Pytorch 2.6.0+cu124
129
+ - Datasets 3.5.0
130
+ - Tokenizers 0.21.1
config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "SegmentationModel"
4
+ ],
5
+ "chunk_duration": 2.25,
6
+ "max_speakers_per_chunk": 2,
7
+ "max_speakers_per_frame": 2,
8
+ "min_duration": null,
9
+ "model_type": "pyannet",
10
+ "sample_rate": 16000,
11
+ "sincnet": {
12
+ "frame_sec": 0.10384,
13
+ "ksize": 251,
14
+ "stride": 10
15
+ },
16
+ "torch_dtype": "float32",
17
+ "transformers_version": "4.51.3",
18
+ "warm_up": [
19
+ 0.0,
20
+ 0.0
21
+ ],
22
+ "weigh_by_cardinality": false
23
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02e53a6171a7d8182cde3e58cfae4dad46981ea21e3c12475ed9b0e05e50b7f9
3
+ size 5897736
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2265f142046980def2b87c39b99ac998ca9ef807e33122c9c188b097d26d89df
3
+ size 5368