speaker-segmentation-fine-tuned-callhome-eng
This model is a fine-tuned version of pyannote/segmentation-3.0 on the diarizers-community/callhome eng dataset. It achieves the following results on the evaluation set:
- Loss: 0.6828
- Der: 0.1936
- False Alarm: 0.0736
- Missed Detection: 0.0727
- Confusion: 0.0473
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.001
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- num_epochs: 100.0
Training results
Training Loss | Epoch | Step | Validation Loss | Der | False Alarm | Missed Detection | Confusion |
---|---|---|---|---|---|---|---|
0.445 | 1.0 | 181 | 0.4932 | 0.1950 | 0.0594 | 0.0752 | 0.0604 |
0.4152 | 2.0 | 362 | 0.4827 | 0.1951 | 0.0659 | 0.0733 | 0.0560 |
0.4013 | 3.0 | 543 | 0.4607 | 0.1857 | 0.0529 | 0.0787 | 0.0542 |
0.3755 | 4.0 | 724 | 0.4658 | 0.1870 | 0.0555 | 0.0764 | 0.0551 |
0.3704 | 5.0 | 905 | 0.4640 | 0.1822 | 0.0582 | 0.0715 | 0.0526 |
0.3563 | 6.0 | 1086 | 0.4640 | 0.1828 | 0.0599 | 0.0701 | 0.0527 |
0.3503 | 7.0 | 1267 | 0.4524 | 0.1794 | 0.0543 | 0.0753 | 0.0498 |
0.3419 | 8.0 | 1448 | 0.4559 | 0.1794 | 0.0554 | 0.0748 | 0.0492 |
0.3349 | 9.0 | 1629 | 0.4705 | 0.1832 | 0.0531 | 0.0784 | 0.0517 |
0.3338 | 10.0 | 1810 | 0.4697 | 0.1822 | 0.0559 | 0.0732 | 0.0531 |
0.3218 | 11.0 | 1991 | 0.4734 | 0.1841 | 0.0652 | 0.0675 | 0.0513 |
0.3182 | 12.0 | 2172 | 0.4732 | 0.1832 | 0.0556 | 0.0754 | 0.0521 |
0.3128 | 13.0 | 2353 | 0.4809 | 0.1837 | 0.0565 | 0.0737 | 0.0535 |
0.3045 | 14.0 | 2534 | 0.4794 | 0.1828 | 0.0571 | 0.0738 | 0.0519 |
0.2969 | 15.0 | 2715 | 0.4890 | 0.1858 | 0.0617 | 0.0751 | 0.0490 |
0.296 | 16.0 | 2896 | 0.4934 | 0.1851 | 0.0640 | 0.0713 | 0.0499 |
0.2885 | 17.0 | 3077 | 0.4840 | 0.1823 | 0.0570 | 0.0752 | 0.0500 |
0.2849 | 18.0 | 3258 | 0.4991 | 0.1870 | 0.0658 | 0.0702 | 0.0509 |
0.2793 | 19.0 | 3439 | 0.4979 | 0.1862 | 0.0633 | 0.0744 | 0.0485 |
0.2763 | 20.0 | 3620 | 0.4953 | 0.1888 | 0.0682 | 0.0714 | 0.0492 |
0.2698 | 21.0 | 3801 | 0.5067 | 0.1835 | 0.0613 | 0.0717 | 0.0505 |
0.2662 | 22.0 | 3982 | 0.4984 | 0.1861 | 0.0681 | 0.0704 | 0.0476 |
0.264 | 23.0 | 4163 | 0.5057 | 0.1862 | 0.0675 | 0.0687 | 0.0500 |
0.2551 | 24.0 | 4344 | 0.5099 | 0.1867 | 0.0662 | 0.0721 | 0.0483 |
0.2597 | 25.0 | 4525 | 0.5166 | 0.1898 | 0.0714 | 0.0676 | 0.0508 |
0.2531 | 26.0 | 4706 | 0.5139 | 0.1885 | 0.0678 | 0.0699 | 0.0508 |
0.2503 | 27.0 | 4887 | 0.5218 | 0.1882 | 0.0667 | 0.0732 | 0.0484 |
0.2446 | 28.0 | 5068 | 0.5182 | 0.1864 | 0.0676 | 0.0709 | 0.0478 |
0.2465 | 29.0 | 5249 | 0.5300 | 0.1877 | 0.0722 | 0.0673 | 0.0482 |
0.2435 | 30.0 | 5430 | 0.5412 | 0.1912 | 0.0757 | 0.0665 | 0.0490 |
0.2395 | 31.0 | 5611 | 0.5311 | 0.1872 | 0.0749 | 0.0667 | 0.0455 |
0.2329 | 32.0 | 5792 | 0.5291 | 0.1871 | 0.0735 | 0.0672 | 0.0464 |
0.2318 | 33.0 | 5973 | 0.5370 | 0.1853 | 0.0713 | 0.0676 | 0.0464 |
0.2331 | 34.0 | 6154 | 0.5558 | 0.1926 | 0.0756 | 0.0684 | 0.0486 |
0.2287 | 35.0 | 6335 | 0.5454 | 0.1864 | 0.0687 | 0.0708 | 0.0469 |
0.2272 | 36.0 | 6516 | 0.5522 | 0.1872 | 0.0673 | 0.0722 | 0.0476 |
0.223 | 37.0 | 6697 | 0.5523 | 0.1903 | 0.0714 | 0.0701 | 0.0488 |
0.2207 | 38.0 | 6878 | 0.5738 | 0.1939 | 0.0739 | 0.0693 | 0.0506 |
0.2192 | 39.0 | 7059 | 0.5616 | 0.1885 | 0.0736 | 0.0680 | 0.0469 |
0.2169 | 40.0 | 7240 | 0.5645 | 0.1876 | 0.0674 | 0.0722 | 0.0480 |
0.2138 | 41.0 | 7421 | 0.5862 | 0.1909 | 0.0693 | 0.0728 | 0.0487 |
0.2112 | 42.0 | 7602 | 0.5898 | 0.1926 | 0.0694 | 0.0735 | 0.0497 |
0.2101 | 43.0 | 7783 | 0.5871 | 0.1956 | 0.0784 | 0.0670 | 0.0502 |
0.2085 | 44.0 | 7964 | 0.5755 | 0.1927 | 0.0757 | 0.0690 | 0.0480 |
0.2073 | 45.0 | 8145 | 0.5843 | 0.1909 | 0.0733 | 0.0704 | 0.0472 |
0.2049 | 46.0 | 8326 | 0.6058 | 0.1906 | 0.0716 | 0.0719 | 0.0471 |
0.2041 | 47.0 | 8507 | 0.6067 | 0.1920 | 0.0723 | 0.0721 | 0.0477 |
0.2009 | 48.0 | 8688 | 0.6101 | 0.1965 | 0.0769 | 0.0705 | 0.0491 |
0.1999 | 49.0 | 8869 | 0.6117 | 0.1921 | 0.0715 | 0.0728 | 0.0478 |
0.2001 | 50.0 | 9050 | 0.6204 | 0.1970 | 0.0770 | 0.0701 | 0.0499 |
0.2 | 51.0 | 9231 | 0.6298 | 0.1951 | 0.0732 | 0.0731 | 0.0488 |
0.1992 | 52.0 | 9412 | 0.6010 | 0.1908 | 0.0725 | 0.0706 | 0.0477 |
0.1953 | 53.0 | 9593 | 0.6342 | 0.1984 | 0.0766 | 0.0713 | 0.0505 |
0.1926 | 54.0 | 9774 | 0.6243 | 0.1935 | 0.0742 | 0.0721 | 0.0472 |
0.19 | 55.0 | 9955 | 0.6162 | 0.1937 | 0.0728 | 0.0721 | 0.0488 |
0.191 | 56.0 | 10136 | 0.6258 | 0.1930 | 0.0718 | 0.0727 | 0.0485 |
0.1898 | 57.0 | 10317 | 0.6357 | 0.1947 | 0.0764 | 0.0707 | 0.0476 |
0.1871 | 58.0 | 10498 | 0.6335 | 0.1927 | 0.0759 | 0.0698 | 0.0470 |
0.1875 | 59.0 | 10679 | 0.6386 | 0.1942 | 0.0739 | 0.0729 | 0.0474 |
0.1876 | 60.0 | 10860 | 0.6346 | 0.1916 | 0.0697 | 0.0735 | 0.0484 |
0.1838 | 61.0 | 11041 | 0.6446 | 0.1936 | 0.0735 | 0.0719 | 0.0482 |
0.1841 | 62.0 | 11222 | 0.6449 | 0.1927 | 0.0748 | 0.0709 | 0.0470 |
0.1827 | 63.0 | 11403 | 0.6476 | 0.1946 | 0.0747 | 0.0721 | 0.0478 |
0.1829 | 64.0 | 11584 | 0.6427 | 0.1914 | 0.0698 | 0.0741 | 0.0476 |
0.181 | 65.0 | 11765 | 0.6463 | 0.1957 | 0.0768 | 0.0713 | 0.0475 |
0.1788 | 66.0 | 11946 | 0.6591 | 0.1943 | 0.0756 | 0.0711 | 0.0476 |
0.1803 | 67.0 | 12127 | 0.6513 | 0.1947 | 0.0746 | 0.0732 | 0.0469 |
0.1789 | 68.0 | 12308 | 0.6617 | 0.1932 | 0.0723 | 0.0740 | 0.0470 |
0.1781 | 69.0 | 12489 | 0.6620 | 0.1934 | 0.0756 | 0.0709 | 0.0468 |
0.1797 | 70.0 | 12670 | 0.6597 | 0.1934 | 0.0739 | 0.0720 | 0.0475 |
0.1773 | 71.0 | 12851 | 0.6671 | 0.1936 | 0.0730 | 0.0728 | 0.0478 |
0.177 | 72.0 | 13032 | 0.6566 | 0.1936 | 0.0746 | 0.0723 | 0.0467 |
0.1741 | 73.0 | 13213 | 0.6712 | 0.1958 | 0.0758 | 0.0718 | 0.0481 |
0.1741 | 74.0 | 13394 | 0.6594 | 0.1921 | 0.0711 | 0.0735 | 0.0475 |
0.1737 | 75.0 | 13575 | 0.6698 | 0.1945 | 0.0739 | 0.0729 | 0.0477 |
0.1737 | 76.0 | 13756 | 0.6729 | 0.1942 | 0.0748 | 0.0723 | 0.0471 |
0.174 | 77.0 | 13937 | 0.6677 | 0.1934 | 0.0735 | 0.0722 | 0.0477 |
0.173 | 78.0 | 14118 | 0.6741 | 0.1944 | 0.0745 | 0.0724 | 0.0474 |
0.1734 | 79.0 | 14299 | 0.6713 | 0.1940 | 0.0746 | 0.0720 | 0.0474 |
0.1716 | 80.0 | 14480 | 0.6789 | 0.1946 | 0.0742 | 0.0723 | 0.0481 |
0.1709 | 81.0 | 14661 | 0.6779 | 0.1954 | 0.0751 | 0.0721 | 0.0483 |
0.1698 | 82.0 | 14842 | 0.6818 | 0.1948 | 0.0746 | 0.0725 | 0.0478 |
0.1725 | 83.0 | 15023 | 0.6802 | 0.1939 | 0.0736 | 0.0732 | 0.0471 |
0.1714 | 84.0 | 15204 | 0.6777 | 0.1928 | 0.0729 | 0.0731 | 0.0468 |
0.1698 | 85.0 | 15385 | 0.6818 | 0.1941 | 0.0756 | 0.0715 | 0.0470 |
0.1713 | 86.0 | 15566 | 0.6783 | 0.1932 | 0.0731 | 0.0734 | 0.0468 |
0.1707 | 87.0 | 15747 | 0.6818 | 0.1941 | 0.0742 | 0.0725 | 0.0474 |
0.1687 | 88.0 | 15928 | 0.6802 | 0.1944 | 0.0742 | 0.0725 | 0.0477 |
0.1717 | 89.0 | 16109 | 0.6814 | 0.1937 | 0.0733 | 0.0730 | 0.0474 |
0.169 | 90.0 | 16290 | 0.6789 | 0.1937 | 0.0729 | 0.0733 | 0.0474 |
0.1713 | 91.0 | 16471 | 0.6811 | 0.1936 | 0.0731 | 0.0729 | 0.0475 |
0.1682 | 92.0 | 16652 | 0.6802 | 0.1938 | 0.0735 | 0.0727 | 0.0476 |
0.1705 | 93.0 | 16833 | 0.6813 | 0.1936 | 0.0735 | 0.0727 | 0.0475 |
0.1679 | 94.0 | 17014 | 0.6818 | 0.1937 | 0.0736 | 0.0725 | 0.0475 |
0.1692 | 95.0 | 17195 | 0.6831 | 0.1940 | 0.0740 | 0.0724 | 0.0475 |
0.1675 | 96.0 | 17376 | 0.6835 | 0.1938 | 0.0737 | 0.0727 | 0.0474 |
0.168 | 97.0 | 17557 | 0.6830 | 0.1937 | 0.0736 | 0.0726 | 0.0474 |
0.1702 | 98.0 | 17738 | 0.6827 | 0.1937 | 0.0736 | 0.0727 | 0.0474 |
0.1676 | 99.0 | 17919 | 0.6828 | 0.1936 | 0.0736 | 0.0727 | 0.0473 |
0.169 | 100.0 | 18100 | 0.6828 | 0.1936 | 0.0736 | 0.0727 | 0.0473 |
Framework versions
- Transformers 4.47.1
- Pytorch 2.3.0+cu118
- Datasets 3.2.0
- Tokenizers 0.21.0
- Downloads last month
- 121
Model tree for vaibhavchavan/speaker-segmentation-fine-tuned-callhome-eng
Base model
pyannote/segmentation-3.0