diff --git a/.ipynb_checkpoints/train-checkpoint.log b/.ipynb_checkpoints/train-checkpoint.log new file mode 100644 index 0000000000000000000000000000000000000000..492a3c12afe34188706a7c31477950d6bb6f7457 --- /dev/null +++ b/.ipynb_checkpoints/train-checkpoint.log @@ -0,0 +1,342 @@ +INFO:2025-06-09 01:00:12,153: Epoch [3/25], Step [50/3970], Mel Loss: 0.63656, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:00:55,689: Epoch [3/25], Step [100/3970], Mel Loss: 0.63674, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:01:45,433: Epoch [3/25], Step [150/3970], Mel Loss: 0.63203, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:02:37,587: Epoch [3/25], Step [200/3970], Mel Loss: 0.62929, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:03:28,825: Epoch [3/25], Step [250/3970], Mel Loss: 0.63209, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:04:18,272: Epoch [3/25], Step [300/3970], Mel Loss: 0.62710, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:05:09,751: Epoch [3/25], Step [350/3970], Mel Loss: 0.62325, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:06:00,396: Epoch [3/25], Step [400/3970], Mel Loss: 0.62540, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:06:51,713: Epoch [3/25], Step [450/3970], Mel Loss: 0.61673, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:08:07,720: Validation loss: 0.568 + + + + +INFO:2025-06-09 01:09:07,489: Epoch [4/25], Step [50/3970], Mel Loss: 0.62133, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:09:59,049: Epoch [4/25], Step [100/3970], Mel Loss: 0.61368, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:10:46,786: Epoch [4/25], Step [150/3970], Mel Loss: 0.61887, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:11:36,393: Epoch [4/25], Step [200/3970], Mel Loss: 0.61688, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:12:21,624: Epoch [4/25], Step [250/3970], Mel Loss: 0.61630, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:13:08,125: Epoch [4/25], Step [300/3970], Mel Loss: 0.61238, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:13:53,747: Epoch [4/25], Step [350/3970], Mel Loss: 0.61566, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:14:42,113: Epoch [4/25], Step [400/3970], Mel Loss: 0.61601, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:15:29,167: Epoch [4/25], Step [450/3970], Mel Loss: 0.61588, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:16:42,289: Validation loss: 0.559 + + + + +INFO:2025-06-09 01:17:33,371: Epoch [5/25], Step [50/3970], Mel Loss: 0.61289, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:18:21,238: Epoch [5/25], Step [100/3970], Mel Loss: 0.61256, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:19:08,129: Epoch [5/25], Step [150/3970], Mel Loss: 0.60756, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:19:55,306: Epoch [5/25], Step [200/3970], Mel Loss: 0.60886, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:20:38,852: Epoch [5/25], Step [250/3970], Mel Loss: 0.61364, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:21:23,920: Epoch [5/25], Step [300/3970], Mel Loss: 0.60994, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:22:13,541: Epoch [5/25], Step [350/3970], Mel Loss: 0.59860, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:22:59,673: Epoch [5/25], Step [400/3970], Mel Loss: 0.61045, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:23:48,982: Epoch [5/25], Step [450/3970], Mel Loss: 0.59750, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 +INFO:2025-06-09 01:25:03,693: Validation loss: 0.554 + + + + +INFO:2025-06-09 01:26:38,581: Epoch [6/25], Step [50/3970], Mel Loss: 0.97984, Gen Loss: 23.17753, Disc Loss: 2.41794, Mono Loss: 0.03876, S2S Loss: 2.62273, SLM Loss: 2.94417 +INFO:2025-06-09 01:28:02,840: Epoch [6/25], Step [100/3970], Mel Loss: 1.07962, Gen Loss: 9.85569, Disc Loss: 3.64044, Mono Loss: 0.04134, S2S Loss: 2.85493, SLM Loss: 2.91264 +INFO:2025-06-09 01:29:28,320: Epoch [6/25], Step [150/3970], Mel Loss: 0.86465, Gen Loss: 7.72572, Disc Loss: 3.54385, Mono Loss: 0.03521, S2S Loss: 2.93994, SLM Loss: 2.88020 +INFO:2025-06-09 01:30:52,824: Epoch [6/25], Step [200/3970], Mel Loss: 0.79431, Gen Loss: 6.71339, Disc Loss: 3.60161, Mono Loss: 0.03136, S2S Loss: 2.63649, SLM Loss: 2.84435 +INFO:2025-06-09 01:32:17,827: Epoch [6/25], Step [250/3970], Mel Loss: 0.69364, Gen Loss: 7.94329, Disc Loss: 3.65054, Mono Loss: 0.02880, S2S Loss: 2.69598, SLM Loss: 2.74597 +INFO:2025-06-09 01:33:40,309: Epoch [6/25], Step [300/3970], Mel Loss: 0.64917, Gen Loss: 6.08137, Disc Loss: 3.72214, Mono Loss: 0.02712, S2S Loss: 2.67866, SLM Loss: 2.67792 +INFO:2025-06-09 01:35:00,338: Epoch [6/25], Step [350/3970], Mel Loss: 0.61861, Gen Loss: 5.89171, Disc Loss: 3.76209, Mono Loss: 0.03497, S2S Loss: 2.41264, SLM Loss: 2.63655 +INFO:2025-06-09 01:36:24,903: Epoch [6/25], Step [400/3970], Mel Loss: 0.60979, Gen Loss: 7.66665, Disc Loss: 3.91850, Mono Loss: 0.03728, S2S Loss: 2.28397, SLM Loss: 2.58636 +INFO:2025-06-09 01:37:48,840: Epoch [6/25], Step [450/3970], Mel Loss: 0.58291, Gen Loss: 6.77857, Disc Loss: 3.68198, Mono Loss: 0.03211, S2S Loss: 2.31591, SLM Loss: 2.60305 +INFO:2025-06-09 01:39:34,991: Validation loss: 0.537 + + + + +INFO:2025-06-09 01:41:06,553: Epoch [7/25], Step [50/3970], Mel Loss: 0.56623, Gen Loss: 9.29693, Disc Loss: 3.64386, Mono Loss: 0.03481, S2S Loss: 2.34774, SLM Loss: 2.54822 +INFO:2025-06-09 01:42:30,275: Epoch [7/25], Step [100/3970], Mel Loss: 0.56485, Gen Loss: 7.86536, Disc Loss: 3.62585, Mono Loss: 0.02787, S2S Loss: 2.28482, SLM Loss: 2.35773 +INFO:2025-06-09 01:43:57,072: Epoch [7/25], Step [150/3970], Mel Loss: 0.55288, Gen Loss: 6.11425, Disc Loss: 3.71416, Mono Loss: 0.03559, S2S Loss: 2.12334, SLM Loss: 2.29561 +INFO:2025-06-09 01:45:19,320: Epoch [7/25], Step [200/3970], Mel Loss: 0.54845, Gen Loss: 8.49720, Disc Loss: 3.60499, Mono Loss: 0.03008, S2S Loss: 2.01824, SLM Loss: 2.54976 +INFO:2025-06-09 01:46:43,392: Epoch [7/25], Step [250/3970], Mel Loss: 0.54677, Gen Loss: 8.23377, Disc Loss: 3.75765, Mono Loss: 0.03607, S2S Loss: 1.90671, SLM Loss: 2.42518 +INFO:2025-06-09 01:48:06,193: Epoch [7/25], Step [300/3970], Mel Loss: 0.53016, Gen Loss: 6.69302, Disc Loss: 3.70597, Mono Loss: 0.02783, S2S Loss: 1.66317, SLM Loss: 2.40360 +INFO:2025-06-09 01:49:31,131: Epoch [7/25], Step [350/3970], Mel Loss: 0.53427, Gen Loss: 9.37859, Disc Loss: 3.71469, Mono Loss: 0.03775, S2S Loss: 1.82596, SLM Loss: 2.32477 +INFO:2025-06-09 01:50:55,050: Epoch [7/25], Step [400/3970], Mel Loss: 0.52960, Gen Loss: 7.75816, Disc Loss: 3.68042, Mono Loss: 0.03405, S2S Loss: 1.99194, SLM Loss: 2.36643 +INFO:2025-06-09 01:52:17,195: Epoch [7/25], Step [450/3970], Mel Loss: 0.54899, Gen Loss: 6.99428, Disc Loss: 3.58416, Mono Loss: 0.02682, S2S Loss: 1.87868, SLM Loss: 2.30144 +INFO:2025-06-09 01:54:03,014: Validation loss: 0.483 + + + + +INFO:2025-06-09 01:55:38,965: Epoch [8/25], Step [50/3970], Mel Loss: 0.52470, Gen Loss: 10.28693, Disc Loss: 3.63978, Mono Loss: 0.03907, S2S Loss: 1.83708, SLM Loss: 2.34067 +INFO:2025-06-09 01:57:02,183: Epoch [8/25], Step [100/3970], Mel Loss: 0.53876, Gen Loss: 7.51378, Disc Loss: 3.64213, Mono Loss: 0.03625, S2S Loss: 1.72809, SLM Loss: 2.16873 +INFO:2025-06-09 01:58:25,854: Epoch [8/25], Step [150/3970], Mel Loss: 0.52859, Gen Loss: 7.03971, Disc Loss: 3.77774, Mono Loss: 0.03803, S2S Loss: 2.06151, SLM Loss: 2.42842 +INFO:2025-06-09 01:59:48,325: Epoch [8/25], Step [200/3970], Mel Loss: 0.52193, Gen Loss: 8.06612, Disc Loss: 3.57948, Mono Loss: 0.02800, S2S Loss: 1.74277, SLM Loss: 2.34817 +INFO:2025-06-09 02:01:13,028: Epoch [8/25], Step [250/3970], Mel Loss: 0.51478, Gen Loss: 8.66409, Disc Loss: 3.49825, Mono Loss: 0.04338, S2S Loss: 1.61835, SLM Loss: 2.15467 +INFO:2025-06-09 02:02:38,248: Epoch [8/25], Step [300/3970], Mel Loss: 0.52305, Gen Loss: 6.17309, Disc Loss: 3.58619, Mono Loss: 0.02633, S2S Loss: 1.72876, SLM Loss: 2.30287 +INFO:2025-06-09 02:03:59,711: Epoch [8/25], Step [350/3970], Mel Loss: 0.52308, Gen Loss: 9.16025, Disc Loss: 3.61594, Mono Loss: 0.03729, S2S Loss: 1.62949, SLM Loss: 2.24222 +INFO:2025-06-09 02:05:25,200: Epoch [8/25], Step [400/3970], Mel Loss: 0.52249, Gen Loss: 10.42249, Disc Loss: 3.28414, Mono Loss: 0.03355, S2S Loss: 1.73984, SLM Loss: 2.41117 +INFO:2025-06-09 02:06:52,106: Epoch [8/25], Step [450/3970], Mel Loss: 0.53768, Gen Loss: 9.69683, Disc Loss: 3.94595, Mono Loss: 0.03386, S2S Loss: 1.51221, SLM Loss: 2.22714 +INFO:2025-06-09 02:08:43,461: Validation loss: 0.576 + + + + +INFO:2025-06-09 02:10:15,610: Epoch [9/25], Step [50/3970], Mel Loss: 0.54530, Gen Loss: 8.99131, Disc Loss: 3.51233, Mono Loss: 0.04022, S2S Loss: 1.83184, SLM Loss: 2.27819 +INFO:2025-06-09 02:11:40,932: Epoch [9/25], Step [100/3970], Mel Loss: 0.52112, Gen Loss: 10.46535, Disc Loss: 3.27846, Mono Loss: 0.04066, S2S Loss: 1.40977, SLM Loss: 2.23133 +INFO:2025-06-09 02:13:05,924: Epoch [9/25], Step [150/3970], Mel Loss: 0.52145, Gen Loss: 7.49124, Disc Loss: 3.61879, Mono Loss: 0.03863, S2S Loss: 1.37945, SLM Loss: 1.97726 +INFO:2025-06-09 02:14:33,751: Epoch [9/25], Step [200/3970], Mel Loss: 0.52140, Gen Loss: 9.70458, Disc Loss: 3.41580, Mono Loss: 0.02937, S2S Loss: 1.11212, SLM Loss: 2.06116 +INFO:2025-06-09 02:16:00,853: Epoch [9/25], Step [250/3970], Mel Loss: 0.51334, Gen Loss: 9.93914, Disc Loss: 3.31054, Mono Loss: 0.03239, S2S Loss: 1.65795, SLM Loss: 2.17712 +INFO:2025-06-09 02:17:34,691: Epoch [9/25], Step [300/3970], Mel Loss: 0.51978, Gen Loss: 8.67425, Disc Loss: 3.50655, Mono Loss: 0.03035, S2S Loss: 1.59474, SLM Loss: 2.07509 +INFO:2025-06-09 02:19:00,242: Epoch [9/25], Step [350/3970], Mel Loss: 0.53707, Gen Loss: 9.08727, Disc Loss: 3.54805, Mono Loss: 0.02839, S2S Loss: 1.37795, SLM Loss: 2.33604 +INFO:2025-06-09 02:20:26,229: Epoch [9/25], Step [400/3970], Mel Loss: 0.53049, Gen Loss: 10.96715, Disc Loss: 3.46047, Mono Loss: 0.01226, S2S Loss: 1.65820, SLM Loss: 2.27651 +INFO:2025-06-09 02:21:51,865: Epoch [9/25], Step [450/3970], Mel Loss: 0.54155, Gen Loss: 12.13123, Disc Loss: 3.59190, Mono Loss: 0.03659, S2S Loss: 1.53226, SLM Loss: 2.27701 +INFO:2025-06-09 02:23:43,778: Validation loss: 0.570 + + + + +INFO:2025-06-09 02:25:19,386: Epoch [10/25], Step [50/3970], Mel Loss: 0.51787, Gen Loss: 8.53281, Disc Loss: 3.29783, Mono Loss: 0.03446, S2S Loss: 1.45484, SLM Loss: 2.05988 +INFO:2025-06-09 02:26:47,838: Epoch [10/25], Step [100/3970], Mel Loss: 0.50870, Gen Loss: 10.23766, Disc Loss: 3.65731, Mono Loss: 0.02026, S2S Loss: 1.49496, SLM Loss: 2.13103 +INFO:2025-06-09 02:28:14,726: Epoch [10/25], Step [150/3970], Mel Loss: 0.54335, Gen Loss: 10.79815, Disc Loss: 3.13370, Mono Loss: 0.03967, S2S Loss: 1.45980, SLM Loss: 2.08032 +INFO:2025-06-09 02:29:40,057: Epoch [10/25], Step [200/3970], Mel Loss: 0.54070, Gen Loss: 16.04053, Disc Loss: 2.25289, Mono Loss: 0.03900, S2S Loss: 1.27599, SLM Loss: 1.92492 +INFO:2025-06-09 02:31:03,537: Epoch [10/25], Step [250/3970], Mel Loss: 0.52524, Gen Loss: 10.16288, Disc Loss: 3.41181, Mono Loss: 0.03559, S2S Loss: 1.43502, SLM Loss: 2.11398 +INFO:2025-06-09 02:32:27,638: Epoch [10/25], Step [300/3970], Mel Loss: 0.50069, Gen Loss: 10.37155, Disc Loss: 3.37344, Mono Loss: 0.02960, S2S Loss: 1.31690, SLM Loss: 2.10520 +INFO:2025-06-09 02:33:55,403: Epoch [10/25], Step [350/3970], Mel Loss: 0.51798, Gen Loss: 8.71367, Disc Loss: 3.56202, Mono Loss: 0.02943, S2S Loss: 1.46521, SLM Loss: 1.94630 +INFO:2025-06-09 02:35:20,742: Epoch [10/25], Step [400/3970], Mel Loss: 0.50403, Gen Loss: 10.07516, Disc Loss: 3.33516, Mono Loss: 0.03924, S2S Loss: 1.33150, SLM Loss: 2.00906 +INFO:2025-06-09 02:36:48,130: Epoch [10/25], Step [450/3970], Mel Loss: 0.51103, Gen Loss: 13.04639, Disc Loss: 2.76208, Mono Loss: 0.03206, S2S Loss: 1.34592, SLM Loss: 1.97846 +INFO:2025-06-09 02:38:35,829: Validation loss: 0.494 + + + + +INFO:2025-06-09 02:40:09,941: Epoch [11/25], Step [50/3970], Mel Loss: 0.50742, Gen Loss: 14.54983, Disc Loss: 3.16779, Mono Loss: 0.03301, S2S Loss: 1.24001, SLM Loss: 2.09207 +INFO:2025-06-09 02:41:31,696: Epoch [11/25], Step [100/3970], Mel Loss: 0.51317, Gen Loss: 10.19775, Disc Loss: 3.39758, Mono Loss: 0.03635, S2S Loss: 1.18542, SLM Loss: 1.87807 +INFO:2025-06-09 02:43:00,303: Epoch [11/25], Step [150/3970], Mel Loss: 0.49535, Gen Loss: 10.78111, Disc Loss: 3.42311, Mono Loss: 0.02923, S2S Loss: 1.30470, SLM Loss: 2.03685 +INFO:2025-06-09 02:44:25,379: Epoch [11/25], Step [200/3970], Mel Loss: 0.52437, Gen Loss: 12.13915, Disc Loss: 3.48997, Mono Loss: 0.03481, S2S Loss: 1.40437, SLM Loss: 2.22972 +INFO:2025-06-09 02:45:48,875: Epoch [11/25], Step [250/3970], Mel Loss: 0.52986, Gen Loss: 10.21078, Disc Loss: 3.29919, Mono Loss: 0.03813, S2S Loss: 1.31922, SLM Loss: 2.12596 +INFO:2025-06-09 02:47:12,758: Epoch [11/25], Step [300/3970], Mel Loss: 0.50854, Gen Loss: 11.45082, Disc Loss: 3.70037, Mono Loss: 0.02635, S2S Loss: 1.30175, SLM Loss: 2.10208 +INFO:2025-06-09 02:48:37,380: Epoch [11/25], Step [350/3970], Mel Loss: 0.51107, Gen Loss: 12.95888, Disc Loss: 2.67743, Mono Loss: 0.03419, S2S Loss: 1.20725, SLM Loss: 1.90229 +INFO:2025-06-09 02:50:00,179: Epoch [11/25], Step [400/3970], Mel Loss: 0.50464, Gen Loss: 14.90269, Disc Loss: 2.48575, Mono Loss: 0.02608, S2S Loss: 1.25147, SLM Loss: 2.09807 +INFO:2025-06-09 02:51:23,892: Epoch [11/25], Step [450/3970], Mel Loss: 0.50813, Gen Loss: 11.16796, Disc Loss: 3.33927, Mono Loss: 0.02247, S2S Loss: 1.23355, SLM Loss: 2.03762 +INFO:2025-06-09 02:53:11,784: Validation loss: 0.500 + + + + +INFO:2025-06-09 02:54:42,588: Epoch [12/25], Step [50/3970], Mel Loss: 0.51559, Gen Loss: 12.73433, Disc Loss: 3.10313, Mono Loss: 0.02916, S2S Loss: 1.15304, SLM Loss: 2.04169 +INFO:2025-06-09 02:56:06,203: Epoch [12/25], Step [100/3970], Mel Loss: 0.53166, Gen Loss: 15.15074, Disc Loss: 2.44015, Mono Loss: 0.02971, S2S Loss: 1.16761, SLM Loss: 2.05299 +INFO:2025-06-09 02:57:34,086: Epoch [12/25], Step [150/3970], Mel Loss: 0.52280, Gen Loss: 13.14780, Disc Loss: 2.62888, Mono Loss: 0.02973, S2S Loss: 1.36648, SLM Loss: 2.20558 +INFO:2025-06-09 02:59:02,556: Epoch [12/25], Step [200/3970], Mel Loss: 0.51851, Gen Loss: 11.34044, Disc Loss: 3.37104, Mono Loss: 0.02984, S2S Loss: 1.27682, SLM Loss: 2.09497 +INFO:2025-06-09 03:00:26,363: Epoch [12/25], Step [250/3970], Mel Loss: 0.50483, Gen Loss: 9.85288, Disc Loss: 3.57650, Mono Loss: 0.02718, S2S Loss: 1.23974, SLM Loss: 2.03210 +INFO:2025-06-09 03:01:51,385: Epoch [12/25], Step [300/3970], Mel Loss: 0.49458, Gen Loss: 11.01429, Disc Loss: 3.23517, Mono Loss: 0.03900, S2S Loss: 1.20380, SLM Loss: 2.05801 +INFO:2025-06-09 03:03:16,092: Epoch [12/25], Step [350/3970], Mel Loss: 0.52576, Gen Loss: 9.24854, Disc Loss: 3.31379, Mono Loss: 0.03915, S2S Loss: 1.13775, SLM Loss: 2.09123 +INFO:2025-06-09 03:04:46,606: Epoch [12/25], Step [400/3970], Mel Loss: 0.50221, Gen Loss: 10.17654, Disc Loss: 3.53851, Mono Loss: 0.03372, S2S Loss: 1.20534, SLM Loss: 2.00393 +INFO:2025-06-09 03:06:09,986: Epoch [12/25], Step [450/3970], Mel Loss: 0.52733, Gen Loss: 14.04951, Disc Loss: 2.97020, Mono Loss: 0.02758, S2S Loss: 1.06250, SLM Loss: 1.99839 +INFO:2025-06-09 03:07:59,553: Validation loss: 0.471 + + + + +INFO:2025-06-09 03:09:33,778: Epoch [13/25], Step [50/3970], Mel Loss: 0.49267, Gen Loss: 11.16026, Disc Loss: 3.05907, Mono Loss: 0.02503, S2S Loss: 1.00237, SLM Loss: 1.82247 +INFO:2025-06-09 03:11:01,558: Epoch [13/25], Step [100/3970], Mel Loss: 0.49195, Gen Loss: 9.90175, Disc Loss: 3.69180, Mono Loss: 0.02045, S2S Loss: 1.30512, SLM Loss: 1.89232 +INFO:2025-06-09 03:12:28,265: Epoch [13/25], Step [150/3970], Mel Loss: 0.49494, Gen Loss: 12.47827, Disc Loss: 2.95689, Mono Loss: 0.02328, S2S Loss: 1.05090, SLM Loss: 1.93738 +INFO:2025-06-09 03:13:50,919: Epoch [13/25], Step [200/3970], Mel Loss: 0.51474, Gen Loss: 11.96166, Disc Loss: 3.02671, Mono Loss: 0.03436, S2S Loss: 1.09315, SLM Loss: 1.92749 +INFO:2025-06-09 03:15:17,888: Epoch [13/25], Step [250/3970], Mel Loss: 0.51118, Gen Loss: 17.41326, Disc Loss: 2.00980, Mono Loss: 0.02579, S2S Loss: 1.27793, SLM Loss: 2.11468 +INFO:2025-06-09 03:16:42,783: Epoch [13/25], Step [300/3970], Mel Loss: 0.52890, Gen Loss: 10.31864, Disc Loss: 3.39953, Mono Loss: 0.03986, S2S Loss: 1.07551, SLM Loss: 1.85846 +INFO:2025-06-09 03:18:10,773: Epoch [13/25], Step [350/3970], Mel Loss: 0.49484, Gen Loss: 9.64177, Disc Loss: 3.54298, Mono Loss: 0.03428, S2S Loss: 1.03875, SLM Loss: 1.94741 +INFO:2025-06-09 03:19:34,116: Epoch [13/25], Step [400/3970], Mel Loss: 0.49129, Gen Loss: 15.92753, Disc Loss: 2.42741, Mono Loss: 0.03079, S2S Loss: 1.08931, SLM Loss: 1.93091 +INFO:2025-06-09 03:20:57,906: Epoch [13/25], Step [450/3970], Mel Loss: 0.50837, Gen Loss: 11.88165, Disc Loss: 3.05828, Mono Loss: 0.02700, S2S Loss: 1.02226, SLM Loss: 2.08510 +INFO:2025-06-09 03:22:45,754: Validation loss: 0.481 + + + + +INFO:2025-06-09 03:24:21,819: Epoch [14/25], Step [50/3970], Mel Loss: 0.48985, Gen Loss: 12.91563, Disc Loss: 2.76847, Mono Loss: 0.04083, S2S Loss: 1.14642, SLM Loss: 1.95628 +INFO:2025-06-09 03:25:47,789: Epoch [14/25], Step [100/3970], Mel Loss: 0.49605, Gen Loss: 12.36551, Disc Loss: 2.68910, Mono Loss: 0.02049, S2S Loss: 1.21105, SLM Loss: 1.98918 +INFO:2025-06-09 03:27:15,432: Epoch [14/25], Step [150/3970], Mel Loss: 0.50257, Gen Loss: 11.92622, Disc Loss: 3.19436, Mono Loss: 0.02369, S2S Loss: 1.07385, SLM Loss: 1.89891 +INFO:2025-06-09 03:28:39,067: Epoch [14/25], Step [200/3970], Mel Loss: 0.49241, Gen Loss: 10.28760, Disc Loss: 3.14552, Mono Loss: 0.02857, S2S Loss: 1.18241, SLM Loss: 1.92434 +INFO:2025-06-09 03:30:02,661: Epoch [14/25], Step [250/3970], Mel Loss: 0.50479, Gen Loss: 14.76222, Disc Loss: 2.38417, Mono Loss: 0.03464, S2S Loss: 1.11464, SLM Loss: 1.90056 +INFO:2025-06-09 03:31:25,232: Epoch [14/25], Step [300/3970], Mel Loss: 0.51008, Gen Loss: 10.93404, Disc Loss: 3.09978, Mono Loss: 0.02636, S2S Loss: 1.06983, SLM Loss: 2.01580 +INFO:2025-06-09 03:32:47,763: Epoch [14/25], Step [350/3970], Mel Loss: 0.49603, Gen Loss: 13.14640, Disc Loss: 2.93398, Mono Loss: 0.03515, S2S Loss: 1.13013, SLM Loss: 1.87904 +INFO:2025-06-09 03:34:09,946: Epoch [14/25], Step [400/3970], Mel Loss: 0.49311, Gen Loss: 15.16063, Disc Loss: 2.26509, Mono Loss: 0.02470, S2S Loss: 1.13779, SLM Loss: 1.98486 +INFO:2025-06-09 03:35:35,944: Epoch [14/25], Step [450/3970], Mel Loss: 0.49960, Gen Loss: 14.25554, Disc Loss: 2.94801, Mono Loss: 0.03173, S2S Loss: 1.06616, SLM Loss: 1.87259 +INFO:2025-06-09 03:37:23,594: Validation loss: 0.474 + + + + +INFO:2025-06-09 03:38:59,186: Epoch [15/25], Step [50/3970], Mel Loss: 0.49733, Gen Loss: 15.70522, Disc Loss: 2.52600, Mono Loss: 0.02580, S2S Loss: 1.16987, SLM Loss: 1.96256 +INFO:2025-06-09 03:40:28,804: Epoch [15/25], Step [100/3970], Mel Loss: 0.50295, Gen Loss: 14.09082, Disc Loss: 2.55250, Mono Loss: 0.02887, S2S Loss: 0.98603, SLM Loss: 1.95487 +INFO:2025-06-09 03:41:51,594: Epoch [15/25], Step [150/3970], Mel Loss: 0.50385, Gen Loss: 9.58022, Disc Loss: 3.57700, Mono Loss: 0.02735, S2S Loss: 1.30176, SLM Loss: 1.89019 +INFO:2025-06-09 03:43:21,769: Epoch [15/25], Step [200/3970], Mel Loss: 0.50567, Gen Loss: 13.83009, Disc Loss: 2.59786, Mono Loss: 0.02999, S2S Loss: 0.97563, SLM Loss: 1.98239 +INFO:2025-06-09 03:44:48,877: Epoch [15/25], Step [250/3970], Mel Loss: 0.50670, Gen Loss: 15.52833, Disc Loss: 2.77497, Mono Loss: 0.03401, S2S Loss: 1.08127, SLM Loss: 1.98682 +INFO:2025-06-09 03:46:14,023: Epoch [15/25], Step [300/3970], Mel Loss: 0.48451, Gen Loss: 10.18036, Disc Loss: 3.13374, Mono Loss: 0.03035, S2S Loss: 1.07857, SLM Loss: 1.76348 +INFO:2025-06-09 03:47:40,871: Epoch [15/25], Step [350/3970], Mel Loss: 0.48213, Gen Loss: 14.40100, Disc Loss: 2.57544, Mono Loss: 0.03224, S2S Loss: 1.27353, SLM Loss: 2.20880 +INFO:2025-06-09 03:49:02,418: Epoch [15/25], Step [400/3970], Mel Loss: 0.49212, Gen Loss: 12.04808, Disc Loss: 3.22365, Mono Loss: 0.04031, S2S Loss: 1.16138, SLM Loss: 1.93176 +INFO:2025-06-09 03:50:29,288: Epoch [15/25], Step [450/3970], Mel Loss: 0.49498, Gen Loss: 15.64781, Disc Loss: 2.42535, Mono Loss: 0.02771, S2S Loss: 1.12834, SLM Loss: 1.92330 +INFO:2025-06-09 03:52:16,555: Validation loss: 0.484 + + + + +INFO:2025-06-09 03:53:54,302: Epoch [16/25], Step [50/3970], Mel Loss: 0.51999, Gen Loss: 14.74118, Disc Loss: 3.12985, Mono Loss: 0.03898, S2S Loss: 1.17255, SLM Loss: 2.04765 +INFO:2025-06-09 03:55:21,475: Epoch [16/25], Step [100/3970], Mel Loss: 0.50182, Gen Loss: 12.08678, Disc Loss: 2.86320, Mono Loss: 0.03554, S2S Loss: 1.07750, SLM Loss: 1.93926 +INFO:2025-06-09 03:56:47,475: Epoch [16/25], Step [150/3970], Mel Loss: 0.50050, Gen Loss: 9.98121, Disc Loss: 3.66234, Mono Loss: 0.04147, S2S Loss: 1.05245, SLM Loss: 1.79418 +INFO:2025-06-09 03:58:18,119: Epoch [16/25], Step [200/3970], Mel Loss: 0.48289, Gen Loss: 11.81459, Disc Loss: 3.18255, Mono Loss: 0.02753, S2S Loss: 1.00464, SLM Loss: 2.07043 +INFO:2025-06-09 03:59:52,641: Epoch [16/25], Step [250/3970], Mel Loss: 0.50084, Gen Loss: 12.98343, Disc Loss: 2.73330, Mono Loss: 0.02163, S2S Loss: 1.01344, SLM Loss: 2.09205 +INFO:2025-06-09 04:01:18,443: Epoch [16/25], Step [300/3970], Mel Loss: 0.48698, Gen Loss: 10.43984, Disc Loss: 3.65968, Mono Loss: 0.02921, S2S Loss: 1.09594, SLM Loss: 1.76368 +INFO:2025-06-09 04:02:43,362: Epoch [16/25], Step [350/3970], Mel Loss: 0.49274, Gen Loss: 14.27684, Disc Loss: 2.40941, Mono Loss: 0.02796, S2S Loss: 1.00817, SLM Loss: 1.93564 +INFO:2025-06-09 04:04:11,908: Epoch [16/25], Step [400/3970], Mel Loss: 0.48388, Gen Loss: 12.83471, Disc Loss: 2.84156, Mono Loss: 0.02278, S2S Loss: 1.02146, SLM Loss: 1.84777 +INFO:2025-06-09 04:05:42,332: Epoch [16/25], Step [450/3970], Mel Loss: 0.47624, Gen Loss: 11.16790, Disc Loss: 3.20130, Mono Loss: 0.03327, S2S Loss: 1.30728, SLM Loss: 2.05759 +INFO:2025-06-09 04:07:30,779: Validation loss: 0.441 + + + + +INFO:2025-06-09 04:09:06,393: Epoch [17/25], Step [50/3970], Mel Loss: 0.47322, Gen Loss: 10.21182, Disc Loss: 3.17271, Mono Loss: 0.03857, S2S Loss: 1.05606, SLM Loss: 1.89386 +INFO:2025-06-09 04:10:36,875: Epoch [17/25], Step [100/3970], Mel Loss: 0.49200, Gen Loss: 11.82951, Disc Loss: 2.96425, Mono Loss: 0.03267, S2S Loss: 0.93129, SLM Loss: 1.86663 +INFO:2025-06-09 04:12:04,591: Epoch [17/25], Step [150/3970], Mel Loss: 0.47137, Gen Loss: 8.62027, Disc Loss: 3.83087, Mono Loss: 0.03580, S2S Loss: 0.96079, SLM Loss: 1.77226 +INFO:2025-06-09 04:13:32,190: Epoch [17/25], Step [200/3970], Mel Loss: 0.47893, Gen Loss: 10.65770, Disc Loss: 3.56187, Mono Loss: 0.02644, S2S Loss: 1.01706, SLM Loss: 2.04362 +INFO:2025-06-09 04:14:56,130: Epoch [17/25], Step [250/3970], Mel Loss: 0.48653, Gen Loss: 8.75337, Disc Loss: 3.41021, Mono Loss: 0.03018, S2S Loss: 1.21360, SLM Loss: 1.96406 +INFO:2025-06-09 04:16:22,783: Epoch [17/25], Step [300/3970], Mel Loss: 0.47912, Gen Loss: 13.35522, Disc Loss: 2.50100, Mono Loss: 0.03506, S2S Loss: 0.97357, SLM Loss: 1.86547 +INFO:2025-06-09 04:17:48,667: Epoch [17/25], Step [350/3970], Mel Loss: 0.48305, Gen Loss: 11.61836, Disc Loss: 2.81784, Mono Loss: 0.03363, S2S Loss: 1.08002, SLM Loss: 2.06560 +INFO:2025-06-09 04:19:14,326: Epoch [17/25], Step [400/3970], Mel Loss: 0.47853, Gen Loss: 12.34376, Disc Loss: 2.69610, Mono Loss: 0.02721, S2S Loss: 1.04334, SLM Loss: 1.70590 +INFO:2025-06-09 04:20:35,887: Epoch [17/25], Step [450/3970], Mel Loss: 0.49008, Gen Loss: 13.10175, Disc Loss: 2.95339, Mono Loss: 0.02577, S2S Loss: 0.94490, SLM Loss: 1.85765 +INFO:2025-06-09 04:22:27,589: Validation loss: 0.452 + + + + +INFO:2025-06-09 04:24:06,620: Epoch [18/25], Step [50/3970], Mel Loss: 0.47998, Gen Loss: 11.54100, Disc Loss: 2.96952, Mono Loss: 0.03144, S2S Loss: 1.16552, SLM Loss: 2.03827 +INFO:2025-06-09 04:25:34,179: Epoch [18/25], Step [100/3970], Mel Loss: 0.47805, Gen Loss: 13.14331, Disc Loss: 2.78030, Mono Loss: 0.03480, S2S Loss: 1.11621, SLM Loss: 1.77017 +INFO:2025-06-09 04:26:56,768: Epoch [18/25], Step [150/3970], Mel Loss: 0.47885, Gen Loss: 13.77032, Disc Loss: 2.75735, Mono Loss: 0.03735, S2S Loss: 1.01647, SLM Loss: 1.84030 +INFO:2025-06-09 04:28:22,414: Epoch [18/25], Step [200/3970], Mel Loss: 0.48841, Gen Loss: 9.09012, Disc Loss: 4.03486, Mono Loss: 0.02713, S2S Loss: 0.98568, SLM Loss: 2.00300 +INFO:2025-06-09 04:29:49,349: Epoch [18/25], Step [250/3970], Mel Loss: 0.47842, Gen Loss: 11.78235, Disc Loss: 3.07922, Mono Loss: 0.03915, S2S Loss: 1.13136, SLM Loss: 1.98326 +INFO:2025-06-09 04:31:14,759: Epoch [18/25], Step [300/3970], Mel Loss: 0.47979, Gen Loss: 11.35575, Disc Loss: 3.51251, Mono Loss: 0.02926, S2S Loss: 0.97950, SLM Loss: 1.76771 +INFO:2025-06-09 04:32:36,423: Epoch [18/25], Step [350/3970], Mel Loss: 0.47505, Gen Loss: 11.21230, Disc Loss: 3.30019, Mono Loss: 0.04272, S2S Loss: 0.82646, SLM Loss: 1.81805 +INFO:2025-06-09 04:34:01,703: Epoch [18/25], Step [400/3970], Mel Loss: 0.47216, Gen Loss: 12.97506, Disc Loss: 2.76475, Mono Loss: 0.02976, S2S Loss: 0.98103, SLM Loss: 1.91380 +INFO:2025-06-09 04:35:26,512: Epoch [18/25], Step [450/3970], Mel Loss: 0.47869, Gen Loss: 14.93690, Disc Loss: 2.70676, Mono Loss: 0.02386, S2S Loss: 0.86181, SLM Loss: 1.86082 +INFO:2025-06-09 04:37:13,388: Validation loss: 0.472 + + + + +INFO:2025-06-09 04:38:47,409: Epoch [19/25], Step [50/3970], Mel Loss: 0.50106, Gen Loss: 10.75373, Disc Loss: 3.20791, Mono Loss: 0.03283, S2S Loss: 1.05680, SLM Loss: 2.04265 +INFO:2025-06-09 04:40:11,614: Epoch [19/25], Step [100/3970], Mel Loss: 0.49229, Gen Loss: 12.93675, Disc Loss: 2.80900, Mono Loss: 0.02687, S2S Loss: 1.00646, SLM Loss: 1.95273 +INFO:2025-06-09 04:41:38,257: Epoch [19/25], Step [150/3970], Mel Loss: 0.48437, Gen Loss: 9.62462, Disc Loss: 3.65687, Mono Loss: 0.03190, S2S Loss: 1.10174, SLM Loss: 1.83353 +INFO:2025-06-09 04:43:10,165: Epoch [19/25], Step [200/3970], Mel Loss: 0.47472, Gen Loss: 14.23377, Disc Loss: 2.64775, Mono Loss: 0.03328, S2S Loss: 0.93959, SLM Loss: 1.89750 +INFO:2025-06-09 04:44:38,022: Epoch [19/25], Step [250/3970], Mel Loss: 0.47779, Gen Loss: 14.43170, Disc Loss: 2.78379, Mono Loss: 0.01906, S2S Loss: 0.99975, SLM Loss: 1.75941 +INFO:2025-06-09 04:46:03,339: Epoch [19/25], Step [300/3970], Mel Loss: 0.48885, Gen Loss: 13.73142, Disc Loss: 2.71020, Mono Loss: 0.03351, S2S Loss: 0.97011, SLM Loss: 1.83246 +INFO:2025-06-09 04:47:31,942: Epoch [19/25], Step [350/3970], Mel Loss: 0.49685, Gen Loss: 11.76912, Disc Loss: 3.36635, Mono Loss: 0.03053, S2S Loss: 1.02552, SLM Loss: 2.14654 +INFO:2025-06-09 04:49:03,189: Epoch [19/25], Step [400/3970], Mel Loss: 0.50997, Gen Loss: 12.43883, Disc Loss: 3.06039, Mono Loss: 0.02553, S2S Loss: 1.06563, SLM Loss: 2.07121 +INFO:2025-06-09 04:50:29,105: Epoch [19/25], Step [450/3970], Mel Loss: 0.47820, Gen Loss: 14.01424, Disc Loss: 2.55081, Mono Loss: 0.02975, S2S Loss: 0.90710, SLM Loss: 1.83058 +INFO:2025-06-09 04:52:19,425: Validation loss: 0.458 + + + + +INFO:2025-06-09 04:53:52,052: Epoch [20/25], Step [50/3970], Mel Loss: 0.46500, Gen Loss: 10.54346, Disc Loss: 3.31023, Mono Loss: 0.02908, S2S Loss: 0.90407, SLM Loss: 1.78864 +INFO:2025-06-09 04:55:17,797: Epoch [20/25], Step [100/3970], Mel Loss: 0.47071, Gen Loss: 12.65996, Disc Loss: 2.62974, Mono Loss: 0.03045, S2S Loss: 0.92329, SLM Loss: 1.80409 +INFO:2025-06-09 04:56:43,658: Epoch [20/25], Step [150/3970], Mel Loss: 0.50137, Gen Loss: 10.70967, Disc Loss: 3.02257, Mono Loss: 0.03576, S2S Loss: 1.06377, SLM Loss: 1.94652 +INFO:2025-06-09 04:58:10,885: Epoch [20/25], Step [200/3970], Mel Loss: 0.48577, Gen Loss: 13.93800, Disc Loss: 2.42582, Mono Loss: 0.03042, S2S Loss: 1.00638, SLM Loss: 1.94606 +INFO:2025-06-09 04:59:36,042: Epoch [20/25], Step [250/3970], Mel Loss: 0.47842, Gen Loss: 11.98050, Disc Loss: 2.85478, Mono Loss: 0.03991, S2S Loss: 0.88393, SLM Loss: 1.78679 +INFO:2025-06-09 05:01:05,055: Epoch [20/25], Step [300/3970], Mel Loss: 0.47698, Gen Loss: 14.17083, Disc Loss: 2.71589, Mono Loss: 0.03090, S2S Loss: 0.95709, SLM Loss: 1.76501 +INFO:2025-06-09 05:02:31,554: Epoch [20/25], Step [350/3970], Mel Loss: 0.47476, Gen Loss: 12.08669, Disc Loss: 2.79200, Mono Loss: 0.03368, S2S Loss: 1.02864, SLM Loss: 1.86428 +INFO:2025-06-09 05:03:58,164: Epoch [20/25], Step [400/3970], Mel Loss: 0.47778, Gen Loss: 16.65605, Disc Loss: 1.92339, Mono Loss: 0.02666, S2S Loss: 1.15662, SLM Loss: 1.86039 +INFO:2025-06-09 05:05:20,371: Epoch [20/25], Step [450/3970], Mel Loss: 0.47590, Gen Loss: 10.66579, Disc Loss: 2.86761, Mono Loss: 0.02992, S2S Loss: 0.91479, SLM Loss: 1.80008 +INFO:2025-06-09 05:07:11,411: Validation loss: 0.443 + + + + +INFO:2025-06-09 05:08:44,874: Epoch [21/25], Step [50/3970], Mel Loss: 0.46999, Gen Loss: 13.12624, Disc Loss: 2.57343, Mono Loss: 0.03491, S2S Loss: 1.04637, SLM Loss: 1.88191 +INFO:2025-06-09 05:10:12,520: Epoch [21/25], Step [100/3970], Mel Loss: 0.48069, Gen Loss: 13.96414, Disc Loss: 2.44672, Mono Loss: 0.02962, S2S Loss: 1.05849, SLM Loss: 1.87176 +INFO:2025-06-09 05:11:34,572: Epoch [21/25], Step [150/3970], Mel Loss: 0.47060, Gen Loss: 9.93725, Disc Loss: 3.41884, Mono Loss: 0.03497, S2S Loss: 0.67804, SLM Loss: 1.80349 +INFO:2025-06-09 05:13:00,537: Epoch [21/25], Step [200/3970], Mel Loss: 0.46755, Gen Loss: 13.78485, Disc Loss: 2.85819, Mono Loss: 0.02449, S2S Loss: 1.11917, SLM Loss: 1.90021 +INFO:2025-06-09 05:14:25,136: Epoch [21/25], Step [250/3970], Mel Loss: 0.47410, Gen Loss: 15.36125, Disc Loss: 2.36199, Mono Loss: 0.02162, S2S Loss: 0.80983, SLM Loss: 1.83971 +INFO:2025-06-09 05:15:52,212: Epoch [21/25], Step [300/3970], Mel Loss: 0.48966, Gen Loss: 10.06612, Disc Loss: 3.59549, Mono Loss: 0.02996, S2S Loss: 0.83736, SLM Loss: 1.80890 +INFO:2025-06-09 05:17:15,903: Epoch [21/25], Step [350/3970], Mel Loss: 0.46942, Gen Loss: 13.45973, Disc Loss: 2.75646, Mono Loss: 0.02494, S2S Loss: 0.99167, SLM Loss: 1.77071 +INFO:2025-06-09 05:18:39,464: Epoch [21/25], Step [400/3970], Mel Loss: 0.50879, Gen Loss: 12.60575, Disc Loss: 2.79759, Mono Loss: 0.02696, S2S Loss: 0.86569, SLM Loss: 1.84605 +INFO:2025-06-09 05:20:04,744: Epoch [21/25], Step [450/3970], Mel Loss: 0.46598, Gen Loss: 11.56731, Disc Loss: 3.28558, Mono Loss: 0.03384, S2S Loss: 1.14885, SLM Loss: 1.89640 +INFO:2025-06-09 05:21:56,848: Validation loss: 0.531 + + + + +INFO:2025-06-09 05:23:29,412: Epoch [22/25], Step [50/3970], Mel Loss: 0.48972, Gen Loss: 9.84740, Disc Loss: 3.27302, Mono Loss: 0.03283, S2S Loss: 0.90118, SLM Loss: 1.72501 +INFO:2025-06-09 05:24:56,674: Epoch [22/25], Step [100/3970], Mel Loss: 0.46670, Gen Loss: 12.74649, Disc Loss: 3.23306, Mono Loss: 0.02487, S2S Loss: 0.86857, SLM Loss: 1.82290 +INFO:2025-06-09 05:26:22,931: Epoch [22/25], Step [150/3970], Mel Loss: 0.47883, Gen Loss: 11.48265, Disc Loss: 2.94219, Mono Loss: 0.02303, S2S Loss: 0.74939, SLM Loss: 1.80694 +INFO:2025-06-09 05:27:48,330: Epoch [22/25], Step [200/3970], Mel Loss: 0.47217, Gen Loss: 13.39269, Disc Loss: 2.95101, Mono Loss: 0.02749, S2S Loss: 1.05043, SLM Loss: 1.96215 +INFO:2025-06-09 05:29:13,642: Epoch [22/25], Step [250/3970], Mel Loss: 0.46282, Gen Loss: 11.55498, Disc Loss: 3.11340, Mono Loss: 0.02747, S2S Loss: 1.03113, SLM Loss: 1.80026 +INFO:2025-06-09 05:30:37,418: Epoch [22/25], Step [300/3970], Mel Loss: 0.46544, Gen Loss: 11.89728, Disc Loss: 3.54280, Mono Loss: 0.02800, S2S Loss: 0.78287, SLM Loss: 1.71039 +INFO:2025-06-09 05:32:00,981: Epoch [22/25], Step [350/3970], Mel Loss: 0.47136, Gen Loss: 10.04337, Disc Loss: 3.48481, Mono Loss: 0.02932, S2S Loss: 0.85002, SLM Loss: 1.70120 +INFO:2025-06-09 05:33:23,173: Epoch [22/25], Step [400/3970], Mel Loss: 0.47503, Gen Loss: 13.56164, Disc Loss: 2.41358, Mono Loss: 0.03078, S2S Loss: 1.02305, SLM Loss: 1.93645 +INFO:2025-06-09 05:34:50,016: Epoch [22/25], Step [450/3970], Mel Loss: 0.46342, Gen Loss: 10.59329, Disc Loss: 3.17809, Mono Loss: 0.02502, S2S Loss: 0.97440, SLM Loss: 1.87193 +INFO:2025-06-09 05:36:40,602: Validation loss: 0.433 + + + + +INFO:2025-06-09 05:38:17,194: Epoch [23/25], Step [50/3970], Mel Loss: 0.46734, Gen Loss: 6.01506, Disc Loss: 3.87277, Mono Loss: 0.03631, S2S Loss: 0.86797, SLM Loss: 1.78308 +INFO:2025-06-09 05:39:45,134: Epoch [23/25], Step [100/3970], Mel Loss: 0.45456, Gen Loss: 10.32937, Disc Loss: 3.33413, Mono Loss: 0.03403, S2S Loss: 0.69475, SLM Loss: 1.64199 +INFO:2025-06-09 05:41:10,925: Epoch [23/25], Step [150/3970], Mel Loss: 0.47284, Gen Loss: 10.40523, Disc Loss: 3.22554, Mono Loss: 0.03001, S2S Loss: 1.01316, SLM Loss: 1.79790 +INFO:2025-06-09 05:42:38,603: Epoch [23/25], Step [200/3970], Mel Loss: 0.46535, Gen Loss: 13.34508, Disc Loss: 3.19096, Mono Loss: 0.03172, S2S Loss: 0.90094, SLM Loss: 1.90261 +INFO:2025-06-09 05:44:04,344: Epoch [23/25], Step [250/3970], Mel Loss: 0.46423, Gen Loss: 13.47969, Disc Loss: 2.72158, Mono Loss: 0.02347, S2S Loss: 0.84805, SLM Loss: 1.84319 +INFO:2025-06-09 05:45:29,913: Epoch [23/25], Step [300/3970], Mel Loss: 0.47467, Gen Loss: 13.43347, Disc Loss: 2.60238, Mono Loss: 0.02359, S2S Loss: 0.73610, SLM Loss: 1.73985 +INFO:2025-06-09 05:46:57,724: Epoch [23/25], Step [350/3970], Mel Loss: 0.47841, Gen Loss: 14.26015, Disc Loss: 2.87291, Mono Loss: 0.02988, S2S Loss: 0.85957, SLM Loss: 1.76982 +INFO:2025-06-09 05:48:22,973: Epoch [23/25], Step [400/3970], Mel Loss: 0.46632, Gen Loss: 11.81628, Disc Loss: 2.90987, Mono Loss: 0.02264, S2S Loss: 0.88753, SLM Loss: 1.71842 +INFO:2025-06-09 05:49:48,069: Epoch [23/25], Step [450/3970], Mel Loss: 0.45516, Gen Loss: 12.15128, Disc Loss: 2.99541, Mono Loss: 0.02988, S2S Loss: 0.89353, SLM Loss: 1.90833 +INFO:2025-06-09 05:51:36,052: Validation loss: 0.414 + + + + +INFO:2025-06-09 05:53:10,255: Epoch [24/25], Step [50/3970], Mel Loss: 0.46737, Gen Loss: 10.23576, Disc Loss: 3.82643, Mono Loss: 0.02651, S2S Loss: 1.06037, SLM Loss: 1.99568 +INFO:2025-06-09 05:54:37,468: Epoch [24/25], Step [100/3970], Mel Loss: 0.47177, Gen Loss: 12.41867, Disc Loss: 2.91510, Mono Loss: 0.02961, S2S Loss: 0.85282, SLM Loss: 1.90385 +INFO:2025-06-09 05:56:04,209: Epoch [24/25], Step [150/3970], Mel Loss: 0.46322, Gen Loss: 11.11626, Disc Loss: 3.14571, Mono Loss: 0.02543, S2S Loss: 1.12388, SLM Loss: 1.97046 +INFO:2025-06-09 05:57:29,491: Epoch [24/25], Step [200/3970], Mel Loss: 0.46526, Gen Loss: 15.59772, Disc Loss: 2.67776, Mono Loss: 0.03296, S2S Loss: 0.79784, SLM Loss: 1.69807 +INFO:2025-06-09 05:58:52,805: Epoch [24/25], Step [250/3970], Mel Loss: 0.46915, Gen Loss: 11.35859, Disc Loss: 2.94053, Mono Loss: 0.03207, S2S Loss: 0.66795, SLM Loss: 1.76326 +INFO:2025-06-09 06:00:20,548: Epoch [24/25], Step [300/3970], Mel Loss: 0.46567, Gen Loss: 11.95895, Disc Loss: 3.04587, Mono Loss: 0.03336, S2S Loss: 1.00292, SLM Loss: 1.68971 +INFO:2025-06-09 06:01:46,336: Epoch [24/25], Step [350/3970], Mel Loss: 0.46672, Gen Loss: 12.67813, Disc Loss: 3.09165, Mono Loss: 0.02409, S2S Loss: 1.00270, SLM Loss: 1.76963 +INFO:2025-06-09 06:03:11,223: Epoch [24/25], Step [400/3970], Mel Loss: 0.47140, Gen Loss: 12.71134, Disc Loss: 2.75087, Mono Loss: 0.03305, S2S Loss: 0.78916, SLM Loss: 1.78080 +INFO:2025-06-09 06:04:34,537: Epoch [24/25], Step [450/3970], Mel Loss: 0.47079, Gen Loss: 8.86274, Disc Loss: 3.47894, Mono Loss: 0.03888, S2S Loss: 0.75467, SLM Loss: 1.68494 +INFO:2025-06-09 06:06:21,549: Validation loss: 0.421 + + + + +INFO:2025-06-09 06:07:56,930: Epoch [25/25], Step [50/3970], Mel Loss: 0.46430, Gen Loss: 10.24589, Disc Loss: 3.36352, Mono Loss: 0.03448, S2S Loss: 0.82375, SLM Loss: 1.89295 +INFO:2025-06-09 06:09:24,068: Epoch [25/25], Step [100/3970], Mel Loss: 0.47254, Gen Loss: 9.77706, Disc Loss: 3.08066, Mono Loss: 0.02397, S2S Loss: 0.90991, SLM Loss: 1.88774 +INFO:2025-06-09 06:10:48,092: Epoch [25/25], Step [150/3970], Mel Loss: 0.46583, Gen Loss: 11.92317, Disc Loss: 3.11952, Mono Loss: 0.02979, S2S Loss: 0.85678, SLM Loss: 1.75618 +INFO:2025-06-09 06:12:13,142: Epoch [25/25], Step [200/3970], Mel Loss: 0.47014, Gen Loss: 11.47270, Disc Loss: 3.22507, Mono Loss: 0.02897, S2S Loss: 1.13971, SLM Loss: 1.97860 +INFO:2025-06-09 06:13:37,846: Epoch [25/25], Step [250/3970], Mel Loss: 0.46148, Gen Loss: 10.36515, Disc Loss: 3.29522, Mono Loss: 0.02770, S2S Loss: 0.90695, SLM Loss: 1.83152 +INFO:2025-06-09 06:15:03,448: Epoch [25/25], Step [300/3970], Mel Loss: 0.47940, Gen Loss: 14.23678, Disc Loss: 2.69026, Mono Loss: 0.03347, S2S Loss: 0.90296, SLM Loss: 1.91125 +INFO:2025-06-09 06:16:31,839: Epoch [25/25], Step [350/3970], Mel Loss: 0.46467, Gen Loss: 12.48178, Disc Loss: 2.69238, Mono Loss: 0.03317, S2S Loss: 0.71102, SLM Loss: 1.83116 +INFO:2025-06-09 06:17:56,540: Epoch [25/25], Step [400/3970], Mel Loss: 0.47247, Gen Loss: 11.85046, Disc Loss: 3.40187, Mono Loss: 0.03009, S2S Loss: 0.67559, SLM Loss: 1.67930 +INFO:2025-06-09 06:19:22,839: Epoch [25/25], Step [450/3970], Mel Loss: 0.46736, Gen Loss: 11.12021, Disc Loss: 2.89506, Mono Loss: 0.03046, S2S Loss: 0.86089, SLM Loss: 1.69023 +INFO:2025-06-09 06:21:10,431: Validation loss: 0.427 + + + + +INFO:2025-06-09 07:59:08,543: Epoch [1/15], Step [50/15883], Loss: 0.59463, Disc Loss: 0.00000, Dur Loss: 1.86396, CE Loss: 0.14515, Norm Loss: 1.16517, F0 Loss: 7.99762, LM Loss: 1.75907, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:00:16,205: Epoch [1/15], Step [100/15883], Loss: 0.58628, Disc Loss: 0.00000, Dur Loss: 1.41872, CE Loss: 0.10803, Norm Loss: 3.45058, F0 Loss: 8.65071, LM Loss: 2.24374, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:03:19,805: Epoch [1/15], Step [50/15883], Loss: 0.62156, Disc Loss: 0.00000, Dur Loss: 1.35592, CE Loss: 0.12068, Norm Loss: 4.53912, F0 Loss: 12.41762, LM Loss: 2.12398, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:04:26,944: Epoch [1/15], Step [100/15883], Loss: 0.57647, Disc Loss: 0.00000, Dur Loss: 1.64274, CE Loss: 0.13532, Norm Loss: 2.14458, F0 Loss: 10.61257, LM Loss: 2.36555, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:05:34,608: Epoch [1/15], Step [150/15883], Loss: 0.57745, Disc Loss: 0.00000, Dur Loss: 1.28145, CE Loss: 0.09792, Norm Loss: 2.62505, F0 Loss: 7.29611, LM Loss: 1.92034, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:06:41,940: Epoch [1/15], Step [200/15883], Loss: 0.56741, Disc Loss: 0.00000, Dur Loss: 1.18127, CE Loss: 0.08711, Norm Loss: 1.55673, F0 Loss: 9.18722, LM Loss: 1.94882, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:07:51,378: Epoch [1/15], Step [250/15883], Loss: 0.58637, Disc Loss: 0.00000, Dur Loss: 1.34848, CE Loss: 0.10076, Norm Loss: 2.22754, F0 Loss: 3.84831, LM Loss: 1.77836, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:09:00,842: Epoch [1/15], Step [300/15883], Loss: 0.52509, Disc Loss: 0.00000, Dur Loss: 2.02432, CE Loss: 0.15954, Norm Loss: 2.11042, F0 Loss: 6.38905, LM Loss: 2.07231, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:10:11,156: Epoch [1/15], Step [350/15883], Loss: 0.52738, Disc Loss: 0.00000, Dur Loss: 1.24347, CE Loss: 0.07718, Norm Loss: 2.49583, F0 Loss: 6.60764, LM Loss: 2.14540, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:11:21,010: Epoch [1/15], Step [400/15883], Loss: 0.53797, Disc Loss: 0.00000, Dur Loss: 1.27058, CE Loss: 0.08413, Norm Loss: 2.08075, F0 Loss: 4.33628, LM Loss: 2.04978, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:12:31,649: Epoch [1/15], Step [450/15883], Loss: 0.53851, Disc Loss: 0.00000, Dur Loss: 1.21631, CE Loss: 0.08424, Norm Loss: 2.76793, F0 Loss: 5.36540, LM Loss: 1.88584, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:13:39,582: Epoch [1/15], Step [500/15883], Loss: 0.53331, Disc Loss: 0.00000, Dur Loss: 1.04547, CE Loss: 0.07928, Norm Loss: 1.58679, F0 Loss: 3.28320, LM Loss: 1.99279, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:14:47,552: Epoch [1/15], Step [550/15883], Loss: 0.53619, Disc Loss: 0.00000, Dur Loss: 1.58742, CE Loss: 0.11445, Norm Loss: 3.74290, F0 Loss: 4.74286, LM Loss: 1.91775, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:15:55,037: Epoch [1/15], Step [600/15883], Loss: 0.54803, Disc Loss: 0.00000, Dur Loss: 1.44735, CE Loss: 0.09927, Norm Loss: 2.30749, F0 Loss: 6.71221, LM Loss: 1.95221, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:17:02,005: Epoch [1/15], Step [650/15883], Loss: 0.53016, Disc Loss: 0.00000, Dur Loss: 1.47494, CE Loss: 0.12974, Norm Loss: 2.32919, F0 Loss: 7.98697, LM Loss: 2.14263, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:18:09,942: Epoch [1/15], Step [700/15883], Loss: 0.53085, Disc Loss: 0.00000, Dur Loss: 1.08772, CE Loss: 0.05854, Norm Loss: 1.36569, F0 Loss: 3.09399, LM Loss: 1.99569, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:19:15,731: Epoch [1/15], Step [750/15883], Loss: 0.53651, Disc Loss: 0.00000, Dur Loss: 0.83910, CE Loss: 0.05060, Norm Loss: 2.25991, F0 Loss: 11.18594, LM Loss: 1.93452, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:20:24,249: Epoch [1/15], Step [800/15883], Loss: 0.52069, Disc Loss: 0.00000, Dur Loss: 1.07455, CE Loss: 0.06327, Norm Loss: 5.22029, F0 Loss: 7.52101, LM Loss: 2.07087, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:21:34,678: Epoch [1/15], Step [850/15883], Loss: 0.51518, Disc Loss: 0.00000, Dur Loss: 1.63769, CE Loss: 0.10630, Norm Loss: 2.12592, F0 Loss: 2.67696, LM Loss: 1.75673, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-09 08:22:43,786: Epoch [1/15], Step [900/15883], Loss: 0.50269, Disc Loss: 0.00000, Dur Loss: 1.29939, CE Loss: 0.07080, Norm Loss: 1.26741, F0 Loss: 2.30947, LM Loss: 1.68987, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 diff --git a/.ipynb_checkpoints/train_second-checkpoint.py b/.ipynb_checkpoints/train_second-checkpoint.py new file mode 100644 index 0000000000000000000000000000000000000000..153ba9e471686cecaba902e1b614f1633e67dff6 --- /dev/null +++ b/.ipynb_checkpoints/train_second-checkpoint.py @@ -0,0 +1,879 @@ +# load packages +import random +import yaml +import time +from munch import Munch +import numpy as np +import torch +from torch import nn +import torch.nn.functional as F +import torchaudio +import librosa +import click +import shutil +import traceback +import warnings +warnings.simplefilter('ignore') +from torch.utils.tensorboard import SummaryWriter + +from meldataset import build_dataloader + +from Utils.ASR.models import ASRCNN +from Utils.JDC.model import JDCNet +from Utils.PLBERT.util import load_plbert + +from models import * +from losses import * +from utils import * + +from Modules.slmadv import SLMAdversarialLoss +from Modules.diffusion.sampler import DiffusionSampler, ADPM2Sampler, KarrasSchedule + +from optimizers import build_optimizer + +def clip_to_bert(texts, mask, max_len: int = 510): + """ + Hard-clip batch to ≤ max_len tokens and return + (texts_clipped, **fresh full-width mask**, new_lengths). + """ + if texts.size(1) > max_len: + texts = texts[:, :max_len] + lengths = (texts != 0).sum(dim=1) # PAD id = 0 + seq_len = texts.size(1) + mask = torch.arange(seq_len, device=texts.device).unsqueeze(0) >= \ + lengths.unsqueeze(1) # shape [B, seq_len] + return texts, mask, lengths + +# simple fix for dataparallel that allows access to class attributes +class MyDataParallel(torch.nn.DataParallel): + def __getattr__(self, name): + try: + return super().__getattr__(name) + except AttributeError: + return getattr(self.module, name) + +import logging +from logging import StreamHandler +logger = logging.getLogger(__name__) +logger.setLevel(logging.DEBUG) +handler = StreamHandler() +handler.setLevel(logging.DEBUG) +logger.addHandler(handler) + + +@click.command() +@click.option('-p', '--config_path', default='Configs/config.yml', type=str) +def main(config_path): + config = yaml.safe_load(open(config_path)) + + log_dir = config['log_dir'] + if not osp.exists(log_dir): os.makedirs(log_dir, exist_ok=True) + shutil.copy(config_path, osp.join(log_dir, osp.basename(config_path))) + writer = SummaryWriter(log_dir + "/tensorboard") + + # write logs + file_handler = logging.FileHandler(osp.join(log_dir, 'train.log')) + file_handler.setLevel(logging.DEBUG) + file_handler.setFormatter(logging.Formatter('%(levelname)s:%(asctime)s: %(message)s')) + logger.addHandler(file_handler) + + + batch_size = config.get('batch_size', 10) + + epochs = config.get('epochs_2nd', 200) + save_freq = config.get('save_freq', 2) + log_interval = config.get('log_interval', 10) + saving_epoch = config.get('save_freq', 2) + + data_params = config.get('data_params', None) + sr = config['preprocess_params'].get('sr', 24000) + train_path = data_params['train_data'] + val_path = data_params['val_data'] + root_path = data_params['root_path'] + min_length = data_params['min_length'] + OOD_data = data_params['OOD_data'] + + max_len = config.get('max_len', 200) + + loss_params = Munch(config['loss_params']) + diff_epoch = loss_params.diff_epoch + joint_epoch = loss_params.joint_epoch + + optimizer_params = Munch(config['optimizer_params']) + + train_list, val_list = get_data_path_list(train_path, val_path) + device = 'cuda' + + train_dataloader = build_dataloader(train_list, + root_path, + OOD_data=OOD_data, + min_length=min_length, + batch_size=batch_size, + num_workers=2, + dataset_config={}, + device=device) + + val_dataloader = build_dataloader(val_list, + root_path, + OOD_data=OOD_data, + min_length=min_length, + batch_size=batch_size, + validation=True, + num_workers=0, + device=device, + dataset_config={}) + + # load pretrained ASR model + ASR_config = config.get('ASR_config', False) + ASR_path = config.get('ASR_path', False) + text_aligner = load_ASR_models(ASR_path, ASR_config) + + # load pretrained F0 model + F0_path = config.get('F0_path', False) + pitch_extractor = load_F0_models(F0_path) + + # load PL-BERT model + BERT_path = config.get('PLBERT_dir', False) + plbert = load_plbert(BERT_path) + + # build model + model_params = recursive_munch(config['model_params']) + multispeaker = model_params.multispeaker + model = build_model(model_params, text_aligner, pitch_extractor, plbert) + _ = [model[key].to(device) for key in model] + + # DP + for key in model: + if key != "mpd" and key != "msd" and key != "wd": + model[key] = MyDataParallel(model[key]) + + start_epoch = 0 + iters = 0 + + load_pretrained = config.get('pretrained_model', '') != '' and config.get('second_stage_load_pretrained', False) + + if not load_pretrained: + if config.get('first_stage_path', '') != '': + first_stage_path = osp.join(log_dir, config.get('first_stage_path', 'first_stage.pth')) + print('Loading the first stage model at %s ...' % first_stage_path) + model, _, start_epoch, iters = load_checkpoint(model, + None, + first_stage_path, + load_only_params=True, + ignore_modules=['bert', 'bert_encoder', 'predictor', 'predictor_encoder', 'msd', 'mpd', 'wd', 'diffusion']) # keep starting epoch for tensorboard log + + # these epochs should be counted from the start epoch + diff_epoch += start_epoch + joint_epoch += start_epoch + epochs += start_epoch + + model.predictor_encoder = copy.deepcopy(model.style_encoder) + else: + raise ValueError('You need to specify the path to the first stage model.') + + gl = GeneratorLoss(model.mpd, model.msd).to(device) + dl = DiscriminatorLoss(model.mpd, model.msd).to(device) + wl = WavLMLoss(model_params.slm.model, + model.wd, + sr, + model_params.slm.sr).to(device) + + gl = MyDataParallel(gl) + dl = MyDataParallel(dl) + wl = MyDataParallel(wl) + + sampler = DiffusionSampler( + model.diffusion.diffusion, + sampler=ADPM2Sampler(), + sigma_schedule=KarrasSchedule(sigma_min=0.0001, sigma_max=3.0, rho=9.0), # empirical parameters + clamp=False + ) + + scheduler_params = { + "max_lr": optimizer_params.lr, + "pct_start": float(0), + "epochs": epochs, + "steps_per_epoch": len(train_dataloader), + } + scheduler_params_dict= {key: scheduler_params.copy() for key in model} + scheduler_params_dict['bert']['max_lr'] = optimizer_params.bert_lr * 2 + scheduler_params_dict['decoder']['max_lr'] = optimizer_params.ft_lr * 2 + scheduler_params_dict['style_encoder']['max_lr'] = optimizer_params.ft_lr * 2 + + optimizer = build_optimizer({key: model[key].parameters() for key in model}, + scheduler_params_dict=scheduler_params_dict, lr=optimizer_params.lr) + + # adjust BERT learning rate + for g in optimizer.optimizers['bert'].param_groups: + g['betas'] = (0.9, 0.99) + g['lr'] = optimizer_params.bert_lr + g['initial_lr'] = optimizer_params.bert_lr + g['min_lr'] = 0 + g['weight_decay'] = 0.01 + + # adjust acoustic module learning rate + for module in ["decoder", "style_encoder"]: + for g in optimizer.optimizers[module].param_groups: + g['betas'] = (0.0, 0.99) + g['lr'] = optimizer_params.ft_lr + g['initial_lr'] = optimizer_params.ft_lr + g['min_lr'] = 0 + g['weight_decay'] = 1e-4 + + # load models if there is a model + if load_pretrained: + model, optimizer, start_epoch, iters = load_checkpoint(model, optimizer, config['pretrained_model'], + load_only_params=config.get('load_only_params', True)) + + n_down = model.text_aligner.n_down + + best_loss = float('inf') # best test loss + loss_train_record = list([]) + loss_test_record = list([]) + iters = 0 + + criterion = nn.L1Loss() # F0 loss (regression) + torch.cuda.empty_cache() + + stft_loss = MultiResolutionSTFTLoss().to(device) + + print('BERT', optimizer.optimizers['bert']) + print('decoder', optimizer.optimizers['decoder']) + + start_ds = False + + running_std = [] + + slmadv_params = Munch(config['slmadv_params']) + slmadv = SLMAdversarialLoss(model, wl, sampler, + slmadv_params.min_len, + slmadv_params.max_len, + batch_percentage=slmadv_params.batch_percentage, + skip_update=slmadv_params.iter, + sig=slmadv_params.sig + ) + + + for epoch in range(start_epoch, epochs): + running_loss = 0 + start_time = time.time() + + _ = [model[key].eval() for key in model] + + model.predictor.train() + model.bert_encoder.train() + model.bert.train() + model.msd.train() + model.mpd.train() + + + if epoch >= diff_epoch: + start_ds = True + + for i, batch in enumerate(train_dataloader): + waves = batch[0] + batch = [b.to(device) for b in batch[1:]] + texts, input_lengths, ref_texts, ref_lengths, mels, mel_input_length, ref_mels = batch + + # --------------- CLIP TEXTS *ONCE* ----------------- + text_mask = length_to_mask(input_lengths).to(texts.device) + texts, text_mask, input_lengths = clip_to_bert(texts, text_mask) + # ── drop rows that became all-PAD after clipping ─────────── + keep = (input_lengths > 0).nonzero(as_tuple=True)[0] + if keep.numel() != texts.size(0): + texts, text_mask, input_lengths = texts[keep], text_mask[keep], input_lengths[keep] + ref_texts, ref_lengths = ref_texts[keep], ref_lengths[keep] + mels, mel_input_length, ref_mels = mels[keep], mel_input_length[keep], ref_mels[keep] + waves = [waves[i] for i in keep.tolist()] + # ---------------------------------------------------- + + with torch.no_grad(): + mask = length_to_mask(mel_input_length // (2 ** n_down)).to(device) + mel_mask = length_to_mask(mel_input_length).to(device) + + try: + _, _, s2s_attn = model.text_aligner(mels, mask, texts) + s2s_attn = s2s_attn.transpose(-1, -2) + s2s_attn = s2s_attn[..., 1:] + s2s_attn = s2s_attn.transpose(-1, -2) + except: + continue + + mask_ST = mask_from_lens(s2s_attn, input_lengths, mel_input_length // (2 ** n_down)) + s2s_attn_mono = maximum_path(s2s_attn, mask_ST) + + # encode + t_en = model.text_encoder(texts, input_lengths, text_mask) + asr = (t_en @ s2s_attn_mono) + + d_gt = s2s_attn_mono.sum(axis=-1).detach() + + # compute reference styles + if multispeaker and epoch >= diff_epoch: + ref_ss = model.style_encoder(ref_mels.unsqueeze(1)) + ref_sp = model.predictor_encoder(ref_mels.unsqueeze(1)) + ref = torch.cat([ref_ss, ref_sp], dim=1) + + # compute the style of the entire utterance + # this operation cannot be done in batch because of the avgpool layer (may need to work on masked avgpool) + ss = [] + gs = [] + for bib in range(len(mel_input_length)): + mel_length = int(mel_input_length[bib].item()) + mel = mels[bib, :, :mel_input_length[bib]] + s = model.predictor_encoder(mel.unsqueeze(0).unsqueeze(1)) + ss.append(s) + s = model.style_encoder(mel.unsqueeze(0).unsqueeze(1)) + gs.append(s) + + s_dur = torch.stack(ss).squeeze() # global prosodic styles + gs = torch.stack(gs).squeeze() # global acoustic styles + s_trg = torch.cat([gs, s_dur], dim=-1).detach() # ground truth for denoiser + + # texts, input_lengths, ref_texts, ref_lengths, mels, mel_input_length, ref_mels = batch + + # # ────── PATCH: keep PL-BERT below 512 tokens ───────── + # MAX_BERT_LEN = 510 # leave room for [CLS] and [SEP] + # if texts.size(1) > MAX_BERT_LEN: # truncate batch-wise + # texts = texts[:, :MAX_BERT_LEN] + # seq_len = texts.size(1) # current padded width + # input_lengths = (texts != 0).sum(1) # 0 is PAD + # arange_row = torch.arange(seq_len, device=texts.device) # shape [L] + # text_mask = arange_row.unsqueeze(0) >= input_lengths.unsqueeze(1) + # # shape [B, L] + + # # keep only rows that still have at least one real token + # keep = (input_lengths > 0).nonzero(as_tuple=True)[0] + # if keep.numel() != texts.size(0): # a row was truncated to length 0 + # texts, text_mask, input_lengths = texts[keep], text_mask[keep], input_lengths[keep] + # ref_texts, ref_lengths = ref_texts[keep], ref_lengths[keep] + # mels, mel_input_length, ref_mels = mels[keep], mel_input_length[keep], ref_mels[keep] + # waves = [waves[i] for i in keep.tolist()] + + # # clip alignments to the *current* width (seq_len) + # s2s_attn_mono = s2s_attn_mono[:, :seq_len, :] + # d_gt = d_gt[:, :seq_len] + # # ───────────────────────────────────────────────────── + + # ------------------------------------------------------------- + # Now build *everything* that depends on token count + with torch.no_grad(): + t_en = model.text_encoder(texts, input_lengths, text_mask) + + _, _, s2s_attn = model.text_aligner(mels, mask, texts) + s2s_attn = s2s_attn.transpose(-1, -2)[..., 1:].transpose(-1, -2) + + mask_ST = mask_from_lens(s2s_attn, input_lengths, + mel_input_length // 2**n_down) + s2s_attn_mono = maximum_path(s2s_attn, mask_ST) + + asr = t_en @ s2s_attn_mono + d_gt = s2s_attn_mono.sum(dim=-1) + + bert_dur = model.bert(texts, attention_mask=(~text_mask).int()) + d_en = model.bert_encoder(bert_dur).transpose(-1, -2) + + # denoiser training + if epoch >= diff_epoch: + num_steps = np.random.randint(3, 5) + + if model_params.diffusion.dist.estimate_sigma_data: + model.diffusion.module.diffusion.sigma_data = s_trg.std(axis=-1).mean().item() # batch-wise std estimation + running_std.append(model.diffusion.module.diffusion.sigma_data) + + if multispeaker: + s_preds = sampler(noise = torch.randn_like(s_trg).unsqueeze(1).to(device), + embedding=bert_dur, + embedding_scale=1, + features=ref, # reference from the same speaker as the embedding + embedding_mask_proba=0.1, + num_steps=num_steps).squeeze(1) + loss_diff = model.diffusion(s_trg.unsqueeze(1), embedding=bert_dur, features=ref).mean() # EDM loss + loss_sty = F.l1_loss(s_preds, s_trg.detach()) # style reconstruction loss + else: + s_preds = sampler(noise = torch.randn_like(s_trg).unsqueeze(1).to(device), + embedding=bert_dur, + embedding_scale=1, + embedding_mask_proba=0.1, + num_steps=num_steps).squeeze(1) + loss_diff = model.diffusion.module.diffusion(s_trg.unsqueeze(1), embedding=bert_dur).mean() # EDM loss + loss_sty = F.l1_loss(s_preds, s_trg.detach()) # style reconstruction loss + else: + loss_sty = 0 + loss_diff = 0 + + d, p = model.predictor(d_en, s_dur, + input_lengths, + s2s_attn_mono, + text_mask) + + mel_len = min(int(mel_input_length.min().item() / 2 - 1), max_len // 2) + mel_len_st = int(mel_input_length.min().item() / 2 - 1) + en = [] + gt = [] + st = [] + p_en = [] + wav = [] + + for bib in range(len(mel_input_length)): + mel_length = int(mel_input_length[bib].item() / 2) + + random_start = np.random.randint(0, mel_length - mel_len) + en.append(asr[bib, :, random_start:random_start+mel_len]) + p_en.append(p[bib, :, random_start:random_start+mel_len]) + gt.append(mels[bib, :, (random_start * 2):((random_start+mel_len) * 2)]) + + y = waves[bib][(random_start * 2) * 300:((random_start+mel_len) * 2) * 300] + wav.append(torch.from_numpy(y).to(device)) + + # style reference (better to be different from the GT) + random_start = np.random.randint(0, mel_length - mel_len_st) + st.append(mels[bib, :, (random_start * 2):((random_start+mel_len_st) * 2)]) + + wav = torch.stack(wav).float().detach() + + en = torch.stack(en) + p_en = torch.stack(p_en) + gt = torch.stack(gt).detach() + st = torch.stack(st).detach() + + if gt.size(-1) < 80: + continue + + s_dur = model.predictor_encoder(st.unsqueeze(1) if multispeaker else gt.unsqueeze(1)) + s = model.style_encoder(st.unsqueeze(1) if multispeaker else gt.unsqueeze(1)) + + with torch.no_grad(): + F0_real, _, F0 = model.pitch_extractor(gt.unsqueeze(1)) + F0 = F0.reshape(F0.shape[0], F0.shape[1] * 2, F0.shape[2], 1).squeeze() + + asr_real = model.text_aligner.get_feature(gt) + + N_real = log_norm(gt.unsqueeze(1)).squeeze(1) + + y_rec_gt = wav.unsqueeze(1) + y_rec_gt_pred = model.decoder(en, F0_real, N_real, s) + + if epoch >= joint_epoch: + # ground truth from recording + wav = y_rec_gt # use recording since decoder is tuned + else: + # ground truth from reconstruction + wav = y_rec_gt_pred # use reconstruction since decoder is fixed + + F0_fake, N_fake = model.predictor.F0Ntrain(p_en, s_dur) + + y_rec = model.decoder(en, F0_fake, N_fake, s) + + loss_F0_rec = (F.smooth_l1_loss(F0_real, F0_fake)) / 10 + loss_norm_rec = F.smooth_l1_loss(N_real, N_fake) + + if start_ds: + optimizer.zero_grad() + d_loss = dl(wav.detach(), y_rec.detach()).mean() + d_loss.backward() + optimizer.step('msd') + optimizer.step('mpd') + else: + d_loss = 0 + + # generator loss + optimizer.zero_grad() + + loss_mel = stft_loss(y_rec, wav) + if start_ds: + loss_gen_all = gl(wav, y_rec).mean() + else: + loss_gen_all = 0 + loss_lm = wl(wav.detach().squeeze(), y_rec.squeeze()).mean() + + loss_ce = 0 + loss_dur = 0 + for _s2s_pred, _text_input, _text_length in zip(d, (d_gt), input_lengths): + _s2s_pred = _s2s_pred[:_text_length, :] + _text_input = _text_input[:_text_length].long() + _s2s_trg = torch.zeros_like(_s2s_pred) + for p in range(_s2s_trg.shape[0]): + _s2s_trg[p, :_text_input[p]] = 1 + _dur_pred = torch.sigmoid(_s2s_pred).sum(axis=1) + + loss_dur += F.l1_loss(_dur_pred[1:_text_length-1], + _text_input[1:_text_length-1]) + loss_ce += F.binary_cross_entropy_with_logits(_s2s_pred.flatten(), _s2s_trg.flatten()) + + loss_ce /= texts.size(0) + loss_dur /= texts.size(0) + + g_loss = loss_params.lambda_mel * loss_mel + \ + loss_params.lambda_F0 * loss_F0_rec + \ + loss_params.lambda_ce * loss_ce + \ + loss_params.lambda_norm * loss_norm_rec + \ + loss_params.lambda_dur * loss_dur + \ + loss_params.lambda_gen * loss_gen_all + \ + loss_params.lambda_slm * loss_lm + \ + loss_params.lambda_sty * loss_sty + \ + loss_params.lambda_diff * loss_diff + + running_loss += loss_mel.item() + g_loss.backward() + if torch.isnan(g_loss): + from IPython.core.debugger import set_trace + set_trace() + + optimizer.step('bert_encoder') + optimizer.step('bert') + optimizer.step('predictor') + optimizer.step('predictor_encoder') + + if epoch >= diff_epoch: + optimizer.step('diffusion') + + if epoch >= joint_epoch: + optimizer.step('style_encoder') + optimizer.step('decoder') + + # randomly pick whether to use in-distribution text + if np.random.rand() < 0.5: + use_ind = True + else: + use_ind = False + + if use_ind: + ref_lengths = input_lengths + ref_texts = texts + + # ---- clip reference text exactly the same way ---- + ref_mask = length_to_mask(ref_lengths).to(ref_texts.device) + ref_texts, ref_mask, ref_lengths = clip_to_bert(ref_texts, ref_mask) + + slm_out = slmadv(i, + y_rec_gt, + y_rec_gt_pred, + waves, + mel_input_length, + ref_texts, + ref_lengths, use_ind, s_trg.detach(), ref if multispeaker else None) + + if slm_out is None: + continue + + d_loss_slm, loss_gen_lm, y_pred = slm_out + + # SLM generator loss + optimizer.zero_grad() + loss_gen_lm.backward() + + # compute the gradient norm + total_norm = {} + for key in model.keys(): + total_norm[key] = 0 + parameters = [p for p in model[key].parameters() if p.grad is not None and p.requires_grad] + for p in parameters: + param_norm = p.grad.detach().data.norm(2) + total_norm[key] += param_norm.item() ** 2 + total_norm[key] = total_norm[key] ** 0.5 + + # gradient scaling + if total_norm['predictor'] > slmadv_params.thresh: + for key in model.keys(): + for p in model[key].parameters(): + if p.grad is not None: + p.grad *= (1 / total_norm['predictor']) + + for p in model.predictor.duration_proj.parameters(): + if p.grad is not None: + p.grad *= slmadv_params.scale + + for p in model.predictor.lstm.parameters(): + if p.grad is not None: + p.grad *= slmadv_params.scale + + for p in model.diffusion.parameters(): + if p.grad is not None: + p.grad *= slmadv_params.scale + + optimizer.step('bert_encoder') + optimizer.step('bert') + optimizer.step('predictor') + optimizer.step('diffusion') + + # SLM discriminator loss + if d_loss_slm != 0: + optimizer.zero_grad() + d_loss_slm.backward(retain_graph=True) + optimizer.step('wd') + + else: + d_loss_slm, loss_gen_lm = 0, 0 + + iters = iters + 1 + + if (i+1)%log_interval == 0: + logger.info ('Epoch [%d/%d], Step [%d/%d], Loss: %.5f, Disc Loss: %.5f, Dur Loss: %.5f, CE Loss: %.5f, Norm Loss: %.5f, F0 Loss: %.5f, LM Loss: %.5f, Gen Loss: %.5f, Sty Loss: %.5f, Diff Loss: %.5f, DiscLM Loss: %.5f, GenLM Loss: %.5f' + %(epoch+1, epochs, i+1, len(train_list)//batch_size, running_loss / log_interval, d_loss, loss_dur, loss_ce, loss_norm_rec, loss_F0_rec, loss_lm, loss_gen_all, loss_sty, loss_diff, d_loss_slm, loss_gen_lm)) + + writer.add_scalar('train/mel_loss', running_loss / log_interval, iters) + writer.add_scalar('train/gen_loss', loss_gen_all, iters) + writer.add_scalar('train/d_loss', d_loss, iters) + writer.add_scalar('train/ce_loss', loss_ce, iters) + writer.add_scalar('train/dur_loss', loss_dur, iters) + writer.add_scalar('train/slm_loss', loss_lm, iters) + writer.add_scalar('train/norm_loss', loss_norm_rec, iters) + writer.add_scalar('train/F0_loss', loss_F0_rec, iters) + writer.add_scalar('train/sty_loss', loss_sty, iters) + writer.add_scalar('train/diff_loss', loss_diff, iters) + writer.add_scalar('train/d_loss_slm', d_loss_slm, iters) + writer.add_scalar('train/gen_loss_slm', loss_gen_lm, iters) + + running_loss = 0 + + print('Time elasped:', time.time()-start_time) + + loss_test = 0 + loss_align = 0 + loss_f = 0 + _ = [model[key].eval() for key in model] + + with torch.no_grad(): + iters_test = 0 + for batch_idx, batch in enumerate(val_dataloader): + optimizer.zero_grad() + + try: + waves = batch[0] + batch = [b.to(device) for b in batch[1:]] + texts, input_lengths, ref_texts, ref_lengths, mels, mel_input_length, ref_mels = batch + + texts, text_mask, input_lengths = clip_to_bert(texts, text_mask) + keep = (input_lengths > 0).nonzero(as_tuple=True)[0] + if keep.numel() != texts.size(0): + texts, text_mask, input_lengths = texts[keep], text_mask[keep], input_lengths[keep] + ref_texts, ref_lengths = ref_texts[keep], ref_lengths[keep] + mels, mel_input_length, ref_mels = mels[keep], mel_input_length[keep], ref_mels[keep] + waves = [waves[i] for i in keep.tolist()] + + with torch.no_grad(): + mask = length_to_mask(mel_input_length // (2 ** n_down)).to(texts.device) + # mask = length_to_mask(mel_input_length // (2 ** n_down)).to('cuda') + + # _, _, s2s_attn = model.text_aligner(mels, mask, texts) + # s2s_attn = s2s_attn.transpose(-1, -2) + # s2s_attn = s2s_attn[..., 1:] + # s2s_attn = s2s_attn.transpose(-1, -2) + + # mask_ST = mask_from_lens(s2s_attn, input_lengths, mel_input_length // (2 ** n_down)) + # s2s_attn_mono = maximum_path(s2s_attn, mask_ST) + + # # encode + # t_en = model.text_encoder(texts, input_lengths, text_mask) + # asr = (t_en @ s2s_attn_mono) + + # d_gt = s2s_attn_mono.sum(axis=-1).detach() + + _, _, s2s_attn = model.text_aligner(mels, mask, texts) + s2s_attn = s2s_attn.transpose(-1, -2)[..., 1:].transpose(-1, -2) + mask_ST = mask_from_lens(s2s_attn, input_lengths, + mel_input_length // 2 ** n_down) + s2s_attn_mono = maximum_path(s2s_attn, mask_ST) + + t_en = model.text_encoder(texts, input_lengths, text_mask) + asr = t_en @ s2s_attn_mono + d_gt = s2s_attn_mono.sum(dim=-1).detach() + + ss = [] + gs = [] + + for bib in range(len(mel_input_length)): + mel_length = int(mel_input_length[bib].item()) + mel = mels[bib, :, :mel_input_length[bib]] + s = model.predictor_encoder(mel.unsqueeze(0).unsqueeze(1)) + ss.append(s) + s = model.style_encoder(mel.unsqueeze(0).unsqueeze(1)) + gs.append(s) + + s = torch.stack(ss).squeeze() + gs = torch.stack(gs).squeeze() + s_trg = torch.cat([s, gs], dim=-1).detach() + + bert_dur = model.bert(texts, attention_mask=(~text_mask).int()) + d_en = model.bert_encoder(bert_dur).transpose(-1, -2) + d, p = model.predictor(d_en, s, + input_lengths, + s2s_attn_mono, + text_mask) + # get clips + mel_len = int(mel_input_length.min().item() / 2 - 1) + en = [] + gt = [] + p_en = [] + wav = [] + + for bib in range(len(mel_input_length)): + mel_length = int(mel_input_length[bib].item() / 2) + + random_start = np.random.randint(0, mel_length - mel_len) + en.append(asr[bib, :, random_start:random_start+mel_len]) + p_en.append(p[bib, :, random_start:random_start+mel_len]) + + gt.append(mels[bib, :, (random_start * 2):((random_start+mel_len) * 2)]) + + y = waves[bib][(random_start * 2) * 300:((random_start+mel_len) * 2) * 300] + wav.append(torch.from_numpy(y).to(device)) + + wav = torch.stack(wav).float().detach() + + en = torch.stack(en) + p_en = torch.stack(p_en) + gt = torch.stack(gt).detach() + + s = model.predictor_encoder(gt.unsqueeze(1)) + + F0_fake, N_fake = model.predictor.F0Ntrain(p_en, s) + + loss_dur = 0 + for _s2s_pred, _text_input, _text_length in zip(d, (d_gt), input_lengths): + _s2s_pred = _s2s_pred[:_text_length, :] + _text_input = _text_input[:_text_length].long() + _s2s_trg = torch.zeros_like(_s2s_pred) + for bib in range(_s2s_trg.shape[0]): + _s2s_trg[bib, :_text_input[bib]] = 1 + _dur_pred = torch.sigmoid(_s2s_pred).sum(axis=1) + loss_dur += F.l1_loss(_dur_pred[1:_text_length-1], + _text_input[1:_text_length-1]) + + loss_dur /= texts.size(0) + + s = model.style_encoder(gt.unsqueeze(1)) + + y_rec = model.decoder(en, F0_fake, N_fake, s) + loss_mel = stft_loss(y_rec.squeeze(), wav.detach()) + + F0_real, _, F0 = model.pitch_extractor(gt.unsqueeze(1)) + + loss_F0 = F.l1_loss(F0_real, F0_fake) / 10 + + loss_test += (loss_mel).mean() + loss_align += (loss_dur).mean() + loss_f += (loss_F0).mean() + + iters_test += 1 + except Exception as e: + print(f"run into exception", e) + traceback.print_exc() + continue + + print('Epochs:', epoch + 1) + logger.info('Validation loss: %.3f, Dur loss: %.3f, F0 loss: %.3f' % (loss_test / iters_test, loss_align / iters_test, loss_f / iters_test) + '\n\n\n') + print('\n\n\n') + writer.add_scalar('eval/mel_loss', loss_test / iters_test, epoch + 1) + writer.add_scalar('eval/dur_loss', loss_align / iters_test, epoch + 1) + writer.add_scalar('eval/F0_loss', loss_f / iters_test, epoch + 1) + + if epoch < joint_epoch: + # generating reconstruction examples with GT duration + + with torch.no_grad(): + for bib in range(len(asr)): + mel_length = int(mel_input_length[bib].item()) + gt = mels[bib, :, :mel_length].unsqueeze(0) + en = asr[bib, :, :mel_length // 2].unsqueeze(0) + + F0_real, _, _ = model.pitch_extractor(gt.unsqueeze(1)) + F0_real = F0_real.unsqueeze(0) + s = model.style_encoder(gt.unsqueeze(1)) + real_norm = log_norm(gt.unsqueeze(1)).squeeze(1) + + y_rec = model.decoder(en, F0_real, real_norm, s) + + writer.add_audio('eval/y' + str(bib), y_rec.cpu().numpy().squeeze(), epoch, sample_rate=sr) + + s_dur = model.predictor_encoder(gt.unsqueeze(1)) + p_en = p[bib, :, :mel_length // 2].unsqueeze(0) + + F0_fake, N_fake = model.predictor.F0Ntrain(p_en, s_dur) + + y_pred = model.decoder(en, F0_fake, N_fake, s) + + writer.add_audio('pred/y' + str(bib), y_pred.cpu().numpy().squeeze(), epoch, sample_rate=sr) + + if epoch == 0: + writer.add_audio('gt/y' + str(bib), waves[bib].squeeze(), epoch, sample_rate=sr) + + if bib >= 5: + break + else: + # generating sampled speech from text directly + with torch.no_grad(): + # compute reference styles + if multispeaker and epoch >= diff_epoch: + ref_ss = model.style_encoder(ref_mels.unsqueeze(1)) + ref_sp = model.predictor_encoder(ref_mels.unsqueeze(1)) + ref_s = torch.cat([ref_ss, ref_sp], dim=1) + + for bib in range(len(d_en)): + if multispeaker: + s_pred = sampler(noise = torch.randn((1, 256)).unsqueeze(1).to(texts.device), + embedding=bert_dur[bib].unsqueeze(0), + embedding_scale=1, + features=ref_s[bib].unsqueeze(0), # reference from the same speaker as the embedding + num_steps=5).squeeze(1) + else: + s_pred = sampler(noise = torch.randn((1, 256)).unsqueeze(1).to(texts.device), + embedding=bert_dur[bib].unsqueeze(0), + embedding_scale=1, + num_steps=5).squeeze(1) + + s = s_pred[:, 128:] + ref = s_pred[:, :128] + + d = model.predictor.text_encoder(d_en[bib, :, :input_lengths[bib]].unsqueeze(0), + s, input_lengths[bib, ...].unsqueeze(0), text_mask[bib, :input_lengths[bib]].unsqueeze(0)) + + x, _ = model.predictor.lstm(d) + duration = model.predictor.duration_proj(x) + + duration = torch.sigmoid(duration).sum(axis=-1) + pred_dur = torch.round(duration.squeeze()).clamp(min=1) + + pred_dur[-1] += 5 + + pred_aln_trg = torch.zeros(input_lengths[bib], int(pred_dur.sum().data)) + c_frame = 0 + for i in range(pred_aln_trg.size(0)): + pred_aln_trg[i, c_frame:c_frame + int(pred_dur[i].data)] = 1 + c_frame += int(pred_dur[i].data) + + # encode prosody + en = (d.transpose(-1, -2) @ pred_aln_trg.unsqueeze(0).to(texts.device)) + F0_pred, N_pred = model.predictor.F0Ntrain(en, s) + out = model.decoder((t_en[bib, :, :input_lengths[bib]].unsqueeze(0) @ pred_aln_trg.unsqueeze(0).to(texts.device)), + F0_pred, N_pred, ref.squeeze().unsqueeze(0)) + + writer.add_audio('pred/y' + str(bib), out.cpu().numpy().squeeze(), epoch, sample_rate=sr) + + if bib >= 5: + break + + if epoch % saving_epoch == 0: + if (loss_test / iters_test) < best_loss: + best_loss = loss_test / iters_test + print('Saving..') + state = { + 'net': {key: model[key].state_dict() for key in model}, + 'optimizer': optimizer.state_dict(), + 'iters': iters, + 'val_loss': loss_test / iters_test, + 'epoch': epoch, + } + save_path = osp.join(log_dir, 'epoch_2nd_%05d.pth' % epoch) + torch.save(state, save_path) + + # if estimate sigma, save the estimated simga + if model_params.diffusion.dist.estimate_sigma_data: + config['model_params']['diffusion']['dist']['sigma_data'] = float(np.mean(running_std)) + + with open(osp.join(log_dir, osp.basename(config_path)), 'w') as outfile: + yaml.dump(config, outfile, default_flow_style=True) + +if __name__=="__main__": + main() \ No newline at end of file diff --git a/logs/pod_90h_30k/config_ft_single.yml b/Configs/.ipynb_checkpoints/config_ft_single-checkpoint.yml similarity index 79% rename from logs/pod_90h_30k/config_ft_single.yml rename to Configs/.ipynb_checkpoints/config_ft_single-checkpoint.yml index b4a544ec9e7a2e79b3e746ff0e2acdfb2bbfee7a..ccc6345c315e1ae14e0ba6fa25ddcbecce189d5e 100644 --- a/logs/pod_90h_30k/config_ft_single.yml +++ b/Configs/.ipynb_checkpoints/config_ft_single-checkpoint.yml @@ -1,18 +1,18 @@ # ─── GLOBAL ────────────────────────────────────────────────────────── -log_dir: logs/pod_90h_30k +log_dir: logs/pod_90h_30k_second_lr1 device: "cuda" -batch_size: 8 # 40 GB A100, fp16 -max_len: 160 # ≈ 8 s (200 × 40 ms) +batch_size: 12 # 40 GB A100, fp16 +max_len: 300 # ≈ 8 s (200 × 40 ms) -epochs_1st: 13 # first-stage schedule -epochs_2nd: 13 # second-stage schedule (later) -save_freq: 2 +epochs_1st: 25 # first-stage schedule +epochs_2nd: 20 # second-stage schedule (later) +save_freq: 1 log_interval: 50 # leave blank on first run -pretrained_model: "" -second_stage_load_pretrained: false +pretrained_model: "" #"/workspace/styletts2/logs/pod_90h_30k/epoch_2nd_00003.pth" +second_stage_load_pretrained: true load_only_params: false # ─── PRE-PROCESS ───────────────────────────────────────────────────── @@ -25,11 +25,11 @@ preprocess_params: # ─── DATA ──────────────────────────────────────────────────────────── data_params: - root_path: /home/ubuntu/styletts2-ft/data/wavs - train_data: /home/ubuntu/styletts2-ft/data/train_list.txt - val_data: /home/ubuntu/styletts2-ft/data/val_list.txt + root_path: /workspace + train_data: /workspace/styletts2/data/train_list.txt + val_data: /workspace/styletts2/data/val_list.txt min_length: 50 # sample until texts with this size are obtained for OOD texts - OOD_data: /home/ubuntu/styletts2-ft/data/OOD_texts.txt + OOD_data: /workspace/styletts2/data/OOD_texts.txt # ─── LOSS SCHEDULE ────────────────────────────────────────────────── loss_params: @@ -39,7 +39,7 @@ loss_params: lambda_mono: 1. # monotonic alignment loss (1st stage, TMA) lambda_s2s: 1. # sequence-to-sequence loss (1st stage, TMA) - TMA_epoch: 50 # TMA starting epoch (1st stage) + TMA_epoch: 14 # TMA starting epoch (1st stage) lambda_F0: 1. # F0 reconstruction loss (2nd stage) lambda_norm: 1. # norm reconstruction loss (2nd stage) @@ -48,14 +48,14 @@ loss_params: lambda_sty: 1. # style reconstruction loss (2nd stage) lambda_diff: 1. # score matching loss (2nd stage) - diff_epoch: 20 # style diffusion starting epoch (2nd stage) - joint_epoch: 50 # joint training starting epoch (2nd stage) + diff_epoch: 0 # style diffusion starting epoch (2nd stage) + joint_epoch: 0 # joint training starting epoch (2nd stage) # ─── OPTIMISER ────────────────────────────────────────────────────── optimizer_params: - lr: 0.0008 - bert_lr: 0.00002 - ft_lr: 0.0002 + lr: 0.0001 + bert_lr: 0.00001 + ft_lr: 0.0001 grad_accum_steps: 2 # ─── MODEL (core network & sub-modules) ───────────────────────────── @@ -105,7 +105,7 @@ F0_path: "Utils/JDC/bst.t7" ASR_config: "Utils/ASR/config.yml" ASR_path: "Utils/ASR/epoch_00080.pth" PLBERT_dir: 'Utils/PLBERT/' -first_stage_path: "" # filled automatically after this run +first_stage_path: "/workspace/styletts2/stage1_final.pth" # filled automatically after this run # ─── SLM ADVERSARIAL (ignored in stage-1, kept default) ───────────── slmadv_params: diff --git a/Configs/.ipynb_checkpoints/config_libritts-checkpoint.yml b/Configs/.ipynb_checkpoints/config_libritts-checkpoint.yml new file mode 100644 index 0000000000000000000000000000000000000000..135d87260aa53cfb18b665333d44744ce5b4152a --- /dev/null +++ b/Configs/.ipynb_checkpoints/config_libritts-checkpoint.yml @@ -0,0 +1,113 @@ +log_dir: "Models/LibriTTS" +first_stage_path: "first_stage.pth" +save_freq: 1 +log_interval: 10 +device: "cuda" +epochs_1st: 50 # number of epochs for first stage training (pre-training) +epochs_2nd: 30 # number of peochs for second stage training (joint training) +batch_size: 16 +max_len: 300 # maximum number of frames +pretrained_model: "" +second_stage_load_pretrained: true # set to true if the pre-trained model is for 2nd stage +load_only_params: false # set to true if do not want to load epoch numbers and optimizer parameters + +F0_path: "Utils/JDC/bst.t7" +ASR_config: "Utils/ASR/config.yml" +ASR_path: "Utils/ASR/epoch_00080.pth" +PLBERT_dir: 'Utils/PLBERT/' + +data_params: + train_data: "Data/train_list.txt" + val_data: "Data/val_list.txt" + root_path: "" + OOD_data: "Data/OOD_texts.txt" + min_length: 50 # sample until texts with this size are obtained for OOD texts + +preprocess_params: + sr: 24000 + spect_params: + n_fft: 2048 + win_length: 1200 + hop_length: 300 + +model_params: + multispeaker: true + + dim_in: 64 + hidden_dim: 512 + max_conv_dim: 512 + n_layer: 3 + n_mels: 80 + + n_token: 178 # number of phoneme tokens + max_dur: 50 # maximum duration of a single phoneme + style_dim: 128 # style vector size + + dropout: 0.2 + + # config for decoder + decoder: + type: 'hifigan' # either hifigan or istftnet + resblock_kernel_sizes: [3,7,11] + upsample_rates : [10,5,3,2] + upsample_initial_channel: 512 + resblock_dilation_sizes: [[1,3,5], [1,3,5], [1,3,5]] + upsample_kernel_sizes: [20,10,6,4] + + # speech language model config + slm: + model: 'microsoft/wavlm-base-plus' + sr: 16000 # sampling rate of SLM + hidden: 768 # hidden size of SLM + nlayers: 13 # number of layers of SLM + initial_channel: 64 # initial channels of SLM discriminator head + + # style diffusion model config + diffusion: + embedding_mask_proba: 0.1 + # transformer config + transformer: + num_layers: 3 + num_heads: 8 + head_features: 64 + multiplier: 2 + + # diffusion distribution config + dist: + sigma_data: 0.2 # placeholder for estimate_sigma_data set to false + estimate_sigma_data: true # estimate sigma_data from the current batch if set to true + mean: -3.0 + std: 1.0 + +loss_params: + lambda_mel: 5. # mel reconstruction loss + lambda_gen: 1. # generator loss + lambda_slm: 1. # slm feature matching loss + + lambda_mono: 1. # monotonic alignment loss (1st stage, TMA) + lambda_s2s: 1. # sequence-to-sequence loss (1st stage, TMA) + TMA_epoch: 5 # TMA starting epoch (1st stage) + + lambda_F0: 1. # F0 reconstruction loss (2nd stage) + lambda_norm: 1. # norm reconstruction loss (2nd stage) + lambda_dur: 1. # duration loss (2nd stage) + lambda_ce: 20. # duration predictor probability output CE loss (2nd stage) + lambda_sty: 1. # style reconstruction loss (2nd stage) + lambda_diff: 1. # score matching loss (2nd stage) + + diff_epoch: 10 # style diffusion starting epoch (2nd stage) + joint_epoch: 15 # joint training starting epoch (2nd stage) + +optimizer_params: + lr: 0.0001 # general learning rate + bert_lr: 0.00001 # learning rate for PLBERT + ft_lr: 0.00001 # learning rate for acoustic modules + +slmadv_params: + min_len: 400 # minimum length of samples + max_len: 500 # maximum length of samples + batch_percentage: 0.5 # to prevent out of memory, only use half of the original batch size + iter: 20 # update the discriminator every this iterations of generator update + thresh: 5 # gradient norm above which the gradient is scaled + scale: 0.01 # gradient scaling factor for predictors from SLM discriminators + sig: 1.5 # sigma for differentiable duration modeling diff --git a/Configs/config_ft_single.yml b/Configs/config_ft_single.yml index 71531f4674a5ff1d2b557011e69d1be7768d90db..c8e9f97627d615c99ad8c35ea59ade6b78f9d644 100644 --- a/Configs/config_ft_single.yml +++ b/Configs/config_ft_single.yml @@ -1,18 +1,18 @@ # ─── GLOBAL ────────────────────────────────────────────────────────── -log_dir: logs/pod_90h_30k +log_dir: logs/pod_90h_30k_second_lr1 device: "cuda" -batch_size: 8 # 40 GB A100, fp16 +batch_size: 12 # 40 GB A100, fp16 max_len: 300 # ≈ 8 s (200 × 40 ms) epochs_1st: 25 # first-stage schedule -epochs_2nd: 15 # second-stage schedule (later) -save_freq: 2 +epochs_2nd: 20 # second-stage schedule (later) +save_freq: 1 log_interval: 50 # leave blank on first run -pretrained_model: /home/ubuntu/styletts2-ft/logs/pod_90h_30k/epoch_1st_0012.pth -second_stage_load_pretrained: false +pretrained_model: "" #"/workspace/styletts2/logs/pod_90h_30k/epoch_2nd_00003.pth" +second_stage_load_pretrained: true load_only_params: false # ─── PRE-PROCESS ───────────────────────────────────────────────────── @@ -25,11 +25,11 @@ preprocess_params: # ─── DATA ──────────────────────────────────────────────────────────── data_params: - root_path: /home/ubuntu/styletts2-ft/data/wavs - train_data: /home/ubuntu/styletts2-ft/data/train_list.txt - val_data: /home/ubuntu/styletts2-ft/data/val_list.txt + root_path: /workspace + train_data: /workspace/styletts2/data/train_list.txt + val_data: /workspace/styletts2/data/val_list.txt min_length: 50 # sample until texts with this size are obtained for OOD texts - OOD_data: /home/ubuntu/styletts2-ft/data/OOD_texts.txt + OOD_data: /workspace/styletts2/data/OOD_texts.txt # ─── LOSS SCHEDULE ────────────────────────────────────────────────── loss_params: @@ -48,14 +48,14 @@ loss_params: lambda_sty: 1. # style reconstruction loss (2nd stage) lambda_diff: 1. # score matching loss (2nd stage) - diff_epoch: 20 # style diffusion starting epoch (2nd stage) - joint_epoch: 50 # joint training starting epoch (2nd stage) + diff_epoch: 0 # style diffusion starting epoch (2nd stage) + joint_epoch: 0 # joint training starting epoch (2nd stage) # ─── OPTIMISER ────────────────────────────────────────────────────── optimizer_params: - lr: 0.0008 - bert_lr: 0.00002 - ft_lr: 0.0002 + lr: 0.0001 + bert_lr: 0.00001 + ft_lr: 0.00001 grad_accum_steps: 2 # ─── MODEL (core network & sub-modules) ───────────────────────────── @@ -105,7 +105,7 @@ F0_path: "Utils/JDC/bst.t7" ASR_config: "Utils/ASR/config.yml" ASR_path: "Utils/ASR/epoch_00080.pth" PLBERT_dir: 'Utils/PLBERT/' -first_stage_path: "" # filled automatically after this run +first_stage_path: "/workspace/styletts2/stage1_final.pth" # filled automatically after this run # ─── SLM ADVERSARIAL (ignored in stage-1, kept default) ───────────── slmadv_params: diff --git a/Demo/.ipynb_checkpoints/Inference_LibriTTS-checkpoint.ipynb b/Demo/.ipynb_checkpoints/Inference_LibriTTS-checkpoint.ipynb new file mode 100644 index 0000000000000000000000000000000000000000..4b85bf5f722d4f95bdeff1235f1873106b1ccf91 --- /dev/null +++ b/Demo/.ipynb_checkpoints/Inference_LibriTTS-checkpoint.ipynb @@ -0,0 +1,1155 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "id": "9adb7bd1", + "metadata": {}, + "source": [ + "# StyleTTS 2 Demo (LibriTTS)\n", + "\n", + "Before you run the following cells, please make sure you have downloaded [reference_audio.zip](https://huggingface.co/yl4579/StyleTTS2-LibriTTS/resolve/main/reference_audio.zip) and unzipped it under the `demo` folder." + ] + }, + { + "cell_type": "markdown", + "id": "6108384d", + "metadata": {}, + "source": [ + "### Utils" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "96e173bf", + "metadata": {}, + "outputs": [], + "source": [ + "import torch\n", + "torch.manual_seed(0)\n", + "torch.backends.cudnn.benchmark = False\n", + "torch.backends.cudnn.deterministic = True\n", + "\n", + "import random\n", + "random.seed(0)\n", + "\n", + "import numpy as np\n", + "np.random.seed(0)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "da84c60f", + "metadata": {}, + "outputs": [], + "source": [ + "%cd .." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "5a3ddcc8", + "metadata": {}, + "outputs": [], + "source": [ + "# load packages\n", + "import time\n", + "import random\n", + "import yaml\n", + "from munch import Munch\n", + "import numpy as np\n", + "import torch\n", + "from torch import nn\n", + "import torch.nn.functional as F\n", + "import torchaudio\n", + "import librosa\n", + "from nltk.tokenize import word_tokenize\n", + "\n", + "from models import *\n", + "from utils import *\n", + "from text_utils import TextCleaner\n", + "textclenaer = TextCleaner()\n", + "\n", + "%matplotlib inline" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "00ee05e1", + "metadata": {}, + "outputs": [], + "source": [ + "to_mel = torchaudio.transforms.MelSpectrogram(\n", + " n_mels=80, n_fft=2048, win_length=1200, hop_length=300)\n", + "mean, std = -4, 4\n", + "\n", + "def length_to_mask(lengths):\n", + " mask = torch.arange(lengths.max()).unsqueeze(0).expand(lengths.shape[0], -1).type_as(lengths)\n", + " mask = torch.gt(mask+1, lengths.unsqueeze(1))\n", + " return mask\n", + "\n", + "def preprocess(wave):\n", + " wave_tensor = torch.from_numpy(wave).float()\n", + " mel_tensor = to_mel(wave_tensor)\n", + " mel_tensor = (torch.log(1e-5 + mel_tensor.unsqueeze(0)) - mean) / std\n", + " return mel_tensor\n", + "\n", + "def compute_style(path):\n", + " wave, sr = librosa.load(path, sr=24000)\n", + " audio, index = librosa.effects.trim(wave, top_db=30)\n", + " if sr != 24000:\n", + " audio = librosa.resample(audio, sr, 24000)\n", + " mel_tensor = preprocess(audio).to(device)\n", + "\n", + " with torch.no_grad():\n", + " ref_s = model.style_encoder(mel_tensor.unsqueeze(1))\n", + " ref_p = model.predictor_encoder(mel_tensor.unsqueeze(1))\n", + "\n", + " return torch.cat([ref_s, ref_p], dim=1)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "bbdc04c0", + "metadata": {}, + "outputs": [], + "source": [ + "device = 'cuda' if torch.cuda.is_available() else 'cpu'" + ] + }, + { + "cell_type": "markdown", + "id": "7b9cecbe", + "metadata": {}, + "source": [ + "### Load models" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "64fc4c0f", + "metadata": {}, + "outputs": [], + "source": [ + "# load phonemizer\n", + "import phonemizer\n", + "global_phonemizer = phonemizer.backend.EspeakBackend(language='en-us', preserve_punctuation=True, with_stress=True)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "48e7b644", + "metadata": {}, + "outputs": [], + "source": [ + "config = yaml.safe_load(open(\"Models/LibriTTS/config.yml\"))\n", + "\n", + "# load pretrained ASR model\n", + "ASR_config = config.get('ASR_config', False)\n", + "ASR_path = config.get('ASR_path', False)\n", + "text_aligner = load_ASR_models(ASR_path, ASR_config)\n", + "\n", + "# load pretrained F0 model\n", + "F0_path = config.get('F0_path', False)\n", + "pitch_extractor = load_F0_models(F0_path)\n", + "\n", + "# load BERT model\n", + "from Utils.PLBERT.util import load_plbert\n", + "BERT_path = config.get('PLBERT_dir', False)\n", + "plbert = load_plbert(BERT_path)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ffc18cf7", + "metadata": {}, + "outputs": [], + "source": [ + "model_params = recursive_munch(config['model_params'])\n", + "model = build_model(model_params, text_aligner, pitch_extractor, plbert)\n", + "_ = [model[key].eval() for key in model]\n", + "_ = [model[key].to(device) for key in model]" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "64529d5c", + "metadata": {}, + "outputs": [], + "source": [ + "params_whole = torch.load(\"Models/LibriTTS/epochs_2nd_00020.pth\", map_location='cpu')\n", + "params = params_whole['net']" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "895d9706", + "metadata": {}, + "outputs": [], + "source": [ + "for key in model:\n", + " if key in params:\n", + " print('%s loaded' % key)\n", + " try:\n", + " model[key].load_state_dict(params[key])\n", + " except:\n", + " from collections import OrderedDict\n", + " state_dict = params[key]\n", + " new_state_dict = OrderedDict()\n", + " for k, v in state_dict.items():\n", + " name = k[7:] # remove `module.`\n", + " new_state_dict[name] = v\n", + " # load params\n", + " model[key].load_state_dict(new_state_dict, strict=False)\n", + "# except:\n", + "# _load(params[key], model[key])\n", + "_ = [model[key].eval() for key in model]" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "c1a59db2", + "metadata": {}, + "outputs": [], + "source": [ + "from Modules.diffusion.sampler import DiffusionSampler, ADPM2Sampler, KarrasSchedule" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "e30985ab", + "metadata": {}, + "outputs": [], + "source": [ + "sampler = DiffusionSampler(\n", + " model.diffusion.diffusion,\n", + " sampler=ADPM2Sampler(),\n", + " sigma_schedule=KarrasSchedule(sigma_min=0.0001, sigma_max=3.0, rho=9.0), # empirical parameters\n", + " clamp=False\n", + ")" + ] + }, + { + "cell_type": "markdown", + "id": "b803110e", + "metadata": {}, + "source": [ + "### Synthesize speech" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ca57469c", + "metadata": {}, + "outputs": [], + "source": [ + "def inference(text, ref_s, alpha = 0.3, beta = 0.7, diffusion_steps=5, embedding_scale=1):\n", + " text = text.strip()\n", + " ps = global_phonemizer.phonemize([text])\n", + " ps = word_tokenize(ps[0])\n", + " ps = ' '.join(ps)\n", + " tokens = textclenaer(ps)\n", + " tokens.insert(0, 0)\n", + " tokens = torch.LongTensor(tokens).to(device).unsqueeze(0)\n", + " \n", + " with torch.no_grad():\n", + " input_lengths = torch.LongTensor([tokens.shape[-1]]).to(device)\n", + " text_mask = length_to_mask(input_lengths).to(device)\n", + "\n", + " t_en = model.text_encoder(tokens, input_lengths, text_mask)\n", + " bert_dur = model.bert(tokens, attention_mask=(~text_mask).int())\n", + " d_en = model.bert_encoder(bert_dur).transpose(-1, -2) \n", + "\n", + " s_pred = sampler(noise = torch.randn((1, 256)).unsqueeze(1).to(device), \n", + " embedding=bert_dur,\n", + " embedding_scale=embedding_scale,\n", + " features=ref_s, # reference from the same speaker as the embedding\n", + " num_steps=diffusion_steps).squeeze(1)\n", + "\n", + "\n", + " s = s_pred[:, 128:]\n", + " ref = s_pred[:, :128]\n", + "\n", + " ref = alpha * ref + (1 - alpha) * ref_s[:, :128]\n", + " s = beta * s + (1 - beta) * ref_s[:, 128:]\n", + "\n", + " d = model.predictor.text_encoder(d_en, \n", + " s, input_lengths, text_mask)\n", + "\n", + " x, _ = model.predictor.lstm(d)\n", + " duration = model.predictor.duration_proj(x)\n", + "\n", + " duration = torch.sigmoid(duration).sum(axis=-1)\n", + " pred_dur = torch.round(duration.squeeze()).clamp(min=1)\n", + "\n", + "\n", + " pred_aln_trg = torch.zeros(input_lengths, int(pred_dur.sum().data))\n", + " c_frame = 0\n", + " for i in range(pred_aln_trg.size(0)):\n", + " pred_aln_trg[i, c_frame:c_frame + int(pred_dur[i].data)] = 1\n", + " c_frame += int(pred_dur[i].data)\n", + "\n", + " # encode prosody\n", + " en = (d.transpose(-1, -2) @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(en)\n", + " asr_new[:, :, 0] = en[:, :, 0]\n", + " asr_new[:, :, 1:] = en[:, :, 0:-1]\n", + " en = asr_new\n", + "\n", + " F0_pred, N_pred = model.predictor.F0Ntrain(en, s)\n", + "\n", + " asr = (t_en @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(asr)\n", + " asr_new[:, :, 0] = asr[:, :, 0]\n", + " asr_new[:, :, 1:] = asr[:, :, 0:-1]\n", + " asr = asr_new\n", + "\n", + " out = model.decoder(asr, \n", + " F0_pred, N_pred, ref.squeeze().unsqueeze(0))\n", + " \n", + " \n", + " return out.squeeze().cpu().numpy()[..., :-50] # weird pulse at the end of the model, need to be fixed later" + ] + }, + { + "cell_type": "markdown", + "id": "d438ef4f", + "metadata": {}, + "source": [ + "#### Basic synthesis (5 diffusion steps, seen speakers)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "cace9787", + "metadata": {}, + "outputs": [], + "source": [ + "text = ''' StyleTTS 2 is a text to speech model that leverages style diffusion and adversarial training with large speech language models to achieve human level text to speech synthesis. '''" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "7c88f461", + "metadata": {}, + "outputs": [], + "source": [ + "reference_dicts = {}\n", + "reference_dicts['696_92939'] = \"Demo/reference_audio/696_92939_000016_000006.wav\"\n", + "reference_dicts['1789_142896'] = \"Demo/reference_audio/1789_142896_000022_000005.wav\"" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "16e8ac60", + "metadata": {}, + "outputs": [], + "source": [ + "start = time.time()\n", + "noise = torch.randn(1,1,256).to(device)\n", + "for k, path in reference_dicts.items():\n", + " ref_s = compute_style(path)\n", + " \n", + " wav = inference(text, ref_s, alpha=0.3, beta=0.7, diffusion_steps=5, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print(f\"RTF = {rtf:5f}\")\n", + " import IPython.display as ipd\n", + " print(k + ' Synthesized:')\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print('Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "14838708", + "metadata": {}, + "source": [ + "#### With higher diffusion steps (more diverse)\n", + "\n", + "Since the sampler is ancestral, the higher the stpes, the more diverse the samples are, with the cost of slower synthesis speed." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "6fbff03b", + "metadata": {}, + "outputs": [], + "source": [ + "noise = torch.randn(1,1,256).to(device)\n", + "for k, path in reference_dicts.items():\n", + " ref_s = compute_style(path)\n", + " start = time.time()\n", + " wav = inference(text, ref_s, alpha=0.3, beta=0.7, diffusion_steps=10, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print(f\"RTF = {rtf:5f}\")\n", + " import IPython.display as ipd\n", + " print(k + ' Synthesized:')\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print(k + ' Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "7e6867fd", + "metadata": {}, + "source": [ + "#### Basic synthesis (5 diffusion steps, umseen speakers)\n", + "The following samples are to reproduce samples in [Section 4](https://styletts2.github.io/#libri) of the demo page. All spsakers are unseen during training. You can compare the generated samples to popular zero-shot TTS models like Vall-E and NaturalSpeech 2." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "f4e8faa0", + "metadata": {}, + "outputs": [], + "source": [ + "reference_dicts = {}\n", + "# format: (path, text)\n", + "reference_dicts['1221-135767'] = (\"Demo/reference_audio/1221-135767-0014.wav\", \"Yea, his honourable worship is within, but he hath a godly minister or two with him, and likewise a leech.\")\n", + "reference_dicts['5639-40744'] = (\"Demo/reference_audio/5639-40744-0020.wav\", \"Thus did this humane and right minded father comfort his unhappy daughter, and her mother embracing her again, did all she could to soothe her feelings.\")\n", + "reference_dicts['908-157963'] = (\"Demo/reference_audio/908-157963-0027.wav\", \"And lay me down in my cold bed and leave my shining lot.\")\n", + "reference_dicts['4077-13754'] = (\"Demo/reference_audio/4077-13754-0000.wav\", \"The army found the people in poverty and left them in comparative wealth.\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "653f1406", + "metadata": {}, + "outputs": [], + "source": [ + "noise = torch.randn(1,1,256).to(device)\n", + "for k, v in reference_dicts.items():\n", + " path, text = v\n", + " ref_s = compute_style(path)\n", + " start = time.time()\n", + " wav = inference(text, ref_s, alpha=0.3, beta=0.7, diffusion_steps=5, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print(f\"RTF = {rtf:5f}\")\n", + " import IPython.display as ipd\n", + " print(k + ' Synthesized: ' + text)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print(k + ' Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "141e91b3", + "metadata": {}, + "source": [ + "### Speech expressiveness\n", + "\n", + "The following section recreates the samples shown in [Section 6](https://styletts2.github.io/#emo) of the demo page. The speaker reference used is `1221-135767-0014.wav`, which is unseen during training. \n", + "\n", + "#### With `embedding_scale=1`\n", + "This is the classifier-free guidance scale. The higher the scale, the more conditional the style is to the input text and hence more emotional.\n", + "\n" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "81addda4", + "metadata": {}, + "outputs": [], + "source": [ + "ref_s = compute_style(\"Demo/reference_audio/1221-135767-0014.wav\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "be1b2a11", + "metadata": {}, + "outputs": [], + "source": [ + "texts = {}\n", + "texts['Happy'] = \"We are happy to invite you to join us on a journey to the past, where we will visit the most amazing monuments ever built by human hands.\"\n", + "texts['Sad'] = \"I am sorry to say that we have suffered a severe setback in our efforts to restore prosperity and confidence.\"\n", + "texts['Angry'] = \"The field of astronomy is a joke! Its theories are based on flawed observations and biased interpretations!\"\n", + "texts['Surprised'] = \"I can't believe it! You mean to tell me that you have discovered a new species of bacteria in this pond?\"\n", + "\n", + "for k,v in texts.items():\n", + " wav = inference(v, ref_s, diffusion_steps=10, alpha=0.3, beta=0.7, embedding_scale=1)\n", + " print(k + \": \")\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "96d262b8", + "metadata": {}, + "source": [ + "#### With `embedding_scale=2`" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "3e7d40b4", + "metadata": {}, + "outputs": [], + "source": [ + "texts = {}\n", + "texts['Happy'] = \"We are happy to invite you to join us on a journey to the past, where we will visit the most amazing monuments ever built by human hands.\"\n", + "texts['Sad'] = \"I am sorry to say that we have suffered a severe setback in our efforts to restore prosperity and confidence.\"\n", + "texts['Angry'] = \"The field of astronomy is a joke! Its theories are based on flawed observations and biased interpretations!\"\n", + "texts['Surprised'] = \"I can't believe it! You mean to tell me that you have discovered a new species of bacteria in this pond?\"\n", + "\n", + "for k,v in texts.items():\n", + " noise = torch.randn(1,1,256).to(device)\n", + " wav = inference(v, ref_s, diffusion_steps=10, alpha=0.3, beta=0.7, embedding_scale=2)\n", + " print(k + \": \")\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "402b2bd6", + "metadata": {}, + "source": [ + "#### With `embedding_scale=2, alpha = 0.5, beta = 0.9`\n", + "`alpha` and `beta` is the factor to determine much we use the style sampled based on the text instead of the reference. The higher the value of `alpha` and `beta`, the more suitable the style it is to the text but less similar to the reference. Using higher beta makes the synthesized speech more emotional, at the cost of lower similarity to the reference. `alpha` determines the timbre of the speaker while `beta` determines the prosody. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "599de5d5", + "metadata": {}, + "outputs": [], + "source": [ + "texts = {}\n", + "texts['Happy'] = \"We are happy to invite you to join us on a journey to the past, where we will visit the most amazing monuments ever built by human hands.\"\n", + "texts['Sad'] = \"I am sorry to say that we have suffered a severe setback in our efforts to restore prosperity and confidence.\"\n", + "texts['Angry'] = \"The field of astronomy is a joke! Its theories are based on flawed observations and biased interpretations!\"\n", + "texts['Surprised'] = \"I can't believe it! You mean to tell me that you have discovered a new species of bacteria in this pond?\"\n", + "\n", + "for k,v in texts.items():\n", + " noise = torch.randn(1,1,256).to(device)\n", + " wav = inference(v, ref_s, diffusion_steps=10, alpha=0.5, beta=0.9, embedding_scale=2)\n", + " print(k + \": \")\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "48548866", + "metadata": {}, + "source": [ + "### Zero-shot speaker adaptation\n", + "This section recreates the \"Acoustic Environment Maintenance\" and \"Speaker’s Emotion Maintenance\" demo in [Section 4](https://styletts2.github.io/#libri) of the demo page. You can compare the generated samples to popular zero-shot TTS models like Vall-E. Note that the model was trained only on LibriTTS, which is about 250 times fewer data compared to those used to trian Vall-E with similar or better effect for these maintainance. " + ] + }, + { + "cell_type": "markdown", + "id": "23e81572", + "metadata": {}, + "source": [ + "#### Acoustic Environment Maintenance\n", + "\n", + "Since we want to maintain the acoustic environment in the speaker (timbre), we set `alpha = 0` to make the speaker as closer to the reference as possible while only changing the prosody according to the text. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "8087bccb", + "metadata": {}, + "outputs": [], + "source": [ + "reference_dicts = {}\n", + "# format: (path, text)\n", + "reference_dicts['3'] = (\"Demo/reference_audio/3.wav\", \"As friends thing I definitely I've got more male friends.\")\n", + "reference_dicts['4'] = (\"Demo/reference_audio/4.wav\", \"Everything is run by computer but you got to know how to think before you can do a computer.\")\n", + "reference_dicts['5'] = (\"Demo/reference_audio/5.wav\", \"Then out in LA you guys got a whole another ball game within California to worry about.\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "1e99c200", + "metadata": {}, + "outputs": [], + "source": [ + "noise = torch.randn(1,1,256).to(device)\n", + "for k, v in reference_dicts.items():\n", + " path, text = v\n", + " ref_s = compute_style(path)\n", + " start = time.time()\n", + " wav = inference(text, ref_s, alpha=0.0, beta=0.5, diffusion_steps=5, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print(f\"RTF = {rtf:5f}\")\n", + " import IPython.display as ipd\n", + " print('Synthesized: ' + text)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print('Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "7d56505d", + "metadata": {}, + "source": [ + "#### Speaker’s Emotion Maintenance\n", + "\n", + "Since we want to maintain the emotion in the speaker (prosody), we set `beta = 0.1` to make the speaker as closer to the reference as possible while having some diversity thruogh the slight timbre change." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "f90179e7", + "metadata": {}, + "outputs": [], + "source": [ + "reference_dicts = {}\n", + "# format: (path, text)\n", + "reference_dicts['Anger'] = (\"Demo/reference_audio/anger.wav\", \"We have to reduce the number of plastic bags.\")\n", + "reference_dicts['Sleepy'] = (\"Demo/reference_audio/sleepy.wav\", \"We have to reduce the number of plastic bags.\")\n", + "reference_dicts['Amused'] = (\"Demo/reference_audio/amused.wav\", \"We have to reduce the number of plastic bags.\")\n", + "reference_dicts['Disgusted'] = (\"Demo/reference_audio/disgusted.wav\", \"We have to reduce the number of plastic bags.\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "2e6bdfed", + "metadata": {}, + "outputs": [], + "source": [ + "noise = torch.randn(1,1,256).to(device)\n", + "for k, v in reference_dicts.items():\n", + " path, text = v\n", + " ref_s = compute_style(path)\n", + " start = time.time()\n", + " wav = inference(text, ref_s, alpha=0.3, beta=0.1, diffusion_steps=10, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print(f\"RTF = {rtf:5f}\")\n", + " import IPython.display as ipd\n", + " print(k + ' Synthesized: ' + text)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print(k + ' Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "37ae3963", + "metadata": {}, + "source": [ + "### Longform Narration\n", + "\n", + "This section includes basic implementation of Algorithm 1 in the paper for consistent longform audio generation. The example passage is taken from [Section 5](https://styletts2.github.io/#long) of the demo page." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "f12a716b", + "metadata": {}, + "outputs": [], + "source": [ + "passage = '''If the supply of fruit is greater than the family needs, it may be made a source of income by sending the fresh fruit to the market if there is one near enough, or by preserving, canning, and making jelly for sale. To make such an enterprise a success the fruit and work must be first class. There is magic in the word \"Homemade,\" when the product appeals to the eye and the palate; but many careless and incompetent people have found to their sorrow that this word has not magic enough to float inferior goods on the market. As a rule large canning and preserving establishments are clean and have the best appliances, and they employ chemists and skilled labor. The home product must be very good to compete with the attractive goods that are sent out from such establishments. Yet for first class home made products there is a market in all large cities. All first-class grocers have customers who purchase such goods.'''" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "a1a38079", + "metadata": {}, + "outputs": [], + "source": [ + "def LFinference(text, s_prev, ref_s, alpha = 0.3, beta = 0.7, t = 0.7, diffusion_steps=5, embedding_scale=1):\n", + " text = text.strip()\n", + " ps = global_phonemizer.phonemize([text])\n", + " ps = word_tokenize(ps[0])\n", + " ps = ' '.join(ps)\n", + " ps = ps.replace('``', '\"')\n", + " ps = ps.replace(\"''\", '\"')\n", + "\n", + " tokens = textclenaer(ps)\n", + " tokens.insert(0, 0)\n", + " tokens = torch.LongTensor(tokens).to(device).unsqueeze(0)\n", + " \n", + " with torch.no_grad():\n", + " input_lengths = torch.LongTensor([tokens.shape[-1]]).to(device)\n", + " text_mask = length_to_mask(input_lengths).to(device)\n", + "\n", + " t_en = model.text_encoder(tokens, input_lengths, text_mask)\n", + " bert_dur = model.bert(tokens, attention_mask=(~text_mask).int())\n", + " d_en = model.bert_encoder(bert_dur).transpose(-1, -2) \n", + "\n", + " s_pred = sampler(noise = torch.randn((1, 256)).unsqueeze(1).to(device), \n", + " embedding=bert_dur,\n", + " embedding_scale=embedding_scale,\n", + " features=ref_s, # reference from the same speaker as the embedding\n", + " num_steps=diffusion_steps).squeeze(1)\n", + " \n", + " if s_prev is not None:\n", + " # convex combination of previous and current style\n", + " s_pred = t * s_prev + (1 - t) * s_pred\n", + " \n", + " s = s_pred[:, 128:]\n", + " ref = s_pred[:, :128]\n", + " \n", + " ref = alpha * ref + (1 - alpha) * ref_s[:, :128]\n", + " s = beta * s + (1 - beta) * ref_s[:, 128:]\n", + "\n", + " s_pred = torch.cat([ref, s], dim=-1)\n", + "\n", + " d = model.predictor.text_encoder(d_en, \n", + " s, input_lengths, text_mask)\n", + "\n", + " x, _ = model.predictor.lstm(d)\n", + " duration = model.predictor.duration_proj(x)\n", + "\n", + " duration = torch.sigmoid(duration).sum(axis=-1)\n", + " pred_dur = torch.round(duration.squeeze()).clamp(min=1)\n", + "\n", + "\n", + " pred_aln_trg = torch.zeros(input_lengths, int(pred_dur.sum().data))\n", + " c_frame = 0\n", + " for i in range(pred_aln_trg.size(0)):\n", + " pred_aln_trg[i, c_frame:c_frame + int(pred_dur[i].data)] = 1\n", + " c_frame += int(pred_dur[i].data)\n", + "\n", + " # encode prosody\n", + " en = (d.transpose(-1, -2) @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(en)\n", + " asr_new[:, :, 0] = en[:, :, 0]\n", + " asr_new[:, :, 1:] = en[:, :, 0:-1]\n", + " en = asr_new\n", + "\n", + " F0_pred, N_pred = model.predictor.F0Ntrain(en, s)\n", + "\n", + " asr = (t_en @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(asr)\n", + " asr_new[:, :, 0] = asr[:, :, 0]\n", + " asr_new[:, :, 1:] = asr[:, :, 0:-1]\n", + " asr = asr_new\n", + "\n", + " out = model.decoder(asr, \n", + " F0_pred, N_pred, ref.squeeze().unsqueeze(0))\n", + " \n", + " \n", + " return out.squeeze().cpu().numpy()[..., :-100], s_pred # weird pulse at the end of the model, need to be fixed later" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "e9088f7a", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "s_ref = compute_style(path)\n", + "sentences = passage.split('.') # simple split by comma\n", + "wavs = []\n", + "s_prev = None\n", + "for text in sentences:\n", + " if text.strip() == \"\": continue\n", + " text += '.' # add it back\n", + " \n", + " wav, s_prev = LFinference(text, \n", + " s_prev, \n", + " s_ref, \n", + " alpha = 0.3, \n", + " beta = 0.9, # make it more suitable for the text\n", + " t = 0.7, \n", + " diffusion_steps=10, embedding_scale=1.5)\n", + " wavs.append(wav)\n", + "print('Synthesized: ')\n", + "display(ipd.Audio(np.concatenate(wavs), rate=24000, normalize=False))\n", + "print('Reference: ')\n", + "display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "7517b657", + "metadata": {}, + "source": [ + "### Style Transfer\n", + "\n", + "The following section demostrates the style transfer capacity for unseen speakers in [Section 6](https://styletts2.github.io/#emo) of the demo page. For this, we set `alpha=0.5, beta = 0.9` for the most pronounced effects (mostly using the sampled style). " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ed95d0f7", + "metadata": {}, + "outputs": [], + "source": [ + "def STinference(text, ref_s, ref_text, alpha = 0.3, beta = 0.7, diffusion_steps=5, embedding_scale=1):\n", + " text = text.strip()\n", + " ps = global_phonemizer.phonemize([text])\n", + " ps = word_tokenize(ps[0])\n", + " ps = ' '.join(ps)\n", + "\n", + " tokens = textclenaer(ps)\n", + " tokens.insert(0, 0)\n", + " tokens = torch.LongTensor(tokens).to(device).unsqueeze(0)\n", + " \n", + " ref_text = ref_text.strip()\n", + " ps = global_phonemizer.phonemize([ref_text])\n", + " ps = word_tokenize(ps[0])\n", + " ps = ' '.join(ps)\n", + "\n", + " ref_tokens = textclenaer(ps)\n", + " ref_tokens.insert(0, 0)\n", + " ref_tokens = torch.LongTensor(ref_tokens).to(device).unsqueeze(0)\n", + " \n", + " \n", + " with torch.no_grad():\n", + " input_lengths = torch.LongTensor([tokens.shape[-1]]).to(device)\n", + " text_mask = length_to_mask(input_lengths).to(device)\n", + "\n", + " t_en = model.text_encoder(tokens, input_lengths, text_mask)\n", + " bert_dur = model.bert(tokens, attention_mask=(~text_mask).int())\n", + " d_en = model.bert_encoder(bert_dur).transpose(-1, -2) \n", + " \n", + " ref_input_lengths = torch.LongTensor([ref_tokens.shape[-1]]).to(device)\n", + " ref_text_mask = length_to_mask(ref_input_lengths).to(device)\n", + " ref_bert_dur = model.bert(ref_tokens, attention_mask=(~ref_text_mask).int())\n", + " s_pred = sampler(noise = torch.randn((1, 256)).unsqueeze(1).to(device), \n", + " embedding=bert_dur,\n", + " embedding_scale=embedding_scale,\n", + " features=ref_s, # reference from the same speaker as the embedding\n", + " num_steps=diffusion_steps).squeeze(1)\n", + "\n", + "\n", + " s = s_pred[:, 128:]\n", + " ref = s_pred[:, :128]\n", + "\n", + " ref = alpha * ref + (1 - alpha) * ref_s[:, :128]\n", + " s = beta * s + (1 - beta) * ref_s[:, 128:]\n", + "\n", + " d = model.predictor.text_encoder(d_en, \n", + " s, input_lengths, text_mask)\n", + "\n", + " x, _ = model.predictor.lstm(d)\n", + " duration = model.predictor.duration_proj(x)\n", + "\n", + " duration = torch.sigmoid(duration).sum(axis=-1)\n", + " pred_dur = torch.round(duration.squeeze()).clamp(min=1)\n", + "\n", + "\n", + " pred_aln_trg = torch.zeros(input_lengths, int(pred_dur.sum().data))\n", + " c_frame = 0\n", + " for i in range(pred_aln_trg.size(0)):\n", + " pred_aln_trg[i, c_frame:c_frame + int(pred_dur[i].data)] = 1\n", + " c_frame += int(pred_dur[i].data)\n", + "\n", + " # encode prosody\n", + " en = (d.transpose(-1, -2) @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(en)\n", + " asr_new[:, :, 0] = en[:, :, 0]\n", + " asr_new[:, :, 1:] = en[:, :, 0:-1]\n", + " en = asr_new\n", + "\n", + " F0_pred, N_pred = model.predictor.F0Ntrain(en, s)\n", + "\n", + " asr = (t_en @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(asr)\n", + " asr_new[:, :, 0] = asr[:, :, 0]\n", + " asr_new[:, :, 1:] = asr[:, :, 0:-1]\n", + " asr = asr_new\n", + "\n", + " out = model.decoder(asr, \n", + " F0_pred, N_pred, ref.squeeze().unsqueeze(0))\n", + " \n", + " \n", + " return out.squeeze().cpu().numpy()[..., :-50] # weird pulse at the end of the model, need to be fixed later" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ec3f0da4", + "metadata": {}, + "outputs": [], + "source": [ + "# reference texts to sample styles\n", + "\n", + "ref_texts = {}\n", + "ref_texts['Happy'] = \"We are happy to invite you to join us on a journey to the past, where we will visit the most amazing monuments ever built by human hands.\"\n", + "ref_texts['Sad'] = \"I am sorry to say that we have suffered a severe setback in our efforts to restore prosperity and confidence.\"\n", + "ref_texts['Angry'] = \"The field of astronomy is a joke! Its theories are based on flawed observations and biased interpretations!\"\n", + "ref_texts['Surprised'] = \"I can't believe it! You mean to tell me that you have discovered a new species of bacteria in this pond?\"" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "6d0a3825", + "metadata": { + "scrolled": false + }, + "outputs": [], + "source": [ + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "s_ref = compute_style(path)\n", + "\n", + "text = \"Yea, his honourable worship is within, but he hath a godly minister or two with him, and likewise a leech.\"\n", + "for k,v in ref_texts.items():\n", + " wav = STinference(text, s_ref, v, diffusion_steps=10, alpha=0.5, beta=0.9, embedding_scale=1.5)\n", + " print(k + \": \")\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "6750aed9", + "metadata": {}, + "source": [ + "### Speech diversity\n", + "\n", + "This section reproduces samples in [Section 7](https://styletts2.github.io/#var) of the demo page. \n", + "\n", + "`alpha` and `beta` determine the diversity of the synthesized speech. There are two extreme cases:\n", + "- If `alpha = 1` and `beta = 1`, the synthesized speech sounds the most dissimilar to the reference speaker, but it is also the most diverse (each time you synthesize a speech it will be totally different). \n", + "- If `alpha = 0` and `beta = 0`, the synthesized speech sounds the most siimlar to the reference speaker, but it is deterministic (i.e., the sampled style is not used for speech synthesis). \n" + ] + }, + { + "cell_type": "markdown", + "id": "f6ae0aa5", + "metadata": {}, + "source": [ + "#### Default setting (`alpha = 0.3, beta=0.7`)\n", + "This setting uses 70% of the reference timbre and 30% of the reference prosody and use the diffusion model to sample them based on the text. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "36dc0148", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "ref_s = compute_style(path)\n", + "\n", + "text = \"How much variation is there?\"\n", + "for _ in range(5):\n", + " wav = inference(text, ref_s, diffusion_steps=10, alpha=0.3, beta=0.7, embedding_scale=1)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "bf9ef421", + "metadata": {}, + "source": [ + "#### Less diverse setting (`alpha = 0.1, beta=0.3`)\n", + "This setting uses 90% of the reference timbre and 70% of the reference prosody. This makes it more similar to the reference speaker at cost of less diverse samples. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "9ba406bd", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "ref_s = compute_style(path)\n", + "\n", + "text = \"How much variation is there?\"\n", + "for _ in range(5):\n", + " wav = inference(text, ref_s, diffusion_steps=10, alpha=0.1, beta=0.3, embedding_scale=1)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "a38fe464", + "metadata": {}, + "source": [ + "#### More diverse setting (`alpha = 0.5, beta=0.95`)\n", + "This setting uses 50% of the reference timbre and 5% of the reference prosody (so it uses 100% of the sampled prosody, which makes it more diverse), but this makes it more dissimilar to the reference speaker. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "5f25bf94", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "ref_s = compute_style(path)\n", + "\n", + "text = \"How much variation is there?\"\n", + "for _ in range(5):\n", + " wav = inference(text, ref_s, diffusion_steps=10, alpha=0.5, beta=0.95, embedding_scale=1)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "21c3a071", + "metadata": {}, + "source": [ + "#### Extreme setting (`alpha = 1, beta=1`)\n", + "This setting uses 0% of the reference timbre and prosody and use the diffusion model to sample the entire style. This makes the speaker very dissimilar to the reference speaker. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "fff8bab1", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "ref_s = compute_style(path)\n", + "\n", + "text = \"How much variation is there?\"\n", + "for _ in range(5):\n", + " wav = inference(text, ref_s, diffusion_steps=10, alpha=1, beta=1, embedding_scale=1)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "a8741e5a", + "metadata": {}, + "source": [ + "#### No variation (`alpha = 0, beta=0`)\n", + "This setting uses 0% of the reference timbre and prosody and use the diffusion model to sample the entire style. This makes the speaker very similar to the reference speaker, but there is no variation. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "e55dd281", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "ref_s = compute_style(path)\n", + "\n", + "text = \"How much variation is there?\"\n", + "for _ in range(5):\n", + " wav = inference(text, ref_s, diffusion_steps=10, alpha=0, beta=0, embedding_scale=1)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "d5e86423", + "metadata": {}, + "source": [ + "### Extra fun!\n", + "\n", + "Here we clone some of the authors' voice of the StyleTTS 2 papers with a few seconds of the recording in the wild. None of the voices is in the dataset and all authors agreed to have their voices cloned here." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "6f558314", + "metadata": {}, + "outputs": [], + "source": [ + "text = ''' StyleTTS 2 is a text to speech model that leverages style diffusion and adversarial training with large speech language models to achieve human level text to speech synthesis. '''" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "caa5747c", + "metadata": {}, + "outputs": [], + "source": [ + "reference_dicts = {}\n", + "reference_dicts['Yinghao'] = \"Demo/reference_audio/Yinghao.wav\"\n", + "reference_dicts['Gavin'] = \"Demo/reference_audio/Gavin.wav\"\n", + "reference_dicts['Vinay'] = \"Demo/reference_audio/Vinay.wav\"\n", + "reference_dicts['Nima'] = \"Demo/reference_audio/Nima.wav\"" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "44a4cea1", + "metadata": { + "scrolled": false + }, + "outputs": [], + "source": [ + "start = time.time()\n", + "noise = torch.randn(1,1,256).to(device)\n", + "for k, path in reference_dicts.items():\n", + " ref_s = compute_style(path)\n", + " \n", + " wav = inference(text, ref_s, alpha=0.1, beta=0.5, diffusion_steps=5, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print('Speaker: ' + k)\n", + " import IPython.display as ipd\n", + " print('Synthesized:')\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print('Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "NLP", + "language": "python", + "name": "nlp" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.9.7" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} diff --git a/Demo/.ipynb_checkpoints/Inference_pod_90h_30k-checkpoint.ipynb b/Demo/.ipynb_checkpoints/Inference_pod_90h_30k-checkpoint.ipynb new file mode 100644 index 0000000000000000000000000000000000000000..4b85bf5f722d4f95bdeff1235f1873106b1ccf91 --- /dev/null +++ b/Demo/.ipynb_checkpoints/Inference_pod_90h_30k-checkpoint.ipynb @@ -0,0 +1,1155 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "id": "9adb7bd1", + "metadata": {}, + "source": [ + "# StyleTTS 2 Demo (LibriTTS)\n", + "\n", + "Before you run the following cells, please make sure you have downloaded [reference_audio.zip](https://huggingface.co/yl4579/StyleTTS2-LibriTTS/resolve/main/reference_audio.zip) and unzipped it under the `demo` folder." + ] + }, + { + "cell_type": "markdown", + "id": "6108384d", + "metadata": {}, + "source": [ + "### Utils" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "96e173bf", + "metadata": {}, + "outputs": [], + "source": [ + "import torch\n", + "torch.manual_seed(0)\n", + "torch.backends.cudnn.benchmark = False\n", + "torch.backends.cudnn.deterministic = True\n", + "\n", + "import random\n", + "random.seed(0)\n", + "\n", + "import numpy as np\n", + "np.random.seed(0)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "da84c60f", + "metadata": {}, + "outputs": [], + "source": [ + "%cd .." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "5a3ddcc8", + "metadata": {}, + "outputs": [], + "source": [ + "# load packages\n", + "import time\n", + "import random\n", + "import yaml\n", + "from munch import Munch\n", + "import numpy as np\n", + "import torch\n", + "from torch import nn\n", + "import torch.nn.functional as F\n", + "import torchaudio\n", + "import librosa\n", + "from nltk.tokenize import word_tokenize\n", + "\n", + "from models import *\n", + "from utils import *\n", + "from text_utils import TextCleaner\n", + "textclenaer = TextCleaner()\n", + "\n", + "%matplotlib inline" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "00ee05e1", + "metadata": {}, + "outputs": [], + "source": [ + "to_mel = torchaudio.transforms.MelSpectrogram(\n", + " n_mels=80, n_fft=2048, win_length=1200, hop_length=300)\n", + "mean, std = -4, 4\n", + "\n", + "def length_to_mask(lengths):\n", + " mask = torch.arange(lengths.max()).unsqueeze(0).expand(lengths.shape[0], -1).type_as(lengths)\n", + " mask = torch.gt(mask+1, lengths.unsqueeze(1))\n", + " return mask\n", + "\n", + "def preprocess(wave):\n", + " wave_tensor = torch.from_numpy(wave).float()\n", + " mel_tensor = to_mel(wave_tensor)\n", + " mel_tensor = (torch.log(1e-5 + mel_tensor.unsqueeze(0)) - mean) / std\n", + " return mel_tensor\n", + "\n", + "def compute_style(path):\n", + " wave, sr = librosa.load(path, sr=24000)\n", + " audio, index = librosa.effects.trim(wave, top_db=30)\n", + " if sr != 24000:\n", + " audio = librosa.resample(audio, sr, 24000)\n", + " mel_tensor = preprocess(audio).to(device)\n", + "\n", + " with torch.no_grad():\n", + " ref_s = model.style_encoder(mel_tensor.unsqueeze(1))\n", + " ref_p = model.predictor_encoder(mel_tensor.unsqueeze(1))\n", + "\n", + " return torch.cat([ref_s, ref_p], dim=1)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "bbdc04c0", + "metadata": {}, + "outputs": [], + "source": [ + "device = 'cuda' if torch.cuda.is_available() else 'cpu'" + ] + }, + { + "cell_type": "markdown", + "id": "7b9cecbe", + "metadata": {}, + "source": [ + "### Load models" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "64fc4c0f", + "metadata": {}, + "outputs": [], + "source": [ + "# load phonemizer\n", + "import phonemizer\n", + "global_phonemizer = phonemizer.backend.EspeakBackend(language='en-us', preserve_punctuation=True, with_stress=True)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "48e7b644", + "metadata": {}, + "outputs": [], + "source": [ + "config = yaml.safe_load(open(\"Models/LibriTTS/config.yml\"))\n", + "\n", + "# load pretrained ASR model\n", + "ASR_config = config.get('ASR_config', False)\n", + "ASR_path = config.get('ASR_path', False)\n", + "text_aligner = load_ASR_models(ASR_path, ASR_config)\n", + "\n", + "# load pretrained F0 model\n", + "F0_path = config.get('F0_path', False)\n", + "pitch_extractor = load_F0_models(F0_path)\n", + "\n", + "# load BERT model\n", + "from Utils.PLBERT.util import load_plbert\n", + "BERT_path = config.get('PLBERT_dir', False)\n", + "plbert = load_plbert(BERT_path)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ffc18cf7", + "metadata": {}, + "outputs": [], + "source": [ + "model_params = recursive_munch(config['model_params'])\n", + "model = build_model(model_params, text_aligner, pitch_extractor, plbert)\n", + "_ = [model[key].eval() for key in model]\n", + "_ = [model[key].to(device) for key in model]" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "64529d5c", + "metadata": {}, + "outputs": [], + "source": [ + "params_whole = torch.load(\"Models/LibriTTS/epochs_2nd_00020.pth\", map_location='cpu')\n", + "params = params_whole['net']" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "895d9706", + "metadata": {}, + "outputs": [], + "source": [ + "for key in model:\n", + " if key in params:\n", + " print('%s loaded' % key)\n", + " try:\n", + " model[key].load_state_dict(params[key])\n", + " except:\n", + " from collections import OrderedDict\n", + " state_dict = params[key]\n", + " new_state_dict = OrderedDict()\n", + " for k, v in state_dict.items():\n", + " name = k[7:] # remove `module.`\n", + " new_state_dict[name] = v\n", + " # load params\n", + " model[key].load_state_dict(new_state_dict, strict=False)\n", + "# except:\n", + "# _load(params[key], model[key])\n", + "_ = [model[key].eval() for key in model]" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "c1a59db2", + "metadata": {}, + "outputs": [], + "source": [ + "from Modules.diffusion.sampler import DiffusionSampler, ADPM2Sampler, KarrasSchedule" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "e30985ab", + "metadata": {}, + "outputs": [], + "source": [ + "sampler = DiffusionSampler(\n", + " model.diffusion.diffusion,\n", + " sampler=ADPM2Sampler(),\n", + " sigma_schedule=KarrasSchedule(sigma_min=0.0001, sigma_max=3.0, rho=9.0), # empirical parameters\n", + " clamp=False\n", + ")" + ] + }, + { + "cell_type": "markdown", + "id": "b803110e", + "metadata": {}, + "source": [ + "### Synthesize speech" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ca57469c", + "metadata": {}, + "outputs": [], + "source": [ + "def inference(text, ref_s, alpha = 0.3, beta = 0.7, diffusion_steps=5, embedding_scale=1):\n", + " text = text.strip()\n", + " ps = global_phonemizer.phonemize([text])\n", + " ps = word_tokenize(ps[0])\n", + " ps = ' '.join(ps)\n", + " tokens = textclenaer(ps)\n", + " tokens.insert(0, 0)\n", + " tokens = torch.LongTensor(tokens).to(device).unsqueeze(0)\n", + " \n", + " with torch.no_grad():\n", + " input_lengths = torch.LongTensor([tokens.shape[-1]]).to(device)\n", + " text_mask = length_to_mask(input_lengths).to(device)\n", + "\n", + " t_en = model.text_encoder(tokens, input_lengths, text_mask)\n", + " bert_dur = model.bert(tokens, attention_mask=(~text_mask).int())\n", + " d_en = model.bert_encoder(bert_dur).transpose(-1, -2) \n", + "\n", + " s_pred = sampler(noise = torch.randn((1, 256)).unsqueeze(1).to(device), \n", + " embedding=bert_dur,\n", + " embedding_scale=embedding_scale,\n", + " features=ref_s, # reference from the same speaker as the embedding\n", + " num_steps=diffusion_steps).squeeze(1)\n", + "\n", + "\n", + " s = s_pred[:, 128:]\n", + " ref = s_pred[:, :128]\n", + "\n", + " ref = alpha * ref + (1 - alpha) * ref_s[:, :128]\n", + " s = beta * s + (1 - beta) * ref_s[:, 128:]\n", + "\n", + " d = model.predictor.text_encoder(d_en, \n", + " s, input_lengths, text_mask)\n", + "\n", + " x, _ = model.predictor.lstm(d)\n", + " duration = model.predictor.duration_proj(x)\n", + "\n", + " duration = torch.sigmoid(duration).sum(axis=-1)\n", + " pred_dur = torch.round(duration.squeeze()).clamp(min=1)\n", + "\n", + "\n", + " pred_aln_trg = torch.zeros(input_lengths, int(pred_dur.sum().data))\n", + " c_frame = 0\n", + " for i in range(pred_aln_trg.size(0)):\n", + " pred_aln_trg[i, c_frame:c_frame + int(pred_dur[i].data)] = 1\n", + " c_frame += int(pred_dur[i].data)\n", + "\n", + " # encode prosody\n", + " en = (d.transpose(-1, -2) @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(en)\n", + " asr_new[:, :, 0] = en[:, :, 0]\n", + " asr_new[:, :, 1:] = en[:, :, 0:-1]\n", + " en = asr_new\n", + "\n", + " F0_pred, N_pred = model.predictor.F0Ntrain(en, s)\n", + "\n", + " asr = (t_en @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(asr)\n", + " asr_new[:, :, 0] = asr[:, :, 0]\n", + " asr_new[:, :, 1:] = asr[:, :, 0:-1]\n", + " asr = asr_new\n", + "\n", + " out = model.decoder(asr, \n", + " F0_pred, N_pred, ref.squeeze().unsqueeze(0))\n", + " \n", + " \n", + " return out.squeeze().cpu().numpy()[..., :-50] # weird pulse at the end of the model, need to be fixed later" + ] + }, + { + "cell_type": "markdown", + "id": "d438ef4f", + "metadata": {}, + "source": [ + "#### Basic synthesis (5 diffusion steps, seen speakers)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "cace9787", + "metadata": {}, + "outputs": [], + "source": [ + "text = ''' StyleTTS 2 is a text to speech model that leverages style diffusion and adversarial training with large speech language models to achieve human level text to speech synthesis. '''" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "7c88f461", + "metadata": {}, + "outputs": [], + "source": [ + "reference_dicts = {}\n", + "reference_dicts['696_92939'] = \"Demo/reference_audio/696_92939_000016_000006.wav\"\n", + "reference_dicts['1789_142896'] = \"Demo/reference_audio/1789_142896_000022_000005.wav\"" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "16e8ac60", + "metadata": {}, + "outputs": [], + "source": [ + "start = time.time()\n", + "noise = torch.randn(1,1,256).to(device)\n", + "for k, path in reference_dicts.items():\n", + " ref_s = compute_style(path)\n", + " \n", + " wav = inference(text, ref_s, alpha=0.3, beta=0.7, diffusion_steps=5, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print(f\"RTF = {rtf:5f}\")\n", + " import IPython.display as ipd\n", + " print(k + ' Synthesized:')\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print('Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "14838708", + "metadata": {}, + "source": [ + "#### With higher diffusion steps (more diverse)\n", + "\n", + "Since the sampler is ancestral, the higher the stpes, the more diverse the samples are, with the cost of slower synthesis speed." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "6fbff03b", + "metadata": {}, + "outputs": [], + "source": [ + "noise = torch.randn(1,1,256).to(device)\n", + "for k, path in reference_dicts.items():\n", + " ref_s = compute_style(path)\n", + " start = time.time()\n", + " wav = inference(text, ref_s, alpha=0.3, beta=0.7, diffusion_steps=10, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print(f\"RTF = {rtf:5f}\")\n", + " import IPython.display as ipd\n", + " print(k + ' Synthesized:')\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print(k + ' Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "7e6867fd", + "metadata": {}, + "source": [ + "#### Basic synthesis (5 diffusion steps, umseen speakers)\n", + "The following samples are to reproduce samples in [Section 4](https://styletts2.github.io/#libri) of the demo page. All spsakers are unseen during training. You can compare the generated samples to popular zero-shot TTS models like Vall-E and NaturalSpeech 2." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "f4e8faa0", + "metadata": {}, + "outputs": [], + "source": [ + "reference_dicts = {}\n", + "# format: (path, text)\n", + "reference_dicts['1221-135767'] = (\"Demo/reference_audio/1221-135767-0014.wav\", \"Yea, his honourable worship is within, but he hath a godly minister or two with him, and likewise a leech.\")\n", + "reference_dicts['5639-40744'] = (\"Demo/reference_audio/5639-40744-0020.wav\", \"Thus did this humane and right minded father comfort his unhappy daughter, and her mother embracing her again, did all she could to soothe her feelings.\")\n", + "reference_dicts['908-157963'] = (\"Demo/reference_audio/908-157963-0027.wav\", \"And lay me down in my cold bed and leave my shining lot.\")\n", + "reference_dicts['4077-13754'] = (\"Demo/reference_audio/4077-13754-0000.wav\", \"The army found the people in poverty and left them in comparative wealth.\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "653f1406", + "metadata": {}, + "outputs": [], + "source": [ + "noise = torch.randn(1,1,256).to(device)\n", + "for k, v in reference_dicts.items():\n", + " path, text = v\n", + " ref_s = compute_style(path)\n", + " start = time.time()\n", + " wav = inference(text, ref_s, alpha=0.3, beta=0.7, diffusion_steps=5, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print(f\"RTF = {rtf:5f}\")\n", + " import IPython.display as ipd\n", + " print(k + ' Synthesized: ' + text)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print(k + ' Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "141e91b3", + "metadata": {}, + "source": [ + "### Speech expressiveness\n", + "\n", + "The following section recreates the samples shown in [Section 6](https://styletts2.github.io/#emo) of the demo page. The speaker reference used is `1221-135767-0014.wav`, which is unseen during training. \n", + "\n", + "#### With `embedding_scale=1`\n", + "This is the classifier-free guidance scale. The higher the scale, the more conditional the style is to the input text and hence more emotional.\n", + "\n" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "81addda4", + "metadata": {}, + "outputs": [], + "source": [ + "ref_s = compute_style(\"Demo/reference_audio/1221-135767-0014.wav\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "be1b2a11", + "metadata": {}, + "outputs": [], + "source": [ + "texts = {}\n", + "texts['Happy'] = \"We are happy to invite you to join us on a journey to the past, where we will visit the most amazing monuments ever built by human hands.\"\n", + "texts['Sad'] = \"I am sorry to say that we have suffered a severe setback in our efforts to restore prosperity and confidence.\"\n", + "texts['Angry'] = \"The field of astronomy is a joke! Its theories are based on flawed observations and biased interpretations!\"\n", + "texts['Surprised'] = \"I can't believe it! You mean to tell me that you have discovered a new species of bacteria in this pond?\"\n", + "\n", + "for k,v in texts.items():\n", + " wav = inference(v, ref_s, diffusion_steps=10, alpha=0.3, beta=0.7, embedding_scale=1)\n", + " print(k + \": \")\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "96d262b8", + "metadata": {}, + "source": [ + "#### With `embedding_scale=2`" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "3e7d40b4", + "metadata": {}, + "outputs": [], + "source": [ + "texts = {}\n", + "texts['Happy'] = \"We are happy to invite you to join us on a journey to the past, where we will visit the most amazing monuments ever built by human hands.\"\n", + "texts['Sad'] = \"I am sorry to say that we have suffered a severe setback in our efforts to restore prosperity and confidence.\"\n", + "texts['Angry'] = \"The field of astronomy is a joke! Its theories are based on flawed observations and biased interpretations!\"\n", + "texts['Surprised'] = \"I can't believe it! You mean to tell me that you have discovered a new species of bacteria in this pond?\"\n", + "\n", + "for k,v in texts.items():\n", + " noise = torch.randn(1,1,256).to(device)\n", + " wav = inference(v, ref_s, diffusion_steps=10, alpha=0.3, beta=0.7, embedding_scale=2)\n", + " print(k + \": \")\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "402b2bd6", + "metadata": {}, + "source": [ + "#### With `embedding_scale=2, alpha = 0.5, beta = 0.9`\n", + "`alpha` and `beta` is the factor to determine much we use the style sampled based on the text instead of the reference. The higher the value of `alpha` and `beta`, the more suitable the style it is to the text but less similar to the reference. Using higher beta makes the synthesized speech more emotional, at the cost of lower similarity to the reference. `alpha` determines the timbre of the speaker while `beta` determines the prosody. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "599de5d5", + "metadata": {}, + "outputs": [], + "source": [ + "texts = {}\n", + "texts['Happy'] = \"We are happy to invite you to join us on a journey to the past, where we will visit the most amazing monuments ever built by human hands.\"\n", + "texts['Sad'] = \"I am sorry to say that we have suffered a severe setback in our efforts to restore prosperity and confidence.\"\n", + "texts['Angry'] = \"The field of astronomy is a joke! Its theories are based on flawed observations and biased interpretations!\"\n", + "texts['Surprised'] = \"I can't believe it! You mean to tell me that you have discovered a new species of bacteria in this pond?\"\n", + "\n", + "for k,v in texts.items():\n", + " noise = torch.randn(1,1,256).to(device)\n", + " wav = inference(v, ref_s, diffusion_steps=10, alpha=0.5, beta=0.9, embedding_scale=2)\n", + " print(k + \": \")\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "48548866", + "metadata": {}, + "source": [ + "### Zero-shot speaker adaptation\n", + "This section recreates the \"Acoustic Environment Maintenance\" and \"Speaker’s Emotion Maintenance\" demo in [Section 4](https://styletts2.github.io/#libri) of the demo page. You can compare the generated samples to popular zero-shot TTS models like Vall-E. Note that the model was trained only on LibriTTS, which is about 250 times fewer data compared to those used to trian Vall-E with similar or better effect for these maintainance. " + ] + }, + { + "cell_type": "markdown", + "id": "23e81572", + "metadata": {}, + "source": [ + "#### Acoustic Environment Maintenance\n", + "\n", + "Since we want to maintain the acoustic environment in the speaker (timbre), we set `alpha = 0` to make the speaker as closer to the reference as possible while only changing the prosody according to the text. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "8087bccb", + "metadata": {}, + "outputs": [], + "source": [ + "reference_dicts = {}\n", + "# format: (path, text)\n", + "reference_dicts['3'] = (\"Demo/reference_audio/3.wav\", \"As friends thing I definitely I've got more male friends.\")\n", + "reference_dicts['4'] = (\"Demo/reference_audio/4.wav\", \"Everything is run by computer but you got to know how to think before you can do a computer.\")\n", + "reference_dicts['5'] = (\"Demo/reference_audio/5.wav\", \"Then out in LA you guys got a whole another ball game within California to worry about.\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "1e99c200", + "metadata": {}, + "outputs": [], + "source": [ + "noise = torch.randn(1,1,256).to(device)\n", + "for k, v in reference_dicts.items():\n", + " path, text = v\n", + " ref_s = compute_style(path)\n", + " start = time.time()\n", + " wav = inference(text, ref_s, alpha=0.0, beta=0.5, diffusion_steps=5, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print(f\"RTF = {rtf:5f}\")\n", + " import IPython.display as ipd\n", + " print('Synthesized: ' + text)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print('Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "7d56505d", + "metadata": {}, + "source": [ + "#### Speaker’s Emotion Maintenance\n", + "\n", + "Since we want to maintain the emotion in the speaker (prosody), we set `beta = 0.1` to make the speaker as closer to the reference as possible while having some diversity thruogh the slight timbre change." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "f90179e7", + "metadata": {}, + "outputs": [], + "source": [ + "reference_dicts = {}\n", + "# format: (path, text)\n", + "reference_dicts['Anger'] = (\"Demo/reference_audio/anger.wav\", \"We have to reduce the number of plastic bags.\")\n", + "reference_dicts['Sleepy'] = (\"Demo/reference_audio/sleepy.wav\", \"We have to reduce the number of plastic bags.\")\n", + "reference_dicts['Amused'] = (\"Demo/reference_audio/amused.wav\", \"We have to reduce the number of plastic bags.\")\n", + "reference_dicts['Disgusted'] = (\"Demo/reference_audio/disgusted.wav\", \"We have to reduce the number of plastic bags.\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "2e6bdfed", + "metadata": {}, + "outputs": [], + "source": [ + "noise = torch.randn(1,1,256).to(device)\n", + "for k, v in reference_dicts.items():\n", + " path, text = v\n", + " ref_s = compute_style(path)\n", + " start = time.time()\n", + " wav = inference(text, ref_s, alpha=0.3, beta=0.1, diffusion_steps=10, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print(f\"RTF = {rtf:5f}\")\n", + " import IPython.display as ipd\n", + " print(k + ' Synthesized: ' + text)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print(k + ' Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "37ae3963", + "metadata": {}, + "source": [ + "### Longform Narration\n", + "\n", + "This section includes basic implementation of Algorithm 1 in the paper for consistent longform audio generation. The example passage is taken from [Section 5](https://styletts2.github.io/#long) of the demo page." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "f12a716b", + "metadata": {}, + "outputs": [], + "source": [ + "passage = '''If the supply of fruit is greater than the family needs, it may be made a source of income by sending the fresh fruit to the market if there is one near enough, or by preserving, canning, and making jelly for sale. To make such an enterprise a success the fruit and work must be first class. There is magic in the word \"Homemade,\" when the product appeals to the eye and the palate; but many careless and incompetent people have found to their sorrow that this word has not magic enough to float inferior goods on the market. As a rule large canning and preserving establishments are clean and have the best appliances, and they employ chemists and skilled labor. The home product must be very good to compete with the attractive goods that are sent out from such establishments. Yet for first class home made products there is a market in all large cities. All first-class grocers have customers who purchase such goods.'''" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "a1a38079", + "metadata": {}, + "outputs": [], + "source": [ + "def LFinference(text, s_prev, ref_s, alpha = 0.3, beta = 0.7, t = 0.7, diffusion_steps=5, embedding_scale=1):\n", + " text = text.strip()\n", + " ps = global_phonemizer.phonemize([text])\n", + " ps = word_tokenize(ps[0])\n", + " ps = ' '.join(ps)\n", + " ps = ps.replace('``', '\"')\n", + " ps = ps.replace(\"''\", '\"')\n", + "\n", + " tokens = textclenaer(ps)\n", + " tokens.insert(0, 0)\n", + " tokens = torch.LongTensor(tokens).to(device).unsqueeze(0)\n", + " \n", + " with torch.no_grad():\n", + " input_lengths = torch.LongTensor([tokens.shape[-1]]).to(device)\n", + " text_mask = length_to_mask(input_lengths).to(device)\n", + "\n", + " t_en = model.text_encoder(tokens, input_lengths, text_mask)\n", + " bert_dur = model.bert(tokens, attention_mask=(~text_mask).int())\n", + " d_en = model.bert_encoder(bert_dur).transpose(-1, -2) \n", + "\n", + " s_pred = sampler(noise = torch.randn((1, 256)).unsqueeze(1).to(device), \n", + " embedding=bert_dur,\n", + " embedding_scale=embedding_scale,\n", + " features=ref_s, # reference from the same speaker as the embedding\n", + " num_steps=diffusion_steps).squeeze(1)\n", + " \n", + " if s_prev is not None:\n", + " # convex combination of previous and current style\n", + " s_pred = t * s_prev + (1 - t) * s_pred\n", + " \n", + " s = s_pred[:, 128:]\n", + " ref = s_pred[:, :128]\n", + " \n", + " ref = alpha * ref + (1 - alpha) * ref_s[:, :128]\n", + " s = beta * s + (1 - beta) * ref_s[:, 128:]\n", + "\n", + " s_pred = torch.cat([ref, s], dim=-1)\n", + "\n", + " d = model.predictor.text_encoder(d_en, \n", + " s, input_lengths, text_mask)\n", + "\n", + " x, _ = model.predictor.lstm(d)\n", + " duration = model.predictor.duration_proj(x)\n", + "\n", + " duration = torch.sigmoid(duration).sum(axis=-1)\n", + " pred_dur = torch.round(duration.squeeze()).clamp(min=1)\n", + "\n", + "\n", + " pred_aln_trg = torch.zeros(input_lengths, int(pred_dur.sum().data))\n", + " c_frame = 0\n", + " for i in range(pred_aln_trg.size(0)):\n", + " pred_aln_trg[i, c_frame:c_frame + int(pred_dur[i].data)] = 1\n", + " c_frame += int(pred_dur[i].data)\n", + "\n", + " # encode prosody\n", + " en = (d.transpose(-1, -2) @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(en)\n", + " asr_new[:, :, 0] = en[:, :, 0]\n", + " asr_new[:, :, 1:] = en[:, :, 0:-1]\n", + " en = asr_new\n", + "\n", + " F0_pred, N_pred = model.predictor.F0Ntrain(en, s)\n", + "\n", + " asr = (t_en @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(asr)\n", + " asr_new[:, :, 0] = asr[:, :, 0]\n", + " asr_new[:, :, 1:] = asr[:, :, 0:-1]\n", + " asr = asr_new\n", + "\n", + " out = model.decoder(asr, \n", + " F0_pred, N_pred, ref.squeeze().unsqueeze(0))\n", + " \n", + " \n", + " return out.squeeze().cpu().numpy()[..., :-100], s_pred # weird pulse at the end of the model, need to be fixed later" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "e9088f7a", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "s_ref = compute_style(path)\n", + "sentences = passage.split('.') # simple split by comma\n", + "wavs = []\n", + "s_prev = None\n", + "for text in sentences:\n", + " if text.strip() == \"\": continue\n", + " text += '.' # add it back\n", + " \n", + " wav, s_prev = LFinference(text, \n", + " s_prev, \n", + " s_ref, \n", + " alpha = 0.3, \n", + " beta = 0.9, # make it more suitable for the text\n", + " t = 0.7, \n", + " diffusion_steps=10, embedding_scale=1.5)\n", + " wavs.append(wav)\n", + "print('Synthesized: ')\n", + "display(ipd.Audio(np.concatenate(wavs), rate=24000, normalize=False))\n", + "print('Reference: ')\n", + "display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "7517b657", + "metadata": {}, + "source": [ + "### Style Transfer\n", + "\n", + "The following section demostrates the style transfer capacity for unseen speakers in [Section 6](https://styletts2.github.io/#emo) of the demo page. For this, we set `alpha=0.5, beta = 0.9` for the most pronounced effects (mostly using the sampled style). " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ed95d0f7", + "metadata": {}, + "outputs": [], + "source": [ + "def STinference(text, ref_s, ref_text, alpha = 0.3, beta = 0.7, diffusion_steps=5, embedding_scale=1):\n", + " text = text.strip()\n", + " ps = global_phonemizer.phonemize([text])\n", + " ps = word_tokenize(ps[0])\n", + " ps = ' '.join(ps)\n", + "\n", + " tokens = textclenaer(ps)\n", + " tokens.insert(0, 0)\n", + " tokens = torch.LongTensor(tokens).to(device).unsqueeze(0)\n", + " \n", + " ref_text = ref_text.strip()\n", + " ps = global_phonemizer.phonemize([ref_text])\n", + " ps = word_tokenize(ps[0])\n", + " ps = ' '.join(ps)\n", + "\n", + " ref_tokens = textclenaer(ps)\n", + " ref_tokens.insert(0, 0)\n", + " ref_tokens = torch.LongTensor(ref_tokens).to(device).unsqueeze(0)\n", + " \n", + " \n", + " with torch.no_grad():\n", + " input_lengths = torch.LongTensor([tokens.shape[-1]]).to(device)\n", + " text_mask = length_to_mask(input_lengths).to(device)\n", + "\n", + " t_en = model.text_encoder(tokens, input_lengths, text_mask)\n", + " bert_dur = model.bert(tokens, attention_mask=(~text_mask).int())\n", + " d_en = model.bert_encoder(bert_dur).transpose(-1, -2) \n", + " \n", + " ref_input_lengths = torch.LongTensor([ref_tokens.shape[-1]]).to(device)\n", + " ref_text_mask = length_to_mask(ref_input_lengths).to(device)\n", + " ref_bert_dur = model.bert(ref_tokens, attention_mask=(~ref_text_mask).int())\n", + " s_pred = sampler(noise = torch.randn((1, 256)).unsqueeze(1).to(device), \n", + " embedding=bert_dur,\n", + " embedding_scale=embedding_scale,\n", + " features=ref_s, # reference from the same speaker as the embedding\n", + " num_steps=diffusion_steps).squeeze(1)\n", + "\n", + "\n", + " s = s_pred[:, 128:]\n", + " ref = s_pred[:, :128]\n", + "\n", + " ref = alpha * ref + (1 - alpha) * ref_s[:, :128]\n", + " s = beta * s + (1 - beta) * ref_s[:, 128:]\n", + "\n", + " d = model.predictor.text_encoder(d_en, \n", + " s, input_lengths, text_mask)\n", + "\n", + " x, _ = model.predictor.lstm(d)\n", + " duration = model.predictor.duration_proj(x)\n", + "\n", + " duration = torch.sigmoid(duration).sum(axis=-1)\n", + " pred_dur = torch.round(duration.squeeze()).clamp(min=1)\n", + "\n", + "\n", + " pred_aln_trg = torch.zeros(input_lengths, int(pred_dur.sum().data))\n", + " c_frame = 0\n", + " for i in range(pred_aln_trg.size(0)):\n", + " pred_aln_trg[i, c_frame:c_frame + int(pred_dur[i].data)] = 1\n", + " c_frame += int(pred_dur[i].data)\n", + "\n", + " # encode prosody\n", + " en = (d.transpose(-1, -2) @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(en)\n", + " asr_new[:, :, 0] = en[:, :, 0]\n", + " asr_new[:, :, 1:] = en[:, :, 0:-1]\n", + " en = asr_new\n", + "\n", + " F0_pred, N_pred = model.predictor.F0Ntrain(en, s)\n", + "\n", + " asr = (t_en @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(asr)\n", + " asr_new[:, :, 0] = asr[:, :, 0]\n", + " asr_new[:, :, 1:] = asr[:, :, 0:-1]\n", + " asr = asr_new\n", + "\n", + " out = model.decoder(asr, \n", + " F0_pred, N_pred, ref.squeeze().unsqueeze(0))\n", + " \n", + " \n", + " return out.squeeze().cpu().numpy()[..., :-50] # weird pulse at the end of the model, need to be fixed later" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ec3f0da4", + "metadata": {}, + "outputs": [], + "source": [ + "# reference texts to sample styles\n", + "\n", + "ref_texts = {}\n", + "ref_texts['Happy'] = \"We are happy to invite you to join us on a journey to the past, where we will visit the most amazing monuments ever built by human hands.\"\n", + "ref_texts['Sad'] = \"I am sorry to say that we have suffered a severe setback in our efforts to restore prosperity and confidence.\"\n", + "ref_texts['Angry'] = \"The field of astronomy is a joke! Its theories are based on flawed observations and biased interpretations!\"\n", + "ref_texts['Surprised'] = \"I can't believe it! You mean to tell me that you have discovered a new species of bacteria in this pond?\"" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "6d0a3825", + "metadata": { + "scrolled": false + }, + "outputs": [], + "source": [ + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "s_ref = compute_style(path)\n", + "\n", + "text = \"Yea, his honourable worship is within, but he hath a godly minister or two with him, and likewise a leech.\"\n", + "for k,v in ref_texts.items():\n", + " wav = STinference(text, s_ref, v, diffusion_steps=10, alpha=0.5, beta=0.9, embedding_scale=1.5)\n", + " print(k + \": \")\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "6750aed9", + "metadata": {}, + "source": [ + "### Speech diversity\n", + "\n", + "This section reproduces samples in [Section 7](https://styletts2.github.io/#var) of the demo page. \n", + "\n", + "`alpha` and `beta` determine the diversity of the synthesized speech. There are two extreme cases:\n", + "- If `alpha = 1` and `beta = 1`, the synthesized speech sounds the most dissimilar to the reference speaker, but it is also the most diverse (each time you synthesize a speech it will be totally different). \n", + "- If `alpha = 0` and `beta = 0`, the synthesized speech sounds the most siimlar to the reference speaker, but it is deterministic (i.e., the sampled style is not used for speech synthesis). \n" + ] + }, + { + "cell_type": "markdown", + "id": "f6ae0aa5", + "metadata": {}, + "source": [ + "#### Default setting (`alpha = 0.3, beta=0.7`)\n", + "This setting uses 70% of the reference timbre and 30% of the reference prosody and use the diffusion model to sample them based on the text. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "36dc0148", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "ref_s = compute_style(path)\n", + "\n", + "text = \"How much variation is there?\"\n", + "for _ in range(5):\n", + " wav = inference(text, ref_s, diffusion_steps=10, alpha=0.3, beta=0.7, embedding_scale=1)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "bf9ef421", + "metadata": {}, + "source": [ + "#### Less diverse setting (`alpha = 0.1, beta=0.3`)\n", + "This setting uses 90% of the reference timbre and 70% of the reference prosody. This makes it more similar to the reference speaker at cost of less diverse samples. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "9ba406bd", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "ref_s = compute_style(path)\n", + "\n", + "text = \"How much variation is there?\"\n", + "for _ in range(5):\n", + " wav = inference(text, ref_s, diffusion_steps=10, alpha=0.1, beta=0.3, embedding_scale=1)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "a38fe464", + "metadata": {}, + "source": [ + "#### More diverse setting (`alpha = 0.5, beta=0.95`)\n", + "This setting uses 50% of the reference timbre and 5% of the reference prosody (so it uses 100% of the sampled prosody, which makes it more diverse), but this makes it more dissimilar to the reference speaker. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "5f25bf94", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "ref_s = compute_style(path)\n", + "\n", + "text = \"How much variation is there?\"\n", + "for _ in range(5):\n", + " wav = inference(text, ref_s, diffusion_steps=10, alpha=0.5, beta=0.95, embedding_scale=1)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "21c3a071", + "metadata": {}, + "source": [ + "#### Extreme setting (`alpha = 1, beta=1`)\n", + "This setting uses 0% of the reference timbre and prosody and use the diffusion model to sample the entire style. This makes the speaker very dissimilar to the reference speaker. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "fff8bab1", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "ref_s = compute_style(path)\n", + "\n", + "text = \"How much variation is there?\"\n", + "for _ in range(5):\n", + " wav = inference(text, ref_s, diffusion_steps=10, alpha=1, beta=1, embedding_scale=1)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "a8741e5a", + "metadata": {}, + "source": [ + "#### No variation (`alpha = 0, beta=0`)\n", + "This setting uses 0% of the reference timbre and prosody and use the diffusion model to sample the entire style. This makes the speaker very similar to the reference speaker, but there is no variation. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "e55dd281", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "ref_s = compute_style(path)\n", + "\n", + "text = \"How much variation is there?\"\n", + "for _ in range(5):\n", + " wav = inference(text, ref_s, diffusion_steps=10, alpha=0, beta=0, embedding_scale=1)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "d5e86423", + "metadata": {}, + "source": [ + "### Extra fun!\n", + "\n", + "Here we clone some of the authors' voice of the StyleTTS 2 papers with a few seconds of the recording in the wild. None of the voices is in the dataset and all authors agreed to have their voices cloned here." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "6f558314", + "metadata": {}, + "outputs": [], + "source": [ + "text = ''' StyleTTS 2 is a text to speech model that leverages style diffusion and adversarial training with large speech language models to achieve human level text to speech synthesis. '''" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "caa5747c", + "metadata": {}, + "outputs": [], + "source": [ + "reference_dicts = {}\n", + "reference_dicts['Yinghao'] = \"Demo/reference_audio/Yinghao.wav\"\n", + "reference_dicts['Gavin'] = \"Demo/reference_audio/Gavin.wav\"\n", + "reference_dicts['Vinay'] = \"Demo/reference_audio/Vinay.wav\"\n", + "reference_dicts['Nima'] = \"Demo/reference_audio/Nima.wav\"" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "44a4cea1", + "metadata": { + "scrolled": false + }, + "outputs": [], + "source": [ + "start = time.time()\n", + "noise = torch.randn(1,1,256).to(device)\n", + "for k, path in reference_dicts.items():\n", + " ref_s = compute_style(path)\n", + " \n", + " wav = inference(text, ref_s, alpha=0.1, beta=0.5, diffusion_steps=5, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print('Speaker: ' + k)\n", + " import IPython.display as ipd\n", + " print('Synthesized:')\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print('Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "NLP", + "language": "python", + "name": "nlp" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.9.7" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} diff --git a/Demo/Inference_pod_90h_30k.ipynb b/Demo/Inference_pod_90h_30k.ipynb new file mode 100644 index 0000000000000000000000000000000000000000..730a833ae72bbacd440054c19a90a4938b74ee50 --- /dev/null +++ b/Demo/Inference_pod_90h_30k.ipynb @@ -0,0 +1,1360 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "id": "9adb7bd1", + "metadata": {}, + "source": [ + "# StyleTTS 2 Demo (LibriTTS)\n", + "\n", + "Before you run the following cells, please make sure you have downloaded [reference_audio.zip](https://huggingface.co/yl4579/StyleTTS2-LibriTTS/resolve/main/reference_audio.zip) and unzipped it under the `demo` folder." + ] + }, + { + "cell_type": "markdown", + "id": "6108384d", + "metadata": {}, + "source": [ + "### Utils" + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "id": "96e173bf", + "metadata": {}, + "outputs": [], + "source": [ + "import torch\n", + "torch.manual_seed(0)\n", + "torch.backends.cudnn.benchmark = False\n", + "torch.backends.cudnn.deterministic = True\n", + "\n", + "import random\n", + "random.seed(0)\n", + "\n", + "import numpy as np\n", + "np.random.seed(0)" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "id": "2458c639-10a0-4b57-8602-22bc893c5176", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Collecting git+https://github.com/resemble-ai/monotonic_align.git (from -r requirements.txt (line 17))\n", + " Cloning https://github.com/resemble-ai/monotonic_align.git to /tmp/pip-req-build-ps9pa2ga\n", + " Running command git clone --filter=blob:none --quiet https://github.com/resemble-ai/monotonic_align.git /tmp/pip-req-build-ps9pa2ga\n", + " Resolved https://github.com/resemble-ai/monotonic_align.git to commit c6e5e6cb19882164027eb6e35118e841eed9298e\n", + " Installing build dependencies ... \u001b[?25ldone\n", + "\u001b[?25h Getting requirements to build wheel ... \u001b[?25ldone\n", + "\u001b[?25h Preparing metadata (pyproject.toml) ... \u001b[?25ldone\n", + "\u001b[?25hCollecting SoundFile (from -r requirements.txt (line 1))\n", + " Using cached soundfile-0.13.1-py2.py3-none-manylinux_2_28_x86_64.whl.metadata (16 kB)\n", + "Requirement already satisfied: torchaudio in /venv/main/lib/python3.12/site-packages (from -r requirements.txt (line 2)) (2.6.0+cu126)\n", + "Collecting munch (from -r requirements.txt (line 3))\n", + " Using cached munch-4.0.0-py2.py3-none-any.whl.metadata (5.9 kB)\n", + "Requirement already satisfied: torch in /venv/main/lib/python3.12/site-packages (from -r requirements.txt (line 4)) (2.6.0+cu126)\n", + "Collecting pydub (from -r requirements.txt (line 5))\n", + " Using cached pydub-0.25.1-py2.py3-none-any.whl.metadata (1.4 kB)\n", + "Requirement already satisfied: pyyaml in /venv/main/lib/python3.12/site-packages (from -r requirements.txt (line 6)) (6.0.2)\n", + "Collecting librosa (from -r requirements.txt (line 7))\n", + " Using cached librosa-0.11.0-py3-none-any.whl.metadata (8.7 kB)\n", + "Collecting nltk (from -r requirements.txt (line 8))\n", + " Using cached nltk-3.9.1-py3-none-any.whl.metadata (2.9 kB)\n", + "Collecting matplotlib (from -r requirements.txt (line 9))\n", + " Downloading matplotlib-3.10.3-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (11 kB)\n", + "Collecting accelerate (from -r requirements.txt (line 10))\n", + " Using cached accelerate-1.7.0-py3-none-any.whl.metadata (19 kB)\n", + "Collecting transformers (from -r requirements.txt (line 11))\n", + " Using cached transformers-4.52.4-py3-none-any.whl.metadata (38 kB)\n", + "Collecting einops (from -r requirements.txt (line 12))\n", + " Using cached einops-0.8.1-py3-none-any.whl.metadata (13 kB)\n", + "Collecting einops-exts (from -r requirements.txt (line 13))\n", + " Using cached einops_exts-0.0.4-py3-none-any.whl.metadata (621 bytes)\n", + "Requirement already satisfied: tqdm in /venv/main/lib/python3.12/site-packages (from -r requirements.txt (line 14)) (4.67.1)\n", + "Collecting typing (from -r requirements.txt (line 15))\n", + " Using cached typing-3.7.4.3.tar.gz (78 kB)\n", + " Preparing metadata (setup.py) ... \u001b[?25ldone\n", + "\u001b[?25hRequirement already satisfied: typing-extensions in /venv/main/lib/python3.12/site-packages (from -r requirements.txt (line 16)) (4.13.2)\n", + "Collecting cffi>=1.0 (from SoundFile->-r requirements.txt (line 1))\n", + " Downloading cffi-1.17.1-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (1.5 kB)\n", + "Requirement already satisfied: numpy in /venv/main/lib/python3.12/site-packages (from SoundFile->-r requirements.txt (line 1)) (2.1.2)\n", + "Requirement already satisfied: filelock in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (3.18.0)\n", + "Requirement already satisfied: setuptools in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (70.2.0)\n", + "Requirement already satisfied: sympy==1.13.1 in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (1.13.1)\n", + "Requirement already satisfied: networkx in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (3.3)\n", + "Requirement already satisfied: jinja2 in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (3.1.4)\n", + "Requirement already satisfied: fsspec in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (2025.3.2)\n", + "Requirement already satisfied: nvidia-cuda-nvrtc-cu12==12.6.77 in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (12.6.77)\n", + "Requirement already satisfied: nvidia-cuda-runtime-cu12==12.6.77 in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (12.6.77)\n", + "Requirement already satisfied: nvidia-cuda-cupti-cu12==12.6.80 in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (12.6.80)\n", + "Requirement already satisfied: nvidia-cudnn-cu12==9.5.1.17 in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (9.5.1.17)\n", + "Requirement already satisfied: nvidia-cublas-cu12==12.6.4.1 in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (12.6.4.1)\n", + "Requirement already satisfied: nvidia-cufft-cu12==11.3.0.4 in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (11.3.0.4)\n", + "Requirement already satisfied: nvidia-curand-cu12==10.3.7.77 in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (10.3.7.77)\n", + "Requirement already satisfied: nvidia-cusolver-cu12==11.7.1.2 in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (11.7.1.2)\n", + "Requirement already satisfied: nvidia-cusparse-cu12==12.5.4.2 in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (12.5.4.2)\n", + "Requirement already satisfied: nvidia-cusparselt-cu12==0.6.3 in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (0.6.3)\n", + "Requirement already satisfied: nvidia-nccl-cu12==2.21.5 in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (2.21.5)\n", + "Requirement already satisfied: nvidia-nvtx-cu12==12.6.77 in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (12.6.77)\n", + "Requirement already satisfied: nvidia-nvjitlink-cu12==12.6.85 in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (12.6.85)\n", + "Requirement already satisfied: triton==3.2.0 in /venv/main/lib/python3.12/site-packages (from torch->-r requirements.txt (line 4)) (3.2.0)\n", + "Requirement already satisfied: mpmath<1.4,>=1.1.0 in /venv/main/lib/python3.12/site-packages (from sympy==1.13.1->torch->-r requirements.txt (line 4)) (1.3.0)\n", + "Collecting audioread>=2.1.9 (from librosa->-r requirements.txt (line 7))\n", + " Using cached audioread-3.0.1-py3-none-any.whl.metadata (8.4 kB)\n", + "Collecting numba>=0.51.0 (from librosa->-r requirements.txt (line 7))\n", + " Downloading numba-0.61.2-cp312-cp312-manylinux2014_x86_64.manylinux_2_17_x86_64.whl.metadata (2.8 kB)\n", + "Collecting scipy>=1.6.0 (from librosa->-r requirements.txt (line 7))\n", + " Downloading scipy-1.15.3-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (61 kB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m62.0/62.0 kB\u001b[0m \u001b[31m1.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0ma \u001b[36m0:00:01\u001b[0m\n", + "\u001b[?25hCollecting scikit-learn>=1.1.0 (from librosa->-r requirements.txt (line 7))\n", + " Downloading scikit_learn-1.7.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (17 kB)\n", + "Collecting joblib>=1.0 (from librosa->-r requirements.txt (line 7))\n", + " Using cached joblib-1.5.1-py3-none-any.whl.metadata (5.6 kB)\n", + "Requirement already satisfied: decorator>=4.3.0 in /venv/main/lib/python3.12/site-packages (from librosa->-r requirements.txt (line 7)) (5.2.1)\n", + "Collecting pooch>=1.1 (from librosa->-r requirements.txt (line 7))\n", + " Using cached pooch-1.8.2-py3-none-any.whl.metadata (10 kB)\n", + "Collecting soxr>=0.3.2 (from librosa->-r requirements.txt (line 7))\n", + " Downloading soxr-0.5.0.post1-cp312-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (5.6 kB)\n", + "Collecting lazy_loader>=0.1 (from librosa->-r requirements.txt (line 7))\n", + " Using cached lazy_loader-0.4-py3-none-any.whl.metadata (7.6 kB)\n", + "Collecting msgpack>=1.0 (from librosa->-r requirements.txt (line 7))\n", + " Downloading msgpack-1.1.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (8.4 kB)\n", + "Collecting click (from nltk->-r requirements.txt (line 8))\n", + " Using cached click-8.2.1-py3-none-any.whl.metadata (2.5 kB)\n", + "Collecting regex>=2021.8.3 (from nltk->-r requirements.txt (line 8))\n", + " Downloading regex-2024.11.6-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (40 kB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m40.5/40.5 kB\u001b[0m \u001b[31m4.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", + "\u001b[?25hCollecting contourpy>=1.0.1 (from matplotlib->-r requirements.txt (line 9))\n", + " Downloading contourpy-1.3.2-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (5.5 kB)\n", + "Collecting cycler>=0.10 (from matplotlib->-r requirements.txt (line 9))\n", + " Using cached cycler-0.12.1-py3-none-any.whl.metadata (3.8 kB)\n", + "Collecting fonttools>=4.22.0 (from matplotlib->-r requirements.txt (line 9))\n", + " Downloading fonttools-4.58.2-cp312-cp312-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (106 kB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m106.3/106.3 kB\u001b[0m \u001b[31m2.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0ma \u001b[36m0:00:01\u001b[0m\n", + "\u001b[?25hCollecting kiwisolver>=1.3.1 (from matplotlib->-r requirements.txt (line 9))\n", + " Downloading kiwisolver-1.4.8-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (6.2 kB)\n", + "Requirement already satisfied: packaging>=20.0 in /venv/main/lib/python3.12/site-packages (from matplotlib->-r requirements.txt (line 9)) (25.0)\n", + "Requirement already satisfied: pillow>=8 in /venv/main/lib/python3.12/site-packages (from matplotlib->-r requirements.txt (line 9)) (11.0.0)\n", + "Collecting pyparsing>=2.3.1 (from matplotlib->-r requirements.txt (line 9))\n", + " Using cached pyparsing-3.2.3-py3-none-any.whl.metadata (5.0 kB)\n", + "Requirement already satisfied: python-dateutil>=2.7 in /venv/main/lib/python3.12/site-packages (from matplotlib->-r requirements.txt (line 9)) (2.9.0.post0)\n", + "Requirement already satisfied: psutil in /venv/main/lib/python3.12/site-packages (from accelerate->-r requirements.txt (line 10)) (7.0.0)\n", + "Requirement already satisfied: huggingface-hub>=0.21.0 in /venv/main/lib/python3.12/site-packages (from accelerate->-r requirements.txt (line 10)) (0.30.2)\n", + "Collecting safetensors>=0.4.3 (from accelerate->-r requirements.txt (line 10))\n", + " Using cached safetensors-0.5.3-cp38-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (3.8 kB)\n", + "Requirement already satisfied: requests in /venv/main/lib/python3.12/site-packages (from transformers->-r requirements.txt (line 11)) (2.32.3)\n", + "Collecting tokenizers<0.22,>=0.21 (from transformers->-r requirements.txt (line 11))\n", + " Using cached tokenizers-0.21.1-cp39-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (6.8 kB)\n", + "Collecting pycparser (from cffi>=1.0->SoundFile->-r requirements.txt (line 1))\n", + " Using cached pycparser-2.22-py3-none-any.whl.metadata (943 bytes)\n", + "Collecting llvmlite<0.45,>=0.44.0dev0 (from numba>=0.51.0->librosa->-r requirements.txt (line 7))\n", + " Downloading llvmlite-0.44.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (5.0 kB)\n", + "Requirement already satisfied: platformdirs>=2.5.0 in /venv/main/lib/python3.12/site-packages (from pooch>=1.1->librosa->-r requirements.txt (line 7)) (4.3.7)\n", + "Requirement already satisfied: six>=1.5 in /venv/main/lib/python3.12/site-packages (from python-dateutil>=2.7->matplotlib->-r requirements.txt (line 9)) (1.17.0)\n", + "Requirement already satisfied: charset-normalizer<4,>=2 in /venv/main/lib/python3.12/site-packages (from requests->transformers->-r requirements.txt (line 11)) (3.4.1)\n", + "Requirement already satisfied: idna<4,>=2.5 in /venv/main/lib/python3.12/site-packages (from requests->transformers->-r requirements.txt (line 11)) (3.10)\n", + "Requirement already satisfied: urllib3<3,>=1.21.1 in /venv/main/lib/python3.12/site-packages (from requests->transformers->-r requirements.txt (line 11)) (2.4.0)\n", + "Requirement already satisfied: certifi>=2017.4.17 in /venv/main/lib/python3.12/site-packages (from requests->transformers->-r requirements.txt (line 11)) (2025.4.26)\n", + "Collecting threadpoolctl>=3.1.0 (from scikit-learn>=1.1.0->librosa->-r requirements.txt (line 7))\n", + " Using cached threadpoolctl-3.6.0-py3-none-any.whl.metadata (13 kB)\n", + "Requirement already satisfied: MarkupSafe>=2.0 in /venv/main/lib/python3.12/site-packages (from jinja2->torch->-r requirements.txt (line 4)) (2.1.5)\n", + "Using cached soundfile-0.13.1-py2.py3-none-manylinux_2_28_x86_64.whl (1.3 MB)\n", + "Using cached munch-4.0.0-py2.py3-none-any.whl (9.9 kB)\n", + "Using cached pydub-0.25.1-py2.py3-none-any.whl (32 kB)\n", + "Using cached librosa-0.11.0-py3-none-any.whl (260 kB)\n", + "Using cached nltk-3.9.1-py3-none-any.whl (1.5 MB)\n", + "Downloading matplotlib-3.10.3-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (8.6 MB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m8.6/8.6 MB\u001b[0m \u001b[31m28.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0ma \u001b[36m0:00:01\u001b[0m\n", + "\u001b[?25hUsing cached accelerate-1.7.0-py3-none-any.whl (362 kB)\n", + "Using cached transformers-4.52.4-py3-none-any.whl (10.5 MB)\n", + "Using cached einops-0.8.1-py3-none-any.whl (64 kB)\n", + "Using cached einops_exts-0.0.4-py3-none-any.whl (3.9 kB)\n", + "Using cached audioread-3.0.1-py3-none-any.whl (23 kB)\n", + "Downloading cffi-1.17.1-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (479 kB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m479.4/479.4 kB\u001b[0m \u001b[31m169.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", + "\u001b[?25hDownloading contourpy-1.3.2-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (323 kB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m323.7/323.7 kB\u001b[0m \u001b[31m127.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", + "\u001b[?25hUsing cached cycler-0.12.1-py3-none-any.whl (8.3 kB)\n", + "Downloading fonttools-4.58.2-cp312-cp312-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl (4.9 MB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m4.9/4.9 MB\u001b[0m \u001b[31m87.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0mta \u001b[36m0:00:01\u001b[0m\n", + "\u001b[?25hUsing cached joblib-1.5.1-py3-none-any.whl (307 kB)\n", + "Downloading kiwisolver-1.4.8-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (1.5 MB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m1.5/1.5 MB\u001b[0m \u001b[31m185.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", + "\u001b[?25hUsing cached lazy_loader-0.4-py3-none-any.whl (12 kB)\n", + "Downloading msgpack-1.1.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (401 kB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m401.4/401.4 kB\u001b[0m \u001b[31m192.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", + "\u001b[?25hDownloading numba-0.61.2-cp312-cp312-manylinux2014_x86_64.manylinux_2_17_x86_64.whl (3.9 MB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m3.9/3.9 MB\u001b[0m \u001b[31m42.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0mta \u001b[36m0:00:01\u001b[0m\n", + "\u001b[?25hUsing cached pooch-1.8.2-py3-none-any.whl (64 kB)\n", + "Using cached pyparsing-3.2.3-py3-none-any.whl (111 kB)\n", + "Downloading regex-2024.11.6-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (796 kB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m796.9/796.9 kB\u001b[0m \u001b[31m125.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", + "\u001b[?25hUsing cached safetensors-0.5.3-cp38-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (471 kB)\n", + "Downloading scikit_learn-1.7.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (12.5 MB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m12.5/12.5 MB\u001b[0m \u001b[31m43.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m00:01\u001b[0m00:01\u001b[0m\n", + "\u001b[?25hDownloading scipy-1.15.3-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (37.3 MB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m37.3/37.3 MB\u001b[0m \u001b[31m26.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m00:01\u001b[0m00:01\u001b[0m\n", + "\u001b[?25hDownloading soxr-0.5.0.post1-cp312-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (248 kB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m248.5/248.5 kB\u001b[0m \u001b[31m36.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", + "\u001b[?25hUsing cached tokenizers-0.21.1-cp39-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (3.0 MB)\n", + "Using cached click-8.2.1-py3-none-any.whl (102 kB)\n", + "Downloading llvmlite-0.44.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (42.4 MB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m42.4/42.4 MB\u001b[0m \u001b[31m14.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m00:01\u001b[0m00:01\u001b[0m\n", + "\u001b[?25hUsing cached threadpoolctl-3.6.0-py3-none-any.whl (18 kB)\n", + "Using cached pycparser-2.22-py3-none-any.whl (117 kB)\n", + "Building wheels for collected packages: typing, monotonic_align\n", + " Building wheel for typing (setup.py) ... \u001b[?25ldone\n", + "\u001b[?25h Created wheel for typing: filename=typing-3.7.4.3-py3-none-any.whl size=26304 sha256=7bd8523fe1f7cb4e20da87ee646956891addbdea2d87074f6bbf77fe282e8720\n", + " Stored in directory: /root/.cache/pip/wheels/12/98/52/2bffe242a9a487f00886e43b8ed8dac46456702e11a0d6abef\n", + " Building wheel for monotonic_align (pyproject.toml) ... \u001b[?25ldone\n", + "\u001b[?25h Created wheel for monotonic_align: filename=monotonic_align-1.2-cp312-cp312-linux_x86_64.whl size=1543517 sha256=dc9566d3e5a0656ebf939e760d934e0926d435f336db84e0019c7391576cd4cc\n", + " Stored in directory: /tmp/pip-ephem-wheel-cache-0gzg26zy/wheels/76/0a/37/00634137cd000799e060087bd1cb49a060ac6a48fc42a15488\n", + "Successfully built typing monotonic_align\n", + "Installing collected packages: pydub, typing, threadpoolctl, soxr, scipy, safetensors, regex, pyparsing, pycparser, munch, msgpack, monotonic_align, llvmlite, lazy_loader, kiwisolver, joblib, fonttools, einops, cycler, contourpy, click, audioread, scikit-learn, pooch, numba, nltk, matplotlib, einops-exts, cffi, tokenizers, SoundFile, transformers, librosa, accelerate\n", + "Successfully installed SoundFile-0.13.1 accelerate-1.7.0 audioread-3.0.1 cffi-1.17.1 click-8.2.1 contourpy-1.3.2 cycler-0.12.1 einops-0.8.1 einops-exts-0.0.4 fonttools-4.58.2 joblib-1.5.1 kiwisolver-1.4.8 lazy_loader-0.4 librosa-0.11.0 llvmlite-0.44.0 matplotlib-3.10.3 monotonic_align-1.2 msgpack-1.1.0 munch-4.0.0 nltk-3.9.1 numba-0.61.2 pooch-1.8.2 pycparser-2.22 pydub-0.25.1 pyparsing-3.2.3 regex-2024.11.6 safetensors-0.5.3 scikit-learn-1.7.0 scipy-1.15.3 soxr-0.5.0.post1 threadpoolctl-3.6.0 tokenizers-0.21.1 transformers-4.52.4 typing-3.7.4.3\n" + ] + } + ], + "source": [ + "!pip install -r requirements.txt" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "id": "da84c60f", + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "/workspace/styletts2\n" + ] + } + ], + "source": [ + "%cd .." + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "id": "5a3ddcc8", + "metadata": {}, + "outputs": [ + { + "ename": "ModuleNotFoundError", + "evalue": "No module named 'munch'", + "output_type": "error", + "traceback": [ + "\u001b[31m---------------------------------------------------------------------------\u001b[39m", + "\u001b[31mModuleNotFoundError\u001b[39m Traceback (most recent call last)", + "\u001b[36mCell\u001b[39m\u001b[36m \u001b[39m\u001b[32mIn[3]\u001b[39m\u001b[32m, line 5\u001b[39m\n\u001b[32m 3\u001b[39m \u001b[38;5;28;01mimport\u001b[39;00m\u001b[38;5;250m \u001b[39m\u001b[34;01mrandom\u001b[39;00m\n\u001b[32m 4\u001b[39m \u001b[38;5;28;01mimport\u001b[39;00m\u001b[38;5;250m \u001b[39m\u001b[34;01myaml\u001b[39;00m\n\u001b[32m----> \u001b[39m\u001b[32m5\u001b[39m \u001b[38;5;28;01mfrom\u001b[39;00m\u001b[38;5;250m \u001b[39m\u001b[34;01mmunch\u001b[39;00m\u001b[38;5;250m \u001b[39m\u001b[38;5;28;01mimport\u001b[39;00m Munch\n\u001b[32m 6\u001b[39m \u001b[38;5;28;01mimport\u001b[39;00m\u001b[38;5;250m \u001b[39m\u001b[34;01mnumpy\u001b[39;00m\u001b[38;5;250m \u001b[39m\u001b[38;5;28;01mas\u001b[39;00m\u001b[38;5;250m \u001b[39m\u001b[34;01mnp\u001b[39;00m\n\u001b[32m 7\u001b[39m \u001b[38;5;28;01mimport\u001b[39;00m\u001b[38;5;250m \u001b[39m\u001b[34;01mtorch\u001b[39;00m\n", + "\u001b[31mModuleNotFoundError\u001b[39m: No module named 'munch'" + ] + } + ], + "source": [ + "# load packages\n", + "import time\n", + "import random\n", + "import yaml\n", + "from munch import Munch\n", + "import numpy as np\n", + "import torch\n", + "from torch import nn\n", + "import torch.nn.functional as F\n", + "import torchaudio\n", + "import librosa\n", + "from nltk.tokenize import word_tokenize\n", + "\n", + "from models import *\n", + "from utils import *\n", + "from text_utils import TextCleaner\n", + "textclenaer = TextCleaner()\n", + "\n", + "%matplotlib inline" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "00ee05e1", + "metadata": {}, + "outputs": [], + "source": [ + "to_mel = torchaudio.transforms.MelSpectrogram(\n", + " n_mels=80, n_fft=2048, win_length=1200, hop_length=300)\n", + "mean, std = -4, 4\n", + "\n", + "def length_to_mask(lengths):\n", + " mask = torch.arange(lengths.max()).unsqueeze(0).expand(lengths.shape[0], -1).type_as(lengths)\n", + " mask = torch.gt(mask+1, lengths.unsqueeze(1))\n", + " return mask\n", + "\n", + "def preprocess(wave):\n", + " wave_tensor = torch.from_numpy(wave).float()\n", + " mel_tensor = to_mel(wave_tensor)\n", + " mel_tensor = (torch.log(1e-5 + mel_tensor.unsqueeze(0)) - mean) / std\n", + " return mel_tensor\n", + "\n", + "def compute_style(path):\n", + " wave, sr = librosa.load(path, sr=24000)\n", + " audio, index = librosa.effects.trim(wave, top_db=30)\n", + " if sr != 24000:\n", + " audio = librosa.resample(audio, sr, 24000)\n", + " mel_tensor = preprocess(audio).to(device)\n", + "\n", + " with torch.no_grad():\n", + " ref_s = model.style_encoder(mel_tensor.unsqueeze(1))\n", + " ref_p = model.predictor_encoder(mel_tensor.unsqueeze(1))\n", + "\n", + " return torch.cat([ref_s, ref_p], dim=1)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "bbdc04c0", + "metadata": {}, + "outputs": [], + "source": [ + "device = 'cuda' if torch.cuda.is_available() else 'cpu'" + ] + }, + { + "cell_type": "markdown", + "id": "7b9cecbe", + "metadata": {}, + "source": [ + "### Load models" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "64fc4c0f", + "metadata": {}, + "outputs": [], + "source": [ + "# load phonemizer\n", + "import phonemizer\n", + "global_phonemizer = phonemizer.backend.EspeakBackend(language='en-us', preserve_punctuation=True, with_stress=True)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "48e7b644", + "metadata": {}, + "outputs": [], + "source": [ + "config = yaml.safe_load(open(\"Models/LibriTTS/config.yml\"))\n", + "\n", + "# load pretrained ASR model\n", + "ASR_config = config.get('ASR_config', False)\n", + "ASR_path = config.get('ASR_path', False)\n", + "text_aligner = load_ASR_models(ASR_path, ASR_config)\n", + "\n", + "# load pretrained F0 model\n", + "F0_path = config.get('F0_path', False)\n", + "pitch_extractor = load_F0_models(F0_path)\n", + "\n", + "# load BERT model\n", + "from Utils.PLBERT.util import load_plbert\n", + "BERT_path = config.get('PLBERT_dir', False)\n", + "plbert = load_plbert(BERT_path)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ffc18cf7", + "metadata": {}, + "outputs": [], + "source": [ + "model_params = recursive_munch(config['model_params'])\n", + "model = build_model(model_params, text_aligner, pitch_extractor, plbert)\n", + "_ = [model[key].eval() for key in model]\n", + "_ = [model[key].to(device) for key in model]" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "64529d5c", + "metadata": {}, + "outputs": [], + "source": [ + "params_whole = torch.load(\"Models/LibriTTS/epochs_2nd_00020.pth\", map_location='cpu')\n", + "params = params_whole['net']" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "895d9706", + "metadata": {}, + "outputs": [], + "source": [ + "for key in model:\n", + " if key in params:\n", + " print('%s loaded' % key)\n", + " try:\n", + " model[key].load_state_dict(params[key])\n", + " except:\n", + " from collections import OrderedDict\n", + " state_dict = params[key]\n", + " new_state_dict = OrderedDict()\n", + " for k, v in state_dict.items():\n", + " name = k[7:] # remove `module.`\n", + " new_state_dict[name] = v\n", + " # load params\n", + " model[key].load_state_dict(new_state_dict, strict=False)\n", + "# except:\n", + "# _load(params[key], model[key])\n", + "_ = [model[key].eval() for key in model]" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "c1a59db2", + "metadata": {}, + "outputs": [], + "source": [ + "from Modules.diffusion.sampler import DiffusionSampler, ADPM2Sampler, KarrasSchedule" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "e30985ab", + "metadata": {}, + "outputs": [], + "source": [ + "sampler = DiffusionSampler(\n", + " model.diffusion.diffusion,\n", + " sampler=ADPM2Sampler(),\n", + " sigma_schedule=KarrasSchedule(sigma_min=0.0001, sigma_max=3.0, rho=9.0), # empirical parameters\n", + " clamp=False\n", + ")" + ] + }, + { + "cell_type": "markdown", + "id": "b803110e", + "metadata": {}, + "source": [ + "### Synthesize speech" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ca57469c", + "metadata": {}, + "outputs": [], + "source": [ + "def inference(text, ref_s, alpha = 0.3, beta = 0.7, diffusion_steps=5, embedding_scale=1):\n", + " text = text.strip()\n", + " ps = global_phonemizer.phonemize([text])\n", + " ps = word_tokenize(ps[0])\n", + " ps = ' '.join(ps)\n", + " tokens = textclenaer(ps)\n", + " tokens.insert(0, 0)\n", + " tokens = torch.LongTensor(tokens).to(device).unsqueeze(0)\n", + " \n", + " with torch.no_grad():\n", + " input_lengths = torch.LongTensor([tokens.shape[-1]]).to(device)\n", + " text_mask = length_to_mask(input_lengths).to(device)\n", + "\n", + " t_en = model.text_encoder(tokens, input_lengths, text_mask)\n", + " bert_dur = model.bert(tokens, attention_mask=(~text_mask).int())\n", + " d_en = model.bert_encoder(bert_dur).transpose(-1, -2) \n", + "\n", + " s_pred = sampler(noise = torch.randn((1, 256)).unsqueeze(1).to(device), \n", + " embedding=bert_dur,\n", + " embedding_scale=embedding_scale,\n", + " features=ref_s, # reference from the same speaker as the embedding\n", + " num_steps=diffusion_steps).squeeze(1)\n", + "\n", + "\n", + " s = s_pred[:, 128:]\n", + " ref = s_pred[:, :128]\n", + "\n", + " ref = alpha * ref + (1 - alpha) * ref_s[:, :128]\n", + " s = beta * s + (1 - beta) * ref_s[:, 128:]\n", + "\n", + " d = model.predictor.text_encoder(d_en, \n", + " s, input_lengths, text_mask)\n", + "\n", + " x, _ = model.predictor.lstm(d)\n", + " duration = model.predictor.duration_proj(x)\n", + "\n", + " duration = torch.sigmoid(duration).sum(axis=-1)\n", + " pred_dur = torch.round(duration.squeeze()).clamp(min=1)\n", + "\n", + "\n", + " pred_aln_trg = torch.zeros(input_lengths, int(pred_dur.sum().data))\n", + " c_frame = 0\n", + " for i in range(pred_aln_trg.size(0)):\n", + " pred_aln_trg[i, c_frame:c_frame + int(pred_dur[i].data)] = 1\n", + " c_frame += int(pred_dur[i].data)\n", + "\n", + " # encode prosody\n", + " en = (d.transpose(-1, -2) @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(en)\n", + " asr_new[:, :, 0] = en[:, :, 0]\n", + " asr_new[:, :, 1:] = en[:, :, 0:-1]\n", + " en = asr_new\n", + "\n", + " F0_pred, N_pred = model.predictor.F0Ntrain(en, s)\n", + "\n", + " asr = (t_en @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(asr)\n", + " asr_new[:, :, 0] = asr[:, :, 0]\n", + " asr_new[:, :, 1:] = asr[:, :, 0:-1]\n", + " asr = asr_new\n", + "\n", + " out = model.decoder(asr, \n", + " F0_pred, N_pred, ref.squeeze().unsqueeze(0))\n", + " \n", + " \n", + " return out.squeeze().cpu().numpy()[..., :-50] # weird pulse at the end of the model, need to be fixed later" + ] + }, + { + "cell_type": "markdown", + "id": "d438ef4f", + "metadata": {}, + "source": [ + "#### Basic synthesis (5 diffusion steps, seen speakers)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "cace9787", + "metadata": {}, + "outputs": [], + "source": [ + "text = ''' StyleTTS 2 is a text to speech model that leverages style diffusion and adversarial training with large speech language models to achieve human level text to speech synthesis. '''" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "7c88f461", + "metadata": {}, + "outputs": [], + "source": [ + "reference_dicts = {}\n", + "reference_dicts['696_92939'] = \"Demo/reference_audio/696_92939_000016_000006.wav\"\n", + "reference_dicts['1789_142896'] = \"Demo/reference_audio/1789_142896_000022_000005.wav\"" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "16e8ac60", + "metadata": {}, + "outputs": [], + "source": [ + "start = time.time()\n", + "noise = torch.randn(1,1,256).to(device)\n", + "for k, path in reference_dicts.items():\n", + " ref_s = compute_style(path)\n", + " \n", + " wav = inference(text, ref_s, alpha=0.3, beta=0.7, diffusion_steps=5, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print(f\"RTF = {rtf:5f}\")\n", + " import IPython.display as ipd\n", + " print(k + ' Synthesized:')\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print('Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "14838708", + "metadata": {}, + "source": [ + "#### With higher diffusion steps (more diverse)\n", + "\n", + "Since the sampler is ancestral, the higher the stpes, the more diverse the samples are, with the cost of slower synthesis speed." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "6fbff03b", + "metadata": {}, + "outputs": [], + "source": [ + "noise = torch.randn(1,1,256).to(device)\n", + "for k, path in reference_dicts.items():\n", + " ref_s = compute_style(path)\n", + " start = time.time()\n", + " wav = inference(text, ref_s, alpha=0.3, beta=0.7, diffusion_steps=10, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print(f\"RTF = {rtf:5f}\")\n", + " import IPython.display as ipd\n", + " print(k + ' Synthesized:')\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print(k + ' Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "7e6867fd", + "metadata": {}, + "source": [ + "#### Basic synthesis (5 diffusion steps, umseen speakers)\n", + "The following samples are to reproduce samples in [Section 4](https://styletts2.github.io/#libri) of the demo page. All spsakers are unseen during training. You can compare the generated samples to popular zero-shot TTS models like Vall-E and NaturalSpeech 2." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "f4e8faa0", + "metadata": {}, + "outputs": [], + "source": [ + "reference_dicts = {}\n", + "# format: (path, text)\n", + "reference_dicts['1221-135767'] = (\"Demo/reference_audio/1221-135767-0014.wav\", \"Yea, his honourable worship is within, but he hath a godly minister or two with him, and likewise a leech.\")\n", + "reference_dicts['5639-40744'] = (\"Demo/reference_audio/5639-40744-0020.wav\", \"Thus did this humane and right minded father comfort his unhappy daughter, and her mother embracing her again, did all she could to soothe her feelings.\")\n", + "reference_dicts['908-157963'] = (\"Demo/reference_audio/908-157963-0027.wav\", \"And lay me down in my cold bed and leave my shining lot.\")\n", + "reference_dicts['4077-13754'] = (\"Demo/reference_audio/4077-13754-0000.wav\", \"The army found the people in poverty and left them in comparative wealth.\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "653f1406", + "metadata": {}, + "outputs": [], + "source": [ + "noise = torch.randn(1,1,256).to(device)\n", + "for k, v in reference_dicts.items():\n", + " path, text = v\n", + " ref_s = compute_style(path)\n", + " start = time.time()\n", + " wav = inference(text, ref_s, alpha=0.3, beta=0.7, diffusion_steps=5, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print(f\"RTF = {rtf:5f}\")\n", + " import IPython.display as ipd\n", + " print(k + ' Synthesized: ' + text)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print(k + ' Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "141e91b3", + "metadata": {}, + "source": [ + "### Speech expressiveness\n", + "\n", + "The following section recreates the samples shown in [Section 6](https://styletts2.github.io/#emo) of the demo page. The speaker reference used is `1221-135767-0014.wav`, which is unseen during training. \n", + "\n", + "#### With `embedding_scale=1`\n", + "This is the classifier-free guidance scale. The higher the scale, the more conditional the style is to the input text and hence more emotional.\n", + "\n" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "81addda4", + "metadata": {}, + "outputs": [], + "source": [ + "ref_s = compute_style(\"Demo/reference_audio/1221-135767-0014.wav\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "be1b2a11", + "metadata": {}, + "outputs": [], + "source": [ + "texts = {}\n", + "texts['Happy'] = \"We are happy to invite you to join us on a journey to the past, where we will visit the most amazing monuments ever built by human hands.\"\n", + "texts['Sad'] = \"I am sorry to say that we have suffered a severe setback in our efforts to restore prosperity and confidence.\"\n", + "texts['Angry'] = \"The field of astronomy is a joke! Its theories are based on flawed observations and biased interpretations!\"\n", + "texts['Surprised'] = \"I can't believe it! You mean to tell me that you have discovered a new species of bacteria in this pond?\"\n", + "\n", + "for k,v in texts.items():\n", + " wav = inference(v, ref_s, diffusion_steps=10, alpha=0.3, beta=0.7, embedding_scale=1)\n", + " print(k + \": \")\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "96d262b8", + "metadata": {}, + "source": [ + "#### With `embedding_scale=2`" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "3e7d40b4", + "metadata": {}, + "outputs": [], + "source": [ + "texts = {}\n", + "texts['Happy'] = \"We are happy to invite you to join us on a journey to the past, where we will visit the most amazing monuments ever built by human hands.\"\n", + "texts['Sad'] = \"I am sorry to say that we have suffered a severe setback in our efforts to restore prosperity and confidence.\"\n", + "texts['Angry'] = \"The field of astronomy is a joke! Its theories are based on flawed observations and biased interpretations!\"\n", + "texts['Surprised'] = \"I can't believe it! You mean to tell me that you have discovered a new species of bacteria in this pond?\"\n", + "\n", + "for k,v in texts.items():\n", + " noise = torch.randn(1,1,256).to(device)\n", + " wav = inference(v, ref_s, diffusion_steps=10, alpha=0.3, beta=0.7, embedding_scale=2)\n", + " print(k + \": \")\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "402b2bd6", + "metadata": {}, + "source": [ + "#### With `embedding_scale=2, alpha = 0.5, beta = 0.9`\n", + "`alpha` and `beta` is the factor to determine much we use the style sampled based on the text instead of the reference. The higher the value of `alpha` and `beta`, the more suitable the style it is to the text but less similar to the reference. Using higher beta makes the synthesized speech more emotional, at the cost of lower similarity to the reference. `alpha` determines the timbre of the speaker while `beta` determines the prosody. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "599de5d5", + "metadata": {}, + "outputs": [], + "source": [ + "texts = {}\n", + "texts['Happy'] = \"We are happy to invite you to join us on a journey to the past, where we will visit the most amazing monuments ever built by human hands.\"\n", + "texts['Sad'] = \"I am sorry to say that we have suffered a severe setback in our efforts to restore prosperity and confidence.\"\n", + "texts['Angry'] = \"The field of astronomy is a joke! Its theories are based on flawed observations and biased interpretations!\"\n", + "texts['Surprised'] = \"I can't believe it! You mean to tell me that you have discovered a new species of bacteria in this pond?\"\n", + "\n", + "for k,v in texts.items():\n", + " noise = torch.randn(1,1,256).to(device)\n", + " wav = inference(v, ref_s, diffusion_steps=10, alpha=0.5, beta=0.9, embedding_scale=2)\n", + " print(k + \": \")\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "48548866", + "metadata": {}, + "source": [ + "### Zero-shot speaker adaptation\n", + "This section recreates the \"Acoustic Environment Maintenance\" and \"Speaker’s Emotion Maintenance\" demo in [Section 4](https://styletts2.github.io/#libri) of the demo page. You can compare the generated samples to popular zero-shot TTS models like Vall-E. Note that the model was trained only on LibriTTS, which is about 250 times fewer data compared to those used to trian Vall-E with similar or better effect for these maintainance. " + ] + }, + { + "cell_type": "markdown", + "id": "23e81572", + "metadata": {}, + "source": [ + "#### Acoustic Environment Maintenance\n", + "\n", + "Since we want to maintain the acoustic environment in the speaker (timbre), we set `alpha = 0` to make the speaker as closer to the reference as possible while only changing the prosody according to the text. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "8087bccb", + "metadata": {}, + "outputs": [], + "source": [ + "reference_dicts = {}\n", + "# format: (path, text)\n", + "reference_dicts['3'] = (\"Demo/reference_audio/3.wav\", \"As friends thing I definitely I've got more male friends.\")\n", + "reference_dicts['4'] = (\"Demo/reference_audio/4.wav\", \"Everything is run by computer but you got to know how to think before you can do a computer.\")\n", + "reference_dicts['5'] = (\"Demo/reference_audio/5.wav\", \"Then out in LA you guys got a whole another ball game within California to worry about.\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "1e99c200", + "metadata": {}, + "outputs": [], + "source": [ + "noise = torch.randn(1,1,256).to(device)\n", + "for k, v in reference_dicts.items():\n", + " path, text = v\n", + " ref_s = compute_style(path)\n", + " start = time.time()\n", + " wav = inference(text, ref_s, alpha=0.0, beta=0.5, diffusion_steps=5, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print(f\"RTF = {rtf:5f}\")\n", + " import IPython.display as ipd\n", + " print('Synthesized: ' + text)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print('Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "7d56505d", + "metadata": {}, + "source": [ + "#### Speaker’s Emotion Maintenance\n", + "\n", + "Since we want to maintain the emotion in the speaker (prosody), we set `beta = 0.1` to make the speaker as closer to the reference as possible while having some diversity thruogh the slight timbre change." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "f90179e7", + "metadata": {}, + "outputs": [], + "source": [ + "reference_dicts = {}\n", + "# format: (path, text)\n", + "reference_dicts['Anger'] = (\"Demo/reference_audio/anger.wav\", \"We have to reduce the number of plastic bags.\")\n", + "reference_dicts['Sleepy'] = (\"Demo/reference_audio/sleepy.wav\", \"We have to reduce the number of plastic bags.\")\n", + "reference_dicts['Amused'] = (\"Demo/reference_audio/amused.wav\", \"We have to reduce the number of plastic bags.\")\n", + "reference_dicts['Disgusted'] = (\"Demo/reference_audio/disgusted.wav\", \"We have to reduce the number of plastic bags.\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "2e6bdfed", + "metadata": {}, + "outputs": [], + "source": [ + "noise = torch.randn(1,1,256).to(device)\n", + "for k, v in reference_dicts.items():\n", + " path, text = v\n", + " ref_s = compute_style(path)\n", + " start = time.time()\n", + " wav = inference(text, ref_s, alpha=0.3, beta=0.1, diffusion_steps=10, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print(f\"RTF = {rtf:5f}\")\n", + " import IPython.display as ipd\n", + " print(k + ' Synthesized: ' + text)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print(k + ' Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "37ae3963", + "metadata": {}, + "source": [ + "### Longform Narration\n", + "\n", + "This section includes basic implementation of Algorithm 1 in the paper for consistent longform audio generation. The example passage is taken from [Section 5](https://styletts2.github.io/#long) of the demo page." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "f12a716b", + "metadata": {}, + "outputs": [], + "source": [ + "passage = '''If the supply of fruit is greater than the family needs, it may be made a source of income by sending the fresh fruit to the market if there is one near enough, or by preserving, canning, and making jelly for sale. To make such an enterprise a success the fruit and work must be first class. There is magic in the word \"Homemade,\" when the product appeals to the eye and the palate; but many careless and incompetent people have found to their sorrow that this word has not magic enough to float inferior goods on the market. As a rule large canning and preserving establishments are clean and have the best appliances, and they employ chemists and skilled labor. The home product must be very good to compete with the attractive goods that are sent out from such establishments. Yet for first class home made products there is a market in all large cities. All first-class grocers have customers who purchase such goods.'''" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "a1a38079", + "metadata": {}, + "outputs": [], + "source": [ + "def LFinference(text, s_prev, ref_s, alpha = 0.3, beta = 0.7, t = 0.7, diffusion_steps=5, embedding_scale=1):\n", + " text = text.strip()\n", + " ps = global_phonemizer.phonemize([text])\n", + " ps = word_tokenize(ps[0])\n", + " ps = ' '.join(ps)\n", + " ps = ps.replace('``', '\"')\n", + " ps = ps.replace(\"''\", '\"')\n", + "\n", + " tokens = textclenaer(ps)\n", + " tokens.insert(0, 0)\n", + " tokens = torch.LongTensor(tokens).to(device).unsqueeze(0)\n", + " \n", + " with torch.no_grad():\n", + " input_lengths = torch.LongTensor([tokens.shape[-1]]).to(device)\n", + " text_mask = length_to_mask(input_lengths).to(device)\n", + "\n", + " t_en = model.text_encoder(tokens, input_lengths, text_mask)\n", + " bert_dur = model.bert(tokens, attention_mask=(~text_mask).int())\n", + " d_en = model.bert_encoder(bert_dur).transpose(-1, -2) \n", + "\n", + " s_pred = sampler(noise = torch.randn((1, 256)).unsqueeze(1).to(device), \n", + " embedding=bert_dur,\n", + " embedding_scale=embedding_scale,\n", + " features=ref_s, # reference from the same speaker as the embedding\n", + " num_steps=diffusion_steps).squeeze(1)\n", + " \n", + " if s_prev is not None:\n", + " # convex combination of previous and current style\n", + " s_pred = t * s_prev + (1 - t) * s_pred\n", + " \n", + " s = s_pred[:, 128:]\n", + " ref = s_pred[:, :128]\n", + " \n", + " ref = alpha * ref + (1 - alpha) * ref_s[:, :128]\n", + " s = beta * s + (1 - beta) * ref_s[:, 128:]\n", + "\n", + " s_pred = torch.cat([ref, s], dim=-1)\n", + "\n", + " d = model.predictor.text_encoder(d_en, \n", + " s, input_lengths, text_mask)\n", + "\n", + " x, _ = model.predictor.lstm(d)\n", + " duration = model.predictor.duration_proj(x)\n", + "\n", + " duration = torch.sigmoid(duration).sum(axis=-1)\n", + " pred_dur = torch.round(duration.squeeze()).clamp(min=1)\n", + "\n", + "\n", + " pred_aln_trg = torch.zeros(input_lengths, int(pred_dur.sum().data))\n", + " c_frame = 0\n", + " for i in range(pred_aln_trg.size(0)):\n", + " pred_aln_trg[i, c_frame:c_frame + int(pred_dur[i].data)] = 1\n", + " c_frame += int(pred_dur[i].data)\n", + "\n", + " # encode prosody\n", + " en = (d.transpose(-1, -2) @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(en)\n", + " asr_new[:, :, 0] = en[:, :, 0]\n", + " asr_new[:, :, 1:] = en[:, :, 0:-1]\n", + " en = asr_new\n", + "\n", + " F0_pred, N_pred = model.predictor.F0Ntrain(en, s)\n", + "\n", + " asr = (t_en @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(asr)\n", + " asr_new[:, :, 0] = asr[:, :, 0]\n", + " asr_new[:, :, 1:] = asr[:, :, 0:-1]\n", + " asr = asr_new\n", + "\n", + " out = model.decoder(asr, \n", + " F0_pred, N_pred, ref.squeeze().unsqueeze(0))\n", + " \n", + " \n", + " return out.squeeze().cpu().numpy()[..., :-100], s_pred # weird pulse at the end of the model, need to be fixed later" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "e9088f7a", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "s_ref = compute_style(path)\n", + "sentences = passage.split('.') # simple split by comma\n", + "wavs = []\n", + "s_prev = None\n", + "for text in sentences:\n", + " if text.strip() == \"\": continue\n", + " text += '.' # add it back\n", + " \n", + " wav, s_prev = LFinference(text, \n", + " s_prev, \n", + " s_ref, \n", + " alpha = 0.3, \n", + " beta = 0.9, # make it more suitable for the text\n", + " t = 0.7, \n", + " diffusion_steps=10, embedding_scale=1.5)\n", + " wavs.append(wav)\n", + "print('Synthesized: ')\n", + "display(ipd.Audio(np.concatenate(wavs), rate=24000, normalize=False))\n", + "print('Reference: ')\n", + "display(ipd.Audio(path, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "7517b657", + "metadata": {}, + "source": [ + "### Style Transfer\n", + "\n", + "The following section demostrates the style transfer capacity for unseen speakers in [Section 6](https://styletts2.github.io/#emo) of the demo page. For this, we set `alpha=0.5, beta = 0.9` for the most pronounced effects (mostly using the sampled style). " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ed95d0f7", + "metadata": {}, + "outputs": [], + "source": [ + "def STinference(text, ref_s, ref_text, alpha = 0.3, beta = 0.7, diffusion_steps=5, embedding_scale=1):\n", + " text = text.strip()\n", + " ps = global_phonemizer.phonemize([text])\n", + " ps = word_tokenize(ps[0])\n", + " ps = ' '.join(ps)\n", + "\n", + " tokens = textclenaer(ps)\n", + " tokens.insert(0, 0)\n", + " tokens = torch.LongTensor(tokens).to(device).unsqueeze(0)\n", + " \n", + " ref_text = ref_text.strip()\n", + " ps = global_phonemizer.phonemize([ref_text])\n", + " ps = word_tokenize(ps[0])\n", + " ps = ' '.join(ps)\n", + "\n", + " ref_tokens = textclenaer(ps)\n", + " ref_tokens.insert(0, 0)\n", + " ref_tokens = torch.LongTensor(ref_tokens).to(device).unsqueeze(0)\n", + " \n", + " \n", + " with torch.no_grad():\n", + " input_lengths = torch.LongTensor([tokens.shape[-1]]).to(device)\n", + " text_mask = length_to_mask(input_lengths).to(device)\n", + "\n", + " t_en = model.text_encoder(tokens, input_lengths, text_mask)\n", + " bert_dur = model.bert(tokens, attention_mask=(~text_mask).int())\n", + " d_en = model.bert_encoder(bert_dur).transpose(-1, -2) \n", + " \n", + " ref_input_lengths = torch.LongTensor([ref_tokens.shape[-1]]).to(device)\n", + " ref_text_mask = length_to_mask(ref_input_lengths).to(device)\n", + " ref_bert_dur = model.bert(ref_tokens, attention_mask=(~ref_text_mask).int())\n", + " s_pred = sampler(noise = torch.randn((1, 256)).unsqueeze(1).to(device), \n", + " embedding=bert_dur,\n", + " embedding_scale=embedding_scale,\n", + " features=ref_s, # reference from the same speaker as the embedding\n", + " num_steps=diffusion_steps).squeeze(1)\n", + "\n", + "\n", + " s = s_pred[:, 128:]\n", + " ref = s_pred[:, :128]\n", + "\n", + " ref = alpha * ref + (1 - alpha) * ref_s[:, :128]\n", + " s = beta * s + (1 - beta) * ref_s[:, 128:]\n", + "\n", + " d = model.predictor.text_encoder(d_en, \n", + " s, input_lengths, text_mask)\n", + "\n", + " x, _ = model.predictor.lstm(d)\n", + " duration = model.predictor.duration_proj(x)\n", + "\n", + " duration = torch.sigmoid(duration).sum(axis=-1)\n", + " pred_dur = torch.round(duration.squeeze()).clamp(min=1)\n", + "\n", + "\n", + " pred_aln_trg = torch.zeros(input_lengths, int(pred_dur.sum().data))\n", + " c_frame = 0\n", + " for i in range(pred_aln_trg.size(0)):\n", + " pred_aln_trg[i, c_frame:c_frame + int(pred_dur[i].data)] = 1\n", + " c_frame += int(pred_dur[i].data)\n", + "\n", + " # encode prosody\n", + " en = (d.transpose(-1, -2) @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(en)\n", + " asr_new[:, :, 0] = en[:, :, 0]\n", + " asr_new[:, :, 1:] = en[:, :, 0:-1]\n", + " en = asr_new\n", + "\n", + " F0_pred, N_pred = model.predictor.F0Ntrain(en, s)\n", + "\n", + " asr = (t_en @ pred_aln_trg.unsqueeze(0).to(device))\n", + " if model_params.decoder.type == \"hifigan\":\n", + " asr_new = torch.zeros_like(asr)\n", + " asr_new[:, :, 0] = asr[:, :, 0]\n", + " asr_new[:, :, 1:] = asr[:, :, 0:-1]\n", + " asr = asr_new\n", + "\n", + " out = model.decoder(asr, \n", + " F0_pred, N_pred, ref.squeeze().unsqueeze(0))\n", + " \n", + " \n", + " return out.squeeze().cpu().numpy()[..., :-50] # weird pulse at the end of the model, need to be fixed later" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "ec3f0da4", + "metadata": {}, + "outputs": [], + "source": [ + "# reference texts to sample styles\n", + "\n", + "ref_texts = {}\n", + "ref_texts['Happy'] = \"We are happy to invite you to join us on a journey to the past, where we will visit the most amazing monuments ever built by human hands.\"\n", + "ref_texts['Sad'] = \"I am sorry to say that we have suffered a severe setback in our efforts to restore prosperity and confidence.\"\n", + "ref_texts['Angry'] = \"The field of astronomy is a joke! Its theories are based on flawed observations and biased interpretations!\"\n", + "ref_texts['Surprised'] = \"I can't believe it! You mean to tell me that you have discovered a new species of bacteria in this pond?\"" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "6d0a3825", + "metadata": {}, + "outputs": [], + "source": [ + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "s_ref = compute_style(path)\n", + "\n", + "text = \"Yea, his honourable worship is within, but he hath a godly minister or two with him, and likewise a leech.\"\n", + "for k,v in ref_texts.items():\n", + " wav = STinference(text, s_ref, v, diffusion_steps=10, alpha=0.5, beta=0.9, embedding_scale=1.5)\n", + " print(k + \": \")\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "6750aed9", + "metadata": {}, + "source": [ + "### Speech diversity\n", + "\n", + "This section reproduces samples in [Section 7](https://styletts2.github.io/#var) of the demo page. \n", + "\n", + "`alpha` and `beta` determine the diversity of the synthesized speech. There are two extreme cases:\n", + "- If `alpha = 1` and `beta = 1`, the synthesized speech sounds the most dissimilar to the reference speaker, but it is also the most diverse (each time you synthesize a speech it will be totally different). \n", + "- If `alpha = 0` and `beta = 0`, the synthesized speech sounds the most siimlar to the reference speaker, but it is deterministic (i.e., the sampled style is not used for speech synthesis). \n" + ] + }, + { + "cell_type": "markdown", + "id": "f6ae0aa5", + "metadata": {}, + "source": [ + "#### Default setting (`alpha = 0.3, beta=0.7`)\n", + "This setting uses 70% of the reference timbre and 30% of the reference prosody and use the diffusion model to sample them based on the text. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "36dc0148", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "ref_s = compute_style(path)\n", + "\n", + "text = \"How much variation is there?\"\n", + "for _ in range(5):\n", + " wav = inference(text, ref_s, diffusion_steps=10, alpha=0.3, beta=0.7, embedding_scale=1)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "bf9ef421", + "metadata": {}, + "source": [ + "#### Less diverse setting (`alpha = 0.1, beta=0.3`)\n", + "This setting uses 90% of the reference timbre and 70% of the reference prosody. This makes it more similar to the reference speaker at cost of less diverse samples. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "9ba406bd", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "ref_s = compute_style(path)\n", + "\n", + "text = \"How much variation is there?\"\n", + "for _ in range(5):\n", + " wav = inference(text, ref_s, diffusion_steps=10, alpha=0.1, beta=0.3, embedding_scale=1)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "a38fe464", + "metadata": {}, + "source": [ + "#### More diverse setting (`alpha = 0.5, beta=0.95`)\n", + "This setting uses 50% of the reference timbre and 5% of the reference prosody (so it uses 100% of the sampled prosody, which makes it more diverse), but this makes it more dissimilar to the reference speaker. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "5f25bf94", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "ref_s = compute_style(path)\n", + "\n", + "text = \"How much variation is there?\"\n", + "for _ in range(5):\n", + " wav = inference(text, ref_s, diffusion_steps=10, alpha=0.5, beta=0.95, embedding_scale=1)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "21c3a071", + "metadata": {}, + "source": [ + "#### Extreme setting (`alpha = 1, beta=1`)\n", + "This setting uses 0% of the reference timbre and prosody and use the diffusion model to sample the entire style. This makes the speaker very dissimilar to the reference speaker. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "fff8bab1", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "ref_s = compute_style(path)\n", + "\n", + "text = \"How much variation is there?\"\n", + "for _ in range(5):\n", + " wav = inference(text, ref_s, diffusion_steps=10, alpha=1, beta=1, embedding_scale=1)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "a8741e5a", + "metadata": {}, + "source": [ + "#### No variation (`alpha = 0, beta=0`)\n", + "This setting uses 0% of the reference timbre and prosody and use the diffusion model to sample the entire style. This makes the speaker very similar to the reference speaker, but there is no variation. " + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "e55dd281", + "metadata": {}, + "outputs": [], + "source": [ + "# unseen speaker\n", + "path = \"Demo/reference_audio/1221-135767-0014.wav\"\n", + "ref_s = compute_style(path)\n", + "\n", + "text = \"How much variation is there?\"\n", + "for _ in range(5):\n", + " wav = inference(text, ref_s, diffusion_steps=10, alpha=0, beta=0, embedding_scale=1)\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))" + ] + }, + { + "cell_type": "markdown", + "id": "d5e86423", + "metadata": {}, + "source": [ + "### Extra fun!\n", + "\n", + "Here we clone some of the authors' voice of the StyleTTS 2 papers with a few seconds of the recording in the wild. None of the voices is in the dataset and all authors agreed to have their voices cloned here." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "6f558314", + "metadata": {}, + "outputs": [], + "source": [ + "text = ''' StyleTTS 2 is a text to speech model that leverages style diffusion and adversarial training with large speech language models to achieve human level text to speech synthesis. '''" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "caa5747c", + "metadata": {}, + "outputs": [], + "source": [ + "reference_dicts = {}\n", + "reference_dicts['Yinghao'] = \"Demo/reference_audio/Yinghao.wav\"\n", + "reference_dicts['Gavin'] = \"Demo/reference_audio/Gavin.wav\"\n", + "reference_dicts['Vinay'] = \"Demo/reference_audio/Vinay.wav\"\n", + "reference_dicts['Nima'] = \"Demo/reference_audio/Nima.wav\"" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "id": "44a4cea1", + "metadata": {}, + "outputs": [], + "source": [ + "start = time.time()\n", + "noise = torch.randn(1,1,256).to(device)\n", + "for k, path in reference_dicts.items():\n", + " ref_s = compute_style(path)\n", + " \n", + " wav = inference(text, ref_s, alpha=0.1, beta=0.5, diffusion_steps=5, embedding_scale=1)\n", + " rtf = (time.time() - start) / (len(wav) / 24000)\n", + " print('Speaker: ' + k)\n", + " import IPython.display as ipd\n", + " print('Synthesized:')\n", + " display(ipd.Audio(wav, rate=24000, normalize=False))\n", + " print('Reference:')\n", + " display(ipd.Audio(path, rate=24000, normalize=False))" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python3 (main venv)", + "language": "python", + "name": "main" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.12.3" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} diff --git a/Modules/.ipynb_checkpoints/slmadv-checkpoint.py b/Modules/.ipynb_checkpoints/slmadv-checkpoint.py new file mode 100644 index 0000000000000000000000000000000000000000..54c560f38a381afe108848863fc1c62b9979bf5a --- /dev/null +++ b/Modules/.ipynb_checkpoints/slmadv-checkpoint.py @@ -0,0 +1,177 @@ +import torch +import numpy as np +import torch.nn.functional as F + + +class SLMAdversarialLoss(torch.nn.Module): + def __init__( + self, + model, + wl, + sampler, + min_len, + max_len, + batch_percentage=0.5, + skip_update=10, + sig=1.5, + ): + super().__init__() + self.model = model + self.wl = wl + self.sampler = sampler + + self.min_len = min_len + self.max_len = max_len + self.batch_percentage = batch_percentage + + self.sig = sig + self.skip_update = skip_update + + # ------------------------------------------------------------------ # + def forward( + self, + iters, + y_rec_gt, + y_rec_gt_pred, + waves, + mel_input_length, + ref_text, + ref_lengths, + use_ind, + s_trg, + ref_s=None, + ): + # ---- full-width mask (matches ref_text.size(1)) ---------------- + seq_len = ref_text.size(1) + text_mask = ( + torch.arange(seq_len, device=ref_text.device) + .unsqueeze(0) + >= ref_lengths.unsqueeze(1) + ) # shape [B, seq_len] + + bert_dur = self.model.bert(ref_text, attention_mask=(~text_mask).int()) + d_en = self.model.bert_encoder(bert_dur).transpose(-1, -2) + + # ----- style / prosody sampling --------------------------------- + if use_ind and np.random.rand() < 0.5: + s_preds = s_trg + else: + num_steps = np.random.randint(3, 5) + noise = torch.randn_like(s_trg).unsqueeze(1).to(ref_text.device) + sampler_kwargs = dict( + noise=noise, + embedding=bert_dur, + embedding_scale=1, + embedding_mask_proba=0.1, + num_steps=num_steps, + ) + if ref_s is not None: + sampler_kwargs["features"] = ref_s + s_preds = self.sampler(**sampler_kwargs).squeeze(1) + + s_dur, s = s_preds[:, 128:], s_preds[:, :128] + + # random alignment placeholder must match the *padded* token width + seq_len = ref_text.size(1) + rand_align = torch.randn(ref_text.size(0), seq_len, 2, device=ref_text.device) + + d, _ = self.model.predictor( + d_en, s_dur, ref_lengths, + rand_align, + text_mask, + ) + + # ----- differentiable duration modelling ----------------------- + attn_preds, output_lengths = [], [] + for _s2s_pred, _len in zip(d, ref_lengths): + _s2s_pred_org = _s2s_pred[: _len] + _s2s_pred_sig = torch.sigmoid(_s2s_pred_org) + _dur_pred = _s2s_pred_sig.sum(dim=-1) + + l = int(torch.round(_s2s_pred_sig.sum()).item()) + t = torch.arange(l, device=ref_text.device).unsqueeze(0).expand(_len, l) + loc = torch.cumsum(_dur_pred, dim=0) - _dur_pred / 2 + h = torch.exp(-0.5 * (t - (l - loc.unsqueeze(-1))) ** 2 / (self.sig**2)) + + out = F.conv1d( + _s2s_pred_org.unsqueeze(0), + h.unsqueeze(1), + padding=h.size(-1) - 1, + groups=int(_len), + )[..., :l] + attn_preds.append(F.softmax(out.squeeze(), dim=0)) + output_lengths.append(l) + + max_len = max(output_lengths) + + # ----- build full-width alignment matrix ----------------------- + with torch.no_grad(): + t_en = self.model.text_encoder(ref_text, ref_lengths, text_mask) + + seq_len = ref_text.size(1) + s2s_attn = torch.zeros( + len(ref_lengths), seq_len, max_len, device=ref_text.device + ) + for bib, (attn, L) in enumerate(zip(attn_preds, output_lengths)): + s2s_attn[bib, : ref_lengths[bib], :L] = attn + + asr_pred = t_en @ s2s_attn + + _, p_pred = self.model.predictor( + d_en, s_dur, ref_lengths, s2s_attn, text_mask + ) + + # ----- clip extraction ----------------------------------------- + mel_len = max(int(min(output_lengths) / 2 - 1), self.min_len // 2) + mel_len = min(mel_len, self.max_len // 2) + + en, p_en, sp, wav = [], [], [], [] + for bib, L_pred in enumerate(output_lengths): + L_gt = int(mel_input_length[bib].item() / 2) + if L_gt <= mel_len or L_pred <= mel_len: + continue + + sp.append(s_preds[bib]) + + start = np.random.randint(0, L_pred - mel_len) + en.append(asr_pred[bib, :, start : start + mel_len]) + p_en.append(p_pred[bib, :, start : start + mel_len]) + + start_gt = np.random.randint(0, L_gt - mel_len) + y = waves[bib][(start_gt * 2) * 300 : ((start_gt + mel_len) * 2) * 300] + wav.append(torch.from_numpy(y).to(ref_text.device)) + + if len(wav) >= self.batch_percentage * len(waves): + break + + if len(sp) <= 1: + return None + + sp = torch.stack(sp) + wav = torch.stack(wav).float() + en = torch.stack(en) + p_en = torch.stack(p_en) + + F0_fake, N_fake = self.model.predictor.F0Ntrain(p_en, sp[:, 128:]) + y_pred = self.model.decoder(en, F0_fake, N_fake, sp[:, :128]) + + # -------------- adversarial losses ----------------------------- + if (iters + 1) % self.skip_update == 0: + d_loss = self.wl.discriminator(wav.squeeze(), y_pred.detach().squeeze()).mean() + else: + d_loss = 0 + + gen_loss = self.wl.generator(y_pred.squeeze()).mean() + return d_loss, gen_loss, y_pred.detach().cpu().numpy() + + +# ------------------------------------------------------------------ # +def length_to_mask(lengths: torch.Tensor) -> torch.Tensor: + """Classic length mask: 1 → PAD, 0 → real token.""" + max_len = lengths.max() + mask = ( + torch.arange(max_len, device=lengths.device) + .unsqueeze(0) + .expand(lengths.size(0), -1) + ) + return mask >= lengths.unsqueeze(1) \ No newline at end of file diff --git a/Modules/slmadv.py b/Modules/slmadv.py index 11acb9149598993be23793925f34b35f659f8d6e..54c560f38a381afe108848863fc1c62b9979bf5a 100644 --- a/Modules/slmadv.py +++ b/Modules/slmadv.py @@ -2,194 +2,176 @@ import torch import numpy as np import torch.nn.functional as F -class SLMAdversarialLoss(torch.nn.Module): - def __init__(self, model, wl, sampler, min_len, max_len, batch_percentage=0.5, skip_update=10, sig=1.5): - super(SLMAdversarialLoss, self).__init__() +class SLMAdversarialLoss(torch.nn.Module): + def __init__( + self, + model, + wl, + sampler, + min_len, + max_len, + batch_percentage=0.5, + skip_update=10, + sig=1.5, + ): + super().__init__() self.model = model self.wl = wl self.sampler = sampler - + self.min_len = min_len self.max_len = max_len self.batch_percentage = batch_percentage - + self.sig = sig self.skip_update = skip_update - - def forward(self, iters, y_rec_gt, y_rec_gt_pred, waves, mel_input_length, ref_text, ref_lengths, use_ind, s_trg, ref_s=None): - text_mask = length_to_mask(ref_lengths).to(ref_text.device) + + # ------------------------------------------------------------------ # + def forward( + self, + iters, + y_rec_gt, + y_rec_gt_pred, + waves, + mel_input_length, + ref_text, + ref_lengths, + use_ind, + s_trg, + ref_s=None, + ): + # ---- full-width mask (matches ref_text.size(1)) ---------------- + seq_len = ref_text.size(1) + text_mask = ( + torch.arange(seq_len, device=ref_text.device) + .unsqueeze(0) + >= ref_lengths.unsqueeze(1) + ) # shape [B, seq_len] + bert_dur = self.model.bert(ref_text, attention_mask=(~text_mask).int()) - d_en = self.model.bert_encoder(bert_dur).transpose(-1, -2) - + d_en = self.model.bert_encoder(bert_dur).transpose(-1, -2) + + # ----- style / prosody sampling --------------------------------- if use_ind and np.random.rand() < 0.5: s_preds = s_trg else: num_steps = np.random.randint(3, 5) + noise = torch.randn_like(s_trg).unsqueeze(1).to(ref_text.device) + sampler_kwargs = dict( + noise=noise, + embedding=bert_dur, + embedding_scale=1, + embedding_mask_proba=0.1, + num_steps=num_steps, + ) if ref_s is not None: - s_preds = self.sampler(noise = torch.randn_like(s_trg).unsqueeze(1).to(ref_text.device), - embedding=bert_dur, - embedding_scale=1, - features=ref_s, # reference from the same speaker as the embedding - embedding_mask_proba=0.1, - num_steps=num_steps).squeeze(1) - else: - s_preds = self.sampler(noise = torch.randn_like(s_trg).unsqueeze(1).to(ref_text.device), - embedding=bert_dur, - embedding_scale=1, - embedding_mask_proba=0.1, - num_steps=num_steps).squeeze(1) - - s_dur = s_preds[:, 128:] - s = s_preds[:, :128] - - d, _ = self.model.predictor(d_en, s_dur, - ref_lengths, - torch.randn(ref_lengths.shape[0], ref_lengths.max(), 2).to(ref_text.device), - text_mask) - - bib = 0 - - output_lengths = [] - attn_preds = [] - - # differentiable duration modeling - for _s2s_pred, _text_length in zip(d, ref_lengths): - - _s2s_pred_org = _s2s_pred[:_text_length, :] - - _s2s_pred = torch.sigmoid(_s2s_pred_org) - _dur_pred = _s2s_pred.sum(axis=-1) - - l = int(torch.round(_s2s_pred.sum()).item()) - t = torch.arange(0, l).expand(l) - - t = torch.arange(0, l).unsqueeze(0).expand((len(_s2s_pred), l)).to(ref_text.device) + sampler_kwargs["features"] = ref_s + s_preds = self.sampler(**sampler_kwargs).squeeze(1) + + s_dur, s = s_preds[:, 128:], s_preds[:, :128] + + # random alignment placeholder must match the *padded* token width + seq_len = ref_text.size(1) + rand_align = torch.randn(ref_text.size(0), seq_len, 2, device=ref_text.device) + + d, _ = self.model.predictor( + d_en, s_dur, ref_lengths, + rand_align, + text_mask, + ) + + # ----- differentiable duration modelling ----------------------- + attn_preds, output_lengths = [], [] + for _s2s_pred, _len in zip(d, ref_lengths): + _s2s_pred_org = _s2s_pred[: _len] + _s2s_pred_sig = torch.sigmoid(_s2s_pred_org) + _dur_pred = _s2s_pred_sig.sum(dim=-1) + + l = int(torch.round(_s2s_pred_sig.sum()).item()) + t = torch.arange(l, device=ref_text.device).unsqueeze(0).expand(_len, l) loc = torch.cumsum(_dur_pred, dim=0) - _dur_pred / 2 - - h = torch.exp(-0.5 * torch.square(t - (l - loc.unsqueeze(-1))) / (self.sig)**2) - - out = torch.nn.functional.conv1d(_s2s_pred_org.unsqueeze(0), - h.unsqueeze(1), - padding=h.shape[-1] - 1, groups=int(_text_length))[..., :l] + h = torch.exp(-0.5 * (t - (l - loc.unsqueeze(-1))) ** 2 / (self.sig**2)) + + out = F.conv1d( + _s2s_pred_org.unsqueeze(0), + h.unsqueeze(1), + padding=h.size(-1) - 1, + groups=int(_len), + )[..., :l] attn_preds.append(F.softmax(out.squeeze(), dim=0)) - output_lengths.append(l) max_len = max(output_lengths) - + + # ----- build full-width alignment matrix ----------------------- with torch.no_grad(): t_en = self.model.text_encoder(ref_text, ref_lengths, text_mask) - - s2s_attn = torch.zeros(len(ref_lengths), int(ref_lengths.max()), max_len).to(ref_text.device) - for bib in range(len(output_lengths)): - s2s_attn[bib, :ref_lengths[bib], :output_lengths[bib]] = attn_preds[bib] + + seq_len = ref_text.size(1) + s2s_attn = torch.zeros( + len(ref_lengths), seq_len, max_len, device=ref_text.device + ) + for bib, (attn, L) in enumerate(zip(attn_preds, output_lengths)): + s2s_attn[bib, : ref_lengths[bib], :L] = attn asr_pred = t_en @ s2s_attn - _, p_pred = self.model.predictor(d_en, s_dur, - ref_lengths, - s2s_attn, - text_mask) - + _, p_pred = self.model.predictor( + d_en, s_dur, ref_lengths, s2s_attn, text_mask + ) + + # ----- clip extraction ----------------------------------------- mel_len = max(int(min(output_lengths) / 2 - 1), self.min_len // 2) mel_len = min(mel_len, self.max_len // 2) - - # get clips - - en = [] - p_en = [] - sp = [] - - F0_fakes = [] - N_fakes = [] - - wav = [] - - for bib in range(len(output_lengths)): - mel_length_pred = output_lengths[bib] - mel_length_gt = int(mel_input_length[bib].item() / 2) - if mel_length_gt <= mel_len or mel_length_pred <= mel_len: + + en, p_en, sp, wav = [], [], [], [] + for bib, L_pred in enumerate(output_lengths): + L_gt = int(mel_input_length[bib].item() / 2) + if L_gt <= mel_len or L_pred <= mel_len: continue sp.append(s_preds[bib]) - random_start = np.random.randint(0, mel_length_pred - mel_len) - en.append(asr_pred[bib, :, random_start:random_start+mel_len]) - p_en.append(p_pred[bib, :, random_start:random_start+mel_len]) + start = np.random.randint(0, L_pred - mel_len) + en.append(asr_pred[bib, :, start : start + mel_len]) + p_en.append(p_pred[bib, :, start : start + mel_len]) - # get ground truth clips - random_start = np.random.randint(0, mel_length_gt - mel_len) - y = waves[bib][(random_start * 2) * 300:((random_start+mel_len) * 2) * 300] + start_gt = np.random.randint(0, L_gt - mel_len) + y = waves[bib][(start_gt * 2) * 300 : ((start_gt + mel_len) * 2) * 300] wav.append(torch.from_numpy(y).to(ref_text.device)) - - if len(wav) >= self.batch_percentage * len(waves): # prevent OOM due to longer lengths + + if len(wav) >= self.batch_percentage * len(waves): break if len(sp) <= 1: return None - + sp = torch.stack(sp) wav = torch.stack(wav).float() en = torch.stack(en) p_en = torch.stack(p_en) - + F0_fake, N_fake = self.model.predictor.F0Ntrain(p_en, sp[:, 128:]) y_pred = self.model.decoder(en, F0_fake, N_fake, sp[:, :128]) - - # discriminator loss + + # -------------- adversarial losses ----------------------------- if (iters + 1) % self.skip_update == 0: - if np.random.randint(0, 2) == 0: - wav = y_rec_gt_pred - use_rec = True - else: - use_rec = False - - crop_size = min(wav.size(-1), y_pred.size(-1)) - if use_rec: # use reconstructed (shorter lengths), do length invariant regularization - if wav.size(-1) > y_pred.size(-1): - real_GP = wav[:, : , :crop_size] - out_crop = self.wl.discriminator_forward(real_GP.detach().squeeze()) - out_org = self.wl.discriminator_forward(wav.detach().squeeze()) - loss_reg = F.l1_loss(out_crop, out_org[..., :out_crop.size(-1)]) - - if np.random.randint(0, 2) == 0: - d_loss = self.wl.discriminator(real_GP.detach().squeeze(), y_pred.detach().squeeze()).mean() - else: - d_loss = self.wl.discriminator(wav.detach().squeeze(), y_pred.detach().squeeze()).mean() - else: - real_GP = y_pred[:, : , :crop_size] - out_crop = self.wl.discriminator_forward(real_GP.detach().squeeze()) - out_org = self.wl.discriminator_forward(y_pred.detach().squeeze()) - loss_reg = F.l1_loss(out_crop, out_org[..., :out_crop.size(-1)]) - - if np.random.randint(0, 2) == 0: - d_loss = self.wl.discriminator(wav.detach().squeeze(), real_GP.detach().squeeze()).mean() - else: - d_loss = self.wl.discriminator(wav.detach().squeeze(), y_pred.detach().squeeze()).mean() - - # regularization (ignore length variation) - d_loss += loss_reg - - out_gt = self.wl.discriminator_forward(y_rec_gt.detach().squeeze()) - out_rec = self.wl.discriminator_forward(y_rec_gt_pred.detach().squeeze()) - - # regularization (ignore reconstruction artifacts) - d_loss += F.l1_loss(out_gt, out_rec) - - else: - d_loss = self.wl.discriminator(wav.detach().squeeze(), y_pred.detach().squeeze()).mean() + d_loss = self.wl.discriminator(wav.squeeze(), y_pred.detach().squeeze()).mean() else: d_loss = 0 - - # generator loss - gen_loss = self.wl.generator(y_pred.squeeze()) - - gen_loss = gen_loss.mean() - + + gen_loss = self.wl.generator(y_pred.squeeze()).mean() return d_loss, gen_loss, y_pred.detach().cpu().numpy() - -def length_to_mask(lengths): - mask = torch.arange(lengths.max()).unsqueeze(0).expand(lengths.shape[0], -1).type_as(lengths) - mask = torch.gt(mask+1, lengths.unsqueeze(1)) - return mask + + +# ------------------------------------------------------------------ # +def length_to_mask(lengths: torch.Tensor) -> torch.Tensor: + """Classic length mask: 1 → PAD, 0 → real token.""" + max_len = lengths.max() + mask = ( + torch.arange(max_len, device=lengths.device) + .unsqueeze(0) + .expand(lengths.size(0), -1) + ) + return mask >= lengths.unsqueeze(1) \ No newline at end of file diff --git a/__pycache__/losses.cpython-310.pyc b/__pycache__/losses.cpython-310.pyc index d2076ddf1697c17e06e0a1bbf3888eff4b5508b6..6ca0cc8217882807796dfbea432dc868c21d45d0 100644 Binary files a/__pycache__/losses.cpython-310.pyc and b/__pycache__/losses.cpython-310.pyc differ diff --git a/__pycache__/meldataset.cpython-310.pyc b/__pycache__/meldataset.cpython-310.pyc index 0963dceffc3104ccb2c90b87f3bc8de3a5ff71ca..ab2c9736368439a3f8e363de6b3b2e1e2a8c6125 100644 Binary files a/__pycache__/meldataset.cpython-310.pyc and b/__pycache__/meldataset.cpython-310.pyc differ diff --git a/__pycache__/models.cpython-310.pyc b/__pycache__/models.cpython-310.pyc index 28e24ce6ea6390000adbf4e093842f173f0fd9ea..9c961a8c58ac0c3637cb2a1bd77bde77d158ec39 100644 Binary files a/__pycache__/models.cpython-310.pyc and b/__pycache__/models.cpython-310.pyc differ diff --git a/__pycache__/optimizers.cpython-310.pyc b/__pycache__/optimizers.cpython-310.pyc index bc03a684b0e89c9168e826d34205dd65933f780e..f648388a1e1d815312097669b11caad9f1a460e4 100644 Binary files a/__pycache__/optimizers.cpython-310.pyc and b/__pycache__/optimizers.cpython-310.pyc differ diff --git a/__pycache__/utils.cpython-310.pyc b/__pycache__/utils.cpython-310.pyc index bc7db9733292808e90ca5eba68534c5f3fadf500..54034498bae0f59b57ec1cd78c717ff6114d8279 100644 Binary files a/__pycache__/utils.cpython-310.pyc and b/__pycache__/utils.cpython-310.pyc differ diff --git a/events.out.tfevents.1749430573.164-152-17-237.34802.0 b/events.out.tfevents.1749430573.164-152-17-237.34802.0 deleted file mode 100644 index 988832c759ce3bcedbde732fc35061752e40c863..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749430573.164-152-17-237.34802.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:a6bc2fbf21ce1532d4722ccec8da0506aaf220a449bcdfe3f09b5c4d78296f1b -size 88 diff --git a/events.out.tfevents.1749430673.164-152-17-237.35024.0 b/events.out.tfevents.1749430673.164-152-17-237.35024.0 deleted file mode 100644 index 424a4a42efdf6fe7ae8aa7cfdeab92aa2e5ddbc4..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749430673.164-152-17-237.35024.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:44458e4bd6dc84c7bb3269df2f7110ef8ac03b77a267f7a3cfdf46b1af7e8828 -size 88 diff --git a/events.out.tfevents.1749430736.164-152-17-237.35371.0 b/events.out.tfevents.1749430736.164-152-17-237.35371.0 deleted file mode 100644 index 2994372120bbf6680194dc2de1a2cfabab1b6044..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749430736.164-152-17-237.35371.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:2afa3088a758e9e2eded1b84d9791755b1e2a8bcb9b0593a7868a12b303f002b -size 34885531 diff --git a/events.out.tfevents.1749451143.164-152-17-237.47705.0 b/events.out.tfevents.1749451143.164-152-17-237.47705.0 deleted file mode 100644 index e10f9a537589122feed4153e0a91fc34f8aeb93e..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749451143.164-152-17-237.47705.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:92575c079c1688160003a6cccb4d83d962533a0aa181260fdc2b5d91e40f6a8e -size 88 diff --git a/events.out.tfevents.1749451143.164-152-17-237.47707.0 b/events.out.tfevents.1749451143.164-152-17-237.47707.0 deleted file mode 100644 index 8707fd5026b81b0776b692d9c99eb3618f2059e1..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749451143.164-152-17-237.47707.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:e5e5fae069eb05128c396152c992005b9e566dec2fe8378cbe7013c1beab53e1 -size 88 diff --git a/events.out.tfevents.1749451143.164-152-17-237.47710.0 b/events.out.tfevents.1749451143.164-152-17-237.47710.0 deleted file mode 100644 index ef9643570dbf96424e14b0cd3aa8875b8b8b3f03..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749451143.164-152-17-237.47710.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:ca5ac7da0de1cd8b2940a042eddfe0f7ea50cc867411a91d90240fa2186962b0 -size 88 diff --git a/events.out.tfevents.1749451143.164-152-17-237.47712.0 b/events.out.tfevents.1749451143.164-152-17-237.47712.0 deleted file mode 100644 index 8b9e0cb633e2cb94ec0e42e3a9d868673b3598a6..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749451143.164-152-17-237.47712.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:24f7b8986b9471590fd70ce3705e31a5b5a97854cdc1887585591ba318c1c150 -size 88 diff --git a/events.out.tfevents.1749451144.164-152-17-237.47706.0 b/events.out.tfevents.1749451144.164-152-17-237.47706.0 deleted file mode 100644 index 0854cbe06484b800fd7ac65c8cd6c91a6bf63db9..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749451144.164-152-17-237.47706.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:a4f87f7a9fa06bc2a39e77d91d3dd4c7d76ee7c9bbbf2f6d6b73f3a9d6836d0a -size 88 diff --git a/events.out.tfevents.1749451144.164-152-17-237.47708.0 b/events.out.tfevents.1749451144.164-152-17-237.47708.0 deleted file mode 100644 index 8047a5a782671db7cb6a79880950dd9cb659b31b..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749451144.164-152-17-237.47708.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:4f57d72c8bb8f7d68c2a16d4e5eea3151d1cd8aa752be3a879c003aa481c19b3 -size 88 diff --git a/events.out.tfevents.1749451144.164-152-17-237.47709.0 b/events.out.tfevents.1749451144.164-152-17-237.47709.0 deleted file mode 100644 index e1a7d070cd41c350431f7cc90e7bbe69f22d1ab0..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749451144.164-152-17-237.47709.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:103d5b2f29512166ac9979033248e0fb344847396ec0ed3dea7e96e5fad84e80 -size 88 diff --git a/events.out.tfevents.1749451144.164-152-17-237.47711.0 b/events.out.tfevents.1749451144.164-152-17-237.47711.0 deleted file mode 100644 index a3ee34762b3628b4cc4e13b2ec16775f19aee03f..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749451144.164-152-17-237.47711.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:7fa0f5d9031c97fbf708f0a40c4e2950dcbd07c683488a659710ab9fcfd1c224 -size 88 diff --git a/events.out.tfevents.1749451220.164-152-17-237.48862.0 b/events.out.tfevents.1749451220.164-152-17-237.48862.0 deleted file mode 100644 index 61b479ab19137be8ccde3411891cdf6abea91275..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749451220.164-152-17-237.48862.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:1f67c96bcdf2b41944f1f6710d3735137dd4254b5d58570d3b304e894de5acc8 -size 88 diff --git a/events.out.tfevents.1749451220.164-152-17-237.48863.0 b/events.out.tfevents.1749451220.164-152-17-237.48863.0 deleted file mode 100644 index 90fec841124dd192f2d5b633eacab8f5e19ae2ba..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749451220.164-152-17-237.48863.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:2e09c7f592ec6d1f20e8a0b1e0fce4ff9b209f4c6d2e466ee6c2a10c761207a4 -size 88 diff --git a/events.out.tfevents.1749451220.164-152-17-237.48864.0 b/events.out.tfevents.1749451220.164-152-17-237.48864.0 deleted file mode 100644 index 2d34d4d4992a370e52df7fe5b1cf3baec1f4e055..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749451220.164-152-17-237.48864.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:6358d1685327b1fc73bcdfc1ba181c900d63e4bc2a679a646aa697446cbcc818 -size 88 diff --git a/events.out.tfevents.1749451220.164-152-17-237.48865.0 b/events.out.tfevents.1749451220.164-152-17-237.48865.0 deleted file mode 100644 index 5d733ebf1ca5e32615b5fc8a3ceb4eebe1bd63e8..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749451220.164-152-17-237.48865.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:0c2cfe09514438496a286074eff7b5d988953c53eca8c446a57b833aca2cd233 -size 88 diff --git a/events.out.tfevents.1749451220.164-152-17-237.48868.0 b/events.out.tfevents.1749451220.164-152-17-237.48868.0 deleted file mode 100644 index 25fc242541124b36a379fd086328da3d04c7ed1d..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749451220.164-152-17-237.48868.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:a2d99a10e442411fc79a40bba5ba012773c90f5be44254d6b872f3e350d0bb98 -size 88 diff --git a/events.out.tfevents.1749451221.164-152-17-237.48861.0 b/events.out.tfevents.1749451221.164-152-17-237.48861.0 deleted file mode 100644 index 43155da781c28756fcfcc5e8795fdb8076d4eb75..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749451221.164-152-17-237.48861.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:72554714ef2293ef47dc12bdc8698f70024abd16bfa736eb3de03d0e8b1c0eee -size 88 diff --git a/events.out.tfevents.1749451221.164-152-17-237.48867.0 b/events.out.tfevents.1749451221.164-152-17-237.48867.0 deleted file mode 100644 index b716cf50fcc51093a02c1eefb8aa61243af2df00..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749451221.164-152-17-237.48867.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:7a882509295dfd1193368519c9cad538370392117ee7b6c483ee939ee7979769 -size 88 diff --git a/events.out.tfevents.1749451222.164-152-17-237.48866.0 b/events.out.tfevents.1749451222.164-152-17-237.48866.0 deleted file mode 100644 index ea075f3837b6a191a7115c198bacbee2d2198781..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749451222.164-152-17-237.48866.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:37a6fd68ad5ef36692d7d5389ba938e318c6287b20c2684a628e5245f186048c -size 88 diff --git a/events.out.tfevents.1749453792.164-152-17-237.51057.0 b/events.out.tfevents.1749453792.164-152-17-237.51057.0 deleted file mode 100644 index d4a35e1869b3f625de6287676d3809dbe35181a2..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453792.164-152-17-237.51057.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:8d42df1e6023c4e593c8479eecb153738ecc7600b94a1c388173708d38fc3688 -size 88 diff --git a/events.out.tfevents.1749453792.164-152-17-237.51059.0 b/events.out.tfevents.1749453792.164-152-17-237.51059.0 deleted file mode 100644 index 24ee3bdd6f1e89d7528bfee41ab065c99b6ebf65..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453792.164-152-17-237.51059.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:3be5b37365ae5d97c1f0573f94f70a082f707c6c7d49926d118111ac7e48a818 -size 88 diff --git a/events.out.tfevents.1749453792.164-152-17-237.51061.0 b/events.out.tfevents.1749453792.164-152-17-237.51061.0 deleted file mode 100644 index 89a89dd31908be41c40e645b15303492c6700189..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453792.164-152-17-237.51061.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:16300e80b48b67bd14ce00c5751e4a6841df3da3be4873c677dc08a99a6c3aae -size 88 diff --git a/events.out.tfevents.1749453792.164-152-17-237.51063.0 b/events.out.tfevents.1749453792.164-152-17-237.51063.0 deleted file mode 100644 index c93fce4ff21be3d3fa10e763fcf1fbbb0172e8f2..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453792.164-152-17-237.51063.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:84bbbebc6a97e89725078bcf6533475d3286f6af25b23f781558e7cb8d8957e3 -size 88 diff --git a/events.out.tfevents.1749453793.164-152-17-237.51056.0 b/events.out.tfevents.1749453793.164-152-17-237.51056.0 deleted file mode 100644 index 09c150d7f62da9e716b817e4808df2b82ee19ad6..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453793.164-152-17-237.51056.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:9a3560e484d3bd2c79ebe1507599d497f261cb3b62d1644697a5a2953d156c0d -size 88 diff --git a/events.out.tfevents.1749453793.164-152-17-237.51058.0 b/events.out.tfevents.1749453793.164-152-17-237.51058.0 deleted file mode 100644 index 7cf26e803ebf76460ec644aa876d5d0107d4eb6d..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453793.164-152-17-237.51058.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:c70617864f6b48f9177c3051cf3d0e857374656601e1f3b2130e91dd6d3090ed -size 88 diff --git a/events.out.tfevents.1749453793.164-152-17-237.51060.0 b/events.out.tfevents.1749453793.164-152-17-237.51060.0 deleted file mode 100644 index 907f12383db044ec6cfc24e6e25449e216328f98..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453793.164-152-17-237.51060.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:e2a2cccc5958217207ffb6d11f98e619f4dedf92682121709eb5870fb3db085d -size 88 diff --git a/events.out.tfevents.1749453794.164-152-17-237.51062.0 b/events.out.tfevents.1749453794.164-152-17-237.51062.0 deleted file mode 100644 index 703649af1b0eef5839ea18d89934c5d151942f12..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453794.164-152-17-237.51062.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:b66ebcaedfeb8d0a1d4a801e5e18783d8dfa41eedf362b2a0733a38c8f0a82fa -size 88 diff --git a/events.out.tfevents.1749453905.164-152-17-237.52357.0 b/events.out.tfevents.1749453905.164-152-17-237.52357.0 deleted file mode 100644 index 2c80982b67e0deebcb64581eda87332c8cff12b2..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453905.164-152-17-237.52357.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:39098f172013dbdfbddcfb78c6c126b7b67671f89989a344d4262fcd433c3e9f -size 88 diff --git a/events.out.tfevents.1749453905.164-152-17-237.52358.0 b/events.out.tfevents.1749453905.164-152-17-237.52358.0 deleted file mode 100644 index bca504263898044c58b21150aa61993a480fd7bc..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453905.164-152-17-237.52358.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:e31a13e3e32ec6f2081c987fd6b1bd6c20bce2ca61312a576eea1aeceea533dc -size 88 diff --git a/events.out.tfevents.1749453905.164-152-17-237.52360.0 b/events.out.tfevents.1749453905.164-152-17-237.52360.0 deleted file mode 100644 index 20057abfe8f1981cd5136c980af106ef51d55010..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453905.164-152-17-237.52360.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:c85710b39b9fb9764094a24aa502a715e94261613d7865f89e037553679ee109 -size 88 diff --git a/events.out.tfevents.1749453905.164-152-17-237.52361.0 b/events.out.tfevents.1749453905.164-152-17-237.52361.0 deleted file mode 100644 index a5e7cd044c149f4c58e54fa502c203700a924e7e..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453905.164-152-17-237.52361.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:521592225e7079f153bc3a7f18f79d9122eedbe6dffc5d342912d69ba4a9a1e1 -size 88 diff --git a/events.out.tfevents.1749453906.164-152-17-237.52355.0 b/events.out.tfevents.1749453906.164-152-17-237.52355.0 deleted file mode 100644 index e6ff18effc3123e2ceb9ca8e6bde393e305388a5..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453906.164-152-17-237.52355.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:914db70df7d7e69fc3396fb49f23b5a6849c9a251533d681c447df73ab81df34 -size 88 diff --git a/events.out.tfevents.1749453906.164-152-17-237.52356.0 b/events.out.tfevents.1749453906.164-152-17-237.52356.0 deleted file mode 100644 index f68d17c5426f5f648e82672c3d4f9a881a299dcc..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453906.164-152-17-237.52356.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:e1c4479d3572d16ab62395082c0c4a300c42b739f724966866a1a7b15c08344e -size 88 diff --git a/events.out.tfevents.1749453906.164-152-17-237.52359.0 b/events.out.tfevents.1749453906.164-152-17-237.52359.0 deleted file mode 100644 index b908378df2ca3cea1b142117664613c6c51c99d0..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453906.164-152-17-237.52359.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:f2daad6b2a7a604a0853f70aee92c15f79557490acd34797d875c670446db7e3 -size 88 diff --git a/events.out.tfevents.1749453906.164-152-17-237.52362.0 b/events.out.tfevents.1749453906.164-152-17-237.52362.0 deleted file mode 100644 index 2cbc940acedcf0fcdcd151fcbe25054fb516d244..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453906.164-152-17-237.52362.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:6abbb8d2b926e553f18af6c4c9203b122736ea0c2a508407150076c1b2842dad -size 88 diff --git a/events.out.tfevents.1749453977.164-152-17-237.53096.0 b/events.out.tfevents.1749453977.164-152-17-237.53096.0 deleted file mode 100644 index f3c656f6fd4f6c2646bb83252fd5a453f052119e..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453977.164-152-17-237.53096.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:7a6eaa7928ccbca758c1c4d170f439c3938a2f300a78a12746749cfca3b997cf -size 88 diff --git a/events.out.tfevents.1749453977.164-152-17-237.53097.0 b/events.out.tfevents.1749453977.164-152-17-237.53097.0 deleted file mode 100644 index 213c398774af1d9e5d24883633032279fcb3da00..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453977.164-152-17-237.53097.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:f90c29e2391c038fb896ced73c9099014dfbff51e889bf6075ea8181a59da78d -size 88 diff --git a/events.out.tfevents.1749453977.164-152-17-237.53098.0 b/events.out.tfevents.1749453977.164-152-17-237.53098.0 deleted file mode 100644 index 1579a281b847c2967fca370eace087372cdb5b48..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453977.164-152-17-237.53098.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:0d82622b483fc183e71198f54b5b10c60851af9752b27ddf4dbaca39e988e15d -size 88 diff --git a/events.out.tfevents.1749453977.164-152-17-237.53099.0 b/events.out.tfevents.1749453977.164-152-17-237.53099.0 deleted file mode 100644 index c4a4b8156d82025e1287ffebb0be399cb5f81d27..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453977.164-152-17-237.53099.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:85fb51c72b78ca2c571570edd1a44288737f426a7e314f865c395d3f1d42d764 -size 88 diff --git a/events.out.tfevents.1749453977.164-152-17-237.53100.0 b/events.out.tfevents.1749453977.164-152-17-237.53100.0 deleted file mode 100644 index c4a849bf3ba7ee56b7d69bafb7f556167d99a82c..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453977.164-152-17-237.53100.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:9d3516306f7671fae64a2df06e433f84e451c5d5e6387a9159d6861958371c75 -size 88 diff --git a/events.out.tfevents.1749453977.164-152-17-237.53101.0 b/events.out.tfevents.1749453977.164-152-17-237.53101.0 deleted file mode 100644 index 35c9bc5bafd51bfa92914ed64f68031dc752ca44..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453977.164-152-17-237.53101.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:06003b110891c57d117b1685afc401a35c65311d6a2b8402d9dbe4ae94f44136 -size 88 diff --git a/events.out.tfevents.1749453977.164-152-17-237.53102.0 b/events.out.tfevents.1749453977.164-152-17-237.53102.0 deleted file mode 100644 index 72f37b6d6d4dc2d90ffb9006a428fdc969e2bfaf..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453977.164-152-17-237.53102.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:a8ece14d3c31836be92fd1f16588df270d2ddd0f929479f9377bb6ba44514830 -size 88 diff --git a/events.out.tfevents.1749453977.164-152-17-237.53103.0 b/events.out.tfevents.1749453977.164-152-17-237.53103.0 deleted file mode 100644 index 91d711359a3a24f31e72382621a09ffea32865ee..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749453977.164-152-17-237.53103.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:40e01ae77f5d81a3dc59b49fadc051857ba39ac213749bdd31a288bd17fec873 -size 88 diff --git a/events.out.tfevents.1749454412.164-152-17-237.54028.0 b/events.out.tfevents.1749454412.164-152-17-237.54028.0 deleted file mode 100644 index 3777a866ba65bbf2001f718487c8c5eba1a48d64..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454412.164-152-17-237.54028.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:c1783e40a76f1a4d143dcccf845dc72a2910adc60e04c67161163b329bc6e97c -size 88 diff --git a/events.out.tfevents.1749454412.164-152-17-237.54029.0 b/events.out.tfevents.1749454412.164-152-17-237.54029.0 deleted file mode 100644 index 45afe46eaf94a0bf2d876b00e609b913e8367dc2..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454412.164-152-17-237.54029.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:9416c85d79ab3ab4c0366fbd4094a01ceb1d203e7b92d469f01efcd5060b7bce -size 88 diff --git a/events.out.tfevents.1749454412.164-152-17-237.54030.0 b/events.out.tfevents.1749454412.164-152-17-237.54030.0 deleted file mode 100644 index 80c406eae50a7ecf6c8839b12b9f5149583ac5f8..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454412.164-152-17-237.54030.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:9a8519fbe0671143a819b616ef581f1d87ef3dbec70515810f8114346ca4baa8 -size 88 diff --git a/events.out.tfevents.1749454412.164-152-17-237.54031.0 b/events.out.tfevents.1749454412.164-152-17-237.54031.0 deleted file mode 100644 index 205305906aac39ffdb939e6a0a9ffc1dff9cdbd0..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454412.164-152-17-237.54031.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:5f2238d64f1ed38b9e7b598de5a704037ac64b1ab066f3e3877cfce0b88f0898 -size 88 diff --git a/events.out.tfevents.1749454412.164-152-17-237.54032.0 b/events.out.tfevents.1749454412.164-152-17-237.54032.0 deleted file mode 100644 index 4cf68a06e930a1950e8bd6d1aafad922f6703b47..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454412.164-152-17-237.54032.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:2def59d36df23bb79a2915d70cd86c0100ca35f8a4afc0164631e9a63d912325 -size 88 diff --git a/events.out.tfevents.1749454412.164-152-17-237.54033.0 b/events.out.tfevents.1749454412.164-152-17-237.54033.0 deleted file mode 100644 index dbac3487895e94c2ca83861714a8da611b726503..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454412.164-152-17-237.54033.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:a90ad82d519b662f0746630d118eea2585dbb408c5305f599089b51878e286e7 -size 88 diff --git a/events.out.tfevents.1749454412.164-152-17-237.54034.0 b/events.out.tfevents.1749454412.164-152-17-237.54034.0 deleted file mode 100644 index f704db996e78f2c6a4e7b556a98f37489e502cbd..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454412.164-152-17-237.54034.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:00aa3953a4f8484598a6d5c6bc608bfc7e226cc228cfbf9303bde11d4046df40 -size 88 diff --git a/events.out.tfevents.1749454412.164-152-17-237.54035.0 b/events.out.tfevents.1749454412.164-152-17-237.54035.0 deleted file mode 100644 index 4f00100571c50de5be83cf8359e7856bb694fc52..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454412.164-152-17-237.54035.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:e7c70004475a6664385b518863f16b666b31b6720c1e360f96f1540f8370f292 -size 88 diff --git a/events.out.tfevents.1749454505.164-152-17-237.55009.0 b/events.out.tfevents.1749454505.164-152-17-237.55009.0 deleted file mode 100644 index 52e3c3f2b2f9ba9eef36a415cdeeb83ab7078ebe..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454505.164-152-17-237.55009.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:c95b2f243fe4c84b2c34b018ad8c173472d7a26d4df3e2f4cd47f9e238f7dd80 -size 88 diff --git a/events.out.tfevents.1749454505.164-152-17-237.55011.0 b/events.out.tfevents.1749454505.164-152-17-237.55011.0 deleted file mode 100644 index 50ffed88511f42e9dbf7070ed9ef5f4de8b67268..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454505.164-152-17-237.55011.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:7b5df5673022433f8e72b6be2163d148e701874515f0397b23cf42594d22ffdb -size 88 diff --git a/events.out.tfevents.1749454505.164-152-17-237.55014.0 b/events.out.tfevents.1749454505.164-152-17-237.55014.0 deleted file mode 100644 index f661da383efcf16244c656af8940f3a86faf0b05..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454505.164-152-17-237.55014.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:7bea4148aa77a7d75121b4dcd2a08f0e6468377bf168d378c9730791bfd6764f -size 88 diff --git a/events.out.tfevents.1749454505.164-152-17-237.55015.0 b/events.out.tfevents.1749454505.164-152-17-237.55015.0 deleted file mode 100644 index f2a0486a08a263fc9e8444cece409926825cb92a..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454505.164-152-17-237.55015.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:73e17164fa7d946c1ae73b182d4727660d306305f2e16c0788afe05082eb7363 -size 88 diff --git a/events.out.tfevents.1749454506.164-152-17-237.55008.0 b/events.out.tfevents.1749454506.164-152-17-237.55008.0 deleted file mode 100644 index bcd0584db214f796fe0cdd9999dd117f4b605df5..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454506.164-152-17-237.55008.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:67f9064ef75fb768d680ce1151a87c20343cc252aa8cc2226f7af6fe5d1d8f94 -size 88 diff --git a/events.out.tfevents.1749454506.164-152-17-237.55010.0 b/events.out.tfevents.1749454506.164-152-17-237.55010.0 deleted file mode 100644 index 0daeb4f27c58e4d7ec90c275c7a0ff63431a1eb3..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454506.164-152-17-237.55010.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:db03da4c173f7327c8f828eaf20cfc7c09842c52eb9ad008273c0281ef0e7922 -size 88 diff --git a/events.out.tfevents.1749454506.164-152-17-237.55012.0 b/events.out.tfevents.1749454506.164-152-17-237.55012.0 deleted file mode 100644 index 338eb1b54a62cab4a6409290dbe5195603416f55..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454506.164-152-17-237.55012.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:469d23fb4d1134c16ee0b5e0d11018c11aba5dd175af0db7f1080cfbbd9aa310 -size 88 diff --git a/events.out.tfevents.1749454506.164-152-17-237.55013.0 b/events.out.tfevents.1749454506.164-152-17-237.55013.0 deleted file mode 100644 index 5086b80a034bdf6ec89338c77522508779fa607f..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454506.164-152-17-237.55013.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:d8ba38f0349115f5da3e62c36227ba3228df4eb17d5bf828b23b6b4e6c8f9fa9 -size 88 diff --git a/events.out.tfevents.1749454562.164-152-17-237.55599.0 b/events.out.tfevents.1749454562.164-152-17-237.55599.0 deleted file mode 100644 index b79f97f9ae8069312afb502f01dcc92214f278ac..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454562.164-152-17-237.55599.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:2e545966b0ca0ea1cd231104932a3c0e118720f250145944009f2b2ee7c7d4ee -size 88 diff --git a/events.out.tfevents.1749454562.164-152-17-237.55600.0 b/events.out.tfevents.1749454562.164-152-17-237.55600.0 deleted file mode 100644 index f65f974a9efb2d77d60158a21e4012bedb621ccf..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454562.164-152-17-237.55600.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:ad6601124f9c29574ad5f3201e7801eca2bc642313f584ca6b7d3d01ca117579 -size 88 diff --git a/events.out.tfevents.1749454562.164-152-17-237.55605.0 b/events.out.tfevents.1749454562.164-152-17-237.55605.0 deleted file mode 100644 index 4f486159ac8b5ff04c3460ce25c4dc6b731dbd0e..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454562.164-152-17-237.55605.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:55286b44c3fbe429e2c2019f0dd36f33a690675c612cb7d0d79ab735af74921b -size 88 diff --git a/events.out.tfevents.1749454563.164-152-17-237.55601.0 b/events.out.tfevents.1749454563.164-152-17-237.55601.0 deleted file mode 100644 index 87d5e63991e7e201360419fd21d937ae885cb09b..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454563.164-152-17-237.55601.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:0ac8619a6b8d955aba094f0971aa00342532332c71a7285d0db3d99368b5d119 -size 88 diff --git a/events.out.tfevents.1749454563.164-152-17-237.55603.0 b/events.out.tfevents.1749454563.164-152-17-237.55603.0 deleted file mode 100644 index b49bc1c79ac7f7c0dadd3534acd8f9d89b87b305..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454563.164-152-17-237.55603.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:4caff0dbe5442befa024b3ec11613da95befea720513dd7db60ce771c816152f -size 88 diff --git a/events.out.tfevents.1749454563.164-152-17-237.55604.0 b/events.out.tfevents.1749454563.164-152-17-237.55604.0 deleted file mode 100644 index 22600e8f33385339cf57bbda9b0f87135256cdd2..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454563.164-152-17-237.55604.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:1868575af9de71d3b68a93f8b592ab7809f3956a10decc89d424e319954f9357 -size 88 diff --git a/events.out.tfevents.1749454564.164-152-17-237.55598.0 b/events.out.tfevents.1749454564.164-152-17-237.55598.0 deleted file mode 100644 index 472acb6110cf91eb03980f917882a5becaff01e6..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454564.164-152-17-237.55598.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:e249828bf52b2e48c221a008b09f85a4f7142504765fc6a3a33d2f3aeced0ebd -size 88 diff --git a/events.out.tfevents.1749454564.164-152-17-237.55602.0 b/events.out.tfevents.1749454564.164-152-17-237.55602.0 deleted file mode 100644 index d352f232b6cc13ecb66d502482917e7ade20d990..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454564.164-152-17-237.55602.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:69323bdcf6fba63d8ef7f6ad1ed41ae142ac45053312c540552461701cf33bc4 -size 88 diff --git a/events.out.tfevents.1749454666.164-152-17-237.56241.0 b/events.out.tfevents.1749454666.164-152-17-237.56241.0 deleted file mode 100644 index 76c505d51c0c3ad985e16125629ed536aa1537fc..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454666.164-152-17-237.56241.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:c7d707573cf2e868d56aaf057a17a961d7e14b387ed768aab60bf3df24eac98b -size 88 diff --git a/events.out.tfevents.1749454666.164-152-17-237.56244.0 b/events.out.tfevents.1749454666.164-152-17-237.56244.0 deleted file mode 100644 index 01cd1e181b1b5985926e85372c71347b97fa0b4f..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454666.164-152-17-237.56244.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:ea8d1be70e90e1773ebbe16d3c69807322a1ec3e0537acb8ba71a8c9ddd2831d -size 88 diff --git a/events.out.tfevents.1749454666.164-152-17-237.56245.0 b/events.out.tfevents.1749454666.164-152-17-237.56245.0 deleted file mode 100644 index 7f0548a53d404590dda231c0be476b62dde31131..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454666.164-152-17-237.56245.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:6aff20e95c2f074c6ddcd16efd16cabb4c9fa9061ef1fd76924e81b77237973b -size 88 diff --git a/events.out.tfevents.1749454666.164-152-17-237.56248.0 b/events.out.tfevents.1749454666.164-152-17-237.56248.0 deleted file mode 100644 index ee92fef8019987b9675a723c404503c5bdc2ae00..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454666.164-152-17-237.56248.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:a4a4abf96bbd9e28a6225e0f4edc8a2dc1a878a6a275cfea413e88afef4cd8ac -size 88 diff --git a/events.out.tfevents.1749454667.164-152-17-237.56242.0 b/events.out.tfevents.1749454667.164-152-17-237.56242.0 deleted file mode 100644 index 8cfa20a13d56ffde75e8ad753136965fa2b7327b..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454667.164-152-17-237.56242.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:bf1edc4d19a9b781a2da3f259d7c1f36a464827beb244e580d7cc354344ca86a -size 88 diff --git a/events.out.tfevents.1749454667.164-152-17-237.56243.0 b/events.out.tfevents.1749454667.164-152-17-237.56243.0 deleted file mode 100644 index 420c5c4962d5cb75174ee6385384160954efe797..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454667.164-152-17-237.56243.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:155e579023d3d9fffe9394f6cd110d2e1b92dc20a105f08b11de40d3195cc1d2 -size 88 diff --git a/events.out.tfevents.1749454667.164-152-17-237.56246.0 b/events.out.tfevents.1749454667.164-152-17-237.56246.0 deleted file mode 100644 index ffdcbdc280749e0f2b960d2cc8dcfda98c8c3ff3..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454667.164-152-17-237.56246.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:d4feb84f3d88e474c7ad56614a47a811163fc50d1e8d3b6f0d8b51c47f167827 -size 88 diff --git a/events.out.tfevents.1749454667.164-152-17-237.56247.0 b/events.out.tfevents.1749454667.164-152-17-237.56247.0 deleted file mode 100644 index 16af6627d614276a8870da5533cb82b3b4d8af4e..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454667.164-152-17-237.56247.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:ade09d0e109132e7cc5477a5f10f5df413b1089728763918fb3c1bb625b22ab6 -size 88 diff --git a/events.out.tfevents.1749454902.164-152-17-237.57126.0 b/events.out.tfevents.1749454902.164-152-17-237.57126.0 deleted file mode 100644 index cdf311b3bbac409dcb1c30600e67230d37f0321d..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454902.164-152-17-237.57126.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:3735dacdc349938c3840428bfc11b014f149fc86781d1b1d1fb6a18e8de8437b -size 88 diff --git a/events.out.tfevents.1749454902.164-152-17-237.57127.0 b/events.out.tfevents.1749454902.164-152-17-237.57127.0 deleted file mode 100644 index 2131059962613d3c81a8e5a73aff83855e2d727c..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454902.164-152-17-237.57127.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:9b08f68ba2e208c41c589483c4b598f441e896e0d2e6ff38a1f12b5c134db706 -size 88 diff --git a/events.out.tfevents.1749454902.164-152-17-237.57128.0 b/events.out.tfevents.1749454902.164-152-17-237.57128.0 deleted file mode 100644 index b00b934cb70b740281dbaeb09fd09f7b6a51f67b..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454902.164-152-17-237.57128.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:a264deb5830bf5ae774b37b345150107351b563c673a3e08b6e1c91875e48ad7 -size 88 diff --git a/events.out.tfevents.1749454902.164-152-17-237.57130.0 b/events.out.tfevents.1749454902.164-152-17-237.57130.0 deleted file mode 100644 index 03da29951610440f8b440bf0c02b17bd7825a17a..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454902.164-152-17-237.57130.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:5a639ee7f6b766f87496d5d8ed429f6579ff9fe9dd8ea88b7b56d0b9a7796e56 -size 88 diff --git a/events.out.tfevents.1749454902.164-152-17-237.57132.0 b/events.out.tfevents.1749454902.164-152-17-237.57132.0 deleted file mode 100644 index f626a482cb6de9ca2f7b059bfd03f647a146496c..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454902.164-152-17-237.57132.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:886c1f8ad541a51ed8cf1bdc54465b836d09fc881d74195ba6fa344b564b1f4b -size 88 diff --git a/events.out.tfevents.1749454902.164-152-17-237.57133.0 b/events.out.tfevents.1749454902.164-152-17-237.57133.0 deleted file mode 100644 index 3a4fe173d33d4a61baaa5af3350aad413bdc9e46..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454902.164-152-17-237.57133.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:9928569691b6d286c7af654eba4f802b47cc02468783699f660101b6aaf60720 -size 88 diff --git a/events.out.tfevents.1749454903.164-152-17-237.57129.0 b/events.out.tfevents.1749454903.164-152-17-237.57129.0 deleted file mode 100644 index 3c6676d92bd6d295c88462f412ad96b2f729f47f..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454903.164-152-17-237.57129.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:0f573254923455717043b7a58f7404be20d242ea2b45a62b475a5aa088afda31 -size 88 diff --git a/events.out.tfevents.1749454903.164-152-17-237.57131.0 b/events.out.tfevents.1749454903.164-152-17-237.57131.0 deleted file mode 100644 index 8622f6d9374a66af631f6b630ce4269dc188952d..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749454903.164-152-17-237.57131.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:e84400baa4a98fe6640a402270ce84e76bf2a4f684d11222d4673516c95d3f23 -size 88 diff --git a/events.out.tfevents.1749455005.164-152-17-237.57873.0 b/events.out.tfevents.1749455005.164-152-17-237.57873.0 deleted file mode 100644 index 46267569829eb8597d27bec8c92b34fe4f518f48..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455005.164-152-17-237.57873.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:bc5bb9cfb59293835f39c1a9fd6e3481747caf489c92691977a37a27dda284cf -size 88 diff --git a/events.out.tfevents.1749455005.164-152-17-237.57874.0 b/events.out.tfevents.1749455005.164-152-17-237.57874.0 deleted file mode 100644 index e796e6916b1a2dd5567be99d8d7cc394dea690b0..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455005.164-152-17-237.57874.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:daa69d34d3d436a7cce70260ec4dddd0446db5aff1844b36a84c4aecfe71adaf -size 88 diff --git a/events.out.tfevents.1749455005.164-152-17-237.57875.0 b/events.out.tfevents.1749455005.164-152-17-237.57875.0 deleted file mode 100644 index 23b8d818504c94282a9bcd1ac56ff8f5cf4668da..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455005.164-152-17-237.57875.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:e08c791b7c2a46aee6f36adf5299ffb6a24143576a2a4c51815fc4535e2583f1 -size 88 diff --git a/events.out.tfevents.1749455005.164-152-17-237.57876.0 b/events.out.tfevents.1749455005.164-152-17-237.57876.0 deleted file mode 100644 index a4b6bf7403a03555c057306af8e5f27a781dbf81..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455005.164-152-17-237.57876.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:9e6ee417b70d9f7c6a124dae923863c793b6f5e078a35480f76339773ba9819e -size 88 diff --git a/events.out.tfevents.1749455005.164-152-17-237.57879.0 b/events.out.tfevents.1749455005.164-152-17-237.57879.0 deleted file mode 100644 index f0bdcb111f3d64cbf924a7c734d1d1174905d3b8..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455005.164-152-17-237.57879.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:166b1dd4f8cc9520a24daf3c8ae606ee4b9839d52c0e261f812fe2970ff3921b -size 88 diff --git a/events.out.tfevents.1749455005.164-152-17-237.57880.0 b/events.out.tfevents.1749455005.164-152-17-237.57880.0 deleted file mode 100644 index a78f1e9b212f19eca735e85c6f3f2fbde712481e..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455005.164-152-17-237.57880.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:18a2cbfc41e2b5424beaf155bf91e044116fe5b0d43361e79c1711d053a326f4 -size 88 diff --git a/events.out.tfevents.1749455006.164-152-17-237.57877.0 b/events.out.tfevents.1749455006.164-152-17-237.57877.0 deleted file mode 100644 index 20a4a69666a28cea48a8660d6aaa3c17b5ed61b9..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455006.164-152-17-237.57877.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:f9be707843a21e3a5fe62f156562793d4b3f05260f0559aeff30b03b42934be4 -size 88 diff --git a/events.out.tfevents.1749455006.164-152-17-237.57878.0 b/events.out.tfevents.1749455006.164-152-17-237.57878.0 deleted file mode 100644 index ba0738af1d53c3019fd14bb55a655812a7e04bed..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455006.164-152-17-237.57878.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:c86c56134f3527e4d84e9e9f9defa27c3ff1c06fe23da76e85b771ab8db167df -size 88 diff --git a/events.out.tfevents.1749455394.164-152-17-237.58875.0 b/events.out.tfevents.1749455394.164-152-17-237.58875.0 deleted file mode 100644 index 37ac9049748d97df52911d2e7a0b77a0441bd21e..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455394.164-152-17-237.58875.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:0dcc50d9a811b1720b1a04f47df4954943f6e3740e2b7d33254e51db00d801d6 -size 88 diff --git a/events.out.tfevents.1749455394.164-152-17-237.58876.0 b/events.out.tfevents.1749455394.164-152-17-237.58876.0 deleted file mode 100644 index bb59b936996e3418a31865326b3a0b10647002ab..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455394.164-152-17-237.58876.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:ff9f5105a6e3a452dd5f2917c83eb00305b3e27b370e871080dd36246aaa795f -size 88 diff --git a/events.out.tfevents.1749455394.164-152-17-237.58880.0 b/events.out.tfevents.1749455394.164-152-17-237.58880.0 deleted file mode 100644 index c2428307980384b45d61d0a89714564a2bbef088..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455394.164-152-17-237.58880.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:d83365e4971895ca2559187d19a63fa56b1e455d0c00b829cc2eaefae762d88c -size 88 diff --git a/events.out.tfevents.1749455394.164-152-17-237.58881.0 b/events.out.tfevents.1749455394.164-152-17-237.58881.0 deleted file mode 100644 index e71554ec36c037a1d7c62d4f45db07d62bed5679..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455394.164-152-17-237.58881.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:f46d512fda8023259a0360169c17cfac146c83298943bc9a96f3ed0c42c1b40c -size 88 diff --git a/events.out.tfevents.1749455394.164-152-17-237.58882.0 b/events.out.tfevents.1749455394.164-152-17-237.58882.0 deleted file mode 100644 index 9463b47d98c528e14d46c0befa3bf279d9aeefcc..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455394.164-152-17-237.58882.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:9ce0706cd931cafa5cf77044ccfcc1e911b9ee3c7a564b4985239416a3c7d34b -size 88 diff --git a/events.out.tfevents.1749455395.164-152-17-237.58877.0 b/events.out.tfevents.1749455395.164-152-17-237.58877.0 deleted file mode 100644 index f3958d68346430aefcbf3853fbacc66c50ebd287..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455395.164-152-17-237.58877.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:8d659923a82ef3daee37053383213e82e3334734f2319b56f5669228ac76afe0 -size 88 diff --git a/events.out.tfevents.1749455395.164-152-17-237.58878.0 b/events.out.tfevents.1749455395.164-152-17-237.58878.0 deleted file mode 100644 index 8e797dcb563f3583dfd7279ce174b837fce13a52..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455395.164-152-17-237.58878.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:c320b4b539696579123428e1d03e32bcccc4db71ff9a4a8883171184d577bfd5 -size 88 diff --git a/events.out.tfevents.1749455395.164-152-17-237.58879.0 b/events.out.tfevents.1749455395.164-152-17-237.58879.0 deleted file mode 100644 index 833198411cd36977875bc5c269ddd9986584d927..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455395.164-152-17-237.58879.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:cb32f91ae4af2e165c0b012fa04f20544f1912b7efa4e83681a3beeccf084b12 -size 88 diff --git a/events.out.tfevents.1749455867.164-152-17-237.59315.0 b/events.out.tfevents.1749455867.164-152-17-237.59315.0 deleted file mode 100644 index f73846a3ea8db28acfc59138455028fc325fe1c7..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749455867.164-152-17-237.59315.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:2373cb699f0b210091fff5a2e139e6daeede77d4ba21ec7d7560a14e206cdeeb -size 1344 diff --git a/events.out.tfevents.1749456114.164-152-17-237.63151.0 b/events.out.tfevents.1749456114.164-152-17-237.63151.0 deleted file mode 100644 index a156c7b98c14b182def96bb6f26f745c14e895f8..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749456114.164-152-17-237.63151.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:a3d684af8a415033d5f39e1ab6d4a02eaf2ae96495764dde1c98b768eda005b1 -size 11584 diff --git a/events.out.tfevents.1749457380.164-152-17-237.85081.0 b/events.out.tfevents.1749457380.164-152-17-237.85081.0 deleted file mode 100644 index 4be7ac0b67533a04c2aaebd377efd942229134dd..0000000000000000000000000000000000000000 --- a/events.out.tfevents.1749457380.164-152-17-237.85081.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:416eb968629b387bfb3057f0e27d6dea47d96e2bca27a2326cf8e3a23b09bbf1 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749338343.104-171-203-10.11888.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749338343.104-171-203-10.11888.0 deleted file mode 100644 index ddefa49aba1d2b1bab725b2bf3fa1a4732bb30f8..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749338343.104-171-203-10.11888.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:2f0b61e4b8751a174bd3e23296859ac8dc0de4f2e2a513995e1d587498a328bc -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749338402.104-171-203-10.12026.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749338402.104-171-203-10.12026.0 deleted file mode 100644 index 8499a035a6071a6f070f62086cc9886859801a6a..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749338402.104-171-203-10.12026.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:32584a00483df136ef4c75afbd1998dff2281ed2ec6004d2626fa9eafcb28813 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749338497.104-171-203-10.12159.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749338497.104-171-203-10.12159.0 deleted file mode 100644 index a4fe8ccecd14ae0fc1e80ed40107e44bd5ceccb6..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749338497.104-171-203-10.12159.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:93bcbc4b41497088ee12a9a15e1b0f1d0ec289550defd61d54c33741674766a1 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749338594.104-171-203-10.12326.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749338594.104-171-203-10.12326.0 deleted file mode 100644 index 4faa2e3f1c5b7755f94ea7cc3e0390f4e0248cf6..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749338594.104-171-203-10.12326.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:17ace6cec0d8005db42957add71a3d9c19777262899cc26e39555b90cc656bb8 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749338898.104-171-203-10.12487.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749338898.104-171-203-10.12487.0 deleted file mode 100644 index 67a60e9ef54a60b468a910f9457a4ab6764efc6b..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749338898.104-171-203-10.12487.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:ecd97af6bf9d1da851ca4fdc1ef686facb6d1c7c152927d9e3300141107c1991 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749339074.104-171-203-10.12619.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749339074.104-171-203-10.12619.0 deleted file mode 100644 index 9cb05c829e6b258bf13ac37a864e67ecb9ce4cb2..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749339074.104-171-203-10.12619.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:b0b289d92d36d43c42352a69fe48271c698fd13fd09aaa4d3f9d2e9d5f98e29d -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749339329.104-171-203-10.12764.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749339329.104-171-203-10.12764.0 deleted file mode 100644 index b60010309b062cedb93527417f181b15c15ddd41..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749339329.104-171-203-10.12764.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:b165151c7b9b653bb7ba41fe178b6ef7570a9659b150af61c7643caf2682d34f -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749339659.104-171-203-10.12929.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749339659.104-171-203-10.12929.0 deleted file mode 100644 index 7c8257683e8d2c4f6065efb51d7283534c904dde..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749339659.104-171-203-10.12929.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:8030ca0650595a6c04761a4747cf2446ef3fac7178fb6e1306fd70dd53cae528 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749339754.104-171-203-10.13058.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749339754.104-171-203-10.13058.0 deleted file mode 100644 index 3f2b6f3102f69a7dfff245b3dcfbc1ffd185281e..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749339754.104-171-203-10.13058.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:8b3c29ea1a0c9c9c3ab3ee67f36a48fa829c54254c359035f7a523a2b4f4a051 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340109.104-171-203-10.13209.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340109.104-171-203-10.13209.0 deleted file mode 100644 index 249c065d20515c11eba870247bc6ed4acccad55c..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340109.104-171-203-10.13209.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:691b073d4910832197a2d8e3c77c9c663c9ca7bb9ed62fcfebcaef52537ab7dd -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340478.104-171-203-10.13349.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340478.104-171-203-10.13349.0 deleted file mode 100644 index f9c2e24ac4a40c50d4341a39f2b751d3edeee5a1..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340478.104-171-203-10.13349.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:45da554feba4a916d6735d89c9a0155fb41a6c77a031c10d3547b0eadad34780 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340544.104-171-203-10.13476.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340544.104-171-203-10.13476.0 deleted file mode 100644 index c799887c211eb9bbc5a3c4cd6911fd80cb6e9d70..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340544.104-171-203-10.13476.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:8caa10448515b83514f75e795d910d1a715d74263674da8b21f512f7ac330a17 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340601.104-171-203-10.13615.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340601.104-171-203-10.13615.0 deleted file mode 100644 index 0bcaa1f5a0c6876c12016ec7ccc28ba10e76a27e..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340601.104-171-203-10.13615.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:e2fb2bf7f7f46595df72be44a5fbe965cc77e0dcde6e9f7cf9e389620f7d82b1 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340754.104-171-203-10.13768.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340754.104-171-203-10.13768.0 deleted file mode 100644 index aff0aa3a949323138ee8c0fb32139314c11fc939..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340754.104-171-203-10.13768.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:f490dddea511c7b1a263ddcea07b5be3c28d2dded9956a4dcbcb74d94fd2c3ba -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340919.104-171-203-10.13990.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340919.104-171-203-10.13990.0 deleted file mode 100644 index 6b36b30219069dc3a52cb2df037c505cb7ba3b83..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749340919.104-171-203-10.13990.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:a9c02005ab536d2da5ce34eb31e9a91d1887751b54bf5a7b83a15afe9e706dfa -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342289.104-171-203-10.14242.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342289.104-171-203-10.14242.0 deleted file mode 100644 index 7694b6e08d8fc5979bfadb9c817aa6634af401ba..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342289.104-171-203-10.14242.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:502a79eb64788404aae6c90f0acb4b64a8aacabd9d0a4dae2fcb28593d3f4de8 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342466.104-171-203-10.14376.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342466.104-171-203-10.14376.0 deleted file mode 100644 index 40ade713651162d8b9cbdbda7559c21471f9bca8..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342466.104-171-203-10.14376.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:54513aad9960f3bb4d987a51dbe7033a4f352e0517983bf90cccea492b6bcc85 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342502.104-171-203-10.14509.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342502.104-171-203-10.14509.0 deleted file mode 100644 index efeebc1163c6874173c5b945f7b8c6c411febe09..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342502.104-171-203-10.14509.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:660ac36101dd46cb7b727439fc96e5a16a0c1e01d1b0eed8d1a9aa5e77457f58 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342694.104-171-203-10.14778.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342694.104-171-203-10.14778.0 deleted file mode 100644 index ea61d12dbeb9752ad6c324e55b28aeb250946654..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342694.104-171-203-10.14778.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:00e5f62cfa06238015eb4d1bff2ceec88191f7ef550b75322c8e7cf643f0756b -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342850.104-171-203-10.15078.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342850.104-171-203-10.15078.0 deleted file mode 100644 index db727d57a5d5ee886d0cd16d06afbf2c550a8080..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342850.104-171-203-10.15078.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:4e7e263c737baf0df2a41e102e96defb53322276b90a45c3edf357a5c9e2f6ad -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342963.104-171-203-10.15342.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342963.104-171-203-10.15342.0 deleted file mode 100644 index 04a3f3ea0f96cdab06c3e2cdb0d762d4029f48f3..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749342963.104-171-203-10.15342.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:8d8cd747ada5ced784a706c2977366a4a5b0d49a8e1eeba9b9d5feaa3426de68 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749343035.104-171-203-10.15598.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749343035.104-171-203-10.15598.0 deleted file mode 100644 index 0112fad8060e635420942d96c85689e084402aed..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749343035.104-171-203-10.15598.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:697bd70c5e0f61071fb047bd3f4e07efce3248af597e54e5c59fff52a0da4fa4 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749343206.104-171-203-10.15868.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749343206.104-171-203-10.15868.0 deleted file mode 100644 index ae9414ffb72a64ef8b07cd919eabf8ca1be917e1..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749343206.104-171-203-10.15868.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:ab6d65d139ab368db65dcf55a34de71995f3c6681c2a032cf3adfb825dff6569 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749344763.104-171-203-10.255334.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749344763.104-171-203-10.255334.0 deleted file mode 100644 index b0b87fc6e15d2882772797f212ea6447b02cf8f9..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749344763.104-171-203-10.255334.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:0c28f178538b2ceca42e4ab97df79cbbbdb001fba7c1127cef30c0ea96e8ec32 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749344965.104-171-203-10.255636.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749344965.104-171-203-10.255636.0 deleted file mode 100644 index 313de8607ede7903c47053379a27557203209cd9..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749344965.104-171-203-10.255636.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:23817f4caaf6087e6f4af10dc7eefda4b7fa0e33f0349cd33b70f3c3d473af95 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345005.104-171-203-10.255894.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345005.104-171-203-10.255894.0 deleted file mode 100644 index 0be8314191d8f8e15bf8787e811d6c5a30bcf0af..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345005.104-171-203-10.255894.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:e9b024b4b2f40f871b55d97bc3b5c5ea6488c9bd27bd9a3ffe4d8781bb96feb9 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345046.104-171-203-10.256146.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345046.104-171-203-10.256146.0 deleted file mode 100644 index 55aaf629e6de6fe0550b22d06d48154f1eff94ae..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345046.104-171-203-10.256146.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:d839df477f9569f37bc4ad563b248e477bf18d2951eaf49714c50e7055a5920a -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345260.104-171-203-10.256432.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345260.104-171-203-10.256432.0 deleted file mode 100644 index 339a94d2efe46c08d29f9887cc9ce3d536a49201..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345260.104-171-203-10.256432.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:7b39a0c6aa441fc88f7b1b49ccae392abe3585789d878d0d344e5db646e1d4e3 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345313.104-171-203-10.256685.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345313.104-171-203-10.256685.0 deleted file mode 100644 index ebe3944336431ee3687d683d7538311fbbb13c12..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345313.104-171-203-10.256685.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:20950b0c52364882eb738d373a0f0c6d32343198873188ecd340501089d0ec48 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345619.104-171-203-10.256968.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345619.104-171-203-10.256968.0 deleted file mode 100644 index aefd836a29df8e5af3370931d5534f627cd4cc18..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345619.104-171-203-10.256968.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:4c667d170a18a87a6adf9e9d87429b6cac6b4bf0898075884d6b3d766cd9c7aa -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345810.104-171-203-10.257276.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345810.104-171-203-10.257276.0 deleted file mode 100644 index a8d95662bbb748c9bdc2a9c34ab342b499d60679..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749345810.104-171-203-10.257276.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:a817dabc528aedd24829fd797c15ac7d39e45112d951219f0625a4075de60f6c -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749346162.104-171-203-10.257574.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749346162.104-171-203-10.257574.0 deleted file mode 100644 index b39e9ea693bfb8b97c73f18c1d8fda43a5613323..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749346162.104-171-203-10.257574.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:209a103fe47cf6f341734bdb2eeb5e3594817b037e2a5b2a69590c33ddb80590 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355010.104-171-203-10.258726.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355010.104-171-203-10.258726.0 deleted file mode 100644 index 732f199da0b68429e97f9a011954a7ce46a7ad77..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355010.104-171-203-10.258726.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:0d36b74d0ae9c3e9308bf2368404c1a43942c101545d24cc8caa29610993c251 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355085.104-171-203-10.258978.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355085.104-171-203-10.258978.0 deleted file mode 100644 index c25c2ede159574901ebb68fe3c7c9e065bf58706..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355085.104-171-203-10.258978.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:c7761b921a8c9a4cdd4de19744827833f2fcf94e57d051aed330f4d24d4c90b7 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355141.104-171-203-10.259238.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355141.104-171-203-10.259238.0 deleted file mode 100644 index 71a62c109f79b038a3d8467c6725013a9eeee916..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355141.104-171-203-10.259238.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:1887280c3d8c469d351763b4fde393d30176ff40c7eda0f387af44baefb46654 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355325.104-171-203-10.259516.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355325.104-171-203-10.259516.0 deleted file mode 100644 index 4b66fd80968a6e514557ca3a2f5d009521c222d2..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355325.104-171-203-10.259516.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:4f73beecf8b5c3d663e88b92d4a5572f80cbf6810a355b9cf0e5de978acf393f -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355622.104-171-203-10.259810.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355622.104-171-203-10.259810.0 deleted file mode 100644 index da314a6730700f0b613c40f2943dd6da30f17316..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355622.104-171-203-10.259810.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:8d775643e7d60c3049f1bd831a1a4f64ca1193b9a6029baf749bb1af1317a66d -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355853.104-171-203-10.260101.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355853.104-171-203-10.260101.0 deleted file mode 100644 index 302768fd831f268a40122fb4ed14ded9d2fca0e5..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749355853.104-171-203-10.260101.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:bb1f8ad541e998341b4ffb72174e6669e24bbd49eda639f7fa2562b63ef96c94 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749356247.104-171-203-10.260392.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749356247.104-171-203-10.260392.0 deleted file mode 100644 index e5693457d1be9d3451b9bf4be95436b43564e14e..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749356247.104-171-203-10.260392.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:7129e599ebe9b1a8f54f8b0140758aae127e5d9670a62e08c316626570f17376 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749356298.104-171-203-10.260652.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749356298.104-171-203-10.260652.0 deleted file mode 100644 index 54c32c001447971716137156a05035fafeac3785..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749356298.104-171-203-10.260652.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:695172a7375bce456532069bc2fd7b0a8fc61de088ac26ebac77d4d1389e6d7d -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749356514.104-171-203-10.260982.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749356514.104-171-203-10.260982.0 deleted file mode 100644 index e657df1be866e753831790db68e123fb549e5335..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749356514.104-171-203-10.260982.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:c1fd51ee32f318c5bf187cb7affec53762dd28c63120db6f94a70e09d50b41ff -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749356545.104-171-203-10.261228.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749356545.104-171-203-10.261228.0 deleted file mode 100644 index 4e6fe0ebd5d12bc3becf3f3c599ea72a3e3c30bf..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749356545.104-171-203-10.261228.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:01516103fb118776a0bc78966c6d1b7279570a58c4a204b7972d39d1c192a052 -size 88 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749356633.104-171-203-10.261492.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749356633.104-171-203-10.261492.0 deleted file mode 100644 index dcbacd9dec6fdf0dfb919611b472ddde6db85a91..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749356633.104-171-203-10.261492.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:20eb59d783b002d8b65171efc7423ad71694aba44f789821e94a93ef33de0816 -size 9269 diff --git a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749357894.104-171-203-10.262507.0 b/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749357894.104-171-203-10.262507.0 deleted file mode 100644 index dce8c717aa014313394670dfd78d2f1d6c635097..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/tensorboard/events.out.tfevents.1749357894.104-171-203-10.262507.0 +++ /dev/null @@ -1,3 +0,0 @@ -version https://git-lfs.github.com/spec/v1 -oid sha256:87637f2383b627d665953c2b1f8c3531fb070d3741be3826445e200e92910976 -size 3192810 diff --git a/logs/pod_90h_30k/train.log b/logs/pod_90h_30k/train.log deleted file mode 100644 index 639ae5a75640c69e6714add14124d59fded946c1..0000000000000000000000000000000000000000 --- a/logs/pod_90h_30k/train.log +++ /dev/null @@ -1,1121 +0,0 @@ -INFO:2025-06-08 04:24:43,065: Epoch [1/13], Step [50/3970], Mel Loss: 1.59229, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:25:18,950: Epoch [1/13], Step [100/3970], Mel Loss: 1.51388, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:25:56,856: Epoch [1/13], Step [150/3970], Mel Loss: 1.52012, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:26:35,935: Epoch [1/13], Step [200/3970], Mel Loss: 1.53224, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:27:16,690: Epoch [1/13], Step [250/3970], Mel Loss: 1.58791, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:27:54,900: Epoch [1/13], Step [300/3970], Mel Loss: 1.59529, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:28:34,030: Epoch [1/13], Step [350/3970], Mel Loss: 1.53330, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:29:12,869: Epoch [1/13], Step [400/3970], Mel Loss: 1.68588, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:29:53,694: Epoch [1/13], Step [450/3970], Mel Loss: 1.86278, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:30:32,105: Epoch [1/13], Step [500/3970], Mel Loss: 1.68792, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:31:13,476: Epoch [1/13], Step [550/3970], Mel Loss: 1.57326, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:31:53,519: Epoch [1/13], Step [600/3970], Mel Loss: 1.68260, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:32:31,388: Epoch [1/13], Step [650/3970], Mel Loss: 1.96622, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:33:09,976: Epoch [1/13], Step [700/3970], Mel Loss: 1.97537, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:33:48,625: Epoch [1/13], Step [750/3970], Mel Loss: 2.01780, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:34:27,213: Epoch [1/13], Step [800/3970], Mel Loss: 2.15039, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:35:04,967: Epoch [1/13], Step [850/3970], Mel Loss: 2.20302, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:35:44,139: Epoch [1/13], Step [900/3970], Mel Loss: 2.18096, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:36:23,589: Epoch [1/13], Step [950/3970], Mel Loss: 2.15651, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:37:01,371: Epoch [1/13], Step [1000/3970], Mel Loss: 1.88864, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:37:40,741: Epoch [1/13], Step [1050/3970], Mel Loss: 1.64650, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:38:19,405: Epoch [1/13], Step [1100/3970], Mel Loss: 1.63909, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:38:57,942: Epoch [1/13], Step [1150/3970], Mel Loss: 1.64249, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:39:35,881: Epoch [1/13], Step [1200/3970], Mel Loss: 1.66006, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:40:12,159: Epoch [1/13], Step [1250/3970], Mel Loss: 1.63626, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:40:51,013: Epoch [1/13], Step [1300/3970], Mel Loss: 1.64158, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:41:30,024: Epoch [1/13], Step [1350/3970], Mel Loss: 1.64270, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:42:07,953: Epoch [1/13], Step [1400/3970], Mel Loss: 1.65131, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:42:46,940: Epoch [1/13], Step [1450/3970], Mel Loss: 1.66478, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:45:42,703: Epoch [1/13], Step [50/3970], Mel Loss: 1.87156, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:46:20,511: Epoch [1/13], Step [100/3970], Mel Loss: 1.86572, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:46:58,127: Epoch [1/13], Step [150/3970], Mel Loss: 1.85973, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:47:37,368: Epoch [1/13], Step [200/3970], Mel Loss: 1.87426, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:48:18,834: Epoch [1/13], Step [250/3970], Mel Loss: 1.86499, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:48:57,910: Epoch [1/13], Step [300/3970], Mel Loss: 1.85298, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:49:38,322: Epoch [1/13], Step [350/3970], Mel Loss: 1.87987, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:50:17,991: Epoch [1/13], Step [400/3970], Mel Loss: 1.87030, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:50:58,463: Epoch [1/13], Step [450/3970], Mel Loss: 1.85640, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:51:37,056: Epoch [1/13], Step [500/3970], Mel Loss: 1.88851, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:52:16,803: Epoch [1/13], Step [550/3970], Mel Loss: 1.87053, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:52:56,367: Epoch [1/13], Step [600/3970], Mel Loss: 1.85791, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:53:34,177: Epoch [1/13], Step [650/3970], Mel Loss: 1.89341, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:54:14,245: Epoch [1/13], Step [700/3970], Mel Loss: 1.93086, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:54:53,074: Epoch [1/13], Step [750/3970], Mel Loss: 1.89002, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:55:32,806: Epoch [1/13], Step [800/3970], Mel Loss: 1.89111, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:56:10,601: Epoch [1/13], Step [850/3970], Mel Loss: 1.88555, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:56:48,851: Epoch [1/13], Step [900/3970], Mel Loss: 1.90091, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:57:28,327: Epoch [1/13], Step [950/3970], Mel Loss: 1.89882, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:58:06,247: Epoch [1/13], Step [1000/3970], Mel Loss: 1.91204, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:58:43,898: Epoch [1/13], Step [1050/3970], Mel Loss: 1.89137, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 04:59:22,898: Epoch [1/13], Step [1100/3970], Mel Loss: 1.89888, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:00:02,476: Epoch [1/13], Step [1150/3970], Mel Loss: 1.89478, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:00:40,076: Epoch [1/13], Step [1200/3970], Mel Loss: 1.90840, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:01:15,889: Epoch [1/13], Step [1250/3970], Mel Loss: 1.91392, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:01:54,820: Epoch [1/13], Step [1300/3970], Mel Loss: 1.87965, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:02:34,432: Epoch [1/13], Step [1350/3970], Mel Loss: 1.89908, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:03:14,209: Epoch [1/13], Step [1400/3970], Mel Loss: 1.88978, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:03:53,291: Epoch [1/13], Step [1450/3970], Mel Loss: 1.89164, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:04:31,454: Epoch [1/13], Step [1500/3970], Mel Loss: 1.90797, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:05:11,570: Epoch [1/13], Step [1550/3970], Mel Loss: 1.90366, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:05:51,912: Epoch [1/13], Step [1600/3970], Mel Loss: 1.89806, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:06:30,280: Epoch [1/13], Step [1650/3970], Mel Loss: 1.89540, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:07:08,555: Epoch [1/13], Step [1700/3970], Mel Loss: 1.88576, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:07:47,060: Epoch [1/13], Step [1750/3970], Mel Loss: 1.89400, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:08:24,708: Epoch [1/13], Step [1800/3970], Mel Loss: 1.89903, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:09:04,667: Epoch [1/13], Step [1850/3970], Mel Loss: 1.88638, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:09:43,369: Epoch [1/13], Step [1900/3970], Mel Loss: 1.90319, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:10:21,100: Epoch [1/13], Step [1950/3970], Mel Loss: 1.88633, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:10:59,813: Epoch [1/13], Step [2000/3970], Mel Loss: 1.90483, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:11:38,391: Epoch [1/13], Step [2050/3970], Mel Loss: 1.87361, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:12:17,228: Epoch [1/13], Step [2100/3970], Mel Loss: 1.88939, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:12:56,391: Epoch [1/13], Step [2150/3970], Mel Loss: 1.88260, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:13:32,458: Epoch [1/13], Step [2200/3970], Mel Loss: 1.89200, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:14:11,361: Epoch [1/13], Step [2250/3970], Mel Loss: 1.88889, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:14:51,853: Epoch [1/13], Step [2300/3970], Mel Loss: 1.89153, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:15:29,900: Epoch [1/13], Step [2350/3970], Mel Loss: 1.89735, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:16:08,763: Epoch [1/13], Step [2400/3970], Mel Loss: 1.91090, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:16:48,700: Epoch [1/13], Step [2450/3970], Mel Loss: 1.91056, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:17:29,699: Epoch [1/13], Step [2500/3970], Mel Loss: 1.91319, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:18:08,583: Epoch [1/13], Step [2550/3970], Mel Loss: 1.88856, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:18:46,949: Epoch [1/13], Step [2600/3970], Mel Loss: 1.91404, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:19:25,927: Epoch [1/13], Step [2650/3970], Mel Loss: 1.88466, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:20:04,979: Epoch [1/13], Step [2700/3970], Mel Loss: 1.89514, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:20:42,625: Epoch [1/13], Step [2750/3970], Mel Loss: 1.88694, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:21:21,829: Epoch [1/13], Step [2800/3970], Mel Loss: 1.89760, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:21:59,970: Epoch [1/13], Step [2850/3970], Mel Loss: 1.88965, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:22:38,974: Epoch [1/13], Step [2900/3970], Mel Loss: 1.89536, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:23:19,441: Epoch [1/13], Step [2950/3970], Mel Loss: 1.91223, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:23:58,887: Epoch [1/13], Step [3000/3970], Mel Loss: 1.86574, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:24:37,018: Epoch [1/13], Step [3050/3970], Mel Loss: 1.91886, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:25:15,354: Epoch [1/13], Step [3100/3970], Mel Loss: 1.89323, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:25:53,457: Epoch [1/13], Step [3150/3970], Mel Loss: 1.91420, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:26:31,921: Epoch [1/13], Step [3200/3970], Mel Loss: 1.90469, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:27:10,057: Epoch [1/13], Step [3250/3970], Mel Loss: 1.87571, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:27:48,425: Epoch [1/13], Step [3300/3970], Mel Loss: 1.89407, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:28:25,055: Epoch [1/13], Step [3350/3970], Mel Loss: 1.88638, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:29:03,003: Epoch [1/13], Step [3400/3970], Mel Loss: 1.88517, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:29:44,440: Epoch [1/13], Step [3450/3970], Mel Loss: 1.90011, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:30:23,921: Epoch [1/13], Step [3500/3970], Mel Loss: 1.88072, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:31:02,985: Epoch [1/13], Step [3550/3970], Mel Loss: 1.89266, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:31:39,614: Epoch [1/13], Step [3600/3970], Mel Loss: 1.89061, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:32:19,125: Epoch [1/13], Step [3650/3970], Mel Loss: 1.88223, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:32:54,778: Epoch [1/13], Step [3700/3970], Mel Loss: 1.90492, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:33:33,081: Epoch [1/13], Step [3750/3970], Mel Loss: 1.90399, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:34:10,780: Epoch [1/13], Step [3800/3970], Mel Loss: 1.90917, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:34:50,605: Epoch [1/13], Step [3850/3970], Mel Loss: 1.90556, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:35:27,502: Epoch [1/13], Step [3900/3970], Mel Loss: 1.89994, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:36:06,842: Epoch [1/13], Step [3950/3970], Mel Loss: 1.89947, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:39:54,318: Validation loss: 1.865 - - - - -INFO:2025-06-08 05:40:40,228: Epoch [2/13], Step [50/3970], Mel Loss: 1.89867, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:41:20,005: Epoch [2/13], Step [100/3970], Mel Loss: 1.91414, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:41:59,747: Epoch [2/13], Step [150/3970], Mel Loss: 1.89090, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:42:38,035: Epoch [2/13], Step [200/3970], Mel Loss: 1.89637, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:43:17,926: Epoch [2/13], Step [250/3970], Mel Loss: 1.90666, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:43:58,311: Epoch [2/13], Step [300/3970], Mel Loss: 1.90334, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:44:37,196: Epoch [2/13], Step [350/3970], Mel Loss: 1.89541, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:45:14,132: Epoch [2/13], Step [400/3970], Mel Loss: 1.88379, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:45:52,285: Epoch [2/13], Step [450/3970], Mel Loss: 1.91129, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:46:30,702: Epoch [2/13], Step [500/3970], Mel Loss: 1.88601, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:47:09,134: Epoch [2/13], Step [550/3970], Mel Loss: 1.89496, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:47:47,597: Epoch [2/13], Step [600/3970], Mel Loss: 1.87840, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:48:25,727: Epoch [2/13], Step [650/3970], Mel Loss: 1.89505, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:49:04,457: Epoch [2/13], Step [700/3970], Mel Loss: 1.87929, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:49:43,274: Epoch [2/13], Step [750/3970], Mel Loss: 1.89702, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:50:21,047: Epoch [2/13], Step [800/3970], Mel Loss: 1.87199, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:50:57,298: Epoch [2/13], Step [850/3970], Mel Loss: 1.90900, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:51:34,226: Epoch [2/13], Step [900/3970], Mel Loss: 1.88943, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:52:13,203: Epoch [2/13], Step [950/3970], Mel Loss: 1.88862, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:52:53,783: Epoch [2/13], Step [1000/3970], Mel Loss: 1.88103, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:53:31,750: Epoch [2/13], Step [1050/3970], Mel Loss: 1.89856, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:54:12,411: Epoch [2/13], Step [1100/3970], Mel Loss: 1.88714, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:54:50,722: Epoch [2/13], Step [1150/3970], Mel Loss: 1.90231, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:55:29,662: Epoch [2/13], Step [1200/3970], Mel Loss: 1.91365, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:56:08,067: Epoch [2/13], Step [1250/3970], Mel Loss: 1.90384, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:56:47,292: Epoch [2/13], Step [1300/3970], Mel Loss: 1.90620, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:57:25,658: Epoch [2/13], Step [1350/3970], Mel Loss: 1.88477, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:58:04,462: Epoch [2/13], Step [1400/3970], Mel Loss: 1.90786, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:58:43,820: Epoch [2/13], Step [1450/3970], Mel Loss: 1.89603, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 05:59:23,164: Epoch [2/13], Step [1500/3970], Mel Loss: 1.89780, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:00:02,126: Epoch [2/13], Step [1550/3970], Mel Loss: 1.89110, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:00:41,462: Epoch [2/13], Step [1600/3970], Mel Loss: 1.89676, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:01:19,399: Epoch [2/13], Step [1650/3970], Mel Loss: 1.92370, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:01:57,858: Epoch [2/13], Step [1700/3970], Mel Loss: 1.89181, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:02:37,087: Epoch [2/13], Step [1750/3970], Mel Loss: 1.89261, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:03:16,517: Epoch [2/13], Step [1800/3970], Mel Loss: 1.89720, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:03:56,882: Epoch [2/13], Step [1850/3970], Mel Loss: 1.89561, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:04:33,900: Epoch [2/13], Step [1900/3970], Mel Loss: 1.88516, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:05:12,720: Epoch [2/13], Step [1950/3970], Mel Loss: 1.89852, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:05:51,722: Epoch [2/13], Step [2000/3970], Mel Loss: 1.90626, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:06:29,803: Epoch [2/13], Step [2050/3970], Mel Loss: 1.89991, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:07:08,911: Epoch [2/13], Step [2100/3970], Mel Loss: 1.89464, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:07:45,253: Epoch [2/13], Step [2150/3970], Mel Loss: 1.90150, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:08:22,992: Epoch [2/13], Step [2200/3970], Mel Loss: 1.88420, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:09:02,234: Epoch [2/13], Step [2250/3970], Mel Loss: 1.90926, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:09:41,179: Epoch [2/13], Step [2300/3970], Mel Loss: 1.87542, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:10:18,695: Epoch [2/13], Step [2350/3970], Mel Loss: 1.89929, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:10:57,414: Epoch [2/13], Step [2400/3970], Mel Loss: 1.87826, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:11:35,626: Epoch [2/13], Step [2450/3970], Mel Loss: 1.90840, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:12:15,316: Epoch [2/13], Step [2500/3970], Mel Loss: 1.90261, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:12:53,686: Epoch [2/13], Step [2550/3970], Mel Loss: 1.89390, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:13:33,080: Epoch [2/13], Step [2600/3970], Mel Loss: 1.91342, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:14:13,360: Epoch [2/13], Step [2650/3970], Mel Loss: 1.91461, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:14:52,125: Epoch [2/13], Step [2700/3970], Mel Loss: 1.90098, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:15:30,608: Epoch [2/13], Step [2750/3970], Mel Loss: 1.89820, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:16:09,477: Epoch [2/13], Step [2800/3970], Mel Loss: 1.90123, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:16:47,755: Epoch [2/13], Step [2850/3970], Mel Loss: 1.88420, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:17:25,802: Epoch [2/13], Step [2900/3970], Mel Loss: 1.88923, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:18:04,212: Epoch [2/13], Step [2950/3970], Mel Loss: 1.91845, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:18:42,282: Epoch [2/13], Step [3000/3970], Mel Loss: 1.88563, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:19:21,074: Epoch [2/13], Step [3050/3970], Mel Loss: 1.90382, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:19:58,279: Epoch [2/13], Step [3100/3970], Mel Loss: 1.88247, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:20:34,297: Epoch [2/13], Step [3150/3970], Mel Loss: 1.89437, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:21:12,787: Epoch [2/13], Step [3200/3970], Mel Loss: 1.89521, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:21:50,685: Epoch [2/13], Step [3250/3970], Mel Loss: 1.89283, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:22:28,944: Epoch [2/13], Step [3300/3970], Mel Loss: 1.87823, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:23:07,857: Epoch [2/13], Step [3350/3970], Mel Loss: 1.89763, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:23:47,012: Epoch [2/13], Step [3400/3970], Mel Loss: 1.89337, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:24:25,292: Epoch [2/13], Step [3450/3970], Mel Loss: 1.88374, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:25:03,472: Epoch [2/13], Step [3500/3970], Mel Loss: 1.88646, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:25:40,560: Epoch [2/13], Step [3550/3970], Mel Loss: 1.89177, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:26:20,468: Epoch [2/13], Step [3600/3970], Mel Loss: 1.92332, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:26:59,415: Epoch [2/13], Step [3650/3970], Mel Loss: 1.89852, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:27:37,511: Epoch [2/13], Step [3700/3970], Mel Loss: 1.90137, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:28:14,789: Epoch [2/13], Step [3750/3970], Mel Loss: 1.90517, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:28:53,740: Epoch [2/13], Step [3800/3970], Mel Loss: 1.91026, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:29:31,372: Epoch [2/13], Step [3850/3970], Mel Loss: 1.88832, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:30:12,109: Epoch [2/13], Step [3900/3970], Mel Loss: 1.87564, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:30:50,119: Epoch [2/13], Step [3950/3970], Mel Loss: 1.88560, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:34:26,526: Validation loss: 1.862 - - - - -INFO:2025-06-08 06:35:07,163: Epoch [3/13], Step [50/3970], Mel Loss: 1.90542, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:35:46,402: Epoch [3/13], Step [100/3970], Mel Loss: 1.89835, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:36:24,654: Epoch [3/13], Step [150/3970], Mel Loss: 1.88893, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:37:04,650: Epoch [3/13], Step [200/3970], Mel Loss: 1.88937, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:37:44,772: Epoch [3/13], Step [250/3970], Mel Loss: 1.90966, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:38:22,095: Epoch [3/13], Step [300/3970], Mel Loss: 1.88757, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:39:00,518: Epoch [3/13], Step [350/3970], Mel Loss: 1.88460, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:39:40,366: Epoch [3/13], Step [400/3970], Mel Loss: 1.89984, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:40:19,661: Epoch [3/13], Step [450/3970], Mel Loss: 1.88539, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:40:57,090: Epoch [3/13], Step [500/3970], Mel Loss: 1.90596, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:41:36,230: Epoch [3/13], Step [550/3970], Mel Loss: 1.88462, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:42:16,529: Epoch [3/13], Step [600/3970], Mel Loss: 1.89333, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:42:56,841: Epoch [3/13], Step [650/3970], Mel Loss: 1.89198, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:43:34,003: Epoch [3/13], Step [700/3970], Mel Loss: 1.89335, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:44:11,961: Epoch [3/13], Step [750/3970], Mel Loss: 1.89752, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:44:50,179: Epoch [3/13], Step [800/3970], Mel Loss: 1.89497, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:45:31,251: Epoch [3/13], Step [850/3970], Mel Loss: 1.89099, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:46:10,973: Epoch [3/13], Step [900/3970], Mel Loss: 1.90122, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:46:49,854: Epoch [3/13], Step [950/3970], Mel Loss: 1.90315, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:47:28,215: Epoch [3/13], Step [1000/3970], Mel Loss: 1.88395, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:48:07,622: Epoch [3/13], Step [1050/3970], Mel Loss: 1.89477, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:48:46,077: Epoch [3/13], Step [1100/3970], Mel Loss: 1.91535, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:49:25,370: Epoch [3/13], Step [1150/3970], Mel Loss: 1.89695, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:50:03,941: Epoch [3/13], Step [1200/3970], Mel Loss: 1.89872, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:50:42,329: Epoch [3/13], Step [1250/3970], Mel Loss: 1.90740, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:51:21,774: Epoch [3/13], Step [1300/3970], Mel Loss: 1.88911, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:52:00,213: Epoch [3/13], Step [1350/3970], Mel Loss: 1.90881, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:52:41,114: Epoch [3/13], Step [1400/3970], Mel Loss: 1.88785, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:53:20,777: Epoch [3/13], Step [1450/3970], Mel Loss: 1.91126, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:54:00,525: Epoch [3/13], Step [1500/3970], Mel Loss: 1.90676, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:54:38,877: Epoch [3/13], Step [1550/3970], Mel Loss: 1.88650, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:55:17,985: Epoch [3/13], Step [1600/3970], Mel Loss: 1.89522, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:55:57,438: Epoch [3/13], Step [1650/3970], Mel Loss: 1.88180, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:56:38,432: Epoch [3/13], Step [1700/3970], Mel Loss: 1.90511, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:57:17,615: Epoch [3/13], Step [1750/3970], Mel Loss: 1.88049, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:57:56,211: Epoch [3/13], Step [1800/3970], Mel Loss: 1.89820, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:58:36,848: Epoch [3/13], Step [1850/3970], Mel Loss: 1.89807, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:59:17,238: Epoch [3/13], Step [1900/3970], Mel Loss: 1.90917, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 06:59:58,465: Epoch [3/13], Step [1950/3970], Mel Loss: 1.89698, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:00:37,267: Epoch [3/13], Step [2000/3970], Mel Loss: 1.89102, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:01:16,948: Epoch [3/13], Step [2050/3970], Mel Loss: 1.89438, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:01:55,817: Epoch [3/13], Step [2100/3970], Mel Loss: 1.88805, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:02:34,961: Epoch [3/13], Step [2150/3970], Mel Loss: 1.87436, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:03:14,641: Epoch [3/13], Step [2200/3970], Mel Loss: 1.89041, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:03:55,026: Epoch [3/13], Step [2250/3970], Mel Loss: 1.88783, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:04:33,409: Epoch [3/13], Step [2300/3970], Mel Loss: 1.89742, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:05:12,463: Epoch [3/13], Step [2350/3970], Mel Loss: 1.89150, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:05:52,865: Epoch [3/13], Step [2400/3970], Mel Loss: 1.88867, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:06:33,442: Epoch [3/13], Step [2450/3970], Mel Loss: 1.89756, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:07:12,556: Epoch [3/13], Step [2500/3970], Mel Loss: 1.87385, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:07:50,376: Epoch [3/13], Step [2550/3970], Mel Loss: 1.91384, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:08:29,972: Epoch [3/13], Step [2600/3970], Mel Loss: 1.89575, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:09:09,647: Epoch [3/13], Step [2650/3970], Mel Loss: 1.88684, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:09:50,778: Epoch [3/13], Step [2700/3970], Mel Loss: 1.91685, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:10:29,167: Epoch [3/13], Step [2750/3970], Mel Loss: 1.90155, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:11:07,208: Epoch [3/13], Step [2800/3970], Mel Loss: 1.90522, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:11:45,643: Epoch [3/13], Step [2850/3970], Mel Loss: 1.89745, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:12:23,106: Epoch [3/13], Step [2900/3970], Mel Loss: 1.90852, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:13:01,589: Epoch [3/13], Step [2950/3970], Mel Loss: 1.90971, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:13:40,282: Epoch [3/13], Step [3000/3970], Mel Loss: 1.90500, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:14:21,740: Epoch [3/13], Step [3050/3970], Mel Loss: 1.89005, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:15:00,396: Epoch [3/13], Step [3100/3970], Mel Loss: 1.91019, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:15:38,660: Epoch [3/13], Step [3150/3970], Mel Loss: 1.90115, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:16:17,899: Epoch [3/13], Step [3200/3970], Mel Loss: 1.90306, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:16:56,002: Epoch [3/13], Step [3250/3970], Mel Loss: 1.88371, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:17:34,313: Epoch [3/13], Step [3300/3970], Mel Loss: 1.88573, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:18:14,446: Epoch [3/13], Step [3350/3970], Mel Loss: 1.88976, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:18:53,487: Epoch [3/13], Step [3400/3970], Mel Loss: 1.89630, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:19:32,236: Epoch [3/13], Step [3450/3970], Mel Loss: 1.88677, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:20:11,023: Epoch [3/13], Step [3500/3970], Mel Loss: 1.89492, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:20:49,578: Epoch [3/13], Step [3550/3970], Mel Loss: 1.88537, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:21:28,166: Epoch [3/13], Step [3600/3970], Mel Loss: 1.90183, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:22:06,712: Epoch [3/13], Step [3650/3970], Mel Loss: 1.88612, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:22:45,718: Epoch [3/13], Step [3700/3970], Mel Loss: 1.88134, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:23:24,160: Epoch [3/13], Step [3750/3970], Mel Loss: 1.89001, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:24:03,262: Epoch [3/13], Step [3800/3970], Mel Loss: 1.89932, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:24:42,718: Epoch [3/13], Step [3850/3970], Mel Loss: 1.91552, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:25:21,912: Epoch [3/13], Step [3900/3970], Mel Loss: 1.90403, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:26:01,618: Epoch [3/13], Step [3950/3970], Mel Loss: 1.90984, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:29:37,538: Validation loss: 1.862 - - - - -INFO:2025-06-08 07:30:23,263: Epoch [4/13], Step [50/3970], Mel Loss: 1.88295, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:31:03,774: Epoch [4/13], Step [100/3970], Mel Loss: 1.91122, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:31:43,991: Epoch [4/13], Step [150/3970], Mel Loss: 1.87954, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:32:22,835: Epoch [4/13], Step [200/3970], Mel Loss: 1.87939, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:33:02,394: Epoch [4/13], Step [250/3970], Mel Loss: 1.90374, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:33:41,381: Epoch [4/13], Step [300/3970], Mel Loss: 1.88488, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:34:19,621: Epoch [4/13], Step [350/3970], Mel Loss: 1.90755, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:34:59,533: Epoch [4/13], Step [400/3970], Mel Loss: 1.89587, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:35:37,253: Epoch [4/13], Step [450/3970], Mel Loss: 1.88485, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:36:15,991: Epoch [4/13], Step [500/3970], Mel Loss: 1.89914, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:36:55,950: Epoch [4/13], Step [550/3970], Mel Loss: 1.90362, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:37:34,573: Epoch [4/13], Step [600/3970], Mel Loss: 1.90583, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:38:15,531: Epoch [4/13], Step [650/3970], Mel Loss: 1.91706, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:38:54,959: Epoch [4/13], Step [700/3970], Mel Loss: 1.89177, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:39:33,695: Epoch [4/13], Step [750/3970], Mel Loss: 1.89544, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:40:13,807: Epoch [4/13], Step [800/3970], Mel Loss: 1.88123, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:40:54,251: Epoch [4/13], Step [850/3970], Mel Loss: 1.88187, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:41:34,090: Epoch [4/13], Step [900/3970], Mel Loss: 1.89633, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:42:13,174: Epoch [4/13], Step [950/3970], Mel Loss: 1.89483, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:42:52,853: Epoch [4/13], Step [1000/3970], Mel Loss: 1.89381, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:43:30,664: Epoch [4/13], Step [1050/3970], Mel Loss: 1.90621, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:44:08,145: Epoch [4/13], Step [1100/3970], Mel Loss: 1.88559, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:44:46,697: Epoch [4/13], Step [1150/3970], Mel Loss: 1.88741, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:45:24,913: Epoch [4/13], Step [1200/3970], Mel Loss: 1.90722, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:46:04,711: Epoch [4/13], Step [1250/3970], Mel Loss: 1.90341, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:46:41,609: Epoch [4/13], Step [1300/3970], Mel Loss: 1.87555, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:47:21,840: Epoch [4/13], Step [1350/3970], Mel Loss: 1.90662, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:47:59,664: Epoch [4/13], Step [1400/3970], Mel Loss: 1.91589, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:48:38,878: Epoch [4/13], Step [1450/3970], Mel Loss: 1.91155, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:49:19,078: Epoch [4/13], Step [1500/3970], Mel Loss: 1.89834, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:49:59,470: Epoch [4/13], Step [1550/3970], Mel Loss: 1.90833, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:50:38,224: Epoch [4/13], Step [1600/3970], Mel Loss: 1.90142, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:51:16,052: Epoch [4/13], Step [1650/3970], Mel Loss: 1.90240, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:51:54,951: Epoch [4/13], Step [1700/3970], Mel Loss: 1.89433, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:52:32,500: Epoch [4/13], Step [1750/3970], Mel Loss: 1.87924, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:53:13,195: Epoch [4/13], Step [1800/3970], Mel Loss: 1.89769, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:53:52,760: Epoch [4/13], Step [1850/3970], Mel Loss: 1.88860, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:54:33,129: Epoch [4/13], Step [1900/3970], Mel Loss: 1.90660, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:55:12,611: Epoch [4/13], Step [1950/3970], Mel Loss: 1.89135, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:55:51,751: Epoch [4/13], Step [2000/3970], Mel Loss: 1.89832, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:56:31,978: Epoch [4/13], Step [2050/3970], Mel Loss: 1.88964, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:57:10,419: Epoch [4/13], Step [2100/3970], Mel Loss: 1.90315, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:57:49,684: Epoch [4/13], Step [2150/3970], Mel Loss: 1.90689, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:58:28,701: Epoch [4/13], Step [2200/3970], Mel Loss: 1.90914, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:59:09,483: Epoch [4/13], Step [2250/3970], Mel Loss: 1.89897, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 07:59:47,701: Epoch [4/13], Step [2300/3970], Mel Loss: 1.88284, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:00:25,262: Epoch [4/13], Step [2350/3970], Mel Loss: 1.86203, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:01:02,821: Epoch [4/13], Step [2400/3970], Mel Loss: 1.89959, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:01:40,746: Epoch [4/13], Step [2450/3970], Mel Loss: 1.87908, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:02:19,535: Epoch [4/13], Step [2500/3970], Mel Loss: 1.86930, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:02:57,886: Epoch [4/13], Step [2550/3970], Mel Loss: 1.91118, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:03:37,970: Epoch [4/13], Step [2600/3970], Mel Loss: 1.90387, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:04:17,671: Epoch [4/13], Step [2650/3970], Mel Loss: 1.90366, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:04:55,964: Epoch [4/13], Step [2700/3970], Mel Loss: 1.90146, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:05:34,551: Epoch [4/13], Step [2750/3970], Mel Loss: 1.89734, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:06:13,384: Epoch [4/13], Step [2800/3970], Mel Loss: 1.90309, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:06:53,095: Epoch [4/13], Step [2850/3970], Mel Loss: 1.87625, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:07:31,590: Epoch [4/13], Step [2900/3970], Mel Loss: 1.88480, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:08:09,605: Epoch [4/13], Step [2950/3970], Mel Loss: 1.91768, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:08:49,032: Epoch [4/13], Step [3000/3970], Mel Loss: 1.88393, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:09:27,033: Epoch [4/13], Step [3050/3970], Mel Loss: 1.89326, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:10:06,155: Epoch [4/13], Step [3100/3970], Mel Loss: 1.89521, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:10:44,429: Epoch [4/13], Step [3150/3970], Mel Loss: 1.89359, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:11:23,341: Epoch [4/13], Step [3200/3970], Mel Loss: 1.88629, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:12:01,564: Epoch [4/13], Step [3250/3970], Mel Loss: 1.89237, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:12:39,461: Epoch [4/13], Step [3300/3970], Mel Loss: 1.87828, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:13:19,364: Epoch [4/13], Step [3350/3970], Mel Loss: 1.90892, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:13:59,705: Epoch [4/13], Step [3400/3970], Mel Loss: 1.89910, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:14:39,591: Epoch [4/13], Step [3450/3970], Mel Loss: 1.90741, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:15:20,182: Epoch [4/13], Step [3500/3970], Mel Loss: 1.88363, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:15:58,278: Epoch [4/13], Step [3550/3970], Mel Loss: 1.88943, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:16:36,226: Epoch [4/13], Step [3600/3970], Mel Loss: 1.89740, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:17:14,825: Epoch [4/13], Step [3650/3970], Mel Loss: 1.89093, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:17:54,992: Epoch [4/13], Step [3700/3970], Mel Loss: 1.88510, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:18:33,522: Epoch [4/13], Step [3750/3970], Mel Loss: 1.89154, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:19:12,352: Epoch [4/13], Step [3800/3970], Mel Loss: 1.91070, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:19:52,372: Epoch [4/13], Step [3850/3970], Mel Loss: 1.89452, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:20:31,604: Epoch [4/13], Step [3900/3970], Mel Loss: 1.89565, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:21:09,779: Epoch [4/13], Step [3950/3970], Mel Loss: 1.87704, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:24:41,288: Validation loss: 1.861 - - - - -INFO:2025-06-08 08:25:21,929: Epoch [5/13], Step [50/3970], Mel Loss: 1.89605, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:26:03,528: Epoch [5/13], Step [100/3970], Mel Loss: 1.88444, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:26:43,453: Epoch [5/13], Step [150/3970], Mel Loss: 1.89979, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:27:22,500: Epoch [5/13], Step [200/3970], Mel Loss: 1.89953, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:28:03,574: Epoch [5/13], Step [250/3970], Mel Loss: 1.89962, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:28:44,283: Epoch [5/13], Step [300/3970], Mel Loss: 1.87168, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:29:24,860: Epoch [5/13], Step [350/3970], Mel Loss: 1.88919, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:30:05,136: Epoch [5/13], Step [400/3970], Mel Loss: 1.88703, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:30:43,893: Epoch [5/13], Step [450/3970], Mel Loss: 1.87984, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:31:24,672: Epoch [5/13], Step [500/3970], Mel Loss: 1.87074, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:32:05,434: Epoch [5/13], Step [550/3970], Mel Loss: 1.90622, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:32:43,933: Epoch [5/13], Step [600/3970], Mel Loss: 1.89154, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:33:23,929: Epoch [5/13], Step [650/3970], Mel Loss: 1.90105, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:34:03,105: Epoch [5/13], Step [700/3970], Mel Loss: 1.89941, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:34:42,556: Epoch [5/13], Step [750/3970], Mel Loss: 1.88425, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:35:21,880: Epoch [5/13], Step [800/3970], Mel Loss: 1.90958, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:36:01,199: Epoch [5/13], Step [850/3970], Mel Loss: 1.89314, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:36:40,462: Epoch [5/13], Step [900/3970], Mel Loss: 1.91613, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:37:20,047: Epoch [5/13], Step [950/3970], Mel Loss: 1.90994, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:37:59,462: Epoch [5/13], Step [1000/3970], Mel Loss: 1.87437, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:38:37,551: Epoch [5/13], Step [1050/3970], Mel Loss: 1.88716, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:39:15,329: Epoch [5/13], Step [1100/3970], Mel Loss: 1.90149, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:39:54,782: Epoch [5/13], Step [1150/3970], Mel Loss: 1.89392, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:40:33,665: Epoch [5/13], Step [1200/3970], Mel Loss: 1.90298, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:41:11,633: Epoch [5/13], Step [1250/3970], Mel Loss: 1.91210, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:41:51,487: Epoch [5/13], Step [1300/3970], Mel Loss: 1.89778, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:42:31,781: Epoch [5/13], Step [1350/3970], Mel Loss: 1.89827, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:43:10,590: Epoch [5/13], Step [1400/3970], Mel Loss: 1.89693, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:43:49,648: Epoch [5/13], Step [1450/3970], Mel Loss: 1.88879, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:44:28,728: Epoch [5/13], Step [1500/3970], Mel Loss: 1.88297, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:45:09,826: Epoch [5/13], Step [1550/3970], Mel Loss: 1.88634, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:45:50,165: Epoch [5/13], Step [1600/3970], Mel Loss: 1.89522, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:46:28,368: Epoch [5/13], Step [1650/3970], Mel Loss: 1.90644, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:47:07,668: Epoch [5/13], Step [1700/3970], Mel Loss: 1.87644, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:47:46,691: Epoch [5/13], Step [1750/3970], Mel Loss: 1.90154, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:48:26,204: Epoch [5/13], Step [1800/3970], Mel Loss: 1.90863, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:49:02,681: Epoch [5/13], Step [1850/3970], Mel Loss: 1.88585, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:49:41,262: Epoch [5/13], Step [1900/3970], Mel Loss: 1.88782, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:50:20,277: Epoch [5/13], Step [1950/3970], Mel Loss: 1.89139, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:50:59,750: Epoch [5/13], Step [2000/3970], Mel Loss: 1.91153, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:51:38,306: Epoch [5/13], Step [2050/3970], Mel Loss: 1.88903, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:52:16,800: Epoch [5/13], Step [2100/3970], Mel Loss: 1.88293, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:52:55,579: Epoch [5/13], Step [2150/3970], Mel Loss: 1.89849, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:53:34,525: Epoch [5/13], Step [2200/3970], Mel Loss: 1.89333, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:54:14,597: Epoch [5/13], Step [2250/3970], Mel Loss: 1.91311, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:54:53,184: Epoch [5/13], Step [2300/3970], Mel Loss: 1.87768, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:55:31,928: Epoch [5/13], Step [2350/3970], Mel Loss: 1.89805, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:56:10,228: Epoch [5/13], Step [2400/3970], Mel Loss: 1.88670, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:56:47,510: Epoch [5/13], Step [2450/3970], Mel Loss: 1.89329, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:57:27,593: Epoch [5/13], Step [2500/3970], Mel Loss: 1.88434, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:58:06,345: Epoch [5/13], Step [2550/3970], Mel Loss: 1.88778, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:58:46,438: Epoch [5/13], Step [2600/3970], Mel Loss: 1.89772, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 08:59:23,894: Epoch [5/13], Step [2650/3970], Mel Loss: 1.91898, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:00:02,349: Epoch [5/13], Step [2700/3970], Mel Loss: 1.89744, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:00:41,519: Epoch [5/13], Step [2750/3970], Mel Loss: 1.90318, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:01:19,949: Epoch [5/13], Step [2800/3970], Mel Loss: 1.88144, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:01:59,617: Epoch [5/13], Step [2850/3970], Mel Loss: 1.89089, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:02:39,091: Epoch [5/13], Step [2900/3970], Mel Loss: 1.90223, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:03:16,791: Epoch [5/13], Step [2950/3970], Mel Loss: 1.89258, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:03:58,035: Epoch [5/13], Step [3000/3970], Mel Loss: 1.91855, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:04:36,680: Epoch [5/13], Step [3050/3970], Mel Loss: 1.89070, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:05:15,158: Epoch [5/13], Step [3100/3970], Mel Loss: 1.88844, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:05:54,212: Epoch [5/13], Step [3150/3970], Mel Loss: 1.89540, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:06:33,157: Epoch [5/13], Step [3200/3970], Mel Loss: 1.89521, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:07:11,966: Epoch [5/13], Step [3250/3970], Mel Loss: 1.90610, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:07:50,475: Epoch [5/13], Step [3300/3970], Mel Loss: 1.91122, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:08:28,304: Epoch [5/13], Step [3350/3970], Mel Loss: 1.89667, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:09:07,061: Epoch [5/13], Step [3400/3970], Mel Loss: 1.90623, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:09:44,787: Epoch [5/13], Step [3450/3970], Mel Loss: 1.88026, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:10:22,449: Epoch [5/13], Step [3500/3970], Mel Loss: 1.86726, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:11:01,211: Epoch [5/13], Step [3550/3970], Mel Loss: 1.89992, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:11:40,547: Epoch [5/13], Step [3600/3970], Mel Loss: 1.88578, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:12:18,643: Epoch [5/13], Step [3650/3970], Mel Loss: 1.90778, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:12:57,548: Epoch [5/13], Step [3700/3970], Mel Loss: 1.90733, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:13:37,449: Epoch [5/13], Step [3750/3970], Mel Loss: 1.89372, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:14:17,487: Epoch [5/13], Step [3800/3970], Mel Loss: 1.89425, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:14:56,366: Epoch [5/13], Step [3850/3970], Mel Loss: 1.90900, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:15:35,441: Epoch [5/13], Step [3900/3970], Mel Loss: 1.90402, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:16:14,013: Epoch [5/13], Step [3950/3970], Mel Loss: 1.89960, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:19:48,922: Validation loss: 1.866 - - - - -INFO:2025-06-08 09:20:34,804: Epoch [6/13], Step [50/3970], Mel Loss: 1.89066, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:21:13,711: Epoch [6/13], Step [100/3970], Mel Loss: 1.89167, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:21:53,540: Epoch [6/13], Step [150/3970], Mel Loss: 1.89372, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:22:32,875: Epoch [6/13], Step [200/3970], Mel Loss: 1.91378, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:23:11,466: Epoch [6/13], Step [250/3970], Mel Loss: 1.88445, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:23:49,902: Epoch [6/13], Step [300/3970], Mel Loss: 1.88759, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:24:28,270: Epoch [6/13], Step [350/3970], Mel Loss: 1.88936, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:25:07,880: Epoch [6/13], Step [400/3970], Mel Loss: 1.91401, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:25:46,660: Epoch [6/13], Step [450/3970], Mel Loss: 1.91462, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:26:26,272: Epoch [6/13], Step [500/3970], Mel Loss: 1.88289, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:27:04,510: Epoch [6/13], Step [550/3970], Mel Loss: 1.89028, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:27:42,579: Epoch [6/13], Step [600/3970], Mel Loss: 1.89515, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:28:20,816: Epoch [6/13], Step [650/3970], Mel Loss: 1.89173, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:29:00,980: Epoch [6/13], Step [700/3970], Mel Loss: 1.89677, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:29:38,590: Epoch [6/13], Step [750/3970], Mel Loss: 1.90326, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:30:16,905: Epoch [6/13], Step [800/3970], Mel Loss: 1.90271, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:30:56,832: Epoch [6/13], Step [850/3970], Mel Loss: 1.89954, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:31:36,891: Epoch [6/13], Step [900/3970], Mel Loss: 1.90614, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:32:14,178: Epoch [6/13], Step [950/3970], Mel Loss: 1.89156, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:32:52,555: Epoch [6/13], Step [1000/3970], Mel Loss: 1.88405, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:33:30,918: Epoch [6/13], Step [1050/3970], Mel Loss: 1.88432, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:34:10,910: Epoch [6/13], Step [1100/3970], Mel Loss: 1.88152, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:34:49,107: Epoch [6/13], Step [1150/3970], Mel Loss: 1.89538, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:35:28,201: Epoch [6/13], Step [1200/3970], Mel Loss: 1.87677, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:36:06,188: Epoch [6/13], Step [1250/3970], Mel Loss: 1.93275, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:36:44,980: Epoch [6/13], Step [1300/3970], Mel Loss: 1.89484, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:37:23,343: Epoch [6/13], Step [1350/3970], Mel Loss: 1.89415, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:38:03,235: Epoch [6/13], Step [1400/3970], Mel Loss: 1.90505, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:38:42,077: Epoch [6/13], Step [1450/3970], Mel Loss: 1.90588, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:39:21,056: Epoch [6/13], Step [1500/3970], Mel Loss: 1.86905, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:40:00,478: Epoch [6/13], Step [1550/3970], Mel Loss: 1.87935, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:40:39,496: Epoch [6/13], Step [1600/3970], Mel Loss: 1.88191, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:41:19,377: Epoch [6/13], Step [1650/3970], Mel Loss: 1.90761, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:42:00,168: Epoch [6/13], Step [1700/3970], Mel Loss: 1.88188, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:42:39,257: Epoch [6/13], Step [1750/3970], Mel Loss: 1.90877, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:43:16,530: Epoch [6/13], Step [1800/3970], Mel Loss: 1.88345, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:43:56,521: Epoch [6/13], Step [1850/3970], Mel Loss: 1.87391, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:44:36,688: Epoch [6/13], Step [1900/3970], Mel Loss: 1.88957, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:45:15,456: Epoch [6/13], Step [1950/3970], Mel Loss: 1.91726, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:45:55,119: Epoch [6/13], Step [2000/3970], Mel Loss: 1.90567, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:46:33,475: Epoch [6/13], Step [2050/3970], Mel Loss: 1.90629, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:47:12,689: Epoch [6/13], Step [2100/3970], Mel Loss: 1.89501, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:47:51,119: Epoch [6/13], Step [2150/3970], Mel Loss: 1.89850, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:48:29,361: Epoch [6/13], Step [2200/3970], Mel Loss: 1.89373, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:49:08,921: Epoch [6/13], Step [2250/3970], Mel Loss: 1.90909, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:49:47,729: Epoch [6/13], Step [2300/3970], Mel Loss: 1.89944, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:50:26,230: Epoch [6/13], Step [2350/3970], Mel Loss: 1.89509, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:51:04,160: Epoch [6/13], Step [2400/3970], Mel Loss: 1.88018, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:51:42,342: Epoch [6/13], Step [2450/3970], Mel Loss: 1.89390, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:52:20,671: Epoch [6/13], Step [2500/3970], Mel Loss: 1.90952, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:53:00,398: Epoch [6/13], Step [2550/3970], Mel Loss: 1.91033, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:53:37,821: Epoch [6/13], Step [2600/3970], Mel Loss: 1.88138, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:54:15,138: Epoch [6/13], Step [2650/3970], Mel Loss: 1.87123, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:54:52,395: Epoch [6/13], Step [2700/3970], Mel Loss: 1.88903, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:55:31,568: Epoch [6/13], Step [2750/3970], Mel Loss: 1.87280, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:56:09,897: Epoch [6/13], Step [2800/3970], Mel Loss: 1.89837, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:56:48,446: Epoch [6/13], Step [2850/3970], Mel Loss: 1.88260, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:57:28,268: Epoch [6/13], Step [2900/3970], Mel Loss: 1.89199, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:58:09,074: Epoch [6/13], Step [2950/3970], Mel Loss: 1.91476, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:58:46,994: Epoch [6/13], Step [3000/3970], Mel Loss: 1.88379, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 09:59:25,169: Epoch [6/13], Step [3050/3970], Mel Loss: 1.90675, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:00:02,048: Epoch [6/13], Step [3100/3970], Mel Loss: 1.87280, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:00:41,879: Epoch [6/13], Step [3150/3970], Mel Loss: 1.90261, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:01:20,620: Epoch [6/13], Step [3200/3970], Mel Loss: 1.90589, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:01:59,709: Epoch [6/13], Step [3250/3970], Mel Loss: 1.89290, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:02:40,360: Epoch [6/13], Step [3300/3970], Mel Loss: 1.90651, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:03:20,821: Epoch [6/13], Step [3350/3970], Mel Loss: 1.88383, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:03:59,958: Epoch [6/13], Step [3400/3970], Mel Loss: 1.89515, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:04:37,903: Epoch [6/13], Step [3450/3970], Mel Loss: 1.89450, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:05:15,965: Epoch [6/13], Step [3500/3970], Mel Loss: 1.90706, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:05:54,892: Epoch [6/13], Step [3550/3970], Mel Loss: 1.90109, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:06:32,892: Epoch [6/13], Step [3600/3970], Mel Loss: 1.90992, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:07:11,037: Epoch [6/13], Step [3650/3970], Mel Loss: 1.90127, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:07:49,224: Epoch [6/13], Step [3700/3970], Mel Loss: 1.90539, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:08:27,102: Epoch [6/13], Step [3750/3970], Mel Loss: 1.89746, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:09:07,038: Epoch [6/13], Step [3800/3970], Mel Loss: 1.91291, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:09:47,399: Epoch [6/13], Step [3850/3970], Mel Loss: 1.87494, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:10:27,641: Epoch [6/13], Step [3900/3970], Mel Loss: 1.88596, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:11:06,967: Epoch [6/13], Step [3950/3970], Mel Loss: 1.88558, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:14:44,338: Validation loss: 1.867 - - - - -INFO:2025-06-08 10:15:25,162: Epoch [7/13], Step [50/3970], Mel Loss: 1.90738, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:16:02,437: Epoch [7/13], Step [100/3970], Mel Loss: 1.88089, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:16:42,435: Epoch [7/13], Step [150/3970], Mel Loss: 1.90015, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:17:19,620: Epoch [7/13], Step [200/3970], Mel Loss: 1.89720, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:17:59,042: Epoch [7/13], Step [250/3970], Mel Loss: 1.89048, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:18:38,963: Epoch [7/13], Step [300/3970], Mel Loss: 1.90259, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:19:16,901: Epoch [7/13], Step [350/3970], Mel Loss: 1.87658, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:19:55,282: Epoch [7/13], Step [400/3970], Mel Loss: 1.89502, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:20:33,516: Epoch [7/13], Step [450/3970], Mel Loss: 1.89300, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:21:13,479: Epoch [7/13], Step [500/3970], Mel Loss: 1.90461, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:21:53,281: Epoch [7/13], Step [550/3970], Mel Loss: 1.87351, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:22:32,403: Epoch [7/13], Step [600/3970], Mel Loss: 1.87229, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:23:11,602: Epoch [7/13], Step [650/3970], Mel Loss: 1.88857, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:23:51,124: Epoch [7/13], Step [700/3970], Mel Loss: 1.88770, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:24:28,829: Epoch [7/13], Step [750/3970], Mel Loss: 1.89847, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:25:09,590: Epoch [7/13], Step [800/3970], Mel Loss: 1.93476, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:25:48,481: Epoch [7/13], Step [850/3970], Mel Loss: 1.88931, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:26:26,940: Epoch [7/13], Step [900/3970], Mel Loss: 1.90834, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:27:04,716: Epoch [7/13], Step [950/3970], Mel Loss: 1.88418, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:27:43,391: Epoch [7/13], Step [1000/3970], Mel Loss: 1.88694, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:28:22,996: Epoch [7/13], Step [1050/3970], Mel Loss: 1.89262, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:29:02,328: Epoch [7/13], Step [1100/3970], Mel Loss: 1.90504, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:29:41,399: Epoch [7/13], Step [1150/3970], Mel Loss: 1.89391, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:30:20,690: Epoch [7/13], Step [1200/3970], Mel Loss: 1.87824, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:31:00,511: Epoch [7/13], Step [1250/3970], Mel Loss: 1.89581, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:31:39,211: Epoch [7/13], Step [1300/3970], Mel Loss: 1.89773, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:32:19,166: Epoch [7/13], Step [1350/3970], Mel Loss: 1.89252, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:32:58,405: Epoch [7/13], Step [1400/3970], Mel Loss: 1.90623, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:33:38,033: Epoch [7/13], Step [1450/3970], Mel Loss: 1.89944, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:34:16,311: Epoch [7/13], Step [1500/3970], Mel Loss: 1.88751, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:34:53,820: Epoch [7/13], Step [1550/3970], Mel Loss: 1.90108, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:35:32,934: Epoch [7/13], Step [1600/3970], Mel Loss: 1.89112, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:36:11,406: Epoch [7/13], Step [1650/3970], Mel Loss: 1.91379, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:36:51,420: Epoch [7/13], Step [1700/3970], Mel Loss: 1.90916, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:37:31,532: Epoch [7/13], Step [1750/3970], Mel Loss: 1.89111, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:38:12,629: Epoch [7/13], Step [1800/3970], Mel Loss: 1.89067, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:38:51,732: Epoch [7/13], Step [1850/3970], Mel Loss: 1.89318, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:39:29,774: Epoch [7/13], Step [1900/3970], Mel Loss: 1.89770, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:40:11,359: Epoch [7/13], Step [1950/3970], Mel Loss: 1.90554, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:40:50,100: Epoch [7/13], Step [2000/3970], Mel Loss: 1.89957, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:41:28,909: Epoch [7/13], Step [2050/3970], Mel Loss: 1.88555, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:42:08,721: Epoch [7/13], Step [2100/3970], Mel Loss: 1.89646, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:42:47,082: Epoch [7/13], Step [2150/3970], Mel Loss: 1.87562, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:43:25,257: Epoch [7/13], Step [2200/3970], Mel Loss: 1.88370, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:44:04,948: Epoch [7/13], Step [2250/3970], Mel Loss: 1.89816, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:44:43,090: Epoch [7/13], Step [2300/3970], Mel Loss: 1.92224, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:45:22,528: Epoch [7/13], Step [2350/3970], Mel Loss: 1.88760, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:46:01,160: Epoch [7/13], Step [2400/3970], Mel Loss: 1.88545, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:46:40,159: Epoch [7/13], Step [2450/3970], Mel Loss: 1.88753, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:47:18,237: Epoch [7/13], Step [2500/3970], Mel Loss: 1.89152, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:47:58,216: Epoch [7/13], Step [2550/3970], Mel Loss: 1.88742, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:48:36,940: Epoch [7/13], Step [2600/3970], Mel Loss: 1.92798, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:49:16,526: Epoch [7/13], Step [2650/3970], Mel Loss: 1.88614, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:49:54,965: Epoch [7/13], Step [2700/3970], Mel Loss: 1.89075, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:50:33,614: Epoch [7/13], Step [2750/3970], Mel Loss: 1.91238, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:51:14,116: Epoch [7/13], Step [2800/3970], Mel Loss: 1.88037, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:51:52,043: Epoch [7/13], Step [2850/3970], Mel Loss: 1.90932, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:52:31,156: Epoch [7/13], Step [2900/3970], Mel Loss: 1.89355, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:53:09,649: Epoch [7/13], Step [2950/3970], Mel Loss: 1.90258, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:53:48,580: Epoch [7/13], Step [3000/3970], Mel Loss: 1.90356, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:54:27,861: Epoch [7/13], Step [3050/3970], Mel Loss: 1.91509, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:55:06,772: Epoch [7/13], Step [3100/3970], Mel Loss: 1.91071, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:55:45,830: Epoch [7/13], Step [3150/3970], Mel Loss: 1.88667, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:56:25,575: Epoch [7/13], Step [3200/3970], Mel Loss: 1.87234, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:57:05,217: Epoch [7/13], Step [3250/3970], Mel Loss: 1.89188, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:57:44,734: Epoch [7/13], Step [3300/3970], Mel Loss: 1.91032, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:58:24,625: Epoch [7/13], Step [3350/3970], Mel Loss: 1.90392, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:59:02,720: Epoch [7/13], Step [3400/3970], Mel Loss: 1.90762, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 10:59:42,328: Epoch [7/13], Step [3450/3970], Mel Loss: 1.88671, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:00:20,785: Epoch [7/13], Step [3500/3970], Mel Loss: 1.89195, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:00:59,364: Epoch [7/13], Step [3550/3970], Mel Loss: 1.88759, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:01:40,147: Epoch [7/13], Step [3600/3970], Mel Loss: 1.88809, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:02:18,700: Epoch [7/13], Step [3650/3970], Mel Loss: 1.90480, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:02:57,105: Epoch [7/13], Step [3700/3970], Mel Loss: 1.90109, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:03:35,714: Epoch [7/13], Step [3750/3970], Mel Loss: 1.88892, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:04:14,175: Epoch [7/13], Step [3800/3970], Mel Loss: 1.90103, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:04:53,090: Epoch [7/13], Step [3850/3970], Mel Loss: 1.88879, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:05:32,994: Epoch [7/13], Step [3900/3970], Mel Loss: 1.88410, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:06:11,256: Epoch [7/13], Step [3950/3970], Mel Loss: 1.89077, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:09:47,475: Validation loss: 1.865 - - - - -INFO:2025-06-08 11:10:32,772: Epoch [8/13], Step [50/3970], Mel Loss: 1.89350, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:11:12,331: Epoch [8/13], Step [100/3970], Mel Loss: 1.90840, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:11:51,666: Epoch [8/13], Step [150/3970], Mel Loss: 1.89228, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:12:32,638: Epoch [8/13], Step [200/3970], Mel Loss: 1.89592, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:13:10,404: Epoch [8/13], Step [250/3970], Mel Loss: 1.88282, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:13:48,709: Epoch [8/13], Step [300/3970], Mel Loss: 1.90377, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:14:28,079: Epoch [8/13], Step [350/3970], Mel Loss: 1.87681, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:15:07,721: Epoch [8/13], Step [400/3970], Mel Loss: 1.87584, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:15:46,052: Epoch [8/13], Step [450/3970], Mel Loss: 1.89294, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:16:24,885: Epoch [8/13], Step [500/3970], Mel Loss: 1.89580, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:17:04,789: Epoch [8/13], Step [550/3970], Mel Loss: 1.89235, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:17:43,955: Epoch [8/13], Step [600/3970], Mel Loss: 1.89649, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:18:22,982: Epoch [8/13], Step [650/3970], Mel Loss: 1.88705, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:19:01,763: Epoch [8/13], Step [700/3970], Mel Loss: 1.89517, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:19:39,926: Epoch [8/13], Step [750/3970], Mel Loss: 1.90923, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:20:19,878: Epoch [8/13], Step [800/3970], Mel Loss: 1.91806, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:20:58,545: Epoch [8/13], Step [850/3970], Mel Loss: 1.89633, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:21:36,988: Epoch [8/13], Step [900/3970], Mel Loss: 1.92855, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:22:16,441: Epoch [8/13], Step [950/3970], Mel Loss: 1.89890, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:22:55,660: Epoch [8/13], Step [1000/3970], Mel Loss: 1.90784, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:23:33,090: Epoch [8/13], Step [1050/3970], Mel Loss: 1.89385, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:24:12,388: Epoch [8/13], Step [1100/3970], Mel Loss: 1.89230, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:24:50,857: Epoch [8/13], Step [1150/3970], Mel Loss: 1.88718, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:25:30,370: Epoch [8/13], Step [1200/3970], Mel Loss: 1.89512, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:26:07,949: Epoch [8/13], Step [1250/3970], Mel Loss: 1.89281, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:26:46,445: Epoch [8/13], Step [1300/3970], Mel Loss: 1.88063, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:27:25,376: Epoch [8/13], Step [1350/3970], Mel Loss: 1.89526, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:28:05,464: Epoch [8/13], Step [1400/3970], Mel Loss: 1.88576, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:28:43,336: Epoch [8/13], Step [1450/3970], Mel Loss: 1.89331, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:29:22,023: Epoch [8/13], Step [1500/3970], Mel Loss: 1.89559, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:30:00,727: Epoch [8/13], Step [1550/3970], Mel Loss: 1.87713, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:30:41,245: Epoch [8/13], Step [1600/3970], Mel Loss: 1.88939, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:31:20,055: Epoch [8/13], Step [1650/3970], Mel Loss: 1.88740, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:31:59,907: Epoch [8/13], Step [1700/3970], Mel Loss: 1.91075, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:32:38,004: Epoch [8/13], Step [1750/3970], Mel Loss: 1.88373, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:33:16,048: Epoch [8/13], Step [1800/3970], Mel Loss: 1.89970, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:33:55,445: Epoch [8/13], Step [1850/3970], Mel Loss: 1.89625, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:34:35,401: Epoch [8/13], Step [1900/3970], Mel Loss: 1.89904, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:35:14,978: Epoch [8/13], Step [1950/3970], Mel Loss: 1.89825, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:35:55,854: Epoch [8/13], Step [2000/3970], Mel Loss: 1.89805, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:36:34,573: Epoch [8/13], Step [2050/3970], Mel Loss: 1.88396, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:37:13,374: Epoch [8/13], Step [2100/3970], Mel Loss: 1.90396, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:37:51,385: Epoch [8/13], Step [2150/3970], Mel Loss: 1.90427, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:38:31,114: Epoch [8/13], Step [2200/3970], Mel Loss: 1.89495, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:39:10,665: Epoch [8/13], Step [2250/3970], Mel Loss: 1.88559, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:39:48,722: Epoch [8/13], Step [2300/3970], Mel Loss: 1.89109, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:40:28,344: Epoch [8/13], Step [2350/3970], Mel Loss: 1.88602, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:41:06,248: Epoch [8/13], Step [2400/3970], Mel Loss: 1.89887, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:41:44,362: Epoch [8/13], Step [2450/3970], Mel Loss: 1.87826, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:42:23,392: Epoch [8/13], Step [2500/3970], Mel Loss: 1.90235, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:43:01,986: Epoch [8/13], Step [2550/3970], Mel Loss: 1.90580, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:43:39,475: Epoch [8/13], Step [2600/3970], Mel Loss: 1.88373, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:44:19,740: Epoch [8/13], Step [2650/3970], Mel Loss: 1.90089, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:44:59,518: Epoch [8/13], Step [2700/3970], Mel Loss: 1.89741, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:45:37,647: Epoch [8/13], Step [2750/3970], Mel Loss: 1.89351, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:46:17,287: Epoch [8/13], Step [2800/3970], Mel Loss: 1.88666, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:46:53,811: Epoch [8/13], Step [2850/3970], Mel Loss: 1.89097, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:47:33,087: Epoch [8/13], Step [2900/3970], Mel Loss: 1.87727, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:48:09,803: Epoch [8/13], Step [2950/3970], Mel Loss: 1.88560, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:48:49,186: Epoch [8/13], Step [3000/3970], Mel Loss: 1.89793, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:49:30,702: Epoch [8/13], Step [3050/3970], Mel Loss: 1.89226, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:50:11,584: Epoch [8/13], Step [3100/3970], Mel Loss: 1.87812, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:50:49,210: Epoch [8/13], Step [3150/3970], Mel Loss: 1.90119, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:51:28,331: Epoch [8/13], Step [3200/3970], Mel Loss: 1.87488, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:52:08,145: Epoch [8/13], Step [3250/3970], Mel Loss: 1.90345, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:52:47,358: Epoch [8/13], Step [3300/3970], Mel Loss: 1.90156, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:53:25,375: Epoch [8/13], Step [3350/3970], Mel Loss: 1.89272, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:54:04,138: Epoch [8/13], Step [3400/3970], Mel Loss: 1.89392, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:54:42,532: Epoch [8/13], Step [3450/3970], Mel Loss: 1.92869, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:55:21,605: Epoch [8/13], Step [3500/3970], Mel Loss: 1.90034, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:55:59,614: Epoch [8/13], Step [3550/3970], Mel Loss: 1.89501, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:56:39,580: Epoch [8/13], Step [3600/3970], Mel Loss: 1.90896, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:57:19,512: Epoch [8/13], Step [3650/3970], Mel Loss: 1.90221, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:57:59,066: Epoch [8/13], Step [3700/3970], Mel Loss: 1.89497, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:58:39,262: Epoch [8/13], Step [3750/3970], Mel Loss: 1.88797, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:59:19,142: Epoch [8/13], Step [3800/3970], Mel Loss: 1.89196, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 11:59:57,407: Epoch [8/13], Step [3850/3970], Mel Loss: 1.89882, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:00:35,755: Epoch [8/13], Step [3900/3970], Mel Loss: 1.89388, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:01:15,748: Epoch [8/13], Step [3950/3970], Mel Loss: 1.91232, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:04:51,693: Validation loss: 1.864 - - - - -INFO:2025-06-08 12:05:34,539: Epoch [9/13], Step [50/3970], Mel Loss: 1.89520, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:06:13,341: Epoch [9/13], Step [100/3970], Mel Loss: 1.88426, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:06:53,294: Epoch [9/13], Step [150/3970], Mel Loss: 1.88484, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:07:30,741: Epoch [9/13], Step [200/3970], Mel Loss: 1.89492, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:08:11,878: Epoch [9/13], Step [250/3970], Mel Loss: 1.89379, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:08:51,177: Epoch [9/13], Step [300/3970], Mel Loss: 1.89941, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:09:28,498: Epoch [9/13], Step [350/3970], Mel Loss: 1.89119, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:10:07,907: Epoch [9/13], Step [400/3970], Mel Loss: 1.90395, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:10:45,311: Epoch [9/13], Step [450/3970], Mel Loss: 1.89731, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:11:24,055: Epoch [9/13], Step [500/3970], Mel Loss: 1.88764, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:12:02,658: Epoch [9/13], Step [550/3970], Mel Loss: 1.90449, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:12:42,594: Epoch [9/13], Step [600/3970], Mel Loss: 1.90585, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:13:21,114: Epoch [9/13], Step [650/3970], Mel Loss: 1.88210, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:14:01,517: Epoch [9/13], Step [700/3970], Mel Loss: 1.89103, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:14:40,904: Epoch [9/13], Step [750/3970], Mel Loss: 1.89526, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:15:20,341: Epoch [9/13], Step [800/3970], Mel Loss: 1.91542, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:15:59,915: Epoch [9/13], Step [850/3970], Mel Loss: 1.89635, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:16:38,098: Epoch [9/13], Step [900/3970], Mel Loss: 1.88783, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:17:16,289: Epoch [9/13], Step [950/3970], Mel Loss: 1.91203, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:17:54,601: Epoch [9/13], Step [1000/3970], Mel Loss: 1.90735, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:18:31,621: Epoch [9/13], Step [1050/3970], Mel Loss: 1.90070, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:19:09,911: Epoch [9/13], Step [1100/3970], Mel Loss: 1.88891, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:19:49,867: Epoch [9/13], Step [1150/3970], Mel Loss: 1.90071, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:20:28,258: Epoch [9/13], Step [1200/3970], Mel Loss: 1.88189, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:21:06,714: Epoch [9/13], Step [1250/3970], Mel Loss: 1.88398, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:21:46,273: Epoch [9/13], Step [1300/3970], Mel Loss: 1.89658, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:22:25,697: Epoch [9/13], Step [1350/3970], Mel Loss: 1.88041, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:23:04,284: Epoch [9/13], Step [1400/3970], Mel Loss: 1.90530, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:23:43,871: Epoch [9/13], Step [1450/3970], Mel Loss: 1.89753, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:24:22,041: Epoch [9/13], Step [1500/3970], Mel Loss: 1.89903, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:25:01,473: Epoch [9/13], Step [1550/3970], Mel Loss: 1.86928, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:25:40,463: Epoch [9/13], Step [1600/3970], Mel Loss: 1.88689, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:26:18,029: Epoch [9/13], Step [1650/3970], Mel Loss: 1.89554, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:26:56,615: Epoch [9/13], Step [1700/3970], Mel Loss: 1.89697, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:27:34,897: Epoch [9/13], Step [1750/3970], Mel Loss: 1.90228, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:28:14,430: Epoch [9/13], Step [1800/3970], Mel Loss: 1.89327, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:28:53,252: Epoch [9/13], Step [1850/3970], Mel Loss: 1.89882, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:29:31,940: Epoch [9/13], Step [1900/3970], Mel Loss: 1.89740, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:30:09,939: Epoch [9/13], Step [1950/3970], Mel Loss: 1.88015, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:30:48,186: Epoch [9/13], Step [2000/3970], Mel Loss: 1.89827, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:31:26,970: Epoch [9/13], Step [2050/3970], Mel Loss: 1.91007, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:32:05,916: Epoch [9/13], Step [2100/3970], Mel Loss: 1.87949, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:32:43,897: Epoch [9/13], Step [2150/3970], Mel Loss: 1.90078, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:33:22,545: Epoch [9/13], Step [2200/3970], Mel Loss: 1.89737, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:34:02,810: Epoch [9/13], Step [2250/3970], Mel Loss: 1.87589, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:34:41,874: Epoch [9/13], Step [2300/3970], Mel Loss: 1.90855, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:35:20,601: Epoch [9/13], Step [2350/3970], Mel Loss: 1.90597, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:36:00,645: Epoch [9/13], Step [2400/3970], Mel Loss: 1.88839, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:36:41,524: Epoch [9/13], Step [2450/3970], Mel Loss: 1.89057, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:37:21,322: Epoch [9/13], Step [2500/3970], Mel Loss: 1.89672, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:38:00,095: Epoch [9/13], Step [2550/3970], Mel Loss: 1.89820, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:38:39,113: Epoch [9/13], Step [2600/3970], Mel Loss: 1.89147, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:39:18,146: Epoch [9/13], Step [2650/3970], Mel Loss: 1.88877, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:39:57,622: Epoch [9/13], Step [2700/3970], Mel Loss: 1.88861, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:40:37,417: Epoch [9/13], Step [2750/3970], Mel Loss: 1.88087, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:41:15,005: Epoch [9/13], Step [2800/3970], Mel Loss: 1.90222, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:41:53,690: Epoch [9/13], Step [2850/3970], Mel Loss: 1.91683, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:42:32,591: Epoch [9/13], Step [2900/3970], Mel Loss: 1.89424, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:43:11,125: Epoch [9/13], Step [2950/3970], Mel Loss: 1.89942, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:43:48,168: Epoch [9/13], Step [3000/3970], Mel Loss: 1.89250, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:44:27,623: Epoch [9/13], Step [3050/3970], Mel Loss: 1.89704, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:45:05,505: Epoch [9/13], Step [3100/3970], Mel Loss: 1.89996, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:45:44,802: Epoch [9/13], Step [3150/3970], Mel Loss: 1.89863, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:46:21,713: Epoch [9/13], Step [3200/3970], Mel Loss: 1.88970, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:47:01,473: Epoch [9/13], Step [3250/3970], Mel Loss: 1.90542, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:47:41,195: Epoch [9/13], Step [3300/3970], Mel Loss: 1.89960, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:48:19,215: Epoch [9/13], Step [3350/3970], Mel Loss: 1.88638, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:48:57,581: Epoch [9/13], Step [3400/3970], Mel Loss: 1.89566, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:49:37,168: Epoch [9/13], Step [3450/3970], Mel Loss: 1.91257, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:50:15,833: Epoch [9/13], Step [3500/3970], Mel Loss: 1.90833, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:50:54,597: Epoch [9/13], Step [3550/3970], Mel Loss: 1.89943, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:51:32,467: Epoch [9/13], Step [3600/3970], Mel Loss: 1.89919, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:52:12,483: Epoch [9/13], Step [3650/3970], Mel Loss: 1.90111, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:52:51,256: Epoch [9/13], Step [3700/3970], Mel Loss: 1.89056, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:53:31,206: Epoch [9/13], Step [3750/3970], Mel Loss: 1.90477, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:54:08,327: Epoch [9/13], Step [3800/3970], Mel Loss: 1.90742, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:54:48,919: Epoch [9/13], Step [3850/3970], Mel Loss: 1.89879, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:55:27,049: Epoch [9/13], Step [3900/3970], Mel Loss: 1.90150, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:56:04,623: Epoch [9/13], Step [3950/3970], Mel Loss: 1.90068, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 12:59:41,806: Validation loss: 1.862 - - - - -INFO:2025-06-08 13:00:27,155: Epoch [10/13], Step [50/3970], Mel Loss: 1.88476, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:01:06,702: Epoch [10/13], Step [100/3970], Mel Loss: 1.90622, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:01:44,652: Epoch [10/13], Step [150/3970], Mel Loss: 1.89560, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:02:23,814: Epoch [10/13], Step [200/3970], Mel Loss: 1.90220, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:03:02,919: Epoch [10/13], Step [250/3970], Mel Loss: 1.89934, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:03:42,451: Epoch [10/13], Step [300/3970], Mel Loss: 1.90119, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:04:23,182: Epoch [10/13], Step [350/3970], Mel Loss: 1.92326, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:05:00,697: Epoch [10/13], Step [400/3970], Mel Loss: 1.92041, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:05:40,465: Epoch [10/13], Step [450/3970], Mel Loss: 1.90204, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:06:19,431: Epoch [10/13], Step [500/3970], Mel Loss: 1.88501, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:06:57,394: Epoch [10/13], Step [550/3970], Mel Loss: 1.89714, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:07:36,496: Epoch [10/13], Step [600/3970], Mel Loss: 1.88207, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:08:15,180: Epoch [10/13], Step [650/3970], Mel Loss: 1.90103, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:08:54,549: Epoch [10/13], Step [700/3970], Mel Loss: 1.87498, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:09:34,377: Epoch [10/13], Step [750/3970], Mel Loss: 1.89821, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:10:11,506: Epoch [10/13], Step [800/3970], Mel Loss: 1.87446, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:10:49,438: Epoch [10/13], Step [850/3970], Mel Loss: 1.89397, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:11:27,218: Epoch [10/13], Step [900/3970], Mel Loss: 1.88012, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:12:05,179: Epoch [10/13], Step [950/3970], Mel Loss: 1.91513, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:12:44,497: Epoch [10/13], Step [1000/3970], Mel Loss: 1.91660, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:13:23,999: Epoch [10/13], Step [1050/3970], Mel Loss: 1.89701, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:14:02,204: Epoch [10/13], Step [1100/3970], Mel Loss: 1.88604, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:14:41,314: Epoch [10/13], Step [1150/3970], Mel Loss: 1.88873, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:15:19,888: Epoch [10/13], Step [1200/3970], Mel Loss: 1.90624, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:15:59,099: Epoch [10/13], Step [1250/3970], Mel Loss: 1.89264, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:16:36,779: Epoch [10/13], Step [1300/3970], Mel Loss: 1.89478, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:17:16,030: Epoch [10/13], Step [1350/3970], Mel Loss: 1.89340, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:17:59,839: Epoch [10/13], Step [1400/3970], Mel Loss: 1.90329, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:18:39,473: Epoch [10/13], Step [1450/3970], Mel Loss: 1.88840, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:19:19,671: Epoch [10/13], Step [1500/3970], Mel Loss: 1.89581, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:19:59,135: Epoch [10/13], Step [1550/3970], Mel Loss: 1.88870, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:20:38,366: Epoch [10/13], Step [1600/3970], Mel Loss: 1.89154, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:21:17,276: Epoch [10/13], Step [1650/3970], Mel Loss: 1.88555, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:21:56,962: Epoch [10/13], Step [1700/3970], Mel Loss: 1.87369, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:22:35,139: Epoch [10/13], Step [1750/3970], Mel Loss: 1.90120, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:23:13,674: Epoch [10/13], Step [1800/3970], Mel Loss: 1.89881, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:23:53,165: Epoch [10/13], Step [1850/3970], Mel Loss: 1.90783, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:24:32,301: Epoch [10/13], Step [1900/3970], Mel Loss: 1.89959, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:25:10,557: Epoch [10/13], Step [1950/3970], Mel Loss: 1.87503, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:25:48,840: Epoch [10/13], Step [2000/3970], Mel Loss: 1.90296, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:26:27,733: Epoch [10/13], Step [2050/3970], Mel Loss: 1.89783, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:27:07,347: Epoch [10/13], Step [2100/3970], Mel Loss: 1.89363, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:27:46,746: Epoch [10/13], Step [2150/3970], Mel Loss: 1.90578, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:28:24,809: Epoch [10/13], Step [2200/3970], Mel Loss: 1.87597, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:29:04,758: Epoch [10/13], Step [2250/3970], Mel Loss: 1.89189, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:29:44,323: Epoch [10/13], Step [2300/3970], Mel Loss: 1.88809, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:30:22,776: Epoch [10/13], Step [2350/3970], Mel Loss: 1.90295, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:31:01,705: Epoch [10/13], Step [2400/3970], Mel Loss: 1.88242, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:31:40,507: Epoch [10/13], Step [2450/3970], Mel Loss: 1.89257, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:32:18,427: Epoch [10/13], Step [2500/3970], Mel Loss: 1.86816, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:32:58,207: Epoch [10/13], Step [2550/3970], Mel Loss: 1.90334, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:33:36,066: Epoch [10/13], Step [2600/3970], Mel Loss: 1.89832, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:34:15,184: Epoch [10/13], Step [2650/3970], Mel Loss: 1.90138, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:34:55,900: Epoch [10/13], Step [2700/3970], Mel Loss: 1.90211, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:35:37,140: Epoch [10/13], Step [2750/3970], Mel Loss: 1.87617, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:36:15,928: Epoch [10/13], Step [2800/3970], Mel Loss: 1.88046, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:36:53,622: Epoch [10/13], Step [2850/3970], Mel Loss: 1.89560, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:37:32,820: Epoch [10/13], Step [2900/3970], Mel Loss: 1.90877, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:38:12,344: Epoch [10/13], Step [2950/3970], Mel Loss: 1.88376, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:38:52,428: Epoch [10/13], Step [3000/3970], Mel Loss: 1.89573, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:39:31,353: Epoch [10/13], Step [3050/3970], Mel Loss: 1.90765, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:40:09,994: Epoch [10/13], Step [3100/3970], Mel Loss: 1.88310, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:40:50,044: Epoch [10/13], Step [3150/3970], Mel Loss: 1.88834, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:41:30,350: Epoch [10/13], Step [3200/3970], Mel Loss: 1.88357, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:42:09,722: Epoch [10/13], Step [3250/3970], Mel Loss: 1.89881, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:42:47,246: Epoch [10/13], Step [3300/3970], Mel Loss: 1.88919, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:43:26,272: Epoch [10/13], Step [3350/3970], Mel Loss: 1.90081, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:44:05,916: Epoch [10/13], Step [3400/3970], Mel Loss: 1.89494, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:44:46,010: Epoch [10/13], Step [3450/3970], Mel Loss: 1.87498, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:45:25,014: Epoch [10/13], Step [3500/3970], Mel Loss: 1.89413, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:46:04,953: Epoch [10/13], Step [3550/3970], Mel Loss: 1.90189, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:46:45,077: Epoch [10/13], Step [3600/3970], Mel Loss: 1.90884, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:47:23,554: Epoch [10/13], Step [3650/3970], Mel Loss: 1.89494, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:48:02,877: Epoch [10/13], Step [3700/3970], Mel Loss: 1.87241, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:48:41,668: Epoch [10/13], Step [3750/3970], Mel Loss: 1.91168, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:49:21,187: Epoch [10/13], Step [3800/3970], Mel Loss: 1.90768, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:50:00,236: Epoch [10/13], Step [3850/3970], Mel Loss: 1.91124, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:50:39,508: Epoch [10/13], Step [3900/3970], Mel Loss: 1.90319, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:51:15,346: Epoch [10/13], Step [3950/3970], Mel Loss: 1.90654, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:54:55,500: Validation loss: 1.867 - - - - -INFO:2025-06-08 13:55:35,931: Epoch [11/13], Step [50/3970], Mel Loss: 1.88614, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:56:16,801: Epoch [11/13], Step [100/3970], Mel Loss: 1.89060, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:56:55,945: Epoch [11/13], Step [150/3970], Mel Loss: 1.87535, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:57:35,613: Epoch [11/13], Step [200/3970], Mel Loss: 1.91457, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:58:15,706: Epoch [11/13], Step [250/3970], Mel Loss: 1.89765, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:58:54,106: Epoch [11/13], Step [300/3970], Mel Loss: 1.89825, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 13:59:33,039: Epoch [11/13], Step [350/3970], Mel Loss: 1.89147, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:00:12,410: Epoch [11/13], Step [400/3970], Mel Loss: 1.91134, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:00:52,417: Epoch [11/13], Step [450/3970], Mel Loss: 1.89009, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:01:30,616: Epoch [11/13], Step [500/3970], Mel Loss: 1.89281, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:02:08,935: Epoch [11/13], Step [550/3970], Mel Loss: 1.90347, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:02:47,982: Epoch [11/13], Step [600/3970], Mel Loss: 1.89988, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:03:28,450: Epoch [11/13], Step [650/3970], Mel Loss: 1.88639, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:04:09,200: Epoch [11/13], Step [700/3970], Mel Loss: 1.88351, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:04:48,545: Epoch [11/13], Step [750/3970], Mel Loss: 1.88620, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:05:27,459: Epoch [11/13], Step [800/3970], Mel Loss: 1.90364, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:06:08,106: Epoch [11/13], Step [850/3970], Mel Loss: 1.90375, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:06:48,145: Epoch [11/13], Step [900/3970], Mel Loss: 1.89880, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:07:27,458: Epoch [11/13], Step [950/3970], Mel Loss: 1.89912, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:08:04,845: Epoch [11/13], Step [1000/3970], Mel Loss: 1.91472, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:08:44,643: Epoch [11/13], Step [1050/3970], Mel Loss: 1.89789, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:09:22,305: Epoch [11/13], Step [1100/3970], Mel Loss: 1.88641, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:10:03,023: Epoch [11/13], Step [1150/3970], Mel Loss: 1.89581, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:10:40,177: Epoch [11/13], Step [1200/3970], Mel Loss: 1.88846, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:11:17,691: Epoch [11/13], Step [1250/3970], Mel Loss: 1.89660, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:11:56,744: Epoch [11/13], Step [1300/3970], Mel Loss: 1.88695, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:12:35,749: Epoch [11/13], Step [1350/3970], Mel Loss: 1.90729, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:13:15,773: Epoch [11/13], Step [1400/3970], Mel Loss: 1.89610, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:13:56,433: Epoch [11/13], Step [1450/3970], Mel Loss: 1.92151, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:14:37,364: Epoch [11/13], Step [1500/3970], Mel Loss: 1.90272, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:15:16,487: Epoch [11/13], Step [1550/3970], Mel Loss: 1.90562, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:15:57,302: Epoch [11/13], Step [1600/3970], Mel Loss: 1.89303, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:16:36,528: Epoch [11/13], Step [1650/3970], Mel Loss: 1.89941, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:17:14,546: Epoch [11/13], Step [1700/3970], Mel Loss: 1.89631, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:17:54,352: Epoch [11/13], Step [1750/3970], Mel Loss: 1.90318, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:18:33,371: Epoch [11/13], Step [1800/3970], Mel Loss: 1.89526, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:19:11,647: Epoch [11/13], Step [1850/3970], Mel Loss: 1.89427, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:19:50,381: Epoch [11/13], Step [1900/3970], Mel Loss: 1.88803, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:20:29,242: Epoch [11/13], Step [1950/3970], Mel Loss: 1.88866, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:21:07,194: Epoch [11/13], Step [2000/3970], Mel Loss: 1.88772, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:21:46,854: Epoch [11/13], Step [2050/3970], Mel Loss: 1.89555, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:22:27,090: Epoch [11/13], Step [2100/3970], Mel Loss: 1.89054, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:23:06,264: Epoch [11/13], Step [2150/3970], Mel Loss: 1.90023, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:23:44,843: Epoch [11/13], Step [2200/3970], Mel Loss: 1.88986, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:24:22,392: Epoch [11/13], Step [2250/3970], Mel Loss: 1.90120, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:25:02,621: Epoch [11/13], Step [2300/3970], Mel Loss: 1.87947, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:25:40,686: Epoch [11/13], Step [2350/3970], Mel Loss: 1.90603, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:26:19,201: Epoch [11/13], Step [2400/3970], Mel Loss: 1.88610, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:26:58,599: Epoch [11/13], Step [2450/3970], Mel Loss: 1.88386, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:27:38,565: Epoch [11/13], Step [2500/3970], Mel Loss: 1.88481, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:28:17,660: Epoch [11/13], Step [2550/3970], Mel Loss: 1.89451, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:28:57,389: Epoch [11/13], Step [2600/3970], Mel Loss: 1.89105, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:29:36,042: Epoch [11/13], Step [2650/3970], Mel Loss: 1.87842, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:30:15,936: Epoch [11/13], Step [2700/3970], Mel Loss: 1.91071, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:30:54,807: Epoch [11/13], Step [2750/3970], Mel Loss: 1.87782, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:31:33,626: Epoch [11/13], Step [2800/3970], Mel Loss: 1.88844, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:32:13,173: Epoch [11/13], Step [2850/3970], Mel Loss: 1.87999, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:32:52,024: Epoch [11/13], Step [2900/3970], Mel Loss: 1.88077, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:33:29,569: Epoch [11/13], Step [2950/3970], Mel Loss: 1.89407, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:34:07,989: Epoch [11/13], Step [3000/3970], Mel Loss: 1.90200, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:34:45,784: Epoch [11/13], Step [3050/3970], Mel Loss: 1.90149, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:35:24,743: Epoch [11/13], Step [3100/3970], Mel Loss: 1.90589, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:36:03,548: Epoch [11/13], Step [3150/3970], Mel Loss: 1.89616, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:36:42,826: Epoch [11/13], Step [3200/3970], Mel Loss: 1.90245, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:37:21,062: Epoch [11/13], Step [3250/3970], Mel Loss: 1.89492, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:38:01,204: Epoch [11/13], Step [3300/3970], Mel Loss: 1.92479, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:38:39,224: Epoch [11/13], Step [3350/3970], Mel Loss: 1.88391, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:39:17,786: Epoch [11/13], Step [3400/3970], Mel Loss: 1.89788, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:39:57,985: Epoch [11/13], Step [3450/3970], Mel Loss: 1.88450, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:40:39,266: Epoch [11/13], Step [3500/3970], Mel Loss: 1.92214, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:41:19,144: Epoch [11/13], Step [3550/3970], Mel Loss: 1.89972, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:41:57,407: Epoch [11/13], Step [3600/3970], Mel Loss: 1.89114, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:42:36,197: Epoch [11/13], Step [3650/3970], Mel Loss: 1.89519, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:43:15,656: Epoch [11/13], Step [3700/3970], Mel Loss: 1.89152, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:43:53,161: Epoch [11/13], Step [3750/3970], Mel Loss: 1.91306, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:44:31,044: Epoch [11/13], Step [3800/3970], Mel Loss: 1.87279, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:45:10,310: Epoch [11/13], Step [3850/3970], Mel Loss: 1.89228, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:45:48,548: Epoch [11/13], Step [3900/3970], Mel Loss: 1.88184, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:46:27,185: Epoch [11/13], Step [3950/3970], Mel Loss: 1.88796, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:50:01,516: Validation loss: 1.865 - - - - -INFO:2025-06-08 14:50:46,693: Epoch [12/13], Step [50/3970], Mel Loss: 1.91149, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:51:23,491: Epoch [12/13], Step [100/3970], Mel Loss: 1.91562, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:52:03,070: Epoch [12/13], Step [150/3970], Mel Loss: 1.88724, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:52:38,528: Epoch [12/13], Step [200/3970], Mel Loss: 1.88756, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:53:19,606: Epoch [12/13], Step [250/3970], Mel Loss: 1.87935, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:53:58,319: Epoch [12/13], Step [300/3970], Mel Loss: 1.90199, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:54:36,372: Epoch [12/13], Step [350/3970], Mel Loss: 1.88902, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:55:14,809: Epoch [12/13], Step [400/3970], Mel Loss: 1.90217, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:55:54,436: Epoch [12/13], Step [450/3970], Mel Loss: 1.89830, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:56:32,817: Epoch [12/13], Step [500/3970], Mel Loss: 1.87969, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:57:13,553: Epoch [12/13], Step [550/3970], Mel Loss: 1.90037, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:57:53,954: Epoch [12/13], Step [600/3970], Mel Loss: 1.88005, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:58:33,262: Epoch [12/13], Step [650/3970], Mel Loss: 1.90369, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:59:12,793: Epoch [12/13], Step [700/3970], Mel Loss: 1.90907, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 14:59:52,852: Epoch [12/13], Step [750/3970], Mel Loss: 1.89484, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:00:32,982: Epoch [12/13], Step [800/3970], Mel Loss: 1.89746, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:01:12,822: Epoch [12/13], Step [850/3970], Mel Loss: 1.88658, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:01:52,027: Epoch [12/13], Step [900/3970], Mel Loss: 1.88946, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:02:30,634: Epoch [12/13], Step [950/3970], Mel Loss: 1.89150, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:03:10,790: Epoch [12/13], Step [1000/3970], Mel Loss: 1.92164, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:03:50,681: Epoch [12/13], Step [1050/3970], Mel Loss: 1.89562, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:04:30,810: Epoch [12/13], Step [1100/3970], Mel Loss: 1.88594, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:05:10,685: Epoch [12/13], Step [1150/3970], Mel Loss: 1.90800, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:05:49,437: Epoch [12/13], Step [1200/3970], Mel Loss: 1.88337, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:06:27,610: Epoch [12/13], Step [1250/3970], Mel Loss: 1.87857, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:07:06,062: Epoch [12/13], Step [1300/3970], Mel Loss: 1.89169, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:07:46,072: Epoch [12/13], Step [1350/3970], Mel Loss: 1.88556, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:08:24,624: Epoch [12/13], Step [1400/3970], Mel Loss: 1.89792, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:09:06,064: Epoch [12/13], Step [1450/3970], Mel Loss: 1.89307, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:09:46,669: Epoch [12/13], Step [1500/3970], Mel Loss: 1.90943, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:10:26,120: Epoch [12/13], Step [1550/3970], Mel Loss: 1.88162, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:11:03,862: Epoch [12/13], Step [1600/3970], Mel Loss: 1.90507, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:11:42,845: Epoch [12/13], Step [1650/3970], Mel Loss: 1.90300, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:12:23,203: Epoch [12/13], Step [1700/3970], Mel Loss: 1.89901, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:13:01,769: Epoch [12/13], Step [1750/3970], Mel Loss: 1.91333, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:13:41,215: Epoch [12/13], Step [1800/3970], Mel Loss: 1.88622, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:14:21,073: Epoch [12/13], Step [1850/3970], Mel Loss: 1.89569, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:14:57,599: Epoch [12/13], Step [1900/3970], Mel Loss: 1.88930, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:15:37,399: Epoch [12/13], Step [1950/3970], Mel Loss: 1.90670, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:16:15,805: Epoch [12/13], Step [2000/3970], Mel Loss: 1.89029, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:16:54,790: Epoch [12/13], Step [2050/3970], Mel Loss: 1.89086, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:17:33,699: Epoch [12/13], Step [2100/3970], Mel Loss: 1.87664, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:18:13,791: Epoch [12/13], Step [2150/3970], Mel Loss: 1.90563, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:18:54,134: Epoch [12/13], Step [2200/3970], Mel Loss: 1.90933, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:19:33,476: Epoch [12/13], Step [2250/3970], Mel Loss: 1.87989, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:20:12,532: Epoch [12/13], Step [2300/3970], Mel Loss: 1.87841, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:20:52,651: Epoch [12/13], Step [2350/3970], Mel Loss: 1.89170, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:21:30,552: Epoch [12/13], Step [2400/3970], Mel Loss: 1.91642, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:22:09,307: Epoch [12/13], Step [2450/3970], Mel Loss: 1.90577, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:22:46,763: Epoch [12/13], Step [2500/3970], Mel Loss: 1.88912, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:23:24,719: Epoch [12/13], Step [2550/3970], Mel Loss: 1.88442, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:24:03,889: Epoch [12/13], Step [2600/3970], Mel Loss: 1.90371, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:24:42,905: Epoch [12/13], Step [2650/3970], Mel Loss: 1.89132, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:25:20,726: Epoch [12/13], Step [2700/3970], Mel Loss: 1.90623, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:25:59,229: Epoch [12/13], Step [2750/3970], Mel Loss: 1.88586, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:26:38,813: Epoch [12/13], Step [2800/3970], Mel Loss: 1.90052, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:27:18,260: Epoch [12/13], Step [2850/3970], Mel Loss: 1.89744, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:27:57,593: Epoch [12/13], Step [2900/3970], Mel Loss: 1.90948, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:28:35,899: Epoch [12/13], Step [2950/3970], Mel Loss: 1.90283, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:29:16,299: Epoch [12/13], Step [3000/3970], Mel Loss: 1.91824, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:29:54,518: Epoch [12/13], Step [3050/3970], Mel Loss: 1.90386, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:30:34,938: Epoch [12/13], Step [3100/3970], Mel Loss: 1.89938, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:31:14,963: Epoch [12/13], Step [3150/3970], Mel Loss: 1.87374, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:31:54,464: Epoch [12/13], Step [3200/3970], Mel Loss: 1.89736, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:32:34,583: Epoch [12/13], Step [3250/3970], Mel Loss: 1.89873, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:33:12,793: Epoch [12/13], Step [3300/3970], Mel Loss: 1.89766, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:33:49,813: Epoch [12/13], Step [3350/3970], Mel Loss: 1.89149, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:34:27,481: Epoch [12/13], Step [3400/3970], Mel Loss: 1.88836, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:35:06,201: Epoch [12/13], Step [3450/3970], Mel Loss: 1.89166, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:35:44,741: Epoch [12/13], Step [3500/3970], Mel Loss: 1.88068, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:36:23,413: Epoch [12/13], Step [3550/3970], Mel Loss: 1.89146, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:37:02,508: Epoch [12/13], Step [3600/3970], Mel Loss: 1.88471, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:37:42,436: Epoch [12/13], Step [3650/3970], Mel Loss: 1.88714, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:38:23,040: Epoch [12/13], Step [3700/3970], Mel Loss: 1.89097, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:39:02,169: Epoch [12/13], Step [3750/3970], Mel Loss: 1.91014, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:39:40,130: Epoch [12/13], Step [3800/3970], Mel Loss: 1.90019, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:40:20,869: Epoch [12/13], Step [3850/3970], Mel Loss: 1.90687, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:41:00,075: Epoch [12/13], Step [3900/3970], Mel Loss: 1.88551, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:41:39,440: Epoch [12/13], Step [3950/3970], Mel Loss: 1.89240, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:45:12,840: Validation loss: 1.865 - - - - -INFO:2025-06-08 15:45:54,923: Epoch [13/13], Step [50/3970], Mel Loss: 1.88450, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:46:34,418: Epoch [13/13], Step [100/3970], Mel Loss: 1.88540, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:47:14,058: Epoch [13/13], Step [150/3970], Mel Loss: 1.89585, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:47:52,759: Epoch [13/13], Step [200/3970], Mel Loss: 1.88536, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:48:33,220: Epoch [13/13], Step [250/3970], Mel Loss: 1.88381, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:49:11,836: Epoch [13/13], Step [300/3970], Mel Loss: 1.90165, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:49:51,629: Epoch [13/13], Step [350/3970], Mel Loss: 1.91351, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:50:31,028: Epoch [13/13], Step [400/3970], Mel Loss: 1.91672, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:51:10,246: Epoch [13/13], Step [450/3970], Mel Loss: 1.90166, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:51:50,143: Epoch [13/13], Step [500/3970], Mel Loss: 1.91098, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:52:31,523: Epoch [13/13], Step [550/3970], Mel Loss: 1.89396, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:53:10,804: Epoch [13/13], Step [600/3970], Mel Loss: 1.91016, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:53:49,858: Epoch [13/13], Step [650/3970], Mel Loss: 1.89070, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:54:30,203: Epoch [13/13], Step [700/3970], Mel Loss: 1.89536, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:55:09,191: Epoch [13/13], Step [750/3970], Mel Loss: 1.90730, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:55:48,804: Epoch [13/13], Step [800/3970], Mel Loss: 1.89983, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:56:29,001: Epoch [13/13], Step [850/3970], Mel Loss: 1.89194, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:57:08,172: Epoch [13/13], Step [900/3970], Mel Loss: 1.91948, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:57:48,206: Epoch [13/13], Step [950/3970], Mel Loss: 1.87996, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:58:29,168: Epoch [13/13], Step [1000/3970], Mel Loss: 1.89659, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:59:08,198: Epoch [13/13], Step [1050/3970], Mel Loss: 1.90218, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 15:59:46,941: Epoch [13/13], Step [1100/3970], Mel Loss: 1.88519, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:00:25,840: Epoch [13/13], Step [1150/3970], Mel Loss: 1.90971, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:01:04,273: Epoch [13/13], Step [1200/3970], Mel Loss: 1.89212, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:01:42,611: Epoch [13/13], Step [1250/3970], Mel Loss: 1.88373, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:02:20,504: Epoch [13/13], Step [1300/3970], Mel Loss: 1.89279, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:02:59,960: Epoch [13/13], Step [1350/3970], Mel Loss: 1.90156, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:03:40,018: Epoch [13/13], Step [1400/3970], Mel Loss: 1.90169, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:04:17,520: Epoch [13/13], Step [1450/3970], Mel Loss: 1.90079, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:04:57,347: Epoch [13/13], Step [1500/3970], Mel Loss: 1.88962, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:05:36,896: Epoch [13/13], Step [1550/3970], Mel Loss: 1.87890, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:06:16,672: Epoch [13/13], Step [1600/3970], Mel Loss: 1.90224, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:06:55,851: Epoch [13/13], Step [1650/3970], Mel Loss: 1.88437, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:07:33,452: Epoch [13/13], Step [1700/3970], Mel Loss: 1.87417, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:08:12,860: Epoch [13/13], Step [1750/3970], Mel Loss: 1.90218, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:08:51,049: Epoch [13/13], Step [1800/3970], Mel Loss: 1.90577, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:09:30,601: Epoch [13/13], Step [1850/3970], Mel Loss: 1.90908, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:10:10,303: Epoch [13/13], Step [1900/3970], Mel Loss: 1.88736, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:10:48,932: Epoch [13/13], Step [1950/3970], Mel Loss: 1.89833, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:11:28,297: Epoch [13/13], Step [2000/3970], Mel Loss: 1.89343, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:12:08,320: Epoch [13/13], Step [2050/3970], Mel Loss: 1.89891, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:12:47,224: Epoch [13/13], Step [2100/3970], Mel Loss: 1.89381, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:13:28,077: Epoch [13/13], Step [2150/3970], Mel Loss: 1.89292, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:14:07,032: Epoch [13/13], Step [2200/3970], Mel Loss: 1.88428, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:14:46,111: Epoch [13/13], Step [2250/3970], Mel Loss: 1.89498, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:15:23,962: Epoch [13/13], Step [2300/3970], Mel Loss: 1.89469, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:16:03,450: Epoch [13/13], Step [2350/3970], Mel Loss: 1.90460, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:16:43,011: Epoch [13/13], Step [2400/3970], Mel Loss: 1.89242, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:17:21,158: Epoch [13/13], Step [2450/3970], Mel Loss: 1.88209, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:18:00,693: Epoch [13/13], Step [2500/3970], Mel Loss: 1.89500, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:18:38,609: Epoch [13/13], Step [2550/3970], Mel Loss: 1.89854, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:19:17,156: Epoch [13/13], Step [2600/3970], Mel Loss: 1.88907, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:19:55,756: Epoch [13/13], Step [2650/3970], Mel Loss: 1.87468, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:20:35,991: Epoch [13/13], Step [2700/3970], Mel Loss: 1.91162, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:21:13,756: Epoch [13/13], Step [2750/3970], Mel Loss: 1.88538, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:21:53,041: Epoch [13/13], Step [2800/3970], Mel Loss: 1.91530, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:22:31,111: Epoch [13/13], Step [2850/3970], Mel Loss: 1.88701, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:23:09,879: Epoch [13/13], Step [2900/3970], Mel Loss: 1.88455, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:23:49,733: Epoch [13/13], Step [2950/3970], Mel Loss: 1.88784, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:24:28,969: Epoch [13/13], Step [3000/3970], Mel Loss: 1.89757, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:25:07,666: Epoch [13/13], Step [3050/3970], Mel Loss: 1.90498, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:25:45,672: Epoch [13/13], Step [3100/3970], Mel Loss: 1.89244, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:26:25,446: Epoch [13/13], Step [3150/3970], Mel Loss: 1.89578, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:27:05,055: Epoch [13/13], Step [3200/3970], Mel Loss: 1.88377, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:27:45,620: Epoch [13/13], Step [3250/3970], Mel Loss: 1.90372, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:28:24,244: Epoch [13/13], Step [3300/3970], Mel Loss: 1.90474, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:29:03,886: Epoch [13/13], Step [3350/3970], Mel Loss: 1.89123, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:29:43,357: Epoch [13/13], Step [3400/3970], Mel Loss: 1.88454, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:30:23,758: Epoch [13/13], Step [3450/3970], Mel Loss: 1.87254, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:31:01,627: Epoch [13/13], Step [3500/3970], Mel Loss: 1.88949, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:31:39,111: Epoch [13/13], Step [3550/3970], Mel Loss: 1.88866, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:32:17,451: Epoch [13/13], Step [3600/3970], Mel Loss: 1.89264, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:32:56,471: Epoch [13/13], Step [3650/3970], Mel Loss: 1.86667, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:33:35,427: Epoch [13/13], Step [3700/3970], Mel Loss: 1.89066, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:34:14,406: Epoch [13/13], Step [3750/3970], Mel Loss: 1.89355, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:34:53,762: Epoch [13/13], Step [3800/3970], Mel Loss: 1.90973, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:35:34,271: Epoch [13/13], Step [3850/3970], Mel Loss: 1.89201, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:36:11,933: Epoch [13/13], Step [3900/3970], Mel Loss: 1.90671, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:36:51,312: Epoch [13/13], Step [3950/3970], Mel Loss: 1.90681, Gen Loss: 0.00000, Disc Loss: 0.00000, Mono Loss: 0.00000, S2S Loss: 0.00000, SLM Loss: 0.00000 -INFO:2025-06-08 16:40:23,787: Validation loss: 1.864 - - - - diff --git a/logs/pod_90h_30k_second_lr1/.ipynb_checkpoints/train-checkpoint.log b/logs/pod_90h_30k_second_lr1/.ipynb_checkpoints/train-checkpoint.log new file mode 100644 index 0000000000000000000000000000000000000000..731016b267302567118d000d6e41a9d17bc66e17 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/.ipynb_checkpoints/train-checkpoint.log @@ -0,0 +1,2 @@ +INFO:2025-06-10 18:43:18,783: Epoch [1/20], Step [50/1985], Loss: 0.62892, Disc Loss: 0.00000, Dur Loss: 1.49694, CE Loss: 0.11438, Norm Loss: 3.04314, F0 Loss: 6.81772, LM Loss: 2.20677, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 18:44:12,918: Epoch [1/20], Step [100/1985], Loss: 0.59547, Disc Loss: 0.00000, Dur Loss: 1.26027, CE Loss: 0.09677, Norm Loss: 3.12705, F0 Loss: 5.71449, LM Loss: 2.14675, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 diff --git a/logs/pod_90h_30k_second_lr1/config_ft_single.yml b/logs/pod_90h_30k_second_lr1/config_ft_single.yml new file mode 100644 index 0000000000000000000000000000000000000000..f7b5fe79dd0e4352d125d2bc6ea170138c2f7ec7 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/config_ft_single.yml @@ -0,0 +1,22 @@ +{ASR_config: Utils/ASR/config.yml, ASR_path: Utils/ASR/epoch_00080.pth, F0_path: Utils/JDC/bst.t7, + PLBERT_dir: Utils/PLBERT/, batch_size: 12, data_params: {OOD_data: /workspace/styletts2/data/OOD_texts.txt, + min_length: 50, root_path: /workspace, train_data: /workspace/styletts2/data/train_list.txt, + val_data: /workspace/styletts2/data/val_list.txt}, device: cuda, epochs_1st: 25, + epochs_2nd: 20, first_stage_path: /workspace/styletts2/stage1_final.pth, load_only_params: false, + log_dir: logs/pod_90h_30k_second_lr1, log_interval: 50, loss_params: {TMA_epoch: 14, + diff_epoch: 0, joint_epoch: 0, lambda_F0: 1.0, lambda_ce: 20.0, lambda_diff: 1.0, + lambda_dur: 1.0, lambda_gen: 1.0, lambda_mel: 5.0, lambda_mono: 1.0, lambda_norm: 1.0, + lambda_s2s: 1.0, lambda_slm: 1.0, lambda_sty: 1.0}, max_len: 300, model_params: { + decoder: {resblock_dilation_sizes: [[1, 3, 5], [1, 3, 5], [1, 3, 5]], resblock_kernel_sizes: [ + 3, 7, 11], type: hifigan, upsample_initial_channel: 512, upsample_kernel_sizes: [ + 20, 10, 6, 4], upsample_rates: [10, 5, 3, 2]}, diffusion: {dist: {estimate_sigma_data: true, + mean: -3.0, sigma_data: 0.4296724944663835, std: 1.0}, embedding_mask_proba: 0.1, + transformer: {head_features: 64, multiplier: 2, num_heads: 8, num_layers: 3}}, + dim_in: 64, dropout: 0.2, hidden_dim: 512, max_conv_dim: 512, max_dur: 50, multispeaker: true, + n_layer: 3, n_mels: 80, n_token: 178, slm: {hidden: 768, initial_channel: 64, + model: microsoft/wavlm-base-plus, nlayers: 13, sr: 16000}, style_dim: 128}, + optimizer_params: {bert_lr: 1.0e-05, ft_lr: 0.0001, grad_accum_steps: 2, lr: 0.0001}, + preprocess_params: {spect_params: {hop_length: 300, n_fft: 2048, win_length: 1200}, + sr: 24000}, pretrained_model: '', save_freq: 1, second_stage_load_pretrained: true, + slmadv_params: {batch_percentage: 0.5, iter: 20, max_len: 500, min_len: 400, scale: 0.01, + sig: 1.5, thresh: 5}} diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00000.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00000.pth new file mode 100644 index 0000000000000000000000000000000000000000..8013fb973ff74ff3929fdb3358de9e15a13c3971 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00000.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:efa1068bbc3a1008faccb16873820844f94d7f14243f4d183de6f9198acc728a +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00001.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00001.pth new file mode 100644 index 0000000000000000000000000000000000000000..4de4868abe1e98dfd8f55afaee1d14f98d274ff6 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00001.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3ce74574465879689a506263fe9e2ea3e61c3726b5b2ad2da4e8abe0592174b5 +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00002.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00002.pth new file mode 100644 index 0000000000000000000000000000000000000000..a6f3a4a9bcc532155fd88fdd4b9edcbc04581b28 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00002.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:317dfe23c36fda281f8475804768d730052cde73b8fc6e8edf2f74aa92ecd3cc +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00003.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00003.pth new file mode 100644 index 0000000000000000000000000000000000000000..dc8042f85d586b2c6d79fe0f6710e9258e74ec24 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00003.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8d4c3e90865539448a69c87395f6dd86eed2c83bf87f3c95508a057a88ec1bd2 +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00004.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00004.pth new file mode 100644 index 0000000000000000000000000000000000000000..465039289c4fe3a14140d637a6b4e364db4b4781 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00004.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1a4990bbdd4338025585f8ff0a79aa5ba741c1647ec8e0f49230b43aca7bad87 +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00005.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00005.pth new file mode 100644 index 0000000000000000000000000000000000000000..ebf1760358a8da490791cd4e258b2354cca6aefa --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00005.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2d37ebae1d1484c91afcc055528893e576603c2a2a276f9fa14f3a59c264ec4d +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00006.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00006.pth new file mode 100644 index 0000000000000000000000000000000000000000..74a70b7cf9a4b1a19fbd783db1230c655f5aa531 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00006.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:69e6bbb0e276fb50b9f9b3e55be7672b3b328b8cb27b050c87a6114ab952e270 +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00007.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00007.pth new file mode 100644 index 0000000000000000000000000000000000000000..a65d86d385eb30cc434bd15c18034daf923eaf6d --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00007.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a3f0cd383880a26a97c69f1ec28d83b1d5bea9e06b09ae7f70953cfaac807afd +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00008.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00008.pth new file mode 100644 index 0000000000000000000000000000000000000000..92b15d3a52a4d0f9e7865bf4fbc7fad985651541 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00008.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:886afb4c287662b46089545efe1be2aab36e65383e4a6a666eabc296a01ea590 +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00009.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00009.pth new file mode 100644 index 0000000000000000000000000000000000000000..63b1dcc7a5c1ddefb864dbc42dbea3a5b675c8a0 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00009.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:109c82d3fdd02e613ef5bec558aa7439568b166b1b5abffb1241b77e0384f6d0 +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00010.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00010.pth new file mode 100644 index 0000000000000000000000000000000000000000..e0624dca1e3f975b04157e823045281e5441ea3e --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00010.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dae323f68e932c838234c234c70ac5c9cc7c070b86b3b205a17d619dc90561fc +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00011.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00011.pth new file mode 100644 index 0000000000000000000000000000000000000000..1541b4432a08bb520d62a419815f3975ec70b07f --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00011.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2befa68fd5e745f57f14689452856e5c1d1389e468f1b34982d2dab6611d6d66 +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00012.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00012.pth new file mode 100644 index 0000000000000000000000000000000000000000..f279c325c95173646d9bb533e243e7b4eaeb9d46 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00012.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:46cd75a179c3a051bbf17e9ddc2c2272a0615a5deb19728241a4691b66b126bb +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00013.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00013.pth new file mode 100644 index 0000000000000000000000000000000000000000..28ed05ecaf081850121786acaadc8f7c75eaf102 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00013.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d20de6f832ab85705b8fc77e1efb45c155997f9c12cb19593050ad7ba1e0f3e8 +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00014.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00014.pth new file mode 100644 index 0000000000000000000000000000000000000000..605793338e724479de79f1ee5581fa1eae1136c5 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00014.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:004bf4ff1c06c8c9266297d1d6826c864b5a1e602591173b57d2792b3d4657ea +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00015.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00015.pth new file mode 100644 index 0000000000000000000000000000000000000000..4b9c44687e40dd62e927ec80f3daf8cbe5b1a1f8 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00015.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9a2ec092601e48e6970043cee561b6ff70722bacd0bce6e9db2e4b4eab61130f +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00016.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00016.pth new file mode 100644 index 0000000000000000000000000000000000000000..977901b0f9de570eaa8bc9435125e2b59436d080 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00016.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2a3d8f488f55ff565060bf264a45da6cb9cf24f5dcc268a8c16a1d7f17f75898 +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00017.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00017.pth new file mode 100644 index 0000000000000000000000000000000000000000..e498dcaaa2964392799c83add80d892f59c6fe57 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00017.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:206949b54256ce828761b2b25ed7f3d55ce6ff145b07ec738a4c629dbe15b860 +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/epoch_2nd_00018.pth b/logs/pod_90h_30k_second_lr1/epoch_2nd_00018.pth new file mode 100644 index 0000000000000000000000000000000000000000..c066d6b92320971e9fa32ccd34c7062b5ad2c358 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/epoch_2nd_00018.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:af00f31df6de0b578b6187ae49c375468d0d35eecca661548373c44f43b000da +size 2144951220 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749580931.7f09b0e2c0b0.7608.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749580931.7f09b0e2c0b0.7608.0 new file mode 100644 index 0000000000000000000000000000000000000000..144070053225a103c175f5f6a193b01888c20f64 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749580931.7f09b0e2c0b0.7608.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:12c39645158efaaf898e75d9493cbad314dbe5d3470ea15d42ccd61756f6a255 +size 1344 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749581195.7f09b0e2c0b0.9068.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749581195.7f09b0e2c0b0.9068.0 new file mode 100644 index 0000000000000000000000000000000000000000..23446b08d466bd42e5f589d28d7d5a3c8016ee35 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749581195.7f09b0e2c0b0.9068.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:13537d6f0863312fb7837f09805d504fd26f2230aee82ccdc8d43ffafd0094cd +size 88 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749581731.7f09b0e2c0b0.11730.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749581731.7f09b0e2c0b0.11730.0 new file mode 100644 index 0000000000000000000000000000000000000000..d00909dc01fcaef381ccac735ca0863aea21a089 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749581731.7f09b0e2c0b0.11730.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dd0260054755f987d300f840a775a2ef4c58391c67100a9882105dc6ad0512f2 +size 716 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749582627.7f09b0e2c0b0.16062.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749582627.7f09b0e2c0b0.16062.0 new file mode 100644 index 0000000000000000000000000000000000000000..8b18bcce20dd9fc36c6f169686da333c4c2f276a --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749582627.7f09b0e2c0b0.16062.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:56f245def2f3e28bee73bc43ee5ef2d2bd0c68a3802c0c55eaba2643b9d40993 +size 28864 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749585252.7f09b0e2c0b0.28252.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749585252.7f09b0e2c0b0.28252.0 new file mode 100644 index 0000000000000000000000000000000000000000..48db1ce08ae8e43e8e526812c81a390700281058 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749585252.7f09b0e2c0b0.28252.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4089eb9ca25900b56b9c51bbc5e3c0010dd8de8bd4324b7156c1588ad93d4f41 +size 88 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749585753.7f09b0e2c0b0.30770.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749585753.7f09b0e2c0b0.30770.0 new file mode 100644 index 0000000000000000000000000000000000000000..844cbf97bbe49f391f7847b0c923d4da420b64f1 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749585753.7f09b0e2c0b0.30770.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:acab989d6b06a88795d08783b1e2c64fbfee4c6f6a1578e557c154948bcaa07c +size 88 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749586138.7f09b0e2c0b0.32741.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749586138.7f09b0e2c0b0.32741.0 new file mode 100644 index 0000000000000000000000000000000000000000..47de4f97ee4e113ba6e8c302a433c13c7c314b04 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749586138.7f09b0e2c0b0.32741.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:296836236f4738d5a3415a4ba087a45f803fb24536dec1ec54de13604d2d1c3f +size 88 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749586253.7f09b0e2c0b0.33535.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749586253.7f09b0e2c0b0.33535.0 new file mode 100644 index 0000000000000000000000000000000000000000..5cadcdf5d52610ca18dfc9968b73294d9c3c2824 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749586253.7f09b0e2c0b0.33535.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8cd4171365035567c85f571cef03e5bcbaddb52af362fc9005d92634e8d81ded +size 88 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749586317.7f09b0e2c0b0.34090.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749586317.7f09b0e2c0b0.34090.0 new file mode 100644 index 0000000000000000000000000000000000000000..d2437fd32cd7c0f6b59037cbaa89ae9828910eb4 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749586317.7f09b0e2c0b0.34090.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f1ad6d9ddb23b26f92b307f2a721dafa4f6a832f636d5f13dd68e37ec1f62761 +size 88 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749586368.7f09b0e2c0b0.34568.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749586368.7f09b0e2c0b0.34568.0 new file mode 100644 index 0000000000000000000000000000000000000000..dffab8a84a02959aea33a42fd7925ba4058d067f --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749586368.7f09b0e2c0b0.34568.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fbe61af0964432ed7c5ea668cdc0109dc05a37f785cdb7ad627b061248c2d76f +size 88 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749586637.7f09b0e2c0b0.36121.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749586637.7f09b0e2c0b0.36121.0 new file mode 100644 index 0000000000000000000000000000000000000000..2fc4ae70aae259d6dfc62bede626ddc657a4301d --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749586637.7f09b0e2c0b0.36121.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c405f07428f0747bee9b0e2306382905ef1562487d679df891c067597cbd82bd +size 88 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749588452.7f09b0e2c0b0.4095.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749588452.7f09b0e2c0b0.4095.0 new file mode 100644 index 0000000000000000000000000000000000000000..fc113a01393437fa3d750692d70a1d704ca81952 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749588452.7f09b0e2c0b0.4095.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bbbeda6a99b7f044f6231877725bbbe50c21d7aab23bc7313f6dce5be0c41cfa +size 88 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749589666.7f09b0e2c0b0.9766.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749589666.7f09b0e2c0b0.9766.0 new file mode 100644 index 0000000000000000000000000000000000000000..7990c796880c3634bd199b87b37e1fab8ce14034 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749589666.7f09b0e2c0b0.9766.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a05c7d55f42d09aa7f9c14764af1713ad8e2d47b242e4af69ed0fa99c350aa1a +size 88 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749589707.7f09b0e2c0b0.10237.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749589707.7f09b0e2c0b0.10237.0 new file mode 100644 index 0000000000000000000000000000000000000000..aede416dfdcabcb4ee5ed5508d1ae554aa060a32 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749589707.7f09b0e2c0b0.10237.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b355ce31a29b82f8793941fa05956a6e6d9d27ed5c89bd28b171f69547e4308e +size 1344 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749593767.7f09b0e2c0b0.2068.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749593767.7f09b0e2c0b0.2068.0 new file mode 100644 index 0000000000000000000000000000000000000000..2cf731f615f94367cc9567d5ad0aa1b3d87a748b --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749593767.7f09b0e2c0b0.2068.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2f0d00b0f8777c491ec04ea5b25e08369037a934fc47554ccedcc664303699b9 +size 88 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749593814.7f09b0e2c0b0.2792.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749593814.7f09b0e2c0b0.2792.0 new file mode 100644 index 0000000000000000000000000000000000000000..3dde029717dc2848efcbc99db0a9ad23e092cb64 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749593814.7f09b0e2c0b0.2792.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:daf31ea237f626aa22e45088dc2a737a488c46418a579db42d36d5d03bb53908 +size 88 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749593921.7f09b0e2c0b0.3723.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749593921.7f09b0e2c0b0.3723.0 new file mode 100644 index 0000000000000000000000000000000000000000..9aa6765d90eb87994fb6e32cd1ce73ccdd2211f7 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749593921.7f09b0e2c0b0.3723.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bf26e72f69e1e0a617bca52d2c72009b975d3e29f7d2680ec8ba7397c708c4a1 +size 2624 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749594162.7f09b0e2c0b0.5079.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749594162.7f09b0e2c0b0.5079.0 new file mode 100644 index 0000000000000000000000000000000000000000..80b8c9b8eff3afd93d6f4cc2785936ad1aac5817 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749594162.7f09b0e2c0b0.5079.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0ea6927fad97a25e8a523722e1c2756e436c2ddc36de37d2b921fd13fa3b6f6e +size 88 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749594217.7f09b0e2c0b0.5581.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749594217.7f09b0e2c0b0.5581.0 new file mode 100644 index 0000000000000000000000000000000000000000..c2f904287a21e6988d8179c645e67874ef189229 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749594217.7f09b0e2c0b0.5581.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c29935e56d7f7fc23af533e4213919900eb6fd067f6f7e0744b37511af27cb97 +size 88 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749594500.7f09b0e2c0b0.1849.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749594500.7f09b0e2c0b0.1849.0 new file mode 100644 index 0000000000000000000000000000000000000000..96c492df184c86357e853e17ce26795da2b30ae6 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749594500.7f09b0e2c0b0.1849.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8fbb9a330ff55bab9a5fe62762153c4738f9d913d8564ca3f4a3654c565ba751 +size 88 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749594636.7f09b0e2c0b0.2782.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749594636.7f09b0e2c0b0.2782.0 new file mode 100644 index 0000000000000000000000000000000000000000..a49bbaf0b414b385a768bacd5b3f6c756fd32f82 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749594636.7f09b0e2c0b0.2782.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d7f21d436739752f5f3f22c6bab745eebce147b0d347f188055b7a2b4763ab01 +size 9664 diff --git a/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749596222.7f09b0e2c0b0.10366.0 b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749596222.7f09b0e2c0b0.10366.0 new file mode 100644 index 0000000000000000000000000000000000000000..5a38d4baf75ec56f6ec4a50ece3efa5fcb6c3ca0 --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/tensorboard/events.out.tfevents.1749596222.7f09b0e2c0b0.10366.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b5e012069ea98c491d356b2944faa79b752e3d66923104e4268aba8c50347914 +size 5842948 diff --git a/logs/pod_90h_30k_second_lr1/train.log b/logs/pod_90h_30k_second_lr1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..b9374b0451f1e656fedad3eacf281dd368e7829c --- /dev/null +++ b/logs/pod_90h_30k_second_lr1/train.log @@ -0,0 +1,1141 @@ +INFO:2025-06-10 18:43:18,783: Epoch [1/20], Step [50/1985], Loss: 0.62892, Disc Loss: 0.00000, Dur Loss: 1.49694, CE Loss: 0.11438, Norm Loss: 3.04314, F0 Loss: 6.81772, LM Loss: 2.20677, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 18:44:12,918: Epoch [1/20], Step [100/1985], Loss: 0.59547, Disc Loss: 0.00000, Dur Loss: 1.26027, CE Loss: 0.09677, Norm Loss: 3.12705, F0 Loss: 5.71449, LM Loss: 2.14675, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 18:56:40,835: Epoch [1/20], Step [50/3970], Loss: 0.63963, Disc Loss: 4.02651, Dur Loss: 1.62412, CE Loss: 0.11296, Norm Loss: 3.39431, F0 Loss: 6.09691, LM Loss: 2.20530, Gen Loss: 3.52588, Sty Loss: 0.65161, Diff Loss: 1.03429, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:11:36,541: Epoch [1/20], Step [50/3970], Loss: 0.64998, Disc Loss: 4.05405, Dur Loss: 1.51091, CE Loss: 0.10766, Norm Loss: 3.04840, F0 Loss: 5.17974, LM Loss: 2.11230, Gen Loss: 3.86755, Sty Loss: 0.67959, Diff Loss: 1.06892, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:12:33,927: Epoch [1/20], Step [100/3970], Loss: 0.61562, Disc Loss: 4.02663, Dur Loss: 1.36480, CE Loss: 0.11650, Norm Loss: 4.07259, F0 Loss: 6.46366, LM Loss: 2.19317, Gen Loss: 3.77569, Sty Loss: 0.66799, Diff Loss: 1.01297, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:13:31,322: Epoch [1/20], Step [150/3970], Loss: 0.59529, Disc Loss: 4.01310, Dur Loss: 1.13463, CE Loss: 0.09169, Norm Loss: 2.66902, F0 Loss: 6.17030, LM Loss: 2.16861, Gen Loss: 3.51550, Sty Loss: 0.63575, Diff Loss: 1.00897, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:14:29,253: Epoch [1/20], Step [200/3970], Loss: 0.56603, Disc Loss: 3.98821, Dur Loss: 1.19692, CE Loss: 0.08294, Norm Loss: 1.96348, F0 Loss: 4.41087, LM Loss: 1.93907, Gen Loss: 3.51368, Sty Loss: 0.45066, Diff Loss: 0.99107, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:15:29,115: Epoch [1/20], Step [250/3970], Loss: 0.55656, Disc Loss: 3.98525, Dur Loss: 1.34390, CE Loss: 0.10599, Norm Loss: 1.89592, F0 Loss: 6.13021, LM Loss: 2.04689, Gen Loss: 3.88413, Sty Loss: 0.45853, Diff Loss: 0.98839, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:16:27,085: Epoch [1/20], Step [300/3970], Loss: 0.55997, Disc Loss: 3.97363, Dur Loss: 1.21027, CE Loss: 0.08877, Norm Loss: 2.08107, F0 Loss: 5.10081, LM Loss: 2.03673, Gen Loss: 3.46515, Sty Loss: 0.36638, Diff Loss: 0.99871, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:17:24,396: Epoch [1/20], Step [350/3970], Loss: 0.54107, Disc Loss: 3.96146, Dur Loss: 1.33028, CE Loss: 0.08496, Norm Loss: 2.13113, F0 Loss: 5.42313, LM Loss: 1.98610, Gen Loss: 3.69403, Sty Loss: 0.35157, Diff Loss: 0.93197, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:18:21,163: Epoch [1/20], Step [400/3970], Loss: 0.53447, Disc Loss: 3.93470, Dur Loss: 1.41631, CE Loss: 0.09236, Norm Loss: 1.97868, F0 Loss: 4.89980, LM Loss: 1.97002, Gen Loss: 3.85118, Sty Loss: 0.26924, Diff Loss: 0.87650, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:19:17,677: Epoch [1/20], Step [450/3970], Loss: 0.53484, Disc Loss: 3.94843, Dur Loss: 1.27066, CE Loss: 0.07282, Norm Loss: 2.59346, F0 Loss: 5.26020, LM Loss: 2.01398, Gen Loss: 4.11224, Sty Loss: 0.29286, Diff Loss: 0.85516, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:20:16,350: Epoch [1/20], Step [500/3970], Loss: 0.52047, Disc Loss: 3.83046, Dur Loss: 1.08408, CE Loss: 0.06640, Norm Loss: 2.07538, F0 Loss: 4.11796, LM Loss: 1.95035, Gen Loss: 3.97281, Sty Loss: 0.22163, Diff Loss: 0.78707, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:21:13,222: Epoch [1/20], Step [550/3970], Loss: 0.51932, Disc Loss: 4.00778, Dur Loss: 1.62378, CE Loss: 0.11239, Norm Loss: 1.99169, F0 Loss: 4.47174, LM Loss: 1.99640, Gen Loss: 4.02283, Sty Loss: 0.27342, Diff Loss: 0.92740, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:22:09,496: Epoch [1/20], Step [600/3970], Loss: 0.51748, Disc Loss: 3.92577, Dur Loss: 1.39971, CE Loss: 0.09339, Norm Loss: 1.99009, F0 Loss: 4.11688, LM Loss: 1.99856, Gen Loss: 4.50186, Sty Loss: 0.23685, Diff Loss: 0.79025, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:23:06,277: Epoch [1/20], Step [650/3970], Loss: 0.52486, Disc Loss: 3.91640, Dur Loss: 1.08502, CE Loss: 0.06107, Norm Loss: 1.83306, F0 Loss: 4.74670, LM Loss: 1.97242, Gen Loss: 4.21311, Sty Loss: 0.20897, Diff Loss: 0.79936, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:24:02,195: Epoch [1/20], Step [700/3970], Loss: 0.51219, Disc Loss: 3.68951, Dur Loss: 1.36272, CE Loss: 0.08991, Norm Loss: 1.89813, F0 Loss: 4.24734, LM Loss: 1.94970, Gen Loss: 5.51013, Sty Loss: 0.17596, Diff Loss: 0.67393, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:24:58,710: Epoch [1/20], Step [750/3970], Loss: 0.50008, Disc Loss: 3.93228, Dur Loss: 1.36094, CE Loss: 0.07939, Norm Loss: 2.28372, F0 Loss: 4.19424, LM Loss: 2.00618, Gen Loss: 4.28721, Sty Loss: 0.18270, Diff Loss: 0.66358, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:25:55,143: Epoch [1/20], Step [800/3970], Loss: 0.49976, Disc Loss: 3.89811, Dur Loss: 1.29906, CE Loss: 0.07065, Norm Loss: 1.75141, F0 Loss: 3.80949, LM Loss: 1.92769, Gen Loss: 4.84365, Sty Loss: 0.18492, Diff Loss: 0.71712, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:26:51,029: Epoch [1/20], Step [850/3970], Loss: 0.50730, Disc Loss: 3.75842, Dur Loss: 1.01579, CE Loss: 0.05694, Norm Loss: 1.99068, F0 Loss: 4.26938, LM Loss: 2.11732, Gen Loss: 5.01011, Sty Loss: 0.20584, Diff Loss: 0.66459, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:27:48,078: Epoch [1/20], Step [900/3970], Loss: 0.51057, Disc Loss: 3.86876, Dur Loss: 1.06412, CE Loss: 0.05662, Norm Loss: 1.47735, F0 Loss: 3.50341, LM Loss: 1.89203, Gen Loss: 4.78802, Sty Loss: 0.17495, Diff Loss: 0.64571, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:28:44,136: Epoch [1/20], Step [950/3970], Loss: 0.50204, Disc Loss: 3.91279, Dur Loss: 1.13988, CE Loss: 0.06612, Norm Loss: 1.63979, F0 Loss: 3.62670, LM Loss: 1.83374, Gen Loss: 4.08289, Sty Loss: 0.16216, Diff Loss: 0.70276, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:29:42,331: Epoch [1/20], Step [1000/3970], Loss: 0.49857, Disc Loss: 3.76260, Dur Loss: 1.18529, CE Loss: 0.06345, Norm Loss: 1.46970, F0 Loss: 4.00451, LM Loss: 1.90543, Gen Loss: 4.31558, Sty Loss: 0.15107, Diff Loss: 0.61285, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:30:41,264: Epoch [1/20], Step [1050/3970], Loss: 0.49288, Disc Loss: 3.85680, Dur Loss: 1.41509, CE Loss: 0.07992, Norm Loss: 1.75398, F0 Loss: 5.42137, LM Loss: 1.85508, Gen Loss: 4.56527, Sty Loss: 0.17338, Diff Loss: 0.59364, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:31:38,527: Epoch [1/20], Step [1100/3970], Loss: 0.48367, Disc Loss: 3.93643, Dur Loss: 1.66103, CE Loss: 0.10444, Norm Loss: 2.16221, F0 Loss: 3.85214, LM Loss: 1.92478, Gen Loss: 4.66435, Sty Loss: 0.16243, Diff Loss: 0.56711, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:32:36,785: Epoch [1/20], Step [1150/3970], Loss: 0.47410, Disc Loss: 3.98996, Dur Loss: 1.40747, CE Loss: 0.08612, Norm Loss: 1.85429, F0 Loss: 4.66532, LM Loss: 1.96277, Gen Loss: 4.58910, Sty Loss: 0.15009, Diff Loss: 0.60066, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:33:33,188: Epoch [1/20], Step [1200/3970], Loss: 0.48618, Disc Loss: 4.05583, Dur Loss: 1.34523, CE Loss: 0.08113, Norm Loss: 1.74998, F0 Loss: 4.49897, LM Loss: 1.76018, Gen Loss: 4.34532, Sty Loss: 0.18752, Diff Loss: 0.55571, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:34:31,505: Epoch [1/20], Step [1250/3970], Loss: 0.47988, Disc Loss: 3.90302, Dur Loss: 1.24098, CE Loss: 0.06994, Norm Loss: 1.62039, F0 Loss: 3.75780, LM Loss: 1.79406, Gen Loss: 4.16349, Sty Loss: 0.15166, Diff Loss: 0.35420, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:35:29,192: Epoch [1/20], Step [1300/3970], Loss: 0.47031, Disc Loss: 3.95793, Dur Loss: 1.03205, CE Loss: 0.05824, Norm Loss: 1.49206, F0 Loss: 3.49897, LM Loss: 1.72684, Gen Loss: 3.98410, Sty Loss: 0.16347, Diff Loss: 0.34756, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:36:26,738: Epoch [1/20], Step [1350/3970], Loss: 0.47702, Disc Loss: 3.98472, Dur Loss: 1.09532, CE Loss: 0.05950, Norm Loss: 1.82446, F0 Loss: 3.80948, LM Loss: 1.91488, Gen Loss: 4.35673, Sty Loss: 0.18350, Diff Loss: 0.42973, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:37:25,097: Epoch [1/20], Step [1400/3970], Loss: 0.47243, Disc Loss: 3.98231, Dur Loss: 1.26794, CE Loss: 0.07467, Norm Loss: 2.01591, F0 Loss: 6.06673, LM Loss: 1.86938, Gen Loss: 4.84629, Sty Loss: 0.17112, Diff Loss: 0.59131, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:38:23,119: Epoch [1/20], Step [1450/3970], Loss: 0.46148, Disc Loss: 3.91551, Dur Loss: 1.14140, CE Loss: 0.05819, Norm Loss: 1.78785, F0 Loss: 4.12295, LM Loss: 1.77045, Gen Loss: 4.49608, Sty Loss: 0.11922, Diff Loss: 0.68242, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:39:23,005: Epoch [1/20], Step [1500/3970], Loss: 0.46605, Disc Loss: 3.93164, Dur Loss: 1.25236, CE Loss: 0.06693, Norm Loss: 1.55835, F0 Loss: 3.82047, LM Loss: 1.66466, Gen Loss: 4.52721, Sty Loss: 0.09777, Diff Loss: 0.46561, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:40:18,814: Epoch [1/20], Step [1550/3970], Loss: 0.47380, Disc Loss: 3.90978, Dur Loss: 1.06693, CE Loss: 0.05722, Norm Loss: 1.58682, F0 Loss: 4.27114, LM Loss: 1.87452, Gen Loss: 3.63996, Sty Loss: 0.14593, Diff Loss: 0.53471, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:41:17,559: Epoch [1/20], Step [1600/3970], Loss: 0.46334, Disc Loss: 3.74228, Dur Loss: 1.04777, CE Loss: 0.05484, Norm Loss: 1.56704, F0 Loss: 4.46790, LM Loss: 1.83545, Gen Loss: 4.84506, Sty Loss: 0.13695, Diff Loss: 0.51344, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:42:17,353: Epoch [1/20], Step [1650/3970], Loss: 0.45598, Disc Loss: 3.87330, Dur Loss: 1.01297, CE Loss: 0.05384, Norm Loss: 1.22133, F0 Loss: 4.74947, LM Loss: 1.81776, Gen Loss: 4.61863, Sty Loss: 0.12068, Diff Loss: 0.48622, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:43:15,440: Epoch [1/20], Step [1700/3970], Loss: 0.46382, Disc Loss: 3.81540, Dur Loss: 1.17084, CE Loss: 0.06989, Norm Loss: 1.32601, F0 Loss: 4.63624, LM Loss: 1.84974, Gen Loss: 4.75014, Sty Loss: 0.11376, Diff Loss: 0.30602, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:44:13,251: Epoch [1/20], Step [1750/3970], Loss: 0.45489, Disc Loss: 3.81526, Dur Loss: 0.98724, CE Loss: 0.05584, Norm Loss: 1.17303, F0 Loss: 3.83023, LM Loss: 1.80512, Gen Loss: 5.09146, Sty Loss: 0.10141, Diff Loss: 0.48264, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:45:08,583: Epoch [1/20], Step [1800/3970], Loss: 0.45012, Disc Loss: 3.81299, Dur Loss: 1.04013, CE Loss: 0.05610, Norm Loss: 1.93821, F0 Loss: 3.53930, LM Loss: 1.75733, Gen Loss: 4.65977, Sty Loss: 0.10423, Diff Loss: 0.51090, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:46:05,417: Epoch [1/20], Step [1850/3970], Loss: 0.45098, Disc Loss: 3.86536, Dur Loss: 1.63563, CE Loss: 0.09989, Norm Loss: 1.86003, F0 Loss: 5.15329, LM Loss: 1.78537, Gen Loss: 3.94153, Sty Loss: 0.15700, Diff Loss: 0.49890, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:47:03,475: Epoch [1/20], Step [1900/3970], Loss: 0.45808, Disc Loss: 3.91990, Dur Loss: 1.22183, CE Loss: 0.06965, Norm Loss: 2.50300, F0 Loss: 4.60850, LM Loss: 2.04333, Gen Loss: 5.00250, Sty Loss: 0.14160, Diff Loss: 0.51222, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:48:00,880: Epoch [1/20], Step [1950/3970], Loss: 0.44669, Disc Loss: 3.85911, Dur Loss: 0.92537, CE Loss: 0.05494, Norm Loss: 1.60406, F0 Loss: 3.96016, LM Loss: 1.77767, Gen Loss: 4.61151, Sty Loss: 0.11126, Diff Loss: 0.36426, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:48:57,176: Epoch [1/20], Step [2000/3970], Loss: 0.44444, Disc Loss: 3.96494, Dur Loss: 1.41998, CE Loss: 0.07200, Norm Loss: 1.62035, F0 Loss: 3.76875, LM Loss: 1.79041, Gen Loss: 4.95324, Sty Loss: 0.14770, Diff Loss: 0.42236, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:49:54,834: Epoch [1/20], Step [2050/3970], Loss: 0.43702, Disc Loss: 3.98989, Dur Loss: 1.10346, CE Loss: 0.06113, Norm Loss: 1.13392, F0 Loss: 6.34397, LM Loss: 1.65671, Gen Loss: 4.27644, Sty Loss: 0.17621, Diff Loss: 0.41278, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:50:52,252: Epoch [1/20], Step [2100/3970], Loss: 0.44322, Disc Loss: 3.86544, Dur Loss: 1.74119, CE Loss: 0.09897, Norm Loss: 1.63533, F0 Loss: 3.76536, LM Loss: 1.68599, Gen Loss: 4.30241, Sty Loss: 0.12328, Diff Loss: 0.40213, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:51:49,342: Epoch [1/20], Step [2150/3970], Loss: 0.44743, Disc Loss: 4.03806, Dur Loss: 1.04294, CE Loss: 0.05176, Norm Loss: 1.15519, F0 Loss: 3.98928, LM Loss: 1.74441, Gen Loss: 5.22170, Sty Loss: 0.15373, Diff Loss: 0.45355, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:52:47,028: Epoch [1/20], Step [2200/3970], Loss: 0.43989, Disc Loss: 3.90921, Dur Loss: 0.99208, CE Loss: 0.04861, Norm Loss: 1.59125, F0 Loss: 4.25682, LM Loss: 1.86831, Gen Loss: 4.87091, Sty Loss: 0.15434, Diff Loss: 0.37106, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 19:53:47,634: Epoch [1/20], Step [2250/3970], Loss: 0.44721, Disc Loss: 3.90209, Dur Loss: 1.02604, CE Loss: 0.05693, Norm Loss: 1.00037, F0 Loss: 3.13139, LM Loss: 1.69111, Gen Loss: 4.58206, Sty Loss: 0.10937, Diff Loss: 0.36809, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 21:09:19,046: Epoch [1/20], Step [50/3970], Loss: 0.63661, Disc Loss: 0.00000, Dur Loss: 1.56116, CE Loss: 0.11892, Norm Loss: 2.32512, F0 Loss: 5.24943, LM Loss: 2.23784, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 21:10:00,349: Epoch [1/20], Step [100/3970], Loss: 0.62613, Disc Loss: 0.00000, Dur Loss: 1.26130, CE Loss: 0.09362, Norm Loss: 2.17488, F0 Loss: 3.68419, LM Loss: 2.06126, Gen Loss: 0.00000, Sty Loss: 0.00000, Diff Loss: 0.00000, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 22:19:52,145: Epoch [1/20], Step [50/3970], Loss: 0.64146, Disc Loss: 4.03037, Dur Loss: 1.35441, CE Loss: 0.09836, Norm Loss: 2.87490, F0 Loss: 7.44204, LM Loss: 2.20048, Gen Loss: 3.92366, Sty Loss: 0.74154, Diff Loss: 0.99704, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 22:20:51,616: Epoch [1/20], Step [100/3970], Loss: 0.61387, Disc Loss: 4.00488, Dur Loss: 1.32640, CE Loss: 0.10046, Norm Loss: 2.31159, F0 Loss: 7.65253, LM Loss: 2.10222, Gen Loss: 3.70114, Sty Loss: 0.73990, Diff Loss: 0.97696, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 22:21:47,972: Epoch [1/20], Step [150/3970], Loss: 0.58609, Disc Loss: 4.01101, Dur Loss: 1.18775, CE Loss: 0.07899, Norm Loss: 2.26675, F0 Loss: 4.91962, LM Loss: 2.01367, Gen Loss: 3.81227, Sty Loss: 0.56582, Diff Loss: 0.98095, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 22:22:47,731: Epoch [1/20], Step [200/3970], Loss: 0.56376, Disc Loss: 4.00948, Dur Loss: 1.61313, CE Loss: 0.11716, Norm Loss: 1.91894, F0 Loss: 4.91463, LM Loss: 1.93063, Gen Loss: 4.57931, Sty Loss: 0.53355, Diff Loss: 1.00780, DiscLM Loss: 0.00000, GenLM Loss: 0.00000 +INFO:2025-06-10 22:32:16,736: Epoch [1/20], Step [50/3970], Loss: 0.65960, Disc Loss: 3.91266, Dur Loss: 1.31799, CE Loss: 0.10695, Norm Loss: 3.15281, F0 Loss: 5.68359, LM Loss: 2.20241, Gen Loss: 4.55058, Sty Loss: 0.61547, Diff Loss: 1.03619, DiscLM Loss: 0.00000, GenLM Loss: 0.83004 +INFO:2025-06-10 22:33:43,859: Epoch [1/20], Step [100/3970], Loss: 0.64852, Disc Loss: 3.94614, Dur Loss: 1.17449, CE Loss: 0.08389, Norm Loss: 2.48182, F0 Loss: 5.18816, LM Loss: 2.52914, Gen Loss: 4.81063, Sty Loss: 0.53424, Diff Loss: 1.08513, DiscLM Loss: 0.74614, GenLM Loss: 0.69258 +INFO:2025-06-10 22:35:08,898: Epoch [1/20], Step [150/3970], Loss: 0.64419, Disc Loss: 3.96315, Dur Loss: 1.29489, CE Loss: 0.09216, Norm Loss: 1.78906, F0 Loss: 5.44122, LM Loss: 2.26565, Gen Loss: 4.17593, Sty Loss: 0.46625, Diff Loss: 0.98490, DiscLM Loss: 0.00000, GenLM Loss: 0.48955 +INFO:2025-06-10 22:36:36,135: Epoch [1/20], Step [200/3970], Loss: 0.62415, Disc Loss: 3.91818, Dur Loss: 1.42751, CE Loss: 0.09730, Norm Loss: 2.45306, F0 Loss: 4.50404, LM Loss: 2.29475, Gen Loss: 4.40923, Sty Loss: 0.30772, Diff Loss: 0.95113, DiscLM Loss: 0.32526, GenLM Loss: 0.38735 +INFO:2025-06-10 22:38:00,436: Epoch [1/20], Step [250/3970], Loss: 0.63085, Disc Loss: 3.90450, Dur Loss: 1.34812, CE Loss: 0.08536, Norm Loss: 1.97972, F0 Loss: 4.87818, LM Loss: 2.39780, Gen Loss: 3.92778, Sty Loss: 0.29208, Diff Loss: 0.96939, DiscLM Loss: 0.00000, GenLM Loss: 0.34464 +INFO:2025-06-10 22:39:27,314: Epoch [1/20], Step [300/3970], Loss: 0.62363, Disc Loss: 3.87172, Dur Loss: 1.24228, CE Loss: 0.07641, Norm Loss: 2.04828, F0 Loss: 5.65067, LM Loss: 2.32423, Gen Loss: 4.66215, Sty Loss: 0.24547, Diff Loss: 0.88361, DiscLM Loss: 0.21198, GenLM Loss: 0.37772 +INFO:2025-06-10 22:40:51,511: Epoch [1/20], Step [350/3970], Loss: 0.62608, Disc Loss: 3.91094, Dur Loss: 1.31073, CE Loss: 0.08699, Norm Loss: 1.81605, F0 Loss: 3.87958, LM Loss: 2.31905, Gen Loss: 4.99893, Sty Loss: 0.20236, Diff Loss: 0.93415, DiscLM Loss: 0.00000, GenLM Loss: 0.54223 +INFO:2025-06-10 22:42:18,909: Epoch [1/20], Step [400/3970], Loss: 0.62661, Disc Loss: 3.86100, Dur Loss: 1.17151, CE Loss: 0.09038, Norm Loss: 2.09602, F0 Loss: 4.79713, LM Loss: 2.31298, Gen Loss: 4.94136, Sty Loss: 0.20259, Diff Loss: 0.84737, DiscLM Loss: 0.12584, GenLM Loss: 0.53977 +INFO:2025-06-10 22:43:45,017: Epoch [1/20], Step [450/3970], Loss: 0.62387, Disc Loss: 3.89230, Dur Loss: 1.12123, CE Loss: 0.06443, Norm Loss: 1.96832, F0 Loss: 4.43496, LM Loss: 2.32850, Gen Loss: 3.99158, Sty Loss: 0.18667, Diff Loss: 0.89072, DiscLM Loss: 0.00000, GenLM Loss: 0.86124 +INFO:2025-06-10 22:45:09,261: Epoch [1/20], Step [500/3970], Loss: 0.62755, Disc Loss: 3.86074, Dur Loss: 1.16746, CE Loss: 0.06907, Norm Loss: 1.74537, F0 Loss: 6.30978, LM Loss: 2.42709, Gen Loss: 5.24256, Sty Loss: 0.21846, Diff Loss: 0.73786, DiscLM Loss: 0.21393, GenLM Loss: 0.89140 +INFO:2025-06-10 22:46:37,033: Epoch [1/20], Step [550/3970], Loss: 0.63895, Disc Loss: 3.81710, Dur Loss: 1.19815, CE Loss: 0.06867, Norm Loss: 2.14595, F0 Loss: 3.61301, LM Loss: 2.34675, Gen Loss: 5.84089, Sty Loss: 0.21555, Diff Loss: 0.73217, DiscLM Loss: 0.00000, GenLM Loss: 0.84883 +INFO:2025-06-10 22:49:28,352: Epoch [1/20], Step [650/3970], Loss: 1.24633, Disc Loss: 3.88302, Dur Loss: 1.69258, CE Loss: 0.10484, Norm Loss: 2.57088, F0 Loss: 6.53094, LM Loss: 2.45969, Gen Loss: 6.12331, Sty Loss: 0.18344, Diff Loss: 0.69724, DiscLM Loss: 0.00000, GenLM Loss: 0.81490 +INFO:2025-06-10 22:52:20,846: Epoch [1/20], Step [750/3970], Loss: 1.24963, Disc Loss: 3.82713, Dur Loss: 1.18809, CE Loss: 0.06549, Norm Loss: 1.70763, F0 Loss: 6.15836, LM Loss: 2.34375, Gen Loss: 5.85760, Sty Loss: 0.16504, Diff Loss: 0.67700, DiscLM Loss: 0.00000, GenLM Loss: 0.93519 +INFO:2025-06-10 22:53:46,631: Epoch [1/20], Step [800/3970], Loss: 0.63207, Disc Loss: 3.98168, Dur Loss: 1.19867, CE Loss: 0.06333, Norm Loss: 1.40222, F0 Loss: 4.85075, LM Loss: 2.37279, Gen Loss: 5.06851, Sty Loss: 0.13838, Diff Loss: 0.68856, DiscLM Loss: 0.06052, GenLM Loss: 0.91701 +INFO:2025-06-10 22:55:13,237: Epoch [1/20], Step [850/3970], Loss: 0.61765, Disc Loss: 3.80917, Dur Loss: 1.58018, CE Loss: 0.09122, Norm Loss: 1.87131, F0 Loss: 4.09471, LM Loss: 2.36387, Gen Loss: 5.70479, Sty Loss: 0.13325, Diff Loss: 0.53769, DiscLM Loss: 0.00000, GenLM Loss: 0.84837 +INFO:2025-06-10 22:59:06,374: Epoch [1/20], Step [50/2647], Loss: 0.64755, Disc Loss: 4.00124, Dur Loss: 1.48040, CE Loss: 0.10335, Norm Loss: 2.48924, F0 Loss: 6.52370, LM Loss: 2.34841, Gen Loss: 4.26980, Sty Loss: 0.65658, Diff Loss: 0.99650, DiscLM Loss: 0.00000, GenLM Loss: 0.92184 +INFO:2025-06-10 23:00:52,966: Epoch [1/20], Step [100/2647], Loss: 0.64385, Disc Loss: 3.91946, Dur Loss: 1.25313, CE Loss: 0.08590, Norm Loss: 2.38393, F0 Loss: 4.86803, LM Loss: 2.36719, Gen Loss: 4.67762, Sty Loss: 0.51777, Diff Loss: 1.01099, DiscLM Loss: 0.83973, GenLM Loss: 0.74562 +INFO:2025-06-10 23:02:38,021: Epoch [1/20], Step [150/2647], Loss: 0.65162, Disc Loss: 3.89482, Dur Loss: 1.21203, CE Loss: 0.08110, Norm Loss: 1.94326, F0 Loss: 5.92914, LM Loss: 2.29995, Gen Loss: 4.32709, Sty Loss: 0.46089, Diff Loss: 0.99174, DiscLM Loss: 0.00000, GenLM Loss: 0.57314 +INFO:2025-06-10 23:04:21,248: Epoch [1/20], Step [200/2647], Loss: 0.63232, Disc Loss: 3.83077, Dur Loss: 1.34033, CE Loss: 0.10509, Norm Loss: 2.36985, F0 Loss: 5.64242, LM Loss: 2.24405, Gen Loss: 4.46949, Sty Loss: 0.38852, Diff Loss: 0.94763, DiscLM Loss: 0.40090, GenLM Loss: 0.31386 +INFO:2025-06-10 23:06:10,376: Epoch [1/20], Step [250/2647], Loss: 0.63473, Disc Loss: 3.97608, Dur Loss: 1.29090, CE Loss: 0.07934, Norm Loss: 2.41112, F0 Loss: 5.70133, LM Loss: 2.21543, Gen Loss: 4.39380, Sty Loss: 0.28630, Diff Loss: 0.94442, DiscLM Loss: 0.00000, GenLM Loss: 0.31711 +INFO:2025-06-10 23:07:56,555: Epoch [1/20], Step [300/2647], Loss: 0.62694, Disc Loss: 4.05135, Dur Loss: 1.21492, CE Loss: 0.07369, Norm Loss: 1.62162, F0 Loss: 3.81344, LM Loss: 2.31092, Gen Loss: 5.25609, Sty Loss: 0.26478, Diff Loss: 0.88248, DiscLM Loss: 0.21629, GenLM Loss: 0.39097 +INFO:2025-06-10 23:09:43,987: Epoch [1/20], Step [350/2647], Loss: 0.62045, Disc Loss: 3.79207, Dur Loss: 1.11382, CE Loss: 0.06862, Norm Loss: 2.00079, F0 Loss: 3.57181, LM Loss: 2.20376, Gen Loss: 5.86240, Sty Loss: 0.23570, Diff Loss: 0.87306, DiscLM Loss: 0.00000, GenLM Loss: 0.61724 +INFO:2025-06-10 23:11:33,128: Epoch [1/20], Step [400/2647], Loss: 0.62804, Disc Loss: 4.13741, Dur Loss: 1.22458, CE Loss: 0.07674, Norm Loss: 1.94336, F0 Loss: 4.58011, LM Loss: 2.24155, Gen Loss: 4.69441, Sty Loss: 0.20099, Diff Loss: 0.74584, DiscLM Loss: 0.11835, GenLM Loss: 0.72402 +INFO:2025-06-10 23:13:17,334: Epoch [1/20], Step [450/2647], Loss: 0.61310, Disc Loss: 3.82978, Dur Loss: 1.61970, CE Loss: 0.12771, Norm Loss: 1.93006, F0 Loss: 5.83418, LM Loss: 2.41949, Gen Loss: 4.67891, Sty Loss: 0.22740, Diff Loss: 0.86771, DiscLM Loss: 0.00000, GenLM Loss: 0.73902 +INFO:2025-06-10 23:15:03,599: Epoch [1/20], Step [500/2647], Loss: 0.62592, Disc Loss: 3.84785, Dur Loss: 1.30113, CE Loss: 0.07538, Norm Loss: 1.83915, F0 Loss: 4.64632, LM Loss: 2.33635, Gen Loss: 4.96465, Sty Loss: 0.14381, Diff Loss: 0.76109, DiscLM Loss: 0.06498, GenLM Loss: 0.87931 +INFO:2025-06-10 23:16:52,383: Epoch [1/20], Step [550/2647], Loss: 0.61466, Disc Loss: 3.92243, Dur Loss: 1.25546, CE Loss: 0.08105, Norm Loss: 2.19770, F0 Loss: 4.80879, LM Loss: 2.27051, Gen Loss: 4.91558, Sty Loss: 0.17542, Diff Loss: 0.76094, DiscLM Loss: 0.00000, GenLM Loss: 0.81056 +INFO:2025-06-10 23:18:41,548: Epoch [1/20], Step [600/2647], Loss: 0.58866, Disc Loss: 3.90473, Dur Loss: 1.20631, CE Loss: 0.07175, Norm Loss: 1.59275, F0 Loss: 4.22395, LM Loss: 2.21583, Gen Loss: 5.19171, Sty Loss: 0.16559, Diff Loss: 0.61715, DiscLM Loss: 0.07970, GenLM Loss: 0.88472 +INFO:2025-06-10 23:20:26,783: Epoch [1/20], Step [650/2647], Loss: 0.61430, Disc Loss: 3.77172, Dur Loss: 1.15394, CE Loss: 0.06428, Norm Loss: 1.69519, F0 Loss: 4.38059, LM Loss: 2.26519, Gen Loss: 5.12754, Sty Loss: 0.17969, Diff Loss: 0.61533, DiscLM Loss: 0.00000, GenLM Loss: 0.90620 +INFO:2025-06-10 23:22:14,111: Epoch [1/20], Step [700/2647], Loss: 0.60016, Disc Loss: 3.82096, Dur Loss: 1.04488, CE Loss: 0.05865, Norm Loss: 1.63734, F0 Loss: 4.00182, LM Loss: 2.33215, Gen Loss: 5.21846, Sty Loss: 0.14021, Diff Loss: 0.49655, DiscLM Loss: 0.03830, GenLM Loss: 0.97805 +INFO:2025-06-10 23:24:03,335: Epoch [1/20], Step [750/2647], Loss: 0.60220, Disc Loss: 3.84842, Dur Loss: 1.31934, CE Loss: 0.07708, Norm Loss: 1.83693, F0 Loss: 4.08663, LM Loss: 2.32928, Gen Loss: 5.08170, Sty Loss: 0.13921, Diff Loss: 0.43914, DiscLM Loss: 0.00000, GenLM Loss: 0.89785 +INFO:2025-06-10 23:25:51,554: Epoch [1/20], Step [800/2647], Loss: 0.60420, Disc Loss: 3.87524, Dur Loss: 1.15806, CE Loss: 0.07243, Norm Loss: 1.34825, F0 Loss: 4.47059, LM Loss: 2.20002, Gen Loss: 5.12199, Sty Loss: 0.13847, Diff Loss: 0.55575, DiscLM Loss: 0.03232, GenLM Loss: 0.94489 +INFO:2025-06-10 23:27:40,531: Epoch [1/20], Step [850/2647], Loss: 0.60409, Disc Loss: 3.79506, Dur Loss: 1.13716, CE Loss: 0.06372, Norm Loss: 1.27143, F0 Loss: 4.59368, LM Loss: 2.27478, Gen Loss: 6.22462, Sty Loss: 0.10187, Diff Loss: 0.53394, DiscLM Loss: 0.00000, GenLM Loss: 1.06942 +INFO:2025-06-10 23:29:28,282: Epoch [1/20], Step [900/2647], Loss: 0.60774, Disc Loss: 3.73534, Dur Loss: 1.11259, CE Loss: 0.06297, Norm Loss: 1.42712, F0 Loss: 3.61090, LM Loss: 2.27773, Gen Loss: 6.48748, Sty Loss: 0.13508, Diff Loss: 0.46527, DiscLM Loss: 0.04626, GenLM Loss: 0.95607 +INFO:2025-06-10 23:31:16,302: Epoch [1/20], Step [950/2647], Loss: 0.61123, Disc Loss: 3.87081, Dur Loss: 1.21734, CE Loss: 0.06698, Norm Loss: 1.69020, F0 Loss: 3.81538, LM Loss: 2.31186, Gen Loss: 5.66418, Sty Loss: 0.13985, Diff Loss: 0.52460, DiscLM Loss: 0.00000, GenLM Loss: 1.04807 +INFO:2025-06-10 23:33:02,430: Epoch [1/20], Step [1000/2647], Loss: 0.60306, Disc Loss: 3.95340, Dur Loss: 1.07669, CE Loss: 0.06853, Norm Loss: 1.85125, F0 Loss: 4.12176, LM Loss: 2.35886, Gen Loss: 5.59112, Sty Loss: 0.12872, Diff Loss: 0.42744, DiscLM Loss: 0.02418, GenLM Loss: 0.96445 +INFO:2025-06-10 23:34:48,220: Epoch [1/20], Step [1050/2647], Loss: 0.59957, Disc Loss: 3.71549, Dur Loss: 1.06826, CE Loss: 0.05949, Norm Loss: 1.40826, F0 Loss: 4.12930, LM Loss: 2.25460, Gen Loss: 5.89467, Sty Loss: 0.13749, Diff Loss: 0.39941, DiscLM Loss: 0.00000, GenLM Loss: 0.94758 +INFO:2025-06-10 23:36:35,857: Epoch [1/20], Step [1100/2647], Loss: 0.61483, Disc Loss: 3.81316, Dur Loss: 1.30494, CE Loss: 0.08229, Norm Loss: 1.91295, F0 Loss: 5.16897, LM Loss: 2.30201, Gen Loss: 5.93011, Sty Loss: 0.11293, Diff Loss: 0.47677, DiscLM Loss: 0.03251, GenLM Loss: 0.91756 +INFO:2025-06-10 23:38:25,179: Epoch [1/20], Step [1150/2647], Loss: 0.59152, Disc Loss: 3.76169, Dur Loss: 1.14162, CE Loss: 0.07186, Norm Loss: 1.59387, F0 Loss: 3.85908, LM Loss: 2.22013, Gen Loss: 5.85028, Sty Loss: 0.12867, Diff Loss: 0.43882, DiscLM Loss: 0.00000, GenLM Loss: 0.93344 +INFO:2025-06-10 23:40:14,158: Epoch [1/20], Step [1200/2647], Loss: 0.59579, Disc Loss: 3.80903, Dur Loss: 1.02478, CE Loss: 0.05092, Norm Loss: 1.29612, F0 Loss: 3.60099, LM Loss: 2.23061, Gen Loss: 5.09183, Sty Loss: 0.10809, Diff Loss: 0.53366, DiscLM Loss: 0.02512, GenLM Loss: 0.97540 +INFO:2025-06-10 23:42:00,703: Epoch [1/20], Step [1250/2647], Loss: 0.58739, Disc Loss: 3.86709, Dur Loss: 1.09719, CE Loss: 0.05681, Norm Loss: 1.16224, F0 Loss: 3.74518, LM Loss: 2.21772, Gen Loss: 5.21801, Sty Loss: 0.10160, Diff Loss: 0.39914, DiscLM Loss: 0.00000, GenLM Loss: 0.96172 +INFO:2025-06-10 23:43:50,837: Epoch [1/20], Step [1300/2647], Loss: 0.58950, Disc Loss: 3.92106, Dur Loss: 1.31094, CE Loss: 0.07262, Norm Loss: 1.66920, F0 Loss: 4.02291, LM Loss: 2.32804, Gen Loss: 6.33301, Sty Loss: 0.12117, Diff Loss: 0.47250, DiscLM Loss: 0.02445, GenLM Loss: 0.98541 +INFO:2025-06-10 23:45:39,925: Epoch [1/20], Step [1350/2647], Loss: 0.59505, Disc Loss: 3.82871, Dur Loss: 0.97202, CE Loss: 0.05074, Norm Loss: 1.37322, F0 Loss: 3.52736, LM Loss: 2.32450, Gen Loss: 6.27945, Sty Loss: 0.14597, Diff Loss: 0.36041, DiscLM Loss: 0.00000, GenLM Loss: 1.03375 +INFO:2025-06-10 23:47:28,081: Epoch [1/20], Step [1400/2647], Loss: 0.60004, Disc Loss: 3.79285, Dur Loss: 1.27125, CE Loss: 0.07962, Norm Loss: 1.35538, F0 Loss: 6.13392, LM Loss: 2.21035, Gen Loss: 5.18545, Sty Loss: 0.10555, Diff Loss: 0.42785, DiscLM Loss: 0.02505, GenLM Loss: 1.02835 +INFO:2025-06-10 23:49:15,619: Epoch [1/20], Step [1450/2647], Loss: 0.59551, Disc Loss: 3.78500, Dur Loss: 1.17627, CE Loss: 0.07112, Norm Loss: 1.60504, F0 Loss: 3.53758, LM Loss: 2.20891, Gen Loss: 6.83264, Sty Loss: 0.11519, Diff Loss: 0.31756, DiscLM Loss: 0.00000, GenLM Loss: 0.95094 +INFO:2025-06-10 23:51:06,854: Epoch [1/20], Step [1500/2647], Loss: 0.59226, Disc Loss: 3.66347, Dur Loss: 1.24261, CE Loss: 0.06549, Norm Loss: 1.51402, F0 Loss: 3.93773, LM Loss: 2.28870, Gen Loss: 6.00873, Sty Loss: 0.12451, Diff Loss: 0.21041, DiscLM Loss: 0.02722, GenLM Loss: 0.85339 +INFO:2025-06-10 23:52:55,191: Epoch [1/20], Step [1550/2647], Loss: 0.60842, Disc Loss: 3.74316, Dur Loss: 1.20849, CE Loss: 0.07174, Norm Loss: 1.68890, F0 Loss: 4.25387, LM Loss: 2.28279, Gen Loss: 5.96578, Sty Loss: 0.10517, Diff Loss: 0.39576, DiscLM Loss: 0.00000, GenLM Loss: 0.97245 +INFO:2025-06-10 23:54:45,069: Epoch [1/20], Step [1600/2647], Loss: 0.60342, Disc Loss: 3.66973, Dur Loss: 1.04094, CE Loss: 0.05318, Norm Loss: 1.62979, F0 Loss: 3.25963, LM Loss: 2.24331, Gen Loss: 5.94857, Sty Loss: 0.11135, Diff Loss: 0.31146, DiscLM Loss: 0.02801, GenLM Loss: 0.98187 +INFO:2025-06-10 23:56:35,322: Epoch [1/20], Step [1650/2647], Loss: 0.59472, Disc Loss: 3.74705, Dur Loss: 0.98171, CE Loss: 0.05398, Norm Loss: 1.17578, F0 Loss: 3.36308, LM Loss: 2.23302, Gen Loss: 6.42123, Sty Loss: 0.10820, Diff Loss: 0.32958, DiscLM Loss: 0.00000, GenLM Loss: 0.95143 +INFO:2025-06-10 23:58:20,214: Epoch [1/20], Step [1700/2647], Loss: 0.60262, Disc Loss: 3.77429, Dur Loss: 0.94902, CE Loss: 0.05451, Norm Loss: 1.81938, F0 Loss: 5.72552, LM Loss: 2.34326, Gen Loss: 6.69601, Sty Loss: 0.14110, Diff Loss: 0.46945, DiscLM Loss: 0.02119, GenLM Loss: 0.98743 +INFO:2025-06-11 00:00:07,771: Epoch [1/20], Step [1750/2647], Loss: 0.59867, Disc Loss: 3.67504, Dur Loss: 1.15199, CE Loss: 0.06486, Norm Loss: 1.53833, F0 Loss: 4.05094, LM Loss: 2.32955, Gen Loss: 7.38573, Sty Loss: 0.12501, Diff Loss: 0.42154, DiscLM Loss: 0.00000, GenLM Loss: 0.99914 +INFO:2025-06-11 00:01:56,805: Epoch [1/20], Step [1800/2647], Loss: 0.59031, Disc Loss: 3.90591, Dur Loss: 1.24378, CE Loss: 0.07287, Norm Loss: 1.53383, F0 Loss: 4.23040, LM Loss: 2.38063, Gen Loss: 5.92487, Sty Loss: 0.13469, Diff Loss: 0.37942, DiscLM Loss: 0.01326, GenLM Loss: 1.01446 +INFO:2025-06-11 00:03:46,760: Epoch [1/20], Step [1850/2647], Loss: 0.57912, Disc Loss: 3.69795, Dur Loss: 1.10993, CE Loss: 0.06502, Norm Loss: 1.20781, F0 Loss: 3.62803, LM Loss: 2.33872, Gen Loss: 6.71583, Sty Loss: 0.12957, Diff Loss: 0.43118, DiscLM Loss: 0.00000, GenLM Loss: 1.04149 +INFO:2025-06-11 00:05:34,796: Epoch [1/20], Step [1900/2647], Loss: 0.58376, Disc Loss: 3.73747, Dur Loss: 1.12708, CE Loss: 0.06318, Norm Loss: 1.80551, F0 Loss: 3.89177, LM Loss: 2.40332, Gen Loss: 6.66965, Sty Loss: 0.12407, Diff Loss: 0.34098, DiscLM Loss: 0.01039, GenLM Loss: 1.01777 +INFO:2025-06-11 00:07:25,154: Epoch [1/20], Step [1950/2647], Loss: 0.58827, Disc Loss: 3.73645, Dur Loss: 1.06168, CE Loss: 0.05307, Norm Loss: 1.30612, F0 Loss: 4.21061, LM Loss: 2.26872, Gen Loss: 6.28787, Sty Loss: 0.13836, Diff Loss: 0.30318, DiscLM Loss: 0.00000, GenLM Loss: 1.00657 +INFO:2025-06-11 00:09:13,651: Epoch [1/20], Step [2000/2647], Loss: 0.59113, Disc Loss: 3.78585, Dur Loss: 1.12138, CE Loss: 0.06412, Norm Loss: 1.23050, F0 Loss: 3.23188, LM Loss: 2.20285, Gen Loss: 6.30126, Sty Loss: 0.10719, Diff Loss: 0.47007, DiscLM Loss: 0.01615, GenLM Loss: 0.98226 +INFO:2025-06-11 00:11:00,914: Epoch [1/20], Step [2050/2647], Loss: 0.59520, Disc Loss: 3.71749, Dur Loss: 1.13055, CE Loss: 0.06112, Norm Loss: 1.83156, F0 Loss: 3.42263, LM Loss: 2.28488, Gen Loss: 6.46515, Sty Loss: 0.12969, Diff Loss: 0.41677, DiscLM Loss: 0.00000, GenLM Loss: 0.98213 +INFO:2025-06-11 00:12:47,948: Epoch [1/20], Step [2100/2647], Loss: 0.58931, Disc Loss: 3.69995, Dur Loss: 0.93082, CE Loss: 0.05032, Norm Loss: 1.40220, F0 Loss: 3.41590, LM Loss: 2.26013, Gen Loss: 6.55380, Sty Loss: 0.12775, Diff Loss: 0.32069, DiscLM Loss: 0.01485, GenLM Loss: 0.94985 +INFO:2025-06-11 00:14:36,854: Epoch [1/20], Step [2150/2647], Loss: 0.57724, Disc Loss: 3.76661, Dur Loss: 1.06378, CE Loss: 0.05523, Norm Loss: 1.36215, F0 Loss: 3.44842, LM Loss: 2.24825, Gen Loss: 6.41037, Sty Loss: 0.10507, Diff Loss: 0.49560, DiscLM Loss: 0.00000, GenLM Loss: 1.01234 +INFO:2025-06-11 00:16:24,014: Epoch [1/20], Step [2200/2647], Loss: 0.58300, Disc Loss: 3.79862, Dur Loss: 1.27889, CE Loss: 0.07811, Norm Loss: 1.39691, F0 Loss: 3.90668, LM Loss: 2.25180, Gen Loss: 5.38687, Sty Loss: 0.09748, Diff Loss: 0.20826, DiscLM Loss: 0.01134, GenLM Loss: 0.97175 +INFO:2025-06-11 00:18:12,272: Epoch [1/20], Step [2250/2647], Loss: 0.58998, Disc Loss: 3.73117, Dur Loss: 1.34615, CE Loss: 0.08349, Norm Loss: 1.11703, F0 Loss: 2.74746, LM Loss: 2.23306, Gen Loss: 5.87906, Sty Loss: 0.12348, Diff Loss: 0.31697, DiscLM Loss: 0.00000, GenLM Loss: 1.02123 +INFO:2025-06-11 00:19:59,915: Epoch [1/20], Step [2300/2647], Loss: 0.57129, Disc Loss: 3.67065, Dur Loss: 1.15248, CE Loss: 0.06495, Norm Loss: 1.44453, F0 Loss: 4.64699, LM Loss: 2.37477, Gen Loss: 7.21674, Sty Loss: 0.10784, Diff Loss: 0.25877, DiscLM Loss: 0.01528, GenLM Loss: 1.00811 +INFO:2025-06-11 00:21:47,085: Epoch [1/20], Step [2350/2647], Loss: 0.57484, Disc Loss: 3.69300, Dur Loss: 1.13333, CE Loss: 0.06594, Norm Loss: 1.49222, F0 Loss: 3.65384, LM Loss: 2.11067, Gen Loss: 6.37711, Sty Loss: 0.11202, Diff Loss: 0.31682, DiscLM Loss: 0.00000, GenLM Loss: 0.98929 +INFO:2025-06-11 00:23:35,395: Epoch [1/20], Step [2400/2647], Loss: 0.58330, Disc Loss: 3.69221, Dur Loss: 1.06042, CE Loss: 0.07141, Norm Loss: 1.67240, F0 Loss: 5.07512, LM Loss: 2.23093, Gen Loss: 6.81950, Sty Loss: 0.10026, Diff Loss: 0.32846, DiscLM Loss: 0.01746, GenLM Loss: 0.97020 +INFO:2025-06-11 00:25:24,618: Epoch [1/20], Step [2450/2647], Loss: 0.57659, Disc Loss: 3.79130, Dur Loss: 1.04118, CE Loss: 0.06466, Norm Loss: 1.39291, F0 Loss: 4.15074, LM Loss: 2.20164, Gen Loss: 5.90006, Sty Loss: 0.11043, Diff Loss: 0.33889, DiscLM Loss: 0.00000, GenLM Loss: 0.97694 +INFO:2025-06-11 00:27:15,063: Epoch [1/20], Step [2500/2647], Loss: 0.57897, Disc Loss: 3.73444, Dur Loss: 1.24121, CE Loss: 0.06691, Norm Loss: 1.60736, F0 Loss: 3.91026, LM Loss: 2.20287, Gen Loss: 7.04520, Sty Loss: 0.09370, Diff Loss: 0.24971, DiscLM Loss: 0.01050, GenLM Loss: 1.01167 +INFO:2025-06-11 00:29:02,760: Epoch [1/20], Step [2550/2647], Loss: 0.57566, Disc Loss: 3.75242, Dur Loss: 1.15930, CE Loss: 0.05801, Norm Loss: 1.53373, F0 Loss: 3.56057, LM Loss: 2.22803, Gen Loss: 6.27098, Sty Loss: 0.13080, Diff Loss: 0.38448, DiscLM Loss: 0.00000, GenLM Loss: 1.00540 +INFO:2025-06-11 00:30:52,751: Epoch [1/20], Step [2600/2647], Loss: 0.57446, Disc Loss: 3.55583, Dur Loss: 0.87542, CE Loss: 0.05175, Norm Loss: 1.08081, F0 Loss: 3.32188, LM Loss: 2.14475, Gen Loss: 8.05530, Sty Loss: 0.10938, Diff Loss: 0.34350, DiscLM Loss: 0.00720, GenLM Loss: 0.99862 +INFO:2025-06-11 00:34:56,211: Validation loss: 0.599, Dur loss: 1.098, F0 loss: 3.926 + + + +INFO:2025-06-11 00:36:47,972: Epoch [2/20], Step [50/2647], Loss: 0.56934, Disc Loss: 3.43541, Dur Loss: 1.16128, CE Loss: 0.07470, Norm Loss: 1.32977, F0 Loss: 3.54997, LM Loss: 2.26009, Gen Loss: 7.14060, Sty Loss: 0.09814, Diff Loss: 0.20112, DiscLM Loss: 0.00000, GenLM Loss: 0.94795 +INFO:2025-06-11 00:38:34,403: Epoch [2/20], Step [100/2647], Loss: 0.57364, Disc Loss: 3.51747, Dur Loss: 0.99972, CE Loss: 0.05364, Norm Loss: 1.49913, F0 Loss: 3.60959, LM Loss: 2.28620, Gen Loss: 7.63709, Sty Loss: 0.09344, Diff Loss: 0.31605, DiscLM Loss: 0.00667, GenLM Loss: 1.00672 +INFO:2025-06-11 00:40:20,425: Epoch [2/20], Step [150/2647], Loss: 0.58286, Disc Loss: 3.84028, Dur Loss: 1.15641, CE Loss: 0.06421, Norm Loss: 1.37146, F0 Loss: 3.32397, LM Loss: 2.15440, Gen Loss: 6.21733, Sty Loss: 0.13281, Diff Loss: 0.26414, DiscLM Loss: 0.00000, GenLM Loss: 0.95543 +INFO:2025-06-11 00:42:10,364: Epoch [2/20], Step [200/2647], Loss: 0.57738, Disc Loss: 3.58448, Dur Loss: 1.50476, CE Loss: 0.08658, Norm Loss: 1.59215, F0 Loss: 4.54305, LM Loss: 2.43954, Gen Loss: 7.99903, Sty Loss: 0.11481, Diff Loss: 0.43789, DiscLM Loss: 0.00914, GenLM Loss: 0.99797 +INFO:2025-06-11 00:44:00,065: Epoch [2/20], Step [250/2647], Loss: 0.58504, Disc Loss: 3.58063, Dur Loss: 1.14259, CE Loss: 0.06630, Norm Loss: 1.45668, F0 Loss: 3.86356, LM Loss: 2.33802, Gen Loss: 6.46900, Sty Loss: 0.08082, Diff Loss: 0.32375, DiscLM Loss: 0.00000, GenLM Loss: 1.01251 +INFO:2025-06-11 00:45:49,605: Epoch [2/20], Step [300/2647], Loss: 0.58409, Disc Loss: 3.54051, Dur Loss: 1.22738, CE Loss: 0.07420, Norm Loss: 1.48974, F0 Loss: 5.34282, LM Loss: 2.35975, Gen Loss: 7.48335, Sty Loss: 0.11749, Diff Loss: 0.21515, DiscLM Loss: 0.01035, GenLM Loss: 0.99185 +INFO:2025-06-11 00:47:37,117: Epoch [2/20], Step [350/2647], Loss: 0.58769, Disc Loss: 3.78112, Dur Loss: 1.17959, CE Loss: 0.06466, Norm Loss: 1.23290, F0 Loss: 4.19075, LM Loss: 2.28664, Gen Loss: 6.18502, Sty Loss: 0.08792, Diff Loss: 0.28590, DiscLM Loss: 0.00000, GenLM Loss: 0.98440 +INFO:2025-06-11 00:49:24,715: Epoch [2/20], Step [400/2647], Loss: 0.57429, Disc Loss: 3.93333, Dur Loss: 0.86613, CE Loss: 0.04551, Norm Loss: 1.14043, F0 Loss: 4.07879, LM Loss: 2.19741, Gen Loss: 6.60381, Sty Loss: 0.10194, Diff Loss: 0.24645, DiscLM Loss: 0.00776, GenLM Loss: 0.99376 +INFO:2025-06-11 00:51:12,984: Epoch [2/20], Step [450/2647], Loss: 0.57995, Disc Loss: 3.68749, Dur Loss: 0.99468, CE Loss: 0.05357, Norm Loss: 1.07584, F0 Loss: 3.23906, LM Loss: 2.27168, Gen Loss: 7.68826, Sty Loss: 0.10742, Diff Loss: 0.26879, DiscLM Loss: 0.00000, GenLM Loss: 0.97665 +INFO:2025-06-11 00:52:56,712: Epoch [2/20], Step [500/2647], Loss: 0.58069, Disc Loss: 3.66627, Dur Loss: 1.08015, CE Loss: 0.06009, Norm Loss: 1.50178, F0 Loss: 3.62313, LM Loss: 2.27645, Gen Loss: 6.30212, Sty Loss: 0.11223, Diff Loss: 0.36820, DiscLM Loss: 0.01065, GenLM Loss: 1.03614 +INFO:2025-06-11 00:54:45,287: Epoch [2/20], Step [550/2647], Loss: 0.57393, Disc Loss: 3.79486, Dur Loss: 0.98779, CE Loss: 0.05747, Norm Loss: 1.52550, F0 Loss: 3.80829, LM Loss: 2.34130, Gen Loss: 6.36980, Sty Loss: 0.13240, Diff Loss: 0.19044, DiscLM Loss: 0.00000, GenLM Loss: 1.01170 +INFO:2025-06-11 00:56:34,775: Epoch [2/20], Step [600/2647], Loss: 0.57080, Disc Loss: 3.80613, Dur Loss: 1.39429, CE Loss: 0.08673, Norm Loss: 1.34077, F0 Loss: 4.27156, LM Loss: 2.24826, Gen Loss: 5.92983, Sty Loss: 0.13567, Diff Loss: 0.34159, DiscLM Loss: 0.00579, GenLM Loss: 0.98956 +INFO:2025-06-11 00:58:20,574: Epoch [2/20], Step [650/2647], Loss: 0.58437, Disc Loss: 3.79891, Dur Loss: 0.89812, CE Loss: 0.04675, Norm Loss: 1.18604, F0 Loss: 3.10389, LM Loss: 2.24702, Gen Loss: 6.48184, Sty Loss: 0.12696, Diff Loss: 0.25244, DiscLM Loss: 0.00000, GenLM Loss: 0.99331 +INFO:2025-06-11 01:00:08,184: Epoch [2/20], Step [700/2647], Loss: 0.57925, Disc Loss: 3.61493, Dur Loss: 0.92019, CE Loss: 0.04863, Norm Loss: 1.19827, F0 Loss: 2.72453, LM Loss: 2.19703, Gen Loss: 6.57531, Sty Loss: 0.09048, Diff Loss: 0.31558, DiscLM Loss: 0.01037, GenLM Loss: 0.99024 +INFO:2025-06-11 01:01:55,948: Epoch [2/20], Step [750/2647], Loss: 0.57223, Disc Loss: 3.55737, Dur Loss: 1.10703, CE Loss: 0.06379, Norm Loss: 1.17487, F0 Loss: 4.81597, LM Loss: 2.29486, Gen Loss: 7.48877, Sty Loss: 0.11958, Diff Loss: 0.34720, DiscLM Loss: 0.00000, GenLM Loss: 0.99925 +INFO:2025-06-11 01:03:44,797: Epoch [2/20], Step [800/2647], Loss: 0.57717, Disc Loss: 3.58542, Dur Loss: 0.99998, CE Loss: 0.05938, Norm Loss: 1.37686, F0 Loss: 4.66941, LM Loss: 2.35301, Gen Loss: 6.81176, Sty Loss: 0.11959, Diff Loss: 0.29831, DiscLM Loss: 0.00556, GenLM Loss: 0.99006 +INFO:2025-06-11 01:05:35,853: Epoch [2/20], Step [850/2647], Loss: 0.56200, Disc Loss: 3.76769, Dur Loss: 1.12081, CE Loss: 0.07017, Norm Loss: 1.19381, F0 Loss: 4.24942, LM Loss: 2.35259, Gen Loss: 6.85444, Sty Loss: 0.09859, Diff Loss: 0.32654, DiscLM Loss: 0.00000, GenLM Loss: 0.99924 +INFO:2025-06-11 01:07:22,773: Epoch [2/20], Step [900/2647], Loss: 0.56545, Disc Loss: 3.54456, Dur Loss: 1.14138, CE Loss: 0.06943, Norm Loss: 1.52144, F0 Loss: 7.03891, LM Loss: 2.34324, Gen Loss: 7.62864, Sty Loss: 0.12882, Diff Loss: 0.28408, DiscLM Loss: 0.01189, GenLM Loss: 0.98360 +INFO:2025-06-11 01:09:10,046: Epoch [2/20], Step [950/2647], Loss: 0.57093, Disc Loss: 3.51483, Dur Loss: 1.01704, CE Loss: 0.05675, Norm Loss: 1.14510, F0 Loss: 4.04532, LM Loss: 2.29781, Gen Loss: 7.26197, Sty Loss: 0.12363, Diff Loss: 0.25112, DiscLM Loss: 0.00000, GenLM Loss: 1.01491 +INFO:2025-06-11 01:10:57,713: Epoch [2/20], Step [1000/2647], Loss: 0.57424, Disc Loss: 3.69525, Dur Loss: 1.30014, CE Loss: 0.07152, Norm Loss: 1.34526, F0 Loss: 3.29854, LM Loss: 2.22664, Gen Loss: 6.29479, Sty Loss: 0.12506, Diff Loss: 0.47032, DiscLM Loss: 0.00482, GenLM Loss: 1.00370 +INFO:2025-06-11 01:12:42,547: Epoch [2/20], Step [1050/2647], Loss: 0.57944, Disc Loss: 3.85555, Dur Loss: 0.83183, CE Loss: 0.04699, Norm Loss: 1.31164, F0 Loss: 3.28694, LM Loss: 2.30125, Gen Loss: 6.36772, Sty Loss: 0.11233, Diff Loss: 0.35931, DiscLM Loss: 0.00000, GenLM Loss: 1.01060 +INFO:2025-06-11 01:14:31,046: Epoch [2/20], Step [1100/2647], Loss: 0.57841, Disc Loss: 3.76438, Dur Loss: 1.23546, CE Loss: 0.06619, Norm Loss: 1.24161, F0 Loss: 3.68991, LM Loss: 2.23649, Gen Loss: 6.02477, Sty Loss: 0.09699, Diff Loss: 0.30588, DiscLM Loss: 0.01079, GenLM Loss: 0.98355 +INFO:2025-06-11 01:16:18,468: Epoch [2/20], Step [1150/2647], Loss: 0.57672, Disc Loss: 3.66415, Dur Loss: 1.37559, CE Loss: 0.07053, Norm Loss: 1.44765, F0 Loss: 3.79584, LM Loss: 2.26985, Gen Loss: 7.18612, Sty Loss: 0.10370, Diff Loss: 0.35780, DiscLM Loss: 0.00000, GenLM Loss: 1.00251 +INFO:2025-06-11 01:18:07,970: Epoch [2/20], Step [1200/2647], Loss: 0.57233, Disc Loss: 3.71244, Dur Loss: 0.97949, CE Loss: 0.05408, Norm Loss: 1.10984, F0 Loss: 4.56106, LM Loss: 2.18274, Gen Loss: 6.40959, Sty Loss: 0.09521, Diff Loss: 0.33878, DiscLM Loss: 0.01043, GenLM Loss: 0.99255 +INFO:2025-06-11 01:19:58,537: Epoch [2/20], Step [1250/2647], Loss: 0.56513, Disc Loss: 3.71529, Dur Loss: 1.16107, CE Loss: 0.06489, Norm Loss: 1.47283, F0 Loss: 3.68115, LM Loss: 2.40889, Gen Loss: 6.55577, Sty Loss: 0.11782, Diff Loss: 0.34093, DiscLM Loss: 0.00000, GenLM Loss: 0.96655 +INFO:2025-06-11 01:21:46,844: Epoch [2/20], Step [1300/2647], Loss: 0.57155, Disc Loss: 3.72565, Dur Loss: 1.05798, CE Loss: 0.06119, Norm Loss: 1.50426, F0 Loss: 3.76022, LM Loss: 2.34931, Gen Loss: 6.26527, Sty Loss: 0.12115, Diff Loss: 0.30836, DiscLM Loss: 0.00503, GenLM Loss: 1.00471 +INFO:2025-06-11 01:23:36,524: Epoch [2/20], Step [1350/2647], Loss: 0.56271, Disc Loss: 3.65057, Dur Loss: 1.06360, CE Loss: 0.07081, Norm Loss: 1.40822, F0 Loss: 3.61317, LM Loss: 2.37813, Gen Loss: 7.00216, Sty Loss: 0.10820, Diff Loss: 0.26398, DiscLM Loss: 0.00000, GenLM Loss: 1.04111 +INFO:2025-06-11 01:25:28,722: Epoch [2/20], Step [1400/2647], Loss: 0.55963, Disc Loss: 3.72062, Dur Loss: 1.08423, CE Loss: 0.06144, Norm Loss: 1.43148, F0 Loss: 4.33578, LM Loss: 2.37032, Gen Loss: 6.26355, Sty Loss: 0.10130, Diff Loss: 0.33786, DiscLM Loss: 0.00495, GenLM Loss: 0.99187 +INFO:2025-06-11 01:27:18,153: Epoch [2/20], Step [1450/2647], Loss: 0.56593, Disc Loss: 3.73313, Dur Loss: 1.29600, CE Loss: 0.06697, Norm Loss: 1.52118, F0 Loss: 3.34377, LM Loss: 2.17466, Gen Loss: 5.85184, Sty Loss: 0.13190, Diff Loss: 0.29608, DiscLM Loss: 0.00000, GenLM Loss: 0.98193 +INFO:2025-06-11 01:29:06,546: Epoch [2/20], Step [1500/2647], Loss: 0.56348, Disc Loss: 3.72354, Dur Loss: 1.33240, CE Loss: 0.07586, Norm Loss: 1.86853, F0 Loss: 3.49084, LM Loss: 2.20734, Gen Loss: 6.47709, Sty Loss: 0.11160, Diff Loss: 0.34364, DiscLM Loss: 0.00666, GenLM Loss: 0.98529 +INFO:2025-06-11 01:30:53,356: Epoch [2/20], Step [1550/2647], Loss: 0.56923, Disc Loss: 3.69038, Dur Loss: 1.03450, CE Loss: 0.06048, Norm Loss: 1.67113, F0 Loss: 3.37054, LM Loss: 2.36943, Gen Loss: 6.12748, Sty Loss: 0.12317, Diff Loss: 0.37621, DiscLM Loss: 0.00000, GenLM Loss: 1.00824 +INFO:2025-06-11 01:32:41,506: Epoch [2/20], Step [1600/2647], Loss: 0.57389, Disc Loss: 3.70401, Dur Loss: 1.08209, CE Loss: 0.05911, Norm Loss: 1.20542, F0 Loss: 3.21055, LM Loss: 2.28073, Gen Loss: 6.01543, Sty Loss: 0.12603, Diff Loss: 0.32342, DiscLM Loss: 0.00331, GenLM Loss: 1.02790 +INFO:2025-06-11 01:34:30,554: Epoch [2/20], Step [1650/2647], Loss: 0.57339, Disc Loss: 3.81974, Dur Loss: 0.97362, CE Loss: 0.05191, Norm Loss: 1.18162, F0 Loss: 3.13128, LM Loss: 2.31051, Gen Loss: 6.01727, Sty Loss: 0.10941, Diff Loss: 0.28507, DiscLM Loss: 0.00000, GenLM Loss: 0.99000 +INFO:2025-06-11 01:36:17,342: Epoch [2/20], Step [1700/2647], Loss: 0.55100, Disc Loss: 3.74394, Dur Loss: 0.95739, CE Loss: 0.05112, Norm Loss: 1.32035, F0 Loss: 3.73458, LM Loss: 2.18306, Gen Loss: 6.25193, Sty Loss: 0.11880, Diff Loss: 0.32273, DiscLM Loss: 0.00626, GenLM Loss: 1.02555 +INFO:2025-06-11 01:38:04,322: Epoch [2/20], Step [1750/2647], Loss: 0.56056, Disc Loss: 3.91990, Dur Loss: 1.11102, CE Loss: 0.05965, Norm Loss: 1.19027, F0 Loss: 3.01247, LM Loss: 2.24616, Gen Loss: 6.68455, Sty Loss: 0.13863, Diff Loss: 0.34840, DiscLM Loss: 0.00000, GenLM Loss: 1.00951 +INFO:2025-06-11 01:39:51,529: Epoch [2/20], Step [1800/2647], Loss: 0.55747, Disc Loss: 3.74523, Dur Loss: 0.99747, CE Loss: 0.05392, Norm Loss: 1.12949, F0 Loss: 3.66129, LM Loss: 2.34537, Gen Loss: 7.38351, Sty Loss: 0.13361, Diff Loss: 0.35213, DiscLM Loss: 0.00206, GenLM Loss: 1.00800 +INFO:2025-06-11 01:41:40,253: Epoch [2/20], Step [1850/2647], Loss: 0.56536, Disc Loss: 3.68469, Dur Loss: 1.20455, CE Loss: 0.06318, Norm Loss: 1.31731, F0 Loss: 3.60251, LM Loss: 2.25131, Gen Loss: 6.61388, Sty Loss: 0.11915, Diff Loss: 0.30980, DiscLM Loss: 0.00000, GenLM Loss: 0.99383 +INFO:2025-06-11 01:43:28,268: Epoch [2/20], Step [1900/2647], Loss: 0.56451, Disc Loss: 3.65711, Dur Loss: 1.09052, CE Loss: 0.06186, Norm Loss: 1.53169, F0 Loss: 3.50574, LM Loss: 2.40083, Gen Loss: 7.47306, Sty Loss: 0.12373, Diff Loss: 0.29454, DiscLM Loss: 0.00294, GenLM Loss: 0.97893 +INFO:2025-06-11 01:45:13,848: Epoch [2/20], Step [1950/2647], Loss: 0.56140, Disc Loss: 3.81027, Dur Loss: 0.88375, CE Loss: 0.04973, Norm Loss: 1.34158, F0 Loss: 2.84393, LM Loss: 2.24061, Gen Loss: 5.81924, Sty Loss: 0.11570, Diff Loss: 0.26343, DiscLM Loss: 0.00000, GenLM Loss: 1.02100 +INFO:2025-06-11 01:47:03,196: Epoch [2/20], Step [2000/2647], Loss: 0.55788, Disc Loss: 3.71894, Dur Loss: 1.08220, CE Loss: 0.05595, Norm Loss: 1.08511, F0 Loss: 3.05574, LM Loss: 2.22241, Gen Loss: 6.24103, Sty Loss: 0.10459, Diff Loss: 0.37435, DiscLM Loss: 0.00302, GenLM Loss: 0.98673 +INFO:2025-06-11 01:48:47,930: Epoch [2/20], Step [2050/2647], Loss: 0.57046, Disc Loss: 3.70507, Dur Loss: 1.01652, CE Loss: 0.05549, Norm Loss: 1.28016, F0 Loss: 3.39538, LM Loss: 2.30493, Gen Loss: 6.40185, Sty Loss: 0.13966, Diff Loss: 0.40374, DiscLM Loss: 0.00000, GenLM Loss: 0.96909 +INFO:2025-06-11 01:50:40,398: Epoch [2/20], Step [2100/2647], Loss: 0.56457, Disc Loss: 3.70912, Dur Loss: 1.08348, CE Loss: 0.06520, Norm Loss: 1.25747, F0 Loss: 3.25451, LM Loss: 2.23852, Gen Loss: 5.98582, Sty Loss: 0.10837, Diff Loss: 0.33497, DiscLM Loss: 0.00369, GenLM Loss: 0.97496 +INFO:2025-06-11 01:52:29,546: Epoch [2/20], Step [2150/2647], Loss: 0.56501, Disc Loss: 3.78248, Dur Loss: 0.96997, CE Loss: 0.05459, Norm Loss: 1.31303, F0 Loss: 5.72722, LM Loss: 2.35157, Gen Loss: 6.53617, Sty Loss: 0.14267, Diff Loss: 0.40501, DiscLM Loss: 0.00000, GenLM Loss: 1.01004 +INFO:2025-06-11 01:54:20,155: Epoch [2/20], Step [2200/2647], Loss: 0.57773, Disc Loss: 3.68629, Dur Loss: 0.91626, CE Loss: 0.04827, Norm Loss: 1.04167, F0 Loss: 3.24749, LM Loss: 2.19344, Gen Loss: 6.21846, Sty Loss: 0.11207, Diff Loss: 0.34261, DiscLM Loss: 0.00367, GenLM Loss: 1.01261 +INFO:2025-06-11 01:56:07,038: Epoch [2/20], Step [2250/2647], Loss: 0.58025, Disc Loss: 3.82214, Dur Loss: 1.38040, CE Loss: 0.08055, Norm Loss: 1.42768, F0 Loss: 4.05550, LM Loss: 2.34692, Gen Loss: 6.48149, Sty Loss: 0.14494, Diff Loss: 0.23376, DiscLM Loss: 0.00000, GenLM Loss: 0.98549 +INFO:2025-06-11 01:57:57,386: Epoch [2/20], Step [2300/2647], Loss: 0.56686, Disc Loss: 3.60845, Dur Loss: 1.03565, CE Loss: 0.06381, Norm Loss: 1.29574, F0 Loss: 3.33304, LM Loss: 2.25307, Gen Loss: 6.68237, Sty Loss: 0.12562, Diff Loss: 0.42042, DiscLM Loss: 0.00261, GenLM Loss: 0.98156 +INFO:2025-06-11 01:59:43,544: Epoch [2/20], Step [2350/2647], Loss: 0.57327, Disc Loss: 3.80446, Dur Loss: 1.05081, CE Loss: 0.05570, Norm Loss: 1.18484, F0 Loss: 3.31708, LM Loss: 2.39623, Gen Loss: 6.70180, Sty Loss: 0.10126, Diff Loss: 0.48355, DiscLM Loss: 0.00000, GenLM Loss: 0.99302 +INFO:2025-06-11 02:01:32,548: Epoch [2/20], Step [2400/2647], Loss: 0.56652, Disc Loss: 3.74050, Dur Loss: 1.08519, CE Loss: 0.06318, Norm Loss: 1.25898, F0 Loss: 3.57880, LM Loss: 2.28287, Gen Loss: 6.65598, Sty Loss: 0.15062, Diff Loss: 0.34168, DiscLM Loss: 0.00382, GenLM Loss: 1.00351 +INFO:2025-06-11 02:03:19,868: Epoch [2/20], Step [2450/2647], Loss: 0.57148, Disc Loss: 3.69951, Dur Loss: 1.07585, CE Loss: 0.05735, Norm Loss: 1.13345, F0 Loss: 3.21422, LM Loss: 2.18922, Gen Loss: 6.32965, Sty Loss: 0.14483, Diff Loss: 0.36881, DiscLM Loss: 0.00000, GenLM Loss: 0.99478 +INFO:2025-06-11 02:05:08,612: Epoch [2/20], Step [2500/2647], Loss: 0.56541, Disc Loss: 3.80012, Dur Loss: 1.10133, CE Loss: 0.06337, Norm Loss: 1.42472, F0 Loss: 3.52199, LM Loss: 2.25779, Gen Loss: 6.52540, Sty Loss: 0.14480, Diff Loss: 0.21098, DiscLM Loss: 0.00318, GenLM Loss: 1.00958 +INFO:2025-06-11 02:06:59,606: Epoch [2/20], Step [2550/2647], Loss: 0.55634, Disc Loss: 3.78196, Dur Loss: 1.12995, CE Loss: 0.05970, Norm Loss: 1.61265, F0 Loss: 3.97470, LM Loss: 2.42614, Gen Loss: 6.80048, Sty Loss: 0.14272, Diff Loss: 0.28055, DiscLM Loss: 0.00000, GenLM Loss: 0.99008 +INFO:2025-06-11 02:08:50,504: Epoch [2/20], Step [2600/2647], Loss: 0.56284, Disc Loss: 3.77976, Dur Loss: 1.04768, CE Loss: 0.05619, Norm Loss: 0.97606, F0 Loss: 3.91988, LM Loss: 2.19384, Gen Loss: 6.28931, Sty Loss: 0.13946, Diff Loss: 0.40807, DiscLM Loss: 0.00305, GenLM Loss: 0.97961 +INFO:2025-06-11 02:12:44,315: Validation loss: 0.599, Dur loss: 1.092, F0 loss: 3.685 + + + +INFO:2025-06-11 02:14:37,517: Epoch [3/20], Step [50/2647], Loss: 0.57383, Disc Loss: 3.68832, Dur Loss: 1.20685, CE Loss: 0.07445, Norm Loss: 1.60192, F0 Loss: 4.05045, LM Loss: 2.33146, Gen Loss: 6.79198, Sty Loss: 0.13189, Diff Loss: 0.31948, DiscLM Loss: 0.00000, GenLM Loss: 0.98958 +INFO:2025-06-11 02:16:28,035: Epoch [3/20], Step [100/2647], Loss: 0.57187, Disc Loss: 3.65674, Dur Loss: 1.05922, CE Loss: 0.06457, Norm Loss: 1.27818, F0 Loss: 3.16829, LM Loss: 2.35976, Gen Loss: 7.38817, Sty Loss: 0.14008, Diff Loss: 0.21225, DiscLM Loss: 0.00175, GenLM Loss: 1.00068 +INFO:2025-06-11 02:18:20,321: Epoch [3/20], Step [150/2647], Loss: 0.56563, Disc Loss: 3.81079, Dur Loss: 0.97336, CE Loss: 0.05737, Norm Loss: 1.19634, F0 Loss: 3.53198, LM Loss: 2.14757, Gen Loss: 5.40026, Sty Loss: 0.10864, Diff Loss: 0.30251, DiscLM Loss: 0.00000, GenLM Loss: 0.98058 +INFO:2025-06-11 02:20:10,044: Epoch [3/20], Step [200/2647], Loss: 0.57279, Disc Loss: 3.70950, Dur Loss: 1.05937, CE Loss: 0.05897, Norm Loss: 1.39933, F0 Loss: 3.65759, LM Loss: 2.33731, Gen Loss: 6.04555, Sty Loss: 0.11578, Diff Loss: 0.18948, DiscLM Loss: 0.00225, GenLM Loss: 0.98122 +INFO:2025-06-11 02:21:58,341: Epoch [3/20], Step [250/2647], Loss: 0.55715, Disc Loss: 3.83176, Dur Loss: 1.08403, CE Loss: 0.06135, Norm Loss: 1.24809, F0 Loss: 3.24004, LM Loss: 2.30865, Gen Loss: 5.30052, Sty Loss: 0.14025, Diff Loss: 0.27473, DiscLM Loss: 0.00000, GenLM Loss: 0.96531 +INFO:2025-06-11 02:23:49,384: Epoch [3/20], Step [300/2647], Loss: 0.55612, Disc Loss: 3.81954, Dur Loss: 0.99147, CE Loss: 0.05497, Norm Loss: 1.18930, F0 Loss: 3.08697, LM Loss: 2.33634, Gen Loss: 6.72639, Sty Loss: 0.12514, Diff Loss: 0.24606, DiscLM Loss: 0.00256, GenLM Loss: 1.04267 +INFO:2025-06-11 02:25:37,744: Epoch [3/20], Step [350/2647], Loss: 0.56640, Disc Loss: 3.73638, Dur Loss: 0.94136, CE Loss: 0.04982, Norm Loss: 1.18150, F0 Loss: 3.62418, LM Loss: 2.24629, Gen Loss: 6.22102, Sty Loss: 0.12850, Diff Loss: 0.23721, DiscLM Loss: 0.00000, GenLM Loss: 0.98556 +INFO:2025-06-11 02:27:26,239: Epoch [3/20], Step [400/2647], Loss: 0.54700, Disc Loss: 3.78198, Dur Loss: 1.16235, CE Loss: 0.05984, Norm Loss: 1.27073, F0 Loss: 3.29738, LM Loss: 2.28774, Gen Loss: 5.70950, Sty Loss: 0.13923, Diff Loss: 0.27791, DiscLM Loss: 0.00704, GenLM Loss: 0.97740 +INFO:2025-06-11 02:29:15,451: Epoch [3/20], Step [450/2647], Loss: 0.56140, Disc Loss: 3.78166, Dur Loss: 1.02316, CE Loss: 0.05249, Norm Loss: 1.14251, F0 Loss: 3.28865, LM Loss: 2.24279, Gen Loss: 6.24230, Sty Loss: 0.10741, Diff Loss: 0.30109, DiscLM Loss: 0.00000, GenLM Loss: 0.99650 +INFO:2025-06-11 02:31:04,643: Epoch [3/20], Step [500/2647], Loss: 0.54395, Disc Loss: 3.80771, Dur Loss: 1.08758, CE Loss: 0.06078, Norm Loss: 1.47247, F0 Loss: 3.44949, LM Loss: 2.40127, Gen Loss: 6.22034, Sty Loss: 0.13428, Diff Loss: 0.38399, DiscLM Loss: 0.00188, GenLM Loss: 1.01304 +INFO:2025-06-11 02:32:53,376: Epoch [3/20], Step [550/2647], Loss: 0.56062, Disc Loss: 3.79515, Dur Loss: 1.08196, CE Loss: 0.05875, Norm Loss: 1.35183, F0 Loss: 3.60523, LM Loss: 2.26795, Gen Loss: 5.81005, Sty Loss: 0.11360, Diff Loss: 0.23100, DiscLM Loss: 0.00000, GenLM Loss: 0.99124 +INFO:2025-06-11 02:34:42,100: Epoch [3/20], Step [600/2647], Loss: 0.55382, Disc Loss: 3.74792, Dur Loss: 1.07560, CE Loss: 0.05884, Norm Loss: 1.49866, F0 Loss: 4.57730, LM Loss: 2.28627, Gen Loss: 6.09046, Sty Loss: 0.12741, Diff Loss: 0.24885, DiscLM Loss: 0.00242, GenLM Loss: 0.99945 +INFO:2025-06-11 02:36:27,928: Epoch [3/20], Step [650/2647], Loss: 0.55336, Disc Loss: 3.77591, Dur Loss: 0.82583, CE Loss: 0.05434, Norm Loss: 1.08938, F0 Loss: 4.33395, LM Loss: 2.24293, Gen Loss: 6.62421, Sty Loss: 0.10441, Diff Loss: 0.44410, DiscLM Loss: 0.00000, GenLM Loss: 0.99442 +INFO:2025-06-11 02:38:17,914: Epoch [3/20], Step [700/2647], Loss: 0.55553, Disc Loss: 3.81558, Dur Loss: 1.13120, CE Loss: 0.06430, Norm Loss: 1.08233, F0 Loss: 3.77269, LM Loss: 2.34511, Gen Loss: 6.48923, Sty Loss: 0.15837, Diff Loss: 0.38252, DiscLM Loss: 0.00328, GenLM Loss: 1.02723 +INFO:2025-06-11 02:40:05,009: Epoch [3/20], Step [750/2647], Loss: 0.55437, Disc Loss: 3.67754, Dur Loss: 0.94887, CE Loss: 0.05139, Norm Loss: 1.15794, F0 Loss: 4.55342, LM Loss: 2.24956, Gen Loss: 6.53768, Sty Loss: 0.13497, Diff Loss: 0.30780, DiscLM Loss: 0.00000, GenLM Loss: 0.98161 +INFO:2025-06-11 02:41:53,141: Epoch [3/20], Step [800/2647], Loss: 0.55342, Disc Loss: 3.77078, Dur Loss: 1.69748, CE Loss: 0.09273, Norm Loss: 1.31683, F0 Loss: 4.78530, LM Loss: 2.18691, Gen Loss: 6.23883, Sty Loss: 0.15393, Diff Loss: 0.40348, DiscLM Loss: 0.00222, GenLM Loss: 0.99545 +INFO:2025-06-11 02:43:39,930: Epoch [3/20], Step [850/2647], Loss: 0.55776, Disc Loss: 3.76118, Dur Loss: 1.13131, CE Loss: 0.06187, Norm Loss: 1.30707, F0 Loss: 3.41065, LM Loss: 2.20075, Gen Loss: 6.33059, Sty Loss: 0.12287, Diff Loss: 0.31062, DiscLM Loss: 0.00000, GenLM Loss: 1.00743 +INFO:2025-06-11 02:45:27,643: Epoch [3/20], Step [900/2647], Loss: 0.55967, Disc Loss: 3.76675, Dur Loss: 1.03221, CE Loss: 0.05909, Norm Loss: 1.13635, F0 Loss: 4.43021, LM Loss: 2.26335, Gen Loss: 6.27203, Sty Loss: 0.11758, Diff Loss: 0.35957, DiscLM Loss: 0.00153, GenLM Loss: 0.98064 +INFO:2025-06-11 02:47:17,560: Epoch [3/20], Step [950/2647], Loss: 0.55951, Disc Loss: 3.80799, Dur Loss: 0.90608, CE Loss: 0.04542, Norm Loss: 0.94607, F0 Loss: 3.27604, LM Loss: 2.24640, Gen Loss: 6.40874, Sty Loss: 0.11910, Diff Loss: 0.24702, DiscLM Loss: 0.00000, GenLM Loss: 1.01068 +INFO:2025-06-11 02:49:07,736: Epoch [3/20], Step [1000/2647], Loss: 0.56017, Disc Loss: 3.69863, Dur Loss: 1.11272, CE Loss: 0.05983, Norm Loss: 1.32116, F0 Loss: 3.54712, LM Loss: 2.34543, Gen Loss: 7.33670, Sty Loss: 0.15755, Diff Loss: 0.30068, DiscLM Loss: 0.00227, GenLM Loss: 0.98785 +INFO:2025-06-11 02:50:55,304: Epoch [3/20], Step [1050/2647], Loss: 0.55201, Disc Loss: 3.76243, Dur Loss: 1.20477, CE Loss: 0.06740, Norm Loss: 1.28667, F0 Loss: 4.00956, LM Loss: 2.26019, Gen Loss: 5.84102, Sty Loss: 0.16912, Diff Loss: 0.41080, DiscLM Loss: 0.00000, GenLM Loss: 0.98073 +INFO:2025-06-11 02:52:43,171: Epoch [3/20], Step [1100/2647], Loss: 0.55195, Disc Loss: 3.78552, Dur Loss: 1.06761, CE Loss: 0.05822, Norm Loss: 0.93801, F0 Loss: 3.21387, LM Loss: 2.16031, Gen Loss: 5.54921, Sty Loss: 0.14850, Diff Loss: 0.29332, DiscLM Loss: 0.00155, GenLM Loss: 0.99545 +INFO:2025-06-11 02:54:31,635: Epoch [3/20], Step [1150/2647], Loss: 0.54933, Disc Loss: 3.72499, Dur Loss: 1.09694, CE Loss: 0.06108, Norm Loss: 1.33195, F0 Loss: 4.73663, LM Loss: 2.26369, Gen Loss: 6.62953, Sty Loss: 0.11643, Diff Loss: 0.24951, DiscLM Loss: 0.00000, GenLM Loss: 0.99759 +INFO:2025-06-11 02:56:24,004: Epoch [3/20], Step [1200/2647], Loss: 0.55032, Disc Loss: 3.76961, Dur Loss: 0.98596, CE Loss: 0.05247, Norm Loss: 1.21121, F0 Loss: 3.57647, LM Loss: 2.20151, Gen Loss: 6.27880, Sty Loss: 0.14797, Diff Loss: 0.40048, DiscLM Loss: 0.00407, GenLM Loss: 0.99694 +INFO:2025-06-11 02:58:10,342: Epoch [3/20], Step [1250/2647], Loss: 0.55409, Disc Loss: 3.66587, Dur Loss: 0.98750, CE Loss: 0.05598, Norm Loss: 1.21548, F0 Loss: 2.96559, LM Loss: 2.21771, Gen Loss: 6.34952, Sty Loss: 0.11484, Diff Loss: 0.28739, DiscLM Loss: 0.00000, GenLM Loss: 0.98856 +INFO:2025-06-11 02:59:58,143: Epoch [3/20], Step [1300/2647], Loss: 0.56521, Disc Loss: 3.78146, Dur Loss: 0.90044, CE Loss: 0.05008, Norm Loss: 1.19008, F0 Loss: 4.08558, LM Loss: 2.21863, Gen Loss: 5.93940, Sty Loss: 0.11532, Diff Loss: 0.31252, DiscLM Loss: 0.00135, GenLM Loss: 0.99564 +INFO:2025-06-11 03:01:47,212: Epoch [3/20], Step [1350/2647], Loss: 0.55155, Disc Loss: 3.71247, Dur Loss: 1.26581, CE Loss: 0.06851, Norm Loss: 1.65130, F0 Loss: 3.69084, LM Loss: 2.21707, Gen Loss: 5.47236, Sty Loss: 0.12958, Diff Loss: 0.32197, DiscLM Loss: 0.00000, GenLM Loss: 1.01861 +INFO:2025-06-11 03:03:35,339: Epoch [3/20], Step [1400/2647], Loss: 0.54667, Disc Loss: 3.83994, Dur Loss: 1.10494, CE Loss: 0.05964, Norm Loss: 1.42262, F0 Loss: 2.88090, LM Loss: 2.21002, Gen Loss: 5.42712, Sty Loss: 0.13972, Diff Loss: 0.28457, DiscLM Loss: 0.01426, GenLM Loss: 0.91887 +INFO:2025-06-11 03:05:23,394: Epoch [3/20], Step [1450/2647], Loss: 0.55296, Disc Loss: 3.71462, Dur Loss: 1.14120, CE Loss: 0.06084, Norm Loss: 1.01267, F0 Loss: 2.74232, LM Loss: 2.16959, Gen Loss: 6.48401, Sty Loss: 0.13593, Diff Loss: 0.27903, DiscLM Loss: 0.00000, GenLM Loss: 0.98485 +INFO:2025-06-11 03:07:11,910: Epoch [3/20], Step [1500/2647], Loss: 0.54673, Disc Loss: 3.88071, Dur Loss: 1.20821, CE Loss: 0.06587, Norm Loss: 1.14343, F0 Loss: 2.98250, LM Loss: 2.19326, Gen Loss: 5.28938, Sty Loss: 0.11911, Diff Loss: 0.28412, DiscLM Loss: 0.03167, GenLM Loss: 0.95931 +INFO:2025-06-11 03:08:58,222: Epoch [3/20], Step [1550/2647], Loss: 0.54752, Disc Loss: 3.79224, Dur Loss: 1.16198, CE Loss: 0.06532, Norm Loss: 0.97427, F0 Loss: 2.96735, LM Loss: 2.26768, Gen Loss: 5.85022, Sty Loss: 0.15891, Diff Loss: 0.43313, DiscLM Loss: 0.00000, GenLM Loss: 1.03450 +INFO:2025-06-11 03:10:49,144: Epoch [3/20], Step [1600/2647], Loss: 0.55990, Disc Loss: 3.80611, Dur Loss: 0.93937, CE Loss: 0.04962, Norm Loss: 1.32402, F0 Loss: 2.86448, LM Loss: 2.21968, Gen Loss: 6.57136, Sty Loss: 0.12257, Diff Loss: 0.19604, DiscLM Loss: 0.00417, GenLM Loss: 0.97046 +INFO:2025-06-11 03:12:38,074: Epoch [3/20], Step [1650/2647], Loss: 0.55267, Disc Loss: 3.74949, Dur Loss: 0.97769, CE Loss: 0.06200, Norm Loss: 1.28467, F0 Loss: 3.62232, LM Loss: 2.40515, Gen Loss: 5.89718, Sty Loss: 0.12120, Diff Loss: 0.26526, DiscLM Loss: 0.00000, GenLM Loss: 0.95618 +INFO:2025-06-11 03:14:28,561: Epoch [3/20], Step [1700/2647], Loss: 0.54865, Disc Loss: 3.79968, Dur Loss: 0.79035, CE Loss: 0.04366, Norm Loss: 0.98468, F0 Loss: 2.68445, LM Loss: 2.22619, Gen Loss: 5.82823, Sty Loss: 0.11722, Diff Loss: 0.24814, DiscLM Loss: 0.00268, GenLM Loss: 1.00727 +INFO:2025-06-11 03:16:17,893: Epoch [3/20], Step [1750/2647], Loss: 0.54934, Disc Loss: 3.83302, Dur Loss: 1.14626, CE Loss: 0.06449, Norm Loss: 1.06486, F0 Loss: 3.27507, LM Loss: 2.23852, Gen Loss: 5.43590, Sty Loss: 0.12254, Diff Loss: 0.29925, DiscLM Loss: 0.00000, GenLM Loss: 1.03987 +INFO:2025-06-11 03:18:10,287: Epoch [3/20], Step [1800/2647], Loss: 0.55585, Disc Loss: 3.73197, Dur Loss: 1.06587, CE Loss: 0.06088, Norm Loss: 1.03593, F0 Loss: 3.46580, LM Loss: 2.20808, Gen Loss: 5.87438, Sty Loss: 0.12622, Diff Loss: 0.21122, DiscLM Loss: 0.00305, GenLM Loss: 0.98423 +INFO:2025-06-11 03:19:58,565: Epoch [3/20], Step [1850/2647], Loss: 0.54337, Disc Loss: 3.83296, Dur Loss: 1.00383, CE Loss: 0.06145, Norm Loss: 1.08092, F0 Loss: 2.95629, LM Loss: 2.13791, Gen Loss: 5.31911, Sty Loss: 0.12529, Diff Loss: 0.27439, DiscLM Loss: 0.00000, GenLM Loss: 0.96799 +INFO:2025-06-11 03:21:47,335: Epoch [3/20], Step [1900/2647], Loss: 0.54208, Disc Loss: 3.79596, Dur Loss: 0.90677, CE Loss: 0.05038, Norm Loss: 1.16335, F0 Loss: 3.18581, LM Loss: 2.26855, Gen Loss: 5.37355, Sty Loss: 0.13323, Diff Loss: 0.33141, DiscLM Loss: 0.00302, GenLM Loss: 1.01107 +INFO:2025-06-11 03:23:36,652: Epoch [3/20], Step [1950/2647], Loss: 0.56691, Disc Loss: 3.82353, Dur Loss: 1.37328, CE Loss: 0.07839, Norm Loss: 1.32571, F0 Loss: 3.30662, LM Loss: 2.30606, Gen Loss: 5.69693, Sty Loss: 0.12014, Diff Loss: 0.30136, DiscLM Loss: 0.00000, GenLM Loss: 0.98701 +INFO:2025-06-11 03:25:24,849: Epoch [3/20], Step [2000/2647], Loss: 0.54182, Disc Loss: 3.70987, Dur Loss: 1.10096, CE Loss: 0.06084, Norm Loss: 1.05389, F0 Loss: 3.03866, LM Loss: 2.14982, Gen Loss: 6.13849, Sty Loss: 0.13311, Diff Loss: 0.19965, DiscLM Loss: 0.00272, GenLM Loss: 0.98390 +INFO:2025-06-11 03:27:12,542: Epoch [3/20], Step [2050/2647], Loss: 0.56463, Disc Loss: 3.78225, Dur Loss: 1.25744, CE Loss: 0.06332, Norm Loss: 1.71586, F0 Loss: 4.32652, LM Loss: 2.32012, Gen Loss: 6.34164, Sty Loss: 0.15461, Diff Loss: 0.36496, DiscLM Loss: 0.00000, GenLM Loss: 0.98596 +INFO:2025-06-11 03:29:00,990: Epoch [3/20], Step [2100/2647], Loss: 0.54566, Disc Loss: 3.81030, Dur Loss: 0.93064, CE Loss: 0.04925, Norm Loss: 1.00868, F0 Loss: 3.11400, LM Loss: 2.21217, Gen Loss: 5.50252, Sty Loss: 0.11463, Diff Loss: 0.25602, DiscLM Loss: 0.00174, GenLM Loss: 1.00544 +INFO:2025-06-11 03:30:49,260: Epoch [3/20], Step [2150/2647], Loss: 0.54978, Disc Loss: 3.82838, Dur Loss: 0.94692, CE Loss: 0.05158, Norm Loss: 1.05384, F0 Loss: 3.21293, LM Loss: 2.23922, Gen Loss: 6.00973, Sty Loss: 0.11928, Diff Loss: 0.22537, DiscLM Loss: 0.00000, GenLM Loss: 1.00591 +INFO:2025-06-11 03:32:39,562: Epoch [3/20], Step [2200/2647], Loss: 0.55296, Disc Loss: 3.78503, Dur Loss: 1.09685, CE Loss: 0.06036, Norm Loss: 1.17392, F0 Loss: 2.77354, LM Loss: 2.24300, Gen Loss: 5.83349, Sty Loss: 0.10894, Diff Loss: 0.24503, DiscLM Loss: 0.00214, GenLM Loss: 0.99626 +INFO:2025-06-11 03:34:28,687: Epoch [3/20], Step [2250/2647], Loss: 0.54721, Disc Loss: 3.84035, Dur Loss: 1.08802, CE Loss: 0.06169, Norm Loss: 1.07619, F0 Loss: 3.15663, LM Loss: 2.26364, Gen Loss: 5.96818, Sty Loss: 0.13514, Diff Loss: 0.33584, DiscLM Loss: 0.00000, GenLM Loss: 0.99372 +INFO:2025-06-11 03:36:19,868: Epoch [3/20], Step [2300/2647], Loss: 0.56332, Disc Loss: 3.74100, Dur Loss: 1.48347, CE Loss: 0.07845, Norm Loss: 1.49415, F0 Loss: 3.89759, LM Loss: 2.33792, Gen Loss: 6.36293, Sty Loss: 0.12715, Diff Loss: 0.33488, DiscLM Loss: 0.00124, GenLM Loss: 0.99196 +INFO:2025-06-11 03:38:08,429: Epoch [3/20], Step [2350/2647], Loss: 0.54344, Disc Loss: 3.78733, Dur Loss: 1.07553, CE Loss: 0.05867, Norm Loss: 1.33687, F0 Loss: 2.83129, LM Loss: 2.26188, Gen Loss: 5.97520, Sty Loss: 0.12885, Diff Loss: 0.38915, DiscLM Loss: 0.00000, GenLM Loss: 1.00175 +INFO:2025-06-11 03:39:59,115: Epoch [3/20], Step [2400/2647], Loss: 0.53809, Disc Loss: 3.82548, Dur Loss: 1.10699, CE Loss: 0.05560, Norm Loss: 1.20159, F0 Loss: 3.89415, LM Loss: 2.24273, Gen Loss: 5.20928, Sty Loss: 0.12296, Diff Loss: 0.30502, DiscLM Loss: 0.00455, GenLM Loss: 0.98341 +INFO:2025-06-11 03:41:45,044: Epoch [3/20], Step [2450/2647], Loss: 0.53856, Disc Loss: 3.79202, Dur Loss: 0.98135, CE Loss: 0.05711, Norm Loss: 1.33362, F0 Loss: 3.88238, LM Loss: 2.24701, Gen Loss: 5.81726, Sty Loss: 0.14846, Diff Loss: 0.35967, DiscLM Loss: 0.00000, GenLM Loss: 0.97857 +INFO:2025-06-11 03:43:33,027: Epoch [3/20], Step [2500/2647], Loss: 0.54016, Disc Loss: 3.80535, Dur Loss: 1.03519, CE Loss: 0.05859, Norm Loss: 1.33022, F0 Loss: 4.31765, LM Loss: 2.35315, Gen Loss: 6.22721, Sty Loss: 0.14334, Diff Loss: 0.27575, DiscLM Loss: 0.00183, GenLM Loss: 1.00626 +INFO:2025-06-11 03:45:23,544: Epoch [3/20], Step [2550/2647], Loss: 0.54407, Disc Loss: 3.80236, Dur Loss: 1.15382, CE Loss: 0.07622, Norm Loss: 1.18777, F0 Loss: 4.35289, LM Loss: 2.18960, Gen Loss: 5.28477, Sty Loss: 0.15989, Diff Loss: 0.23380, DiscLM Loss: 0.00000, GenLM Loss: 1.00804 +INFO:2025-06-11 03:47:13,858: Epoch [3/20], Step [2600/2647], Loss: 0.55267, Disc Loss: 3.77739, Dur Loss: 0.97143, CE Loss: 0.04735, Norm Loss: 1.11965, F0 Loss: 3.35983, LM Loss: 2.19505, Gen Loss: 5.95682, Sty Loss: 0.09918, Diff Loss: 0.24308, DiscLM Loss: 0.00194, GenLM Loss: 0.98533 +INFO:2025-06-11 03:51:15,310: Validation loss: 0.568, Dur loss: 1.100, F0 loss: 3.586 + + + +INFO:2025-06-11 03:53:06,732: Epoch [4/20], Step [50/2647], Loss: 0.54797, Disc Loss: 3.73915, Dur Loss: 0.81316, CE Loss: 0.04335, Norm Loss: 0.81716, F0 Loss: 3.28349, LM Loss: 2.10930, Gen Loss: 6.23630, Sty Loss: 0.11181, Diff Loss: 0.29591, DiscLM Loss: 0.00000, GenLM Loss: 0.99057 +INFO:2025-06-11 03:54:57,669: Epoch [4/20], Step [100/2647], Loss: 0.55378, Disc Loss: 3.82495, Dur Loss: 1.36386, CE Loss: 0.07648, Norm Loss: 1.43716, F0 Loss: 3.66591, LM Loss: 2.31515, Gen Loss: 5.16973, Sty Loss: 0.14488, Diff Loss: 0.38957, DiscLM Loss: 0.00243, GenLM Loss: 0.98108 +INFO:2025-06-11 03:56:46,530: Epoch [4/20], Step [150/2647], Loss: 0.54165, Disc Loss: 3.72688, Dur Loss: 0.93236, CE Loss: 0.04911, Norm Loss: 1.38544, F0 Loss: 3.57424, LM Loss: 2.26009, Gen Loss: 6.18594, Sty Loss: 0.12384, Diff Loss: 0.26616, DiscLM Loss: 0.00000, GenLM Loss: 0.99579 +INFO:2025-06-11 03:58:34,926: Epoch [4/20], Step [200/2647], Loss: 0.54929, Disc Loss: 3.75587, Dur Loss: 0.98017, CE Loss: 0.05764, Norm Loss: 1.00657, F0 Loss: 3.77600, LM Loss: 2.33130, Gen Loss: 6.97970, Sty Loss: 0.14410, Diff Loss: 0.18884, DiscLM Loss: 0.00233, GenLM Loss: 1.00481 +INFO:2025-06-11 04:00:22,763: Epoch [4/20], Step [250/2647], Loss: 0.54486, Disc Loss: 3.70760, Dur Loss: 1.05699, CE Loss: 0.06244, Norm Loss: 1.54042, F0 Loss: 3.53339, LM Loss: 2.28950, Gen Loss: 5.96082, Sty Loss: 0.12015, Diff Loss: 0.31363, DiscLM Loss: 0.00000, GenLM Loss: 1.00614 +INFO:2025-06-11 04:02:11,420: Epoch [4/20], Step [300/2647], Loss: 0.55553, Disc Loss: 3.79089, Dur Loss: 0.79137, CE Loss: 0.04609, Norm Loss: 1.05764, F0 Loss: 2.76642, LM Loss: 2.25130, Gen Loss: 5.86828, Sty Loss: 0.11782, Diff Loss: 0.24508, DiscLM Loss: 0.00225, GenLM Loss: 0.98811 +INFO:2025-06-11 04:03:57,331: Epoch [4/20], Step [350/2647], Loss: 0.53785, Disc Loss: 3.82228, Dur Loss: 1.05612, CE Loss: 0.05579, Norm Loss: 1.30473, F0 Loss: 3.24657, LM Loss: 2.22592, Gen Loss: 6.45432, Sty Loss: 0.10769, Diff Loss: 0.31922, DiscLM Loss: 0.00000, GenLM Loss: 1.00037 +INFO:2025-06-11 04:05:45,508: Epoch [4/20], Step [400/2647], Loss: 0.54451, Disc Loss: 3.86480, Dur Loss: 0.98617, CE Loss: 0.04966, Norm Loss: 0.99363, F0 Loss: 3.33191, LM Loss: 2.22243, Gen Loss: 5.35754, Sty Loss: 0.11112, Diff Loss: 0.24239, DiscLM Loss: 0.00267, GenLM Loss: 0.99916 +INFO:2025-06-11 04:07:37,578: Epoch [4/20], Step [450/2647], Loss: 0.54630, Disc Loss: 3.69522, Dur Loss: 1.27364, CE Loss: 0.07001, Norm Loss: 1.63517, F0 Loss: 3.90797, LM Loss: 2.31845, Gen Loss: 6.17666, Sty Loss: 0.14613, Diff Loss: 0.25111, DiscLM Loss: 0.00000, GenLM Loss: 1.00406 +INFO:2025-06-11 04:09:27,343: Epoch [4/20], Step [500/2647], Loss: 0.54414, Disc Loss: 3.73573, Dur Loss: 1.12839, CE Loss: 0.06322, Norm Loss: 1.19184, F0 Loss: 3.55610, LM Loss: 2.27781, Gen Loss: 6.03450, Sty Loss: 0.14725, Diff Loss: 0.29478, DiscLM Loss: 0.00174, GenLM Loss: 0.99546 +INFO:2025-06-11 04:11:17,238: Epoch [4/20], Step [550/2647], Loss: 0.56262, Disc Loss: 3.71184, Dur Loss: 1.03371, CE Loss: 0.05284, Norm Loss: 1.24805, F0 Loss: 3.04783, LM Loss: 2.28106, Gen Loss: 6.06963, Sty Loss: 0.13599, Diff Loss: 0.31706, DiscLM Loss: 0.00000, GenLM Loss: 0.99704 +INFO:2025-06-11 04:13:04,162: Epoch [4/20], Step [600/2647], Loss: 0.55049, Disc Loss: 3.71238, Dur Loss: 1.03288, CE Loss: 0.05538, Norm Loss: 1.13053, F0 Loss: 3.72598, LM Loss: 2.25225, Gen Loss: 6.30487, Sty Loss: 0.13482, Diff Loss: 0.30850, DiscLM Loss: 0.00150, GenLM Loss: 0.99000 +INFO:2025-06-11 04:14:51,858: Epoch [4/20], Step [650/2647], Loss: 0.55335, Disc Loss: 3.76557, Dur Loss: 1.05257, CE Loss: 0.05694, Norm Loss: 1.20168, F0 Loss: 3.01602, LM Loss: 2.22685, Gen Loss: 5.47819, Sty Loss: 0.14288, Diff Loss: 0.43769, DiscLM Loss: 0.00000, GenLM Loss: 1.00670 +INFO:2025-06-11 04:16:38,663: Epoch [4/20], Step [700/2647], Loss: 0.55112, Disc Loss: 3.75076, Dur Loss: 0.99355, CE Loss: 0.05614, Norm Loss: 0.97793, F0 Loss: 3.23017, LM Loss: 2.18306, Gen Loss: 6.10682, Sty Loss: 0.11103, Diff Loss: 0.26672, DiscLM Loss: 0.00091, GenLM Loss: 1.00169 +INFO:2025-06-11 04:18:26,025: Epoch [4/20], Step [750/2647], Loss: 0.55536, Disc Loss: 3.82135, Dur Loss: 0.95103, CE Loss: 0.04943, Norm Loss: 0.86966, F0 Loss: 4.19975, LM Loss: 2.15283, Gen Loss: 6.00525, Sty Loss: 0.12755, Diff Loss: 0.27480, DiscLM Loss: 0.00000, GenLM Loss: 1.00006 +INFO:2025-06-11 04:20:14,264: Epoch [4/20], Step [800/2647], Loss: 0.54691, Disc Loss: 3.71251, Dur Loss: 1.08012, CE Loss: 0.06092, Norm Loss: 1.19125, F0 Loss: 2.72295, LM Loss: 2.27261, Gen Loss: 6.07689, Sty Loss: 0.12631, Diff Loss: 0.35677, DiscLM Loss: 0.00117, GenLM Loss: 1.00970 +INFO:2025-06-11 04:22:03,543: Epoch [4/20], Step [850/2647], Loss: 0.55172, Disc Loss: 3.87173, Dur Loss: 0.84799, CE Loss: 0.04503, Norm Loss: 1.30381, F0 Loss: 2.90561, LM Loss: 2.18174, Gen Loss: 5.88949, Sty Loss: 0.11090, Diff Loss: 0.33777, DiscLM Loss: 0.00000, GenLM Loss: 0.98709 +INFO:2025-06-11 04:23:52,227: Epoch [4/20], Step [900/2647], Loss: 0.55170, Disc Loss: 3.74848, Dur Loss: 0.91416, CE Loss: 0.05118, Norm Loss: 1.11664, F0 Loss: 4.12050, LM Loss: 2.25910, Gen Loss: 6.18688, Sty Loss: 0.13439, Diff Loss: 0.26950, DiscLM Loss: 0.04129, GenLM Loss: 1.00854 +INFO:2025-06-11 04:25:42,267: Epoch [4/20], Step [950/2647], Loss: 0.53657, Disc Loss: 3.72899, Dur Loss: 0.98657, CE Loss: 0.05251, Norm Loss: 1.00277, F0 Loss: 3.87458, LM Loss: 2.24786, Gen Loss: 5.51950, Sty Loss: 0.11595, Diff Loss: 0.19189, DiscLM Loss: 0.00000, GenLM Loss: 0.80921 +INFO:2025-06-11 04:27:32,873: Epoch [4/20], Step [1000/2647], Loss: 0.54665, Disc Loss: 3.79481, Dur Loss: 0.98282, CE Loss: 0.05541, Norm Loss: 0.95978, F0 Loss: 2.60073, LM Loss: 2.16843, Gen Loss: 6.15912, Sty Loss: 0.12163, Diff Loss: 0.23303, DiscLM Loss: 0.04295, GenLM Loss: 1.08199 +INFO:2025-06-11 04:29:21,635: Epoch [4/20], Step [1050/2647], Loss: 0.54199, Disc Loss: 3.75226, Dur Loss: 1.18961, CE Loss: 0.07253, Norm Loss: 0.97613, F0 Loss: 3.43700, LM Loss: 2.30638, Gen Loss: 6.64394, Sty Loss: 0.14143, Diff Loss: 0.24404, DiscLM Loss: 0.00000, GenLM Loss: 1.01395 +INFO:2025-06-11 04:31:08,539: Epoch [4/20], Step [1100/2647], Loss: 0.54030, Disc Loss: 3.95795, Dur Loss: 1.00398, CE Loss: 0.05662, Norm Loss: 1.51919, F0 Loss: 3.42364, LM Loss: 2.40352, Gen Loss: 6.12959, Sty Loss: 0.11860, Diff Loss: 0.20824, DiscLM Loss: 0.02162, GenLM Loss: 0.91731 +INFO:2025-06-11 04:33:00,146: Epoch [4/20], Step [1150/2647], Loss: 0.53881, Disc Loss: 3.79952, Dur Loss: 1.12433, CE Loss: 0.06558, Norm Loss: 1.24991, F0 Loss: 3.35716, LM Loss: 2.15402, Gen Loss: 5.91607, Sty Loss: 0.13044, Diff Loss: 0.38931, DiscLM Loss: 0.00000, GenLM Loss: 0.98505 +INFO:2025-06-11 04:34:49,442: Epoch [4/20], Step [1200/2647], Loss: 0.55052, Disc Loss: 3.84526, Dur Loss: 1.19648, CE Loss: 0.06720, Norm Loss: 1.07357, F0 Loss: 3.44003, LM Loss: 2.25199, Gen Loss: 4.95437, Sty Loss: 0.15286, Diff Loss: 0.33088, DiscLM Loss: 0.00527, GenLM Loss: 1.02620 +INFO:2025-06-11 04:36:38,144: Epoch [4/20], Step [1250/2647], Loss: 0.54315, Disc Loss: 3.84298, Dur Loss: 0.96696, CE Loss: 0.05392, Norm Loss: 0.91833, F0 Loss: 3.03036, LM Loss: 2.17655, Gen Loss: 5.84037, Sty Loss: 0.12660, Diff Loss: 0.25785, DiscLM Loss: 0.00000, GenLM Loss: 1.01540 +INFO:2025-06-11 04:38:29,288: Epoch [4/20], Step [1300/2647], Loss: 0.54209, Disc Loss: 3.82975, Dur Loss: 0.99451, CE Loss: 0.05365, Norm Loss: 0.92731, F0 Loss: 3.45715, LM Loss: 2.23013, Gen Loss: 6.44346, Sty Loss: 0.11404, Diff Loss: 0.36408, DiscLM Loss: 0.00541, GenLM Loss: 0.99058 +INFO:2025-06-11 04:40:18,316: Epoch [4/20], Step [1350/2647], Loss: 0.54048, Disc Loss: 3.76939, Dur Loss: 0.91543, CE Loss: 0.04880, Norm Loss: 0.88789, F0 Loss: 2.50542, LM Loss: 2.17605, Gen Loss: 6.22761, Sty Loss: 0.11452, Diff Loss: 0.29974, DiscLM Loss: 0.00000, GenLM Loss: 0.98773 +INFO:2025-06-11 04:42:05,589: Epoch [4/20], Step [1400/2647], Loss: 0.54760, Disc Loss: 3.74385, Dur Loss: 1.21229, CE Loss: 0.07599, Norm Loss: 1.34528, F0 Loss: 3.02909, LM Loss: 2.40194, Gen Loss: 6.06253, Sty Loss: 0.12964, Diff Loss: 0.33928, DiscLM Loss: 0.00678, GenLM Loss: 0.96261 +INFO:2025-06-11 04:43:54,580: Epoch [4/20], Step [1450/2647], Loss: 0.54463, Disc Loss: 3.83318, Dur Loss: 1.05404, CE Loss: 0.05578, Norm Loss: 1.42366, F0 Loss: 2.63290, LM Loss: 2.14612, Gen Loss: 5.55874, Sty Loss: 0.10319, Diff Loss: 0.34021, DiscLM Loss: 0.00000, GenLM Loss: 0.97351 +INFO:2025-06-11 04:45:42,873: Epoch [4/20], Step [1500/2647], Loss: 0.54632, Disc Loss: 3.69954, Dur Loss: 1.06347, CE Loss: 0.05746, Norm Loss: 1.20473, F0 Loss: 2.81327, LM Loss: 2.23047, Gen Loss: 6.05170, Sty Loss: 0.14315, Diff Loss: 0.18758, DiscLM Loss: 0.00447, GenLM Loss: 0.99608 +INFO:2025-06-11 04:47:31,510: Epoch [4/20], Step [1550/2647], Loss: 0.53499, Disc Loss: 3.80301, Dur Loss: 0.87437, CE Loss: 0.04684, Norm Loss: 0.97317, F0 Loss: 3.25978, LM Loss: 2.28280, Gen Loss: 5.68796, Sty Loss: 0.11020, Diff Loss: 0.28879, DiscLM Loss: 0.00000, GenLM Loss: 0.99019 +INFO:2025-06-11 04:49:20,871: Epoch [4/20], Step [1600/2647], Loss: 0.54641, Disc Loss: 3.82697, Dur Loss: 1.11595, CE Loss: 0.06061, Norm Loss: 1.45468, F0 Loss: 3.06132, LM Loss: 2.26116, Gen Loss: 5.57898, Sty Loss: 0.10590, Diff Loss: 0.20528, DiscLM Loss: 0.00156, GenLM Loss: 0.99942 +INFO:2025-06-11 04:51:12,676: Epoch [4/20], Step [1650/2647], Loss: 0.54537, Disc Loss: 3.75519, Dur Loss: 1.08333, CE Loss: 0.06107, Norm Loss: 1.02180, F0 Loss: 2.83754, LM Loss: 2.18865, Gen Loss: 5.89812, Sty Loss: 0.12375, Diff Loss: 0.25459, DiscLM Loss: 0.00000, GenLM Loss: 1.00146 +INFO:2025-06-11 04:53:02,502: Epoch [4/20], Step [1700/2647], Loss: 0.54640, Disc Loss: 3.74257, Dur Loss: 1.63180, CE Loss: 0.09175, Norm Loss: 1.49429, F0 Loss: 3.87718, LM Loss: 2.38786, Gen Loss: 6.35189, Sty Loss: 0.13950, Diff Loss: 0.23179, DiscLM Loss: 0.00134, GenLM Loss: 0.99900 +INFO:2025-06-11 04:54:48,661: Epoch [4/20], Step [1750/2647], Loss: 0.54361, Disc Loss: 3.73654, Dur Loss: 0.84565, CE Loss: 0.04853, Norm Loss: 0.92630, F0 Loss: 2.53537, LM Loss: 2.18326, Gen Loss: 6.33246, Sty Loss: 0.14004, Diff Loss: 0.27119, DiscLM Loss: 0.00000, GenLM Loss: 0.99086 +INFO:2025-06-11 04:56:37,538: Epoch [4/20], Step [1800/2647], Loss: 0.53965, Disc Loss: 3.67636, Dur Loss: 1.11843, CE Loss: 0.06131, Norm Loss: 1.41174, F0 Loss: 4.48281, LM Loss: 2.16811, Gen Loss: 6.25507, Sty Loss: 0.12373, Diff Loss: 0.26681, DiscLM Loss: 0.00099, GenLM Loss: 0.99752 +INFO:2025-06-11 04:58:24,470: Epoch [4/20], Step [1850/2647], Loss: 0.54173, Disc Loss: 3.73203, Dur Loss: 1.09778, CE Loss: 0.06381, Norm Loss: 1.09382, F0 Loss: 3.47473, LM Loss: 2.23490, Gen Loss: 5.61410, Sty Loss: 0.11398, Diff Loss: 0.21905, DiscLM Loss: 0.00000, GenLM Loss: 1.00900 +INFO:2025-06-11 05:00:12,647: Epoch [4/20], Step [1900/2647], Loss: 0.54843, Disc Loss: 3.82111, Dur Loss: 1.13029, CE Loss: 0.06385, Norm Loss: 1.10421, F0 Loss: 3.31739, LM Loss: 2.27248, Gen Loss: 5.60212, Sty Loss: 0.11491, Diff Loss: 0.28333, DiscLM Loss: 0.00096, GenLM Loss: 1.00075 +INFO:2025-06-11 05:02:04,684: Epoch [4/20], Step [1950/2647], Loss: 0.54459, Disc Loss: 3.81770, Dur Loss: 1.10183, CE Loss: 0.05979, Norm Loss: 1.13386, F0 Loss: 3.76188, LM Loss: 2.24478, Gen Loss: 6.19476, Sty Loss: 0.12907, Diff Loss: 0.26062, DiscLM Loss: 0.00000, GenLM Loss: 0.99671 +INFO:2025-06-11 05:03:52,197: Epoch [4/20], Step [2000/2647], Loss: 0.53789, Disc Loss: 3.74910, Dur Loss: 1.00845, CE Loss: 0.05517, Norm Loss: 1.06168, F0 Loss: 3.55443, LM Loss: 2.25611, Gen Loss: 6.12405, Sty Loss: 0.12231, Diff Loss: 0.21815, DiscLM Loss: 0.00091, GenLM Loss: 0.99970 +INFO:2025-06-11 05:05:41,080: Epoch [4/20], Step [2050/2647], Loss: 0.54185, Disc Loss: 3.70801, Dur Loss: 0.80584, CE Loss: 0.04633, Norm Loss: 1.09930, F0 Loss: 3.29653, LM Loss: 2.28425, Gen Loss: 5.92864, Sty Loss: 0.13437, Diff Loss: 0.42109, DiscLM Loss: 0.00000, GenLM Loss: 0.97505 +INFO:2025-06-11 05:07:29,854: Epoch [4/20], Step [2100/2647], Loss: 0.55023, Disc Loss: 3.72150, Dur Loss: 1.08014, CE Loss: 0.05867, Norm Loss: 1.26316, F0 Loss: 2.97170, LM Loss: 2.21401, Gen Loss: 5.90165, Sty Loss: 0.11781, Diff Loss: 0.19606, DiscLM Loss: 0.00099, GenLM Loss: 1.00278 +INFO:2025-06-11 05:09:20,937: Epoch [4/20], Step [2150/2647], Loss: 0.53558, Disc Loss: 3.71650, Dur Loss: 1.06254, CE Loss: 0.05991, Norm Loss: 1.03151, F0 Loss: 3.86988, LM Loss: 2.22348, Gen Loss: 6.33039, Sty Loss: 0.09641, Diff Loss: 0.30964, DiscLM Loss: 0.00000, GenLM Loss: 1.00162 +INFO:2025-06-11 05:11:10,824: Epoch [4/20], Step [2200/2647], Loss: 0.53874, Disc Loss: 3.75627, Dur Loss: 1.08751, CE Loss: 0.05996, Norm Loss: 0.98060, F0 Loss: 3.62160, LM Loss: 2.18406, Gen Loss: 5.75817, Sty Loss: 0.12184, Diff Loss: 0.23073, DiscLM Loss: 0.00132, GenLM Loss: 0.99645 +INFO:2025-06-11 05:12:59,163: Epoch [4/20], Step [2250/2647], Loss: 0.53200, Disc Loss: 3.88545, Dur Loss: 1.15711, CE Loss: 0.06624, Norm Loss: 1.00718, F0 Loss: 3.50137, LM Loss: 2.22748, Gen Loss: 5.39056, Sty Loss: 0.11257, Diff Loss: 0.16199, DiscLM Loss: 0.00000, GenLM Loss: 1.00168 +INFO:2025-06-11 05:14:48,342: Epoch [4/20], Step [2300/2647], Loss: 0.54328, Disc Loss: 3.80571, Dur Loss: 1.03046, CE Loss: 0.06303, Norm Loss: 1.39903, F0 Loss: 3.92478, LM Loss: 2.29408, Gen Loss: 6.18720, Sty Loss: 0.11774, Diff Loss: 0.26986, DiscLM Loss: 0.00140, GenLM Loss: 0.99047 +INFO:2025-06-11 05:16:36,710: Epoch [4/20], Step [2350/2647], Loss: 0.53968, Disc Loss: 3.81982, Dur Loss: 1.10068, CE Loss: 0.05774, Norm Loss: 1.16935, F0 Loss: 3.34761, LM Loss: 2.10971, Gen Loss: 5.65810, Sty Loss: 0.12850, Diff Loss: 0.28717, DiscLM Loss: 0.00000, GenLM Loss: 0.99986 +INFO:2025-06-11 05:18:25,182: Epoch [4/20], Step [2400/2647], Loss: 0.54291, Disc Loss: 3.72265, Dur Loss: 0.90312, CE Loss: 0.05218, Norm Loss: 1.18468, F0 Loss: 3.29547, LM Loss: 2.24150, Gen Loss: 5.96925, Sty Loss: 0.13606, Diff Loss: 0.21066, DiscLM Loss: 0.00094, GenLM Loss: 1.00953 +INFO:2025-06-11 05:20:13,752: Epoch [4/20], Step [2450/2647], Loss: 0.54148, Disc Loss: 3.71902, Dur Loss: 0.92178, CE Loss: 0.04722, Norm Loss: 1.20730, F0 Loss: 3.16251, LM Loss: 2.17607, Gen Loss: 6.24093, Sty Loss: 0.12280, Diff Loss: 0.23636, DiscLM Loss: 0.00000, GenLM Loss: 0.99869 +INFO:2025-06-11 05:22:02,415: Epoch [4/20], Step [2500/2647], Loss: 0.54723, Disc Loss: 3.71402, Dur Loss: 0.97070, CE Loss: 0.05450, Norm Loss: 1.31777, F0 Loss: 2.89750, LM Loss: 2.28485, Gen Loss: 6.35604, Sty Loss: 0.11961, Diff Loss: 0.38758, DiscLM Loss: 0.03528, GenLM Loss: 0.90107 +INFO:2025-06-11 05:23:53,787: Epoch [4/20], Step [2550/2647], Loss: 0.53617, Disc Loss: 3.73913, Dur Loss: 0.99235, CE Loss: 0.05118, Norm Loss: 0.93283, F0 Loss: 3.28269, LM Loss: 2.14267, Gen Loss: 5.56781, Sty Loss: 0.11996, Diff Loss: 0.29841, DiscLM Loss: 0.00000, GenLM Loss: 1.01902 +INFO:2025-06-11 05:25:43,338: Epoch [4/20], Step [2600/2647], Loss: 0.54355, Disc Loss: 3.83134, Dur Loss: 1.22644, CE Loss: 0.06546, Norm Loss: 1.17534, F0 Loss: 3.70416, LM Loss: 2.23345, Gen Loss: 5.81797, Sty Loss: 0.14837, Diff Loss: 0.27634, DiscLM Loss: 0.00234, GenLM Loss: 0.99044 +INFO:2025-06-11 05:29:44,326: Validation loss: 0.556, Dur loss: 1.065, F0 loss: 3.605 + + + +INFO:2025-06-11 05:31:37,050: Epoch [5/20], Step [50/2647], Loss: 0.52968, Disc Loss: 3.79236, Dur Loss: 1.04705, CE Loss: 0.05537, Norm Loss: 1.36570, F0 Loss: 3.28027, LM Loss: 2.15568, Gen Loss: 5.29051, Sty Loss: 0.11486, Diff Loss: 0.20350, DiscLM Loss: 0.00000, GenLM Loss: 1.00418 +INFO:2025-06-11 05:33:27,122: Epoch [5/20], Step [100/2647], Loss: 0.54586, Disc Loss: 3.75091, Dur Loss: 1.29702, CE Loss: 0.07460, Norm Loss: 1.14135, F0 Loss: 2.48134, LM Loss: 2.20483, Gen Loss: 5.81949, Sty Loss: 0.12746, Diff Loss: 0.27405, DiscLM Loss: 0.00601, GenLM Loss: 0.99193 +INFO:2025-06-11 05:35:15,971: Epoch [5/20], Step [150/2647], Loss: 0.52965, Disc Loss: 3.74895, Dur Loss: 1.00825, CE Loss: 0.05100, Norm Loss: 0.98949, F0 Loss: 2.45608, LM Loss: 2.18374, Gen Loss: 5.78370, Sty Loss: 0.13692, Diff Loss: 0.19973, DiscLM Loss: 0.00000, GenLM Loss: 0.98436 +INFO:2025-06-11 05:37:09,876: Epoch [5/20], Step [200/2647], Loss: 0.54313, Disc Loss: 3.78463, Dur Loss: 1.04913, CE Loss: 0.06976, Norm Loss: 1.12117, F0 Loss: 3.13175, LM Loss: 2.19514, Gen Loss: 5.01472, Sty Loss: 0.13028, Diff Loss: 0.33882, DiscLM Loss: 0.00151, GenLM Loss: 1.00045 +INFO:2025-06-11 05:38:59,862: Epoch [5/20], Step [250/2647], Loss: 0.54142, Disc Loss: 3.75177, Dur Loss: 0.88823, CE Loss: 0.04346, Norm Loss: 0.93881, F0 Loss: 2.91240, LM Loss: 2.27450, Gen Loss: 6.17555, Sty Loss: 0.11340, Diff Loss: 0.30096, DiscLM Loss: 0.00000, GenLM Loss: 1.02118 +INFO:2025-06-11 05:40:49,493: Epoch [5/20], Step [300/2647], Loss: 0.54459, Disc Loss: 3.71064, Dur Loss: 1.04471, CE Loss: 0.05727, Norm Loss: 1.05580, F0 Loss: 3.48903, LM Loss: 2.23098, Gen Loss: 6.16871, Sty Loss: 0.11206, Diff Loss: 0.21475, DiscLM Loss: 0.00392, GenLM Loss: 1.00597 +INFO:2025-06-11 05:42:38,157: Epoch [5/20], Step [350/2647], Loss: 0.54232, Disc Loss: 3.77387, Dur Loss: 1.10749, CE Loss: 0.06787, Norm Loss: 1.26597, F0 Loss: 3.46836, LM Loss: 2.33181, Gen Loss: 6.16971, Sty Loss: 0.14208, Diff Loss: 0.25473, DiscLM Loss: 0.00000, GenLM Loss: 0.99066 +INFO:2025-06-11 05:44:25,554: Epoch [5/20], Step [400/2647], Loss: 0.55092, Disc Loss: 3.84630, Dur Loss: 1.32518, CE Loss: 0.07941, Norm Loss: 0.76643, F0 Loss: 2.94955, LM Loss: 2.30473, Gen Loss: 6.57199, Sty Loss: 0.14073, Diff Loss: 0.25440, DiscLM Loss: 0.00126, GenLM Loss: 0.97908 +INFO:2025-06-11 05:46:14,336: Epoch [5/20], Step [450/2647], Loss: 0.54621, Disc Loss: 3.85976, Dur Loss: 0.96968, CE Loss: 0.05220, Norm Loss: 1.58975, F0 Loss: 4.21639, LM Loss: 2.31106, Gen Loss: 5.77161, Sty Loss: 0.15366, Diff Loss: 0.22905, DiscLM Loss: 0.00000, GenLM Loss: 0.99783 +INFO:2025-06-11 05:48:01,961: Epoch [5/20], Step [500/2647], Loss: 0.54585, Disc Loss: 3.79424, Dur Loss: 0.93891, CE Loss: 0.05089, Norm Loss: 1.08298, F0 Loss: 2.77977, LM Loss: 2.24613, Gen Loss: 6.13445, Sty Loss: 0.12854, Diff Loss: 0.29729, DiscLM Loss: 0.00095, GenLM Loss: 1.01590 +INFO:2025-06-11 05:49:51,348: Epoch [5/20], Step [550/2647], Loss: 0.53667, Disc Loss: 3.69810, Dur Loss: 0.85307, CE Loss: 0.04555, Norm Loss: 1.20421, F0 Loss: 4.33564, LM Loss: 2.21430, Gen Loss: 5.98989, Sty Loss: 0.15654, Diff Loss: 0.24469, DiscLM Loss: 0.00000, GenLM Loss: 1.00029 +INFO:2025-06-11 05:51:40,748: Epoch [5/20], Step [600/2647], Loss: 0.54642, Disc Loss: 3.74013, Dur Loss: 0.87124, CE Loss: 0.04594, Norm Loss: 0.98156, F0 Loss: 3.06004, LM Loss: 2.18041, Gen Loss: 6.12929, Sty Loss: 0.14036, Diff Loss: 0.39155, DiscLM Loss: 0.00178, GenLM Loss: 0.99406 +INFO:2025-06-11 05:53:30,424: Epoch [5/20], Step [650/2647], Loss: 0.54382, Disc Loss: 3.85001, Dur Loss: 1.02313, CE Loss: 0.05940, Norm Loss: 0.99890, F0 Loss: 3.04021, LM Loss: 2.31772, Gen Loss: 5.54534, Sty Loss: 0.12705, Diff Loss: 0.23124, DiscLM Loss: 0.00000, GenLM Loss: 1.00131 +INFO:2025-06-11 05:55:19,655: Epoch [5/20], Step [700/2647], Loss: 0.54043, Disc Loss: 3.83904, Dur Loss: 1.32423, CE Loss: 0.07261, Norm Loss: 1.18042, F0 Loss: 2.94463, LM Loss: 2.18229, Gen Loss: 5.35243, Sty Loss: 0.11703, Diff Loss: 0.28311, DiscLM Loss: 0.00404, GenLM Loss: 1.01193 +INFO:2025-06-11 05:57:08,547: Epoch [5/20], Step [750/2647], Loss: 0.55082, Disc Loss: 3.81409, Dur Loss: 0.94726, CE Loss: 0.04941, Norm Loss: 1.07637, F0 Loss: 3.13539, LM Loss: 2.20044, Gen Loss: 5.63708, Sty Loss: 0.11152, Diff Loss: 0.37034, DiscLM Loss: 0.00000, GenLM Loss: 0.99014 +INFO:2025-06-11 05:58:57,977: Epoch [5/20], Step [800/2647], Loss: 0.54093, Disc Loss: 3.77480, Dur Loss: 1.03989, CE Loss: 0.06038, Norm Loss: 0.87646, F0 Loss: 3.51947, LM Loss: 2.13573, Gen Loss: 5.88369, Sty Loss: 0.12813, Diff Loss: 0.27732, DiscLM Loss: 0.00104, GenLM Loss: 0.99513 +INFO:2025-06-11 06:00:46,327: Epoch [5/20], Step [850/2647], Loss: 0.53447, Disc Loss: 3.82507, Dur Loss: 0.88877, CE Loss: 0.05005, Norm Loss: 1.00712, F0 Loss: 3.50083, LM Loss: 2.27304, Gen Loss: 5.13675, Sty Loss: 0.11714, Diff Loss: 0.28389, DiscLM Loss: 0.00000, GenLM Loss: 1.00636 +INFO:2025-06-11 06:02:38,553: Epoch [5/20], Step [900/2647], Loss: 0.53702, Disc Loss: 3.74772, Dur Loss: 1.28693, CE Loss: 0.06757, Norm Loss: 0.79753, F0 Loss: 2.26073, LM Loss: 2.08942, Gen Loss: 5.50879, Sty Loss: 0.14151, Diff Loss: 0.22997, DiscLM Loss: 0.00131, GenLM Loss: 0.99651 +INFO:2025-06-11 06:04:28,119: Epoch [5/20], Step [950/2647], Loss: 0.53536, Disc Loss: 3.77932, Dur Loss: 0.76395, CE Loss: 0.03797, Norm Loss: 1.01766, F0 Loss: 3.45553, LM Loss: 2.19888, Gen Loss: 6.79509, Sty Loss: 0.10850, Diff Loss: 0.23163, DiscLM Loss: 0.00000, GenLM Loss: 0.98701 +INFO:2025-06-11 06:06:16,824: Epoch [5/20], Step [1000/2647], Loss: 0.53683, Disc Loss: 3.79756, Dur Loss: 1.00234, CE Loss: 0.05351, Norm Loss: 1.27587, F0 Loss: 2.84793, LM Loss: 2.19381, Gen Loss: 5.22794, Sty Loss: 0.12857, Diff Loss: 0.31474, DiscLM Loss: 0.00634, GenLM Loss: 0.99346 +INFO:2025-06-11 06:08:07,154: Epoch [5/20], Step [1050/2647], Loss: 0.52788, Disc Loss: 3.75714, Dur Loss: 1.07980, CE Loss: 0.05817, Norm Loss: 1.38636, F0 Loss: 3.62797, LM Loss: 2.16181, Gen Loss: 5.17870, Sty Loss: 0.12440, Diff Loss: 0.22065, DiscLM Loss: 0.00000, GenLM Loss: 0.99342 +INFO:2025-06-11 06:09:54,067: Epoch [5/20], Step [1100/2647], Loss: 0.54556, Disc Loss: 3.84310, Dur Loss: 1.29502, CE Loss: 0.07129, Norm Loss: 1.14865, F0 Loss: 3.23472, LM Loss: 2.24283, Gen Loss: 5.23616, Sty Loss: 0.11256, Diff Loss: 0.24425, DiscLM Loss: 0.00125, GenLM Loss: 0.99975 +INFO:2025-06-11 06:11:42,220: Epoch [5/20], Step [1150/2647], Loss: 0.53787, Disc Loss: 3.87424, Dur Loss: 1.13351, CE Loss: 0.06319, Norm Loss: 1.04389, F0 Loss: 3.15452, LM Loss: 2.23614, Gen Loss: 6.68355, Sty Loss: 0.14245, Diff Loss: 0.25303, DiscLM Loss: 0.00000, GenLM Loss: 0.96349 +INFO:2025-06-11 06:13:31,653: Epoch [5/20], Step [1200/2647], Loss: 0.54556, Disc Loss: 3.71108, Dur Loss: 1.17089, CE Loss: 0.06097, Norm Loss: 1.21254, F0 Loss: 3.48944, LM Loss: 2.18825, Gen Loss: 6.02571, Sty Loss: 0.09680, Diff Loss: 0.23713, DiscLM Loss: 0.00155, GenLM Loss: 0.98899 +INFO:2025-06-11 06:15:23,764: Epoch [5/20], Step [1250/2647], Loss: 0.53932, Disc Loss: 3.65335, Dur Loss: 0.76638, CE Loss: 0.04186, Norm Loss: 1.00985, F0 Loss: 3.24668, LM Loss: 2.15032, Gen Loss: 6.01190, Sty Loss: 0.14413, Diff Loss: 0.36240, DiscLM Loss: 0.00000, GenLM Loss: 1.00370 +INFO:2025-06-11 06:17:11,339: Epoch [5/20], Step [1300/2647], Loss: 0.54638, Disc Loss: 3.72806, Dur Loss: 1.00393, CE Loss: 0.05645, Norm Loss: 0.97390, F0 Loss: 3.09106, LM Loss: 2.18683, Gen Loss: 5.95672, Sty Loss: 0.14177, Diff Loss: 0.30790, DiscLM Loss: 0.00060, GenLM Loss: 1.00869 +INFO:2025-06-11 06:18:59,573: Epoch [5/20], Step [1350/2647], Loss: 0.54732, Disc Loss: 3.75682, Dur Loss: 1.02947, CE Loss: 0.05731, Norm Loss: 1.37634, F0 Loss: 3.16354, LM Loss: 2.20697, Gen Loss: 6.15801, Sty Loss: 0.10279, Diff Loss: 0.32542, DiscLM Loss: 0.00000, GenLM Loss: 0.99678 +INFO:2025-06-11 06:20:47,498: Epoch [5/20], Step [1400/2647], Loss: 0.54078, Disc Loss: 3.79541, Dur Loss: 0.97340, CE Loss: 0.05450, Norm Loss: 1.06638, F0 Loss: 2.97478, LM Loss: 2.21980, Gen Loss: 5.32549, Sty Loss: 0.11584, Diff Loss: 0.16245, DiscLM Loss: 0.00078, GenLM Loss: 0.99989 +INFO:2025-06-11 06:22:35,758: Epoch [5/20], Step [1450/2647], Loss: 0.54500, Disc Loss: 3.71118, Dur Loss: 0.92641, CE Loss: 0.05093, Norm Loss: 1.18308, F0 Loss: 3.10055, LM Loss: 2.33965, Gen Loss: 5.80539, Sty Loss: 0.11420, Diff Loss: 0.27305, DiscLM Loss: 0.00000, GenLM Loss: 1.00313 +INFO:2025-06-11 06:24:24,041: Epoch [5/20], Step [1500/2647], Loss: 0.54350, Disc Loss: 3.79055, Dur Loss: 1.01183, CE Loss: 0.05773, Norm Loss: 1.32545, F0 Loss: 3.44767, LM Loss: 2.26559, Gen Loss: 5.43247, Sty Loss: 0.10937, Diff Loss: 0.28471, DiscLM Loss: 0.00127, GenLM Loss: 1.00110 +INFO:2025-06-11 06:26:16,864: Epoch [5/20], Step [1550/2647], Loss: 0.52450, Disc Loss: 3.77158, Dur Loss: 1.22182, CE Loss: 0.06699, Norm Loss: 1.20853, F0 Loss: 3.26363, LM Loss: 2.18970, Gen Loss: 5.43073, Sty Loss: 0.11732, Diff Loss: 0.27279, DiscLM Loss: 0.00000, GenLM Loss: 0.99682 +INFO:2025-06-11 06:28:06,452: Epoch [5/20], Step [1600/2647], Loss: 0.54703, Disc Loss: 3.83462, Dur Loss: 0.91710, CE Loss: 0.04695, Norm Loss: 0.92895, F0 Loss: 3.22610, LM Loss: 2.07942, Gen Loss: 5.77718, Sty Loss: 0.11308, Diff Loss: 0.21536, DiscLM Loss: 0.00088, GenLM Loss: 1.00405 +INFO:2025-06-11 06:29:53,035: Epoch [5/20], Step [1650/2647], Loss: 0.54028, Disc Loss: 3.81960, Dur Loss: 1.10517, CE Loss: 0.06838, Norm Loss: 1.30553, F0 Loss: 3.60554, LM Loss: 2.23960, Gen Loss: 5.64172, Sty Loss: 0.13860, Diff Loss: 0.19738, DiscLM Loss: 0.00000, GenLM Loss: 1.00805 +INFO:2025-06-11 06:31:40,806: Epoch [5/20], Step [1700/2647], Loss: 0.55806, Disc Loss: 3.67049, Dur Loss: 0.92904, CE Loss: 0.05402, Norm Loss: 1.19283, F0 Loss: 3.40712, LM Loss: 2.38816, Gen Loss: 6.55513, Sty Loss: 0.17182, Diff Loss: 0.29286, DiscLM Loss: 0.00068, GenLM Loss: 1.00389 +INFO:2025-06-11 06:33:28,819: Epoch [5/20], Step [1750/2647], Loss: 0.54186, Disc Loss: 3.73531, Dur Loss: 0.98862, CE Loss: 0.05450, Norm Loss: 1.15110, F0 Loss: 2.78691, LM Loss: 2.14514, Gen Loss: 5.73542, Sty Loss: 0.14197, Diff Loss: 0.22364, DiscLM Loss: 0.00000, GenLM Loss: 0.99326 +INFO:2025-06-11 06:35:19,270: Epoch [5/20], Step [1800/2647], Loss: 0.54028, Disc Loss: 3.74079, Dur Loss: 1.25303, CE Loss: 0.07017, Norm Loss: 1.15686, F0 Loss: 3.42664, LM Loss: 2.20084, Gen Loss: 5.90421, Sty Loss: 0.12211, Diff Loss: 0.30564, DiscLM Loss: 0.00147, GenLM Loss: 1.00235 +INFO:2025-06-11 06:37:07,755: Epoch [5/20], Step [1850/2647], Loss: 0.54894, Disc Loss: 3.71944, Dur Loss: 0.92902, CE Loss: 0.04997, Norm Loss: 1.13726, F0 Loss: 2.69249, LM Loss: 2.19567, Gen Loss: 5.95951, Sty Loss: 0.11503, Diff Loss: 0.27592, DiscLM Loss: 0.00000, GenLM Loss: 0.99884 +INFO:2025-06-11 06:38:55,742: Epoch [5/20], Step [1900/2647], Loss: 0.54554, Disc Loss: 3.71643, Dur Loss: 0.80460, CE Loss: 0.04102, Norm Loss: 0.91277, F0 Loss: 2.20011, LM Loss: 2.19510, Gen Loss: 6.19870, Sty Loss: 0.13030, Diff Loss: 0.35538, DiscLM Loss: 0.00162, GenLM Loss: 0.99461 +INFO:2025-06-11 06:40:44,890: Epoch [5/20], Step [1950/2647], Loss: 0.54404, Disc Loss: 3.72163, Dur Loss: 0.88170, CE Loss: 0.04735, Norm Loss: 0.84375, F0 Loss: 3.55350, LM Loss: 2.26628, Gen Loss: 5.62077, Sty Loss: 0.11926, Diff Loss: 0.35492, DiscLM Loss: 0.00000, GenLM Loss: 1.00486 +INFO:2025-06-11 06:42:30,875: Epoch [5/20], Step [2000/2647], Loss: 0.53944, Disc Loss: 3.78409, Dur Loss: 1.03856, CE Loss: 0.05508, Norm Loss: 1.12782, F0 Loss: 2.23691, LM Loss: 2.18233, Gen Loss: 6.08278, Sty Loss: 0.14764, Diff Loss: 0.34734, DiscLM Loss: 0.00199, GenLM Loss: 0.99750 +INFO:2025-06-11 06:44:20,632: Epoch [5/20], Step [2050/2647], Loss: 0.54087, Disc Loss: 3.77111, Dur Loss: 1.00133, CE Loss: 0.05934, Norm Loss: 1.21236, F0 Loss: 3.10963, LM Loss: 2.36938, Gen Loss: 6.24650, Sty Loss: 0.11725, Diff Loss: 0.43295, DiscLM Loss: 0.00000, GenLM Loss: 0.99672 +INFO:2025-06-11 06:46:08,972: Epoch [5/20], Step [2100/2647], Loss: 0.54123, Disc Loss: 3.84173, Dur Loss: 0.88333, CE Loss: 0.04509, Norm Loss: 1.17972, F0 Loss: 3.39491, LM Loss: 2.30948, Gen Loss: 6.29190, Sty Loss: 0.17849, Diff Loss: 0.33585, DiscLM Loss: 0.00126, GenLM Loss: 0.99934 +INFO:2025-06-11 06:47:57,629: Epoch [5/20], Step [2150/2647], Loss: 0.53379, Disc Loss: 3.80626, Dur Loss: 1.22885, CE Loss: 0.07042, Norm Loss: 1.51979, F0 Loss: 2.22605, LM Loss: 2.26614, Gen Loss: 6.43723, Sty Loss: 0.14639, Diff Loss: 0.23872, DiscLM Loss: 0.00000, GenLM Loss: 0.99611 +INFO:2025-06-11 06:49:48,143: Epoch [5/20], Step [2200/2647], Loss: 0.53430, Disc Loss: 3.80636, Dur Loss: 1.08357, CE Loss: 0.06173, Norm Loss: 1.28021, F0 Loss: 2.78769, LM Loss: 2.24372, Gen Loss: 5.71898, Sty Loss: 0.13945, Diff Loss: 0.19299, DiscLM Loss: 0.00078, GenLM Loss: 1.00528 +INFO:2025-06-11 06:51:36,294: Epoch [5/20], Step [2250/2647], Loss: 0.54421, Disc Loss: 3.75309, Dur Loss: 1.03121, CE Loss: 0.05166, Norm Loss: 1.16090, F0 Loss: 3.76659, LM Loss: 2.20444, Gen Loss: 5.38645, Sty Loss: 0.10740, Diff Loss: 0.27125, DiscLM Loss: 0.00000, GenLM Loss: 0.98034 +INFO:2025-06-11 06:53:27,185: Epoch [5/20], Step [2300/2647], Loss: 0.53756, Disc Loss: 3.76503, Dur Loss: 0.99284, CE Loss: 0.05118, Norm Loss: 1.33033, F0 Loss: 3.40540, LM Loss: 2.22051, Gen Loss: 5.48914, Sty Loss: 0.11650, Diff Loss: 0.27629, DiscLM Loss: 0.01199, GenLM Loss: 0.94893 +INFO:2025-06-11 06:55:16,189: Epoch [5/20], Step [2350/2647], Loss: 0.54068, Disc Loss: 3.77803, Dur Loss: 0.90567, CE Loss: 0.04768, Norm Loss: 0.99794, F0 Loss: 2.79714, LM Loss: 2.17797, Gen Loss: 5.74234, Sty Loss: 0.12310, Diff Loss: 0.25104, DiscLM Loss: 0.00000, GenLM Loss: 1.01179 +INFO:2025-06-11 06:57:03,240: Epoch [5/20], Step [2400/2647], Loss: 0.53323, Disc Loss: 3.88373, Dur Loss: 0.94783, CE Loss: 0.05103, Norm Loss: 0.99581, F0 Loss: 2.28547, LM Loss: 2.17560, Gen Loss: 5.81881, Sty Loss: 0.10820, Diff Loss: 0.34976, DiscLM Loss: 0.00217, GenLM Loss: 0.98007 +INFO:2025-06-11 06:58:49,949: Epoch [5/20], Step [2450/2647], Loss: 0.54647, Disc Loss: 3.90181, Dur Loss: 0.90792, CE Loss: 0.05107, Norm Loss: 1.04417, F0 Loss: 3.58478, LM Loss: 2.09628, Gen Loss: 4.82529, Sty Loss: 0.12350, Diff Loss: 0.27110, DiscLM Loss: 0.00000, GenLM Loss: 1.01094 +INFO:2025-06-11 07:00:37,968: Epoch [5/20], Step [2500/2647], Loss: 0.53127, Disc Loss: 3.85212, Dur Loss: 1.02897, CE Loss: 0.05414, Norm Loss: 1.07251, F0 Loss: 3.45573, LM Loss: 2.23802, Gen Loss: 5.69668, Sty Loss: 0.16778, Diff Loss: 0.34797, DiscLM Loss: 0.00083, GenLM Loss: 0.99286 +INFO:2025-06-11 07:02:26,715: Epoch [5/20], Step [2550/2647], Loss: 0.54276, Disc Loss: 3.92866, Dur Loss: 0.94406, CE Loss: 0.05271, Norm Loss: 1.26690, F0 Loss: 2.69756, LM Loss: 2.18951, Gen Loss: 5.98507, Sty Loss: 0.12433, Diff Loss: 0.43547, DiscLM Loss: 0.00000, GenLM Loss: 1.00212 +INFO:2025-06-11 07:04:18,335: Epoch [5/20], Step [2600/2647], Loss: 0.53121, Disc Loss: 3.78072, Dur Loss: 0.87729, CE Loss: 0.05102, Norm Loss: 0.99139, F0 Loss: 3.19746, LM Loss: 2.18884, Gen Loss: 5.87993, Sty Loss: 0.11013, Diff Loss: 0.29458, DiscLM Loss: 0.00397, GenLM Loss: 1.00719 +INFO:2025-06-11 07:08:22,022: Validation loss: 0.554, Dur loss: 1.032, F0 loss: 3.394 + + + +INFO:2025-06-11 07:10:13,485: Epoch [6/20], Step [50/2647], Loss: 0.54128, Disc Loss: 3.69048, Dur Loss: 1.22414, CE Loss: 0.06463, Norm Loss: 1.26511, F0 Loss: 3.12388, LM Loss: 2.19790, Gen Loss: 5.84457, Sty Loss: 0.11039, Diff Loss: 0.28775, DiscLM Loss: 0.00000, GenLM Loss: 0.99499 +INFO:2025-06-11 07:12:03,705: Epoch [6/20], Step [100/2647], Loss: 0.53719, Disc Loss: 3.86415, Dur Loss: 1.04448, CE Loss: 0.05518, Norm Loss: 1.23746, F0 Loss: 2.99668, LM Loss: 2.19885, Gen Loss: 5.45074, Sty Loss: 0.16208, Diff Loss: 0.34264, DiscLM Loss: 0.00113, GenLM Loss: 0.99959 +INFO:2025-06-11 07:13:51,377: Epoch [6/20], Step [150/2647], Loss: 0.53879, Disc Loss: 3.71756, Dur Loss: 0.89916, CE Loss: 0.04896, Norm Loss: 1.14006, F0 Loss: 2.75372, LM Loss: 2.16036, Gen Loss: 5.77602, Sty Loss: 0.10816, Diff Loss: 0.23913, DiscLM Loss: 0.00000, GenLM Loss: 0.99434 +INFO:2025-06-11 07:15:41,149: Epoch [6/20], Step [200/2647], Loss: 0.53184, Disc Loss: 3.68452, Dur Loss: 0.96509, CE Loss: 0.05675, Norm Loss: 1.23817, F0 Loss: 3.43885, LM Loss: 2.22395, Gen Loss: 5.60741, Sty Loss: 0.09329, Diff Loss: 0.35721, DiscLM Loss: 0.00104, GenLM Loss: 0.99000 +INFO:2025-06-11 07:17:31,472: Epoch [6/20], Step [250/2647], Loss: 0.53539, Disc Loss: 3.85561, Dur Loss: 1.21764, CE Loss: 0.07205, Norm Loss: 1.23570, F0 Loss: 3.33773, LM Loss: 2.30557, Gen Loss: 6.41103, Sty Loss: 0.12935, Diff Loss: 0.25454, DiscLM Loss: 0.00000, GenLM Loss: 1.01330 +INFO:2025-06-11 07:19:20,893: Epoch [6/20], Step [300/2647], Loss: 0.53298, Disc Loss: 3.77920, Dur Loss: 1.06169, CE Loss: 0.05637, Norm Loss: 1.12775, F0 Loss: 3.58253, LM Loss: 2.14718, Gen Loss: 5.97809, Sty Loss: 0.14021, Diff Loss: 0.29148, DiscLM Loss: 0.00106, GenLM Loss: 1.00976 +INFO:2025-06-11 07:21:07,946: Epoch [6/20], Step [350/2647], Loss: 0.53529, Disc Loss: 3.88462, Dur Loss: 0.90949, CE Loss: 0.05000, Norm Loss: 0.92128, F0 Loss: 3.03988, LM Loss: 2.28068, Gen Loss: 6.16736, Sty Loss: 0.12246, Diff Loss: 0.28002, DiscLM Loss: 0.00000, GenLM Loss: 0.98503 +INFO:2025-06-11 07:22:54,539: Epoch [6/20], Step [400/2647], Loss: 0.53420, Disc Loss: 3.74630, Dur Loss: 1.19347, CE Loss: 0.06815, Norm Loss: 1.24853, F0 Loss: 3.22200, LM Loss: 2.28040, Gen Loss: 5.95453, Sty Loss: 0.12523, Diff Loss: 0.29615, DiscLM Loss: 0.00342, GenLM Loss: 1.00090 +INFO:2025-06-11 07:24:41,862: Epoch [6/20], Step [450/2647], Loss: 0.52670, Disc Loss: 3.80268, Dur Loss: 0.97782, CE Loss: 0.05587, Norm Loss: 1.00842, F0 Loss: 3.27118, LM Loss: 2.21522, Gen Loss: 5.23582, Sty Loss: 0.14518, Diff Loss: 0.31242, DiscLM Loss: 0.00000, GenLM Loss: 0.97955 +INFO:2025-06-11 07:26:31,774: Epoch [6/20], Step [500/2647], Loss: 0.52986, Disc Loss: 3.79197, Dur Loss: 1.47106, CE Loss: 0.09409, Norm Loss: 1.42862, F0 Loss: 3.42685, LM Loss: 2.18725, Gen Loss: 5.18455, Sty Loss: 0.12598, Diff Loss: 0.22552, DiscLM Loss: 0.00248, GenLM Loss: 0.98943 +INFO:2025-06-11 07:28:21,883: Epoch [6/20], Step [550/2647], Loss: 0.53155, Disc Loss: 3.73918, Dur Loss: 1.21107, CE Loss: 0.06940, Norm Loss: 1.09221, F0 Loss: 3.57044, LM Loss: 2.20259, Gen Loss: 5.85515, Sty Loss: 0.09759, Diff Loss: 0.37929, DiscLM Loss: 0.00000, GenLM Loss: 1.00553 +INFO:2025-06-11 07:30:10,318: Epoch [6/20], Step [600/2647], Loss: 0.54283, Disc Loss: 3.71367, Dur Loss: 1.14094, CE Loss: 0.06089, Norm Loss: 1.17809, F0 Loss: 4.10530, LM Loss: 2.23487, Gen Loss: 6.02763, Sty Loss: 0.13757, Diff Loss: 0.34970, DiscLM Loss: 0.00252, GenLM Loss: 0.99780 +INFO:2025-06-11 07:31:59,650: Epoch [6/20], Step [650/2647], Loss: 0.52498, Disc Loss: 3.74107, Dur Loss: 0.96815, CE Loss: 0.05020, Norm Loss: 1.24288, F0 Loss: 3.76768, LM Loss: 2.14653, Gen Loss: 6.69323, Sty Loss: 0.11956, Diff Loss: 0.25847, DiscLM Loss: 0.00000, GenLM Loss: 1.00622 +INFO:2025-06-11 07:33:48,540: Epoch [6/20], Step [700/2647], Loss: 0.52776, Disc Loss: 3.82828, Dur Loss: 1.02567, CE Loss: 0.05872, Norm Loss: 1.22875, F0 Loss: 2.99398, LM Loss: 2.20367, Gen Loss: 5.74341, Sty Loss: 0.11301, Diff Loss: 0.25629, DiscLM Loss: 0.00192, GenLM Loss: 0.98232 +INFO:2025-06-11 07:35:37,221: Epoch [6/20], Step [750/2647], Loss: 0.53310, Disc Loss: 3.70338, Dur Loss: 1.13975, CE Loss: 0.05982, Norm Loss: 1.22468, F0 Loss: 3.13207, LM Loss: 2.30185, Gen Loss: 6.14510, Sty Loss: 0.11567, Diff Loss: 0.18525, DiscLM Loss: 0.00000, GenLM Loss: 1.00867 +INFO:2025-06-11 07:37:26,629: Epoch [6/20], Step [800/2647], Loss: 0.55163, Disc Loss: 3.86960, Dur Loss: 0.97188, CE Loss: 0.05801, Norm Loss: 1.32417, F0 Loss: 4.65139, LM Loss: 2.27124, Gen Loss: 5.70242, Sty Loss: 0.11310, Diff Loss: 0.24289, DiscLM Loss: 0.00481, GenLM Loss: 0.99909 +INFO:2025-06-11 07:39:13,910: Epoch [6/20], Step [850/2647], Loss: 0.53174, Disc Loss: 3.89635, Dur Loss: 0.74672, CE Loss: 0.04116, Norm Loss: 1.37879, F0 Loss: 2.72593, LM Loss: 2.19695, Gen Loss: 5.82518, Sty Loss: 0.11740, Diff Loss: 0.28909, DiscLM Loss: 0.00000, GenLM Loss: 0.99801 +INFO:2025-06-11 07:41:03,659: Epoch [6/20], Step [900/2647], Loss: 0.53171, Disc Loss: 3.75153, Dur Loss: 0.96633, CE Loss: 0.05013, Norm Loss: 0.79310, F0 Loss: 3.13307, LM Loss: 2.14319, Gen Loss: 5.85317, Sty Loss: 0.11742, Diff Loss: 0.27464, DiscLM Loss: 0.00254, GenLM Loss: 1.00059 +INFO:2025-06-11 07:42:51,169: Epoch [6/20], Step [950/2647], Loss: 0.54135, Disc Loss: 3.67224, Dur Loss: 1.05125, CE Loss: 0.06644, Norm Loss: 0.94606, F0 Loss: 2.71900, LM Loss: 2.22903, Gen Loss: 6.33727, Sty Loss: 0.15476, Diff Loss: 0.27566, DiscLM Loss: 0.00000, GenLM Loss: 1.00827 +INFO:2025-06-11 07:44:37,985: Epoch [6/20], Step [1000/2647], Loss: 0.53057, Disc Loss: 3.79709, Dur Loss: 0.88341, CE Loss: 0.04993, Norm Loss: 0.93054, F0 Loss: 3.87809, LM Loss: 2.12727, Gen Loss: 6.04496, Sty Loss: 0.11862, Diff Loss: 0.30733, DiscLM Loss: 0.00240, GenLM Loss: 0.99123 +INFO:2025-06-11 07:46:27,060: Epoch [6/20], Step [1050/2647], Loss: 0.54104, Disc Loss: 3.73038, Dur Loss: 0.97923, CE Loss: 0.04849, Norm Loss: 1.03804, F0 Loss: 3.22764, LM Loss: 2.13028, Gen Loss: 5.49049, Sty Loss: 0.11619, Diff Loss: 0.21831, DiscLM Loss: 0.00000, GenLM Loss: 0.99942 +INFO:2025-06-11 07:48:16,241: Epoch [6/20], Step [1100/2647], Loss: 0.52477, Disc Loss: 3.83513, Dur Loss: 0.95125, CE Loss: 0.05306, Norm Loss: 1.09554, F0 Loss: 3.07717, LM Loss: 2.24020, Gen Loss: 5.64526, Sty Loss: 0.11816, Diff Loss: 0.30640, DiscLM Loss: 0.00104, GenLM Loss: 0.99433 +INFO:2025-06-11 07:50:03,948: Epoch [6/20], Step [1150/2647], Loss: 0.52917, Disc Loss: 3.76101, Dur Loss: 1.08367, CE Loss: 0.05889, Norm Loss: 1.10269, F0 Loss: 3.24438, LM Loss: 2.15471, Gen Loss: 5.04449, Sty Loss: 0.10354, Diff Loss: 0.31392, DiscLM Loss: 0.00000, GenLM Loss: 0.99918 +INFO:2025-06-11 07:51:54,552: Epoch [6/20], Step [1200/2647], Loss: 0.52414, Disc Loss: 3.75421, Dur Loss: 1.00541, CE Loss: 0.05321, Norm Loss: 1.00939, F0 Loss: 3.25090, LM Loss: 2.16358, Gen Loss: 6.05053, Sty Loss: 0.11283, Diff Loss: 0.33502, DiscLM Loss: 0.01694, GenLM Loss: 0.99797 +INFO:2025-06-11 07:53:41,575: Epoch [6/20], Step [1250/2647], Loss: 0.52790, Disc Loss: 3.72902, Dur Loss: 0.90798, CE Loss: 0.04662, Norm Loss: 0.89514, F0 Loss: 2.62521, LM Loss: 2.14160, Gen Loss: 5.90371, Sty Loss: 0.11642, Diff Loss: 0.21441, DiscLM Loss: 0.00000, GenLM Loss: 1.00005 +INFO:2025-06-11 07:55:29,741: Epoch [6/20], Step [1300/2647], Loss: 0.54206, Disc Loss: 3.75695, Dur Loss: 0.98923, CE Loss: 0.05760, Norm Loss: 0.93186, F0 Loss: 3.06510, LM Loss: 2.14697, Gen Loss: 5.91675, Sty Loss: 0.11716, Diff Loss: 0.32815, DiscLM Loss: 0.00313, GenLM Loss: 1.00042 +INFO:2025-06-11 07:57:16,438: Epoch [6/20], Step [1350/2647], Loss: 0.53939, Disc Loss: 3.81696, Dur Loss: 0.99786, CE Loss: 0.05483, Norm Loss: 1.03810, F0 Loss: 4.26151, LM Loss: 2.23765, Gen Loss: 5.75811, Sty Loss: 0.13271, Diff Loss: 0.22209, DiscLM Loss: 0.00000, GenLM Loss: 0.99031 +INFO:2025-06-11 07:59:03,236: Epoch [6/20], Step [1400/2647], Loss: 0.53453, Disc Loss: 3.79073, Dur Loss: 0.96507, CE Loss: 0.05148, Norm Loss: 1.06086, F0 Loss: 2.68425, LM Loss: 2.15061, Gen Loss: 5.76885, Sty Loss: 0.11420, Diff Loss: 0.27536, DiscLM Loss: 0.00188, GenLM Loss: 1.00013 +INFO:2025-06-11 08:00:51,665: Epoch [6/20], Step [1450/2647], Loss: 0.53370, Disc Loss: 3.69033, Dur Loss: 0.95737, CE Loss: 0.05849, Norm Loss: 0.89421, F0 Loss: 2.93956, LM Loss: 2.16150, Gen Loss: 5.88668, Sty Loss: 0.10819, Diff Loss: 0.19888, DiscLM Loss: 0.00000, GenLM Loss: 0.99656 +INFO:2025-06-11 08:02:41,942: Epoch [6/20], Step [1500/2647], Loss: 0.53714, Disc Loss: 3.71462, Dur Loss: 1.04135, CE Loss: 0.05918, Norm Loss: 1.10397, F0 Loss: 4.22278, LM Loss: 2.33946, Gen Loss: 6.39830, Sty Loss: 0.10802, Diff Loss: 0.21620, DiscLM Loss: 0.00197, GenLM Loss: 0.99269 +INFO:2025-06-11 08:04:31,156: Epoch [6/20], Step [1550/2647], Loss: 0.53419, Disc Loss: 3.86184, Dur Loss: 1.21903, CE Loss: 0.07580, Norm Loss: 1.29938, F0 Loss: 3.49931, LM Loss: 2.32411, Gen Loss: 5.39686, Sty Loss: 0.11764, Diff Loss: 0.28840, DiscLM Loss: 0.00000, GenLM Loss: 1.00829 +INFO:2025-06-11 08:06:17,539: Epoch [6/20], Step [1600/2647], Loss: 0.52513, Disc Loss: 3.74224, Dur Loss: 1.32451, CE Loss: 0.07019, Norm Loss: 1.06743, F0 Loss: 3.09952, LM Loss: 2.18474, Gen Loss: 5.84305, Sty Loss: 0.10312, Diff Loss: 0.22805, DiscLM Loss: 0.00235, GenLM Loss: 0.98847 +INFO:2025-06-11 08:08:05,475: Epoch [6/20], Step [1650/2647], Loss: 0.53809, Disc Loss: 3.81916, Dur Loss: 1.07866, CE Loss: 0.06293, Norm Loss: 1.31305, F0 Loss: 3.65949, LM Loss: 2.18386, Gen Loss: 6.56027, Sty Loss: 0.12724, Diff Loss: 0.25566, DiscLM Loss: 0.00000, GenLM Loss: 1.00169 +INFO:2025-06-11 08:09:54,255: Epoch [6/20], Step [1700/2647], Loss: 0.54078, Disc Loss: 3.81987, Dur Loss: 0.90574, CE Loss: 0.04642, Norm Loss: 1.02062, F0 Loss: 3.01345, LM Loss: 2.11049, Gen Loss: 5.20098, Sty Loss: 0.10561, Diff Loss: 0.25074, DiscLM Loss: 0.00065, GenLM Loss: 1.00070 +INFO:2025-06-11 08:11:42,333: Epoch [6/20], Step [1750/2647], Loss: 0.53095, Disc Loss: 3.75200, Dur Loss: 0.79368, CE Loss: 0.04291, Norm Loss: 0.92029, F0 Loss: 2.70564, LM Loss: 2.01786, Gen Loss: 5.51749, Sty Loss: 0.10967, Diff Loss: 0.26573, DiscLM Loss: 0.00000, GenLM Loss: 1.00018 +INFO:2025-06-11 08:13:30,601: Epoch [6/20], Step [1800/2647], Loss: 0.52970, Disc Loss: 3.73374, Dur Loss: 0.99076, CE Loss: 0.05517, Norm Loss: 1.24305, F0 Loss: 3.17711, LM Loss: 2.11952, Gen Loss: 5.65479, Sty Loss: 0.13550, Diff Loss: 0.25593, DiscLM Loss: 0.00097, GenLM Loss: 1.00845 +INFO:2025-06-11 08:15:19,367: Epoch [6/20], Step [1850/2647], Loss: 0.53758, Disc Loss: 3.71935, Dur Loss: 1.01131, CE Loss: 0.05341, Norm Loss: 1.16702, F0 Loss: 3.30291, LM Loss: 2.18715, Gen Loss: 5.86098, Sty Loss: 0.12380, Diff Loss: 0.29631, DiscLM Loss: 0.00000, GenLM Loss: 0.98925 +INFO:2025-06-11 08:17:09,692: Epoch [6/20], Step [1900/2647], Loss: 0.52594, Disc Loss: 3.69960, Dur Loss: 0.96559, CE Loss: 0.05427, Norm Loss: 0.92996, F0 Loss: 4.03694, LM Loss: 2.18648, Gen Loss: 6.36531, Sty Loss: 0.11384, Diff Loss: 0.35254, DiscLM Loss: 0.00117, GenLM Loss: 0.99125 +INFO:2025-06-11 08:18:58,120: Epoch [6/20], Step [1950/2647], Loss: 0.53398, Disc Loss: 3.74281, Dur Loss: 0.96872, CE Loss: 0.05054, Norm Loss: 1.07910, F0 Loss: 3.11973, LM Loss: 2.15452, Gen Loss: 5.89391, Sty Loss: 0.09441, Diff Loss: 0.28890, DiscLM Loss: 0.00000, GenLM Loss: 1.00156 +INFO:2025-06-11 08:20:48,773: Epoch [6/20], Step [2000/2647], Loss: 0.54094, Disc Loss: 3.71644, Dur Loss: 0.94773, CE Loss: 0.04888, Norm Loss: 1.13545, F0 Loss: 3.06197, LM Loss: 2.21458, Gen Loss: 6.05456, Sty Loss: 0.10247, Diff Loss: 0.24757, DiscLM Loss: 0.00198, GenLM Loss: 0.98192 +INFO:2025-06-11 08:22:41,534: Epoch [6/20], Step [2050/2647], Loss: 0.53157, Disc Loss: 3.77948, Dur Loss: 0.99940, CE Loss: 0.05353, Norm Loss: 0.92897, F0 Loss: 2.66893, LM Loss: 2.10806, Gen Loss: 6.00700, Sty Loss: 0.11717, Diff Loss: 0.20429, DiscLM Loss: 0.00000, GenLM Loss: 1.00450 +INFO:2025-06-11 08:24:31,278: Epoch [6/20], Step [2100/2647], Loss: 0.53927, Disc Loss: 3.71435, Dur Loss: 0.95663, CE Loss: 0.05167, Norm Loss: 0.93590, F0 Loss: 2.93823, LM Loss: 2.15662, Gen Loss: 5.92463, Sty Loss: 0.09462, Diff Loss: 0.25946, DiscLM Loss: 0.00166, GenLM Loss: 0.98453 +INFO:2025-06-11 08:26:19,923: Epoch [6/20], Step [2150/2647], Loss: 0.53286, Disc Loss: 3.79308, Dur Loss: 1.16072, CE Loss: 0.06357, Norm Loss: 1.18812, F0 Loss: 4.08183, LM Loss: 2.16442, Gen Loss: 5.83793, Sty Loss: 0.12115, Diff Loss: 0.19012, DiscLM Loss: 0.00000, GenLM Loss: 1.00685 +INFO:2025-06-11 08:28:09,241: Epoch [6/20], Step [2200/2647], Loss: 0.53361, Disc Loss: 3.76550, Dur Loss: 1.00483, CE Loss: 0.05516, Norm Loss: 1.19271, F0 Loss: 2.89756, LM Loss: 2.23428, Gen Loss: 5.79526, Sty Loss: 0.10699, Diff Loss: 0.30911, DiscLM Loss: 0.00164, GenLM Loss: 0.99837 +INFO:2025-06-11 08:29:56,152: Epoch [6/20], Step [2250/2647], Loss: 0.53500, Disc Loss: 3.71153, Dur Loss: 0.89372, CE Loss: 0.04895, Norm Loss: 1.23149, F0 Loss: 3.29314, LM Loss: 2.21835, Gen Loss: 6.37681, Sty Loss: 0.12672, Diff Loss: 0.24163, DiscLM Loss: 0.00000, GenLM Loss: 0.99137 +INFO:2025-06-11 08:31:45,651: Epoch [6/20], Step [2300/2647], Loss: 0.52468, Disc Loss: 3.81851, Dur Loss: 1.03611, CE Loss: 0.05482, Norm Loss: 0.99772, F0 Loss: 3.50847, LM Loss: 2.15843, Gen Loss: 4.83728, Sty Loss: 0.09098, Diff Loss: 0.31529, DiscLM Loss: 0.02523, GenLM Loss: 0.87267 +INFO:2025-06-11 08:33:35,258: Epoch [6/20], Step [2350/2647], Loss: 0.53433, Disc Loss: 3.82740, Dur Loss: 1.01092, CE Loss: 0.06537, Norm Loss: 1.08623, F0 Loss: 3.87632, LM Loss: 2.19313, Gen Loss: 6.33916, Sty Loss: 0.12392, Diff Loss: 0.25357, DiscLM Loss: 0.00000, GenLM Loss: 1.04532 +INFO:2025-06-11 08:35:24,287: Epoch [6/20], Step [2400/2647], Loss: 0.53432, Disc Loss: 3.83245, Dur Loss: 0.95614, CE Loss: 0.05011, Norm Loss: 0.97162, F0 Loss: 3.19107, LM Loss: 2.20780, Gen Loss: 6.17443, Sty Loss: 0.13276, Diff Loss: 0.37820, DiscLM Loss: 0.00919, GenLM Loss: 1.00036 +INFO:2025-06-11 08:37:14,220: Epoch [6/20], Step [2450/2647], Loss: 0.52264, Disc Loss: 3.81221, Dur Loss: 0.99204, CE Loss: 0.05179, Norm Loss: 0.79781, F0 Loss: 2.69391, LM Loss: 2.22358, Gen Loss: 5.63915, Sty Loss: 0.12992, Diff Loss: 0.22191, DiscLM Loss: 0.00000, GenLM Loss: 0.96288 +INFO:2025-06-11 08:39:04,147: Epoch [6/20], Step [2500/2647], Loss: 0.53380, Disc Loss: 3.85827, Dur Loss: 1.37512, CE Loss: 0.08037, Norm Loss: 1.18212, F0 Loss: 2.91848, LM Loss: 2.06628, Gen Loss: 4.96094, Sty Loss: 0.10998, Diff Loss: 0.23504, DiscLM Loss: 0.01891, GenLM Loss: 1.02451 +INFO:2025-06-11 08:40:52,218: Epoch [6/20], Step [2550/2647], Loss: 0.52910, Disc Loss: 3.77185, Dur Loss: 1.14159, CE Loss: 0.05873, Norm Loss: 0.82044, F0 Loss: 3.10250, LM Loss: 2.04605, Gen Loss: 5.48754, Sty Loss: 0.10962, Diff Loss: 0.29330, DiscLM Loss: 0.00000, GenLM Loss: 0.99723 +INFO:2025-06-11 08:42:40,411: Epoch [6/20], Step [2600/2647], Loss: 0.52532, Disc Loss: 3.64850, Dur Loss: 1.01839, CE Loss: 0.05721, Norm Loss: 1.10209, F0 Loss: 3.04980, LM Loss: 2.17563, Gen Loss: 6.47452, Sty Loss: 0.12269, Diff Loss: 0.17577, DiscLM Loss: 0.00376, GenLM Loss: 0.97028 +INFO:2025-06-11 08:46:40,703: Validation loss: 0.559, Dur loss: 1.024, F0 loss: 3.388 + + + +INFO:2025-06-11 08:48:32,845: Epoch [7/20], Step [50/2647], Loss: 0.52891, Disc Loss: 3.75742, Dur Loss: 1.10500, CE Loss: 0.05871, Norm Loss: 1.30886, F0 Loss: 3.83201, LM Loss: 2.28703, Gen Loss: 5.88922, Sty Loss: 0.13089, Diff Loss: 0.26302, DiscLM Loss: 0.00000, GenLM Loss: 0.96225 +INFO:2025-06-11 08:50:21,584: Epoch [7/20], Step [100/2647], Loss: 0.53519, Disc Loss: 3.77354, Dur Loss: 1.01115, CE Loss: 0.05473, Norm Loss: 0.79520, F0 Loss: 3.23942, LM Loss: 2.15372, Gen Loss: 6.05389, Sty Loss: 0.11683, Diff Loss: 0.27546, DiscLM Loss: 0.06937, GenLM Loss: 0.85893 +INFO:2025-06-11 08:52:10,914: Epoch [7/20], Step [150/2647], Loss: 0.53390, Disc Loss: 3.91810, Dur Loss: 1.01516, CE Loss: 0.06853, Norm Loss: 1.32817, F0 Loss: 3.25543, LM Loss: 2.24879, Gen Loss: 5.64048, Sty Loss: 0.14710, Diff Loss: 0.31995, DiscLM Loss: 0.00000, GenLM Loss: 0.98728 +INFO:2025-06-11 08:53:56,235: Epoch [7/20], Step [200/2647], Loss: 0.52531, Disc Loss: 3.73598, Dur Loss: 1.04300, CE Loss: 0.06008, Norm Loss: 0.92603, F0 Loss: 2.59465, LM Loss: 2.11728, Gen Loss: 5.90186, Sty Loss: 0.11129, Diff Loss: 0.21453, DiscLM Loss: 0.01705, GenLM Loss: 0.98772 +INFO:2025-06-11 08:55:45,872: Epoch [7/20], Step [250/2647], Loss: 0.53257, Disc Loss: 3.71688, Dur Loss: 0.96134, CE Loss: 0.05395, Norm Loss: 1.04913, F0 Loss: 3.10090, LM Loss: 2.22863, Gen Loss: 6.12528, Sty Loss: 0.09225, Diff Loss: 0.28894, DiscLM Loss: 0.00000, GenLM Loss: 1.00500 +INFO:2025-06-11 08:57:35,040: Epoch [7/20], Step [300/2647], Loss: 0.53245, Disc Loss: 3.74931, Dur Loss: 1.32066, CE Loss: 0.06855, Norm Loss: 1.05855, F0 Loss: 2.98445, LM Loss: 2.19107, Gen Loss: 5.71340, Sty Loss: 0.14254, Diff Loss: 0.26876, DiscLM Loss: 0.00301, GenLM Loss: 0.98416 +INFO:2025-06-11 08:59:23,230: Epoch [7/20], Step [350/2647], Loss: 0.52080, Disc Loss: 3.83856, Dur Loss: 1.01889, CE Loss: 0.05208, Norm Loss: 1.29176, F0 Loss: 3.84710, LM Loss: 2.26217, Gen Loss: 5.31000, Sty Loss: 0.12392, Diff Loss: 0.24563, DiscLM Loss: 0.00000, GenLM Loss: 0.93979 +INFO:2025-06-11 09:01:11,706: Epoch [7/20], Step [400/2647], Loss: 0.52913, Disc Loss: 3.77724, Dur Loss: 1.01581, CE Loss: 0.06954, Norm Loss: 1.50710, F0 Loss: 4.34267, LM Loss: 2.18503, Gen Loss: 5.66468, Sty Loss: 0.12464, Diff Loss: 0.18859, DiscLM Loss: 0.01149, GenLM Loss: 0.99421 +INFO:2025-06-11 09:03:01,538: Epoch [7/20], Step [450/2647], Loss: 0.52846, Disc Loss: 3.74639, Dur Loss: 0.81534, CE Loss: 0.04140, Norm Loss: 1.09515, F0 Loss: 2.93570, LM Loss: 2.15962, Gen Loss: 5.91105, Sty Loss: 0.11363, Diff Loss: 0.31345, DiscLM Loss: 0.00000, GenLM Loss: 1.01382 +INFO:2025-06-11 09:04:51,231: Epoch [7/20], Step [500/2647], Loss: 0.52221, Disc Loss: 3.78002, Dur Loss: 0.98707, CE Loss: 0.05790, Norm Loss: 0.88904, F0 Loss: 2.94631, LM Loss: 2.11326, Gen Loss: 5.68890, Sty Loss: 0.14051, Diff Loss: 0.28998, DiscLM Loss: 0.00240, GenLM Loss: 0.99926 +INFO:2025-06-11 09:06:40,936: Epoch [7/20], Step [550/2647], Loss: 0.52996, Disc Loss: 3.82611, Dur Loss: 0.92305, CE Loss: 0.04899, Norm Loss: 1.00984, F0 Loss: 3.09756, LM Loss: 2.14973, Gen Loss: 5.38374, Sty Loss: 0.12052, Diff Loss: 0.20457, DiscLM Loss: 0.00000, GenLM Loss: 1.00078 +INFO:2025-06-11 09:08:31,705: Epoch [7/20], Step [600/2647], Loss: 0.53485, Disc Loss: 3.79245, Dur Loss: 0.93076, CE Loss: 0.04853, Norm Loss: 1.17961, F0 Loss: 3.73361, LM Loss: 2.10226, Gen Loss: 5.39135, Sty Loss: 0.12772, Diff Loss: 0.27704, DiscLM Loss: 0.00444, GenLM Loss: 0.97640 +INFO:2025-06-11 09:10:18,791: Epoch [7/20], Step [650/2647], Loss: 0.52392, Disc Loss: 3.78147, Dur Loss: 1.01779, CE Loss: 0.05745, Norm Loss: 1.19601, F0 Loss: 2.23660, LM Loss: 2.09007, Gen Loss: 5.60089, Sty Loss: 0.11379, Diff Loss: 0.24353, DiscLM Loss: 0.00000, GenLM Loss: 1.00895 +INFO:2025-06-11 09:12:08,939: Epoch [7/20], Step [700/2647], Loss: 0.52517, Disc Loss: 3.71482, Dur Loss: 0.92641, CE Loss: 0.04974, Norm Loss: 0.84285, F0 Loss: 2.59039, LM Loss: 2.02705, Gen Loss: 6.32539, Sty Loss: 0.10546, Diff Loss: 0.26454, DiscLM Loss: 0.00480, GenLM Loss: 0.99175 +INFO:2025-06-11 09:13:56,438: Epoch [7/20], Step [750/2647], Loss: 0.53280, Disc Loss: 3.76368, Dur Loss: 1.10010, CE Loss: 0.06290, Norm Loss: 1.15279, F0 Loss: 3.03793, LM Loss: 2.24657, Gen Loss: 6.08901, Sty Loss: 0.11727, Diff Loss: 0.20180, DiscLM Loss: 0.00000, GenLM Loss: 0.99271 +INFO:2025-06-11 09:15:45,153: Epoch [7/20], Step [800/2647], Loss: 0.52427, Disc Loss: 3.88053, Dur Loss: 1.00085, CE Loss: 0.05774, Norm Loss: 1.22419, F0 Loss: 2.69666, LM Loss: 2.23107, Gen Loss: 6.97542, Sty Loss: 0.15403, Diff Loss: 0.24728, DiscLM Loss: 0.00205, GenLM Loss: 0.99235 +INFO:2025-06-11 09:17:34,672: Epoch [7/20], Step [850/2647], Loss: 0.52795, Disc Loss: 3.88418, Dur Loss: 0.90426, CE Loss: 0.04867, Norm Loss: 1.29589, F0 Loss: 3.50338, LM Loss: 2.16784, Gen Loss: 6.53102, Sty Loss: 0.11191, Diff Loss: 0.25492, DiscLM Loss: 0.00000, GenLM Loss: 1.00447 +INFO:2025-06-11 09:19:23,936: Epoch [7/20], Step [900/2647], Loss: 0.53285, Disc Loss: 3.65344, Dur Loss: 0.91924, CE Loss: 0.05549, Norm Loss: 1.04718, F0 Loss: 3.46420, LM Loss: 2.14494, Gen Loss: 6.14266, Sty Loss: 0.12405, Diff Loss: 0.35572, DiscLM Loss: 0.00087, GenLM Loss: 0.99124 +INFO:2025-06-11 09:21:11,565: Epoch [7/20], Step [950/2647], Loss: 0.53275, Disc Loss: 3.89744, Dur Loss: 0.97516, CE Loss: 0.05689, Norm Loss: 1.08720, F0 Loss: 3.15767, LM Loss: 2.11875, Gen Loss: 6.74692, Sty Loss: 0.09233, Diff Loss: 0.37821, DiscLM Loss: 0.00000, GenLM Loss: 1.01124 +INFO:2025-06-11 09:22:59,870: Epoch [7/20], Step [1000/2647], Loss: 0.53289, Disc Loss: 3.71543, Dur Loss: 0.90812, CE Loss: 0.05093, Norm Loss: 1.02133, F0 Loss: 3.25154, LM Loss: 2.18206, Gen Loss: 6.23914, Sty Loss: 0.09930, Diff Loss: 0.38558, DiscLM Loss: 0.00102, GenLM Loss: 1.01250 +INFO:2025-06-11 09:24:49,007: Epoch [7/20], Step [1050/2647], Loss: 0.53108, Disc Loss: 3.83479, Dur Loss: 0.91561, CE Loss: 0.04691, Norm Loss: 1.17994, F0 Loss: 3.39584, LM Loss: 2.25108, Gen Loss: 5.67874, Sty Loss: 0.09946, Diff Loss: 0.26378, DiscLM Loss: 0.00000, GenLM Loss: 0.98167 +INFO:2025-06-11 09:26:36,170: Epoch [7/20], Step [1100/2647], Loss: 0.53211, Disc Loss: 3.66318, Dur Loss: 1.03930, CE Loss: 0.05954, Norm Loss: 1.18663, F0 Loss: 3.93905, LM Loss: 2.17846, Gen Loss: 6.10638, Sty Loss: 0.13149, Diff Loss: 0.22501, DiscLM Loss: 0.01190, GenLM Loss: 0.99764 +INFO:2025-06-11 09:28:25,835: Epoch [7/20], Step [1150/2647], Loss: 0.52159, Disc Loss: 3.69172, Dur Loss: 1.00726, CE Loss: 0.05276, Norm Loss: 0.96472, F0 Loss: 2.64857, LM Loss: 2.12101, Gen Loss: 6.45294, Sty Loss: 0.13298, Diff Loss: 0.25845, DiscLM Loss: 0.00000, GenLM Loss: 1.00062 +INFO:2025-06-11 09:30:14,851: Epoch [7/20], Step [1200/2647], Loss: 0.51392, Disc Loss: 3.82119, Dur Loss: 1.28827, CE Loss: 0.08193, Norm Loss: 1.41800, F0 Loss: 3.67724, LM Loss: 2.35291, Gen Loss: 5.34297, Sty Loss: 0.11358, Diff Loss: 0.33340, DiscLM Loss: 0.00075, GenLM Loss: 1.00662 +INFO:2025-06-11 09:32:05,354: Epoch [7/20], Step [1250/2647], Loss: 0.52526, Disc Loss: 3.83302, Dur Loss: 1.17196, CE Loss: 0.06807, Norm Loss: 1.18860, F0 Loss: 3.67342, LM Loss: 2.28566, Gen Loss: 5.78821, Sty Loss: 0.11784, Diff Loss: 0.30572, DiscLM Loss: 0.00000, GenLM Loss: 0.99612 +INFO:2025-06-11 09:33:54,259: Epoch [7/20], Step [1300/2647], Loss: 0.51933, Disc Loss: 3.75984, Dur Loss: 0.74971, CE Loss: 0.04044, Norm Loss: 0.90221, F0 Loss: 4.17756, LM Loss: 2.17162, Gen Loss: 5.45715, Sty Loss: 0.12420, Diff Loss: 0.31960, DiscLM Loss: 0.00373, GenLM Loss: 1.00288 +INFO:2025-06-11 09:35:43,004: Epoch [7/20], Step [1350/2647], Loss: 0.52723, Disc Loss: 3.75591, Dur Loss: 1.10730, CE Loss: 0.07433, Norm Loss: 1.44382, F0 Loss: 3.06362, LM Loss: 2.29078, Gen Loss: 5.84367, Sty Loss: 0.10211, Diff Loss: 0.24821, DiscLM Loss: 0.00000, GenLM Loss: 0.97549 +INFO:2025-06-11 09:37:32,547: Epoch [7/20], Step [1400/2647], Loss: 0.53850, Disc Loss: 3.68107, Dur Loss: 1.05039, CE Loss: 0.05585, Norm Loss: 1.05091, F0 Loss: 2.99668, LM Loss: 2.10634, Gen Loss: 5.92512, Sty Loss: 0.14036, Diff Loss: 0.24018, DiscLM Loss: 0.00156, GenLM Loss: 1.00744 +INFO:2025-06-11 09:39:21,727: Epoch [7/20], Step [1450/2647], Loss: 0.52174, Disc Loss: 3.71752, Dur Loss: 0.95742, CE Loss: 0.05013, Norm Loss: 1.13116, F0 Loss: 3.17455, LM Loss: 2.23914, Gen Loss: 6.00335, Sty Loss: 0.10371, Diff Loss: 0.18188, DiscLM Loss: 0.00000, GenLM Loss: 1.00020 +INFO:2025-06-11 09:41:08,628: Epoch [7/20], Step [1500/2647], Loss: 0.52285, Disc Loss: 3.76624, Dur Loss: 1.39562, CE Loss: 0.08251, Norm Loss: 1.27747, F0 Loss: 3.60337, LM Loss: 2.34449, Gen Loss: 6.14970, Sty Loss: 0.10103, Diff Loss: 0.30567, DiscLM Loss: 0.00122, GenLM Loss: 0.99177 +INFO:2025-06-11 09:42:56,923: Epoch [7/20], Step [1550/2647], Loss: 0.51617, Disc Loss: 3.71259, Dur Loss: 0.80754, CE Loss: 0.04178, Norm Loss: 1.03340, F0 Loss: 2.72174, LM Loss: 2.08157, Gen Loss: 5.95317, Sty Loss: 0.10285, Diff Loss: 0.27341, DiscLM Loss: 0.00000, GenLM Loss: 1.00097 +INFO:2025-06-11 09:44:45,843: Epoch [7/20], Step [1600/2647], Loss: 0.52801, Disc Loss: 3.84529, Dur Loss: 0.99760, CE Loss: 0.05966, Norm Loss: 0.98870, F0 Loss: 3.31025, LM Loss: 2.33127, Gen Loss: 6.31903, Sty Loss: 0.11068, Diff Loss: 0.19390, DiscLM Loss: 0.00121, GenLM Loss: 1.00183 +INFO:2025-06-11 09:46:33,920: Epoch [7/20], Step [1650/2647], Loss: 0.52054, Disc Loss: 3.68767, Dur Loss: 0.97016, CE Loss: 0.04951, Norm Loss: 1.41589, F0 Loss: 3.01524, LM Loss: 2.24581, Gen Loss: 6.13108, Sty Loss: 0.09107, Diff Loss: 0.27382, DiscLM Loss: 0.00000, GenLM Loss: 0.97874 +INFO:2025-06-11 09:48:21,861: Epoch [7/20], Step [1700/2647], Loss: 0.53115, Disc Loss: 3.78806, Dur Loss: 0.92264, CE Loss: 0.04857, Norm Loss: 0.92713, F0 Loss: 3.58584, LM Loss: 2.20302, Gen Loss: 5.94841, Sty Loss: 0.11357, Diff Loss: 0.22479, DiscLM Loss: 0.00160, GenLM Loss: 0.99984 +INFO:2025-06-11 09:50:07,677: Epoch [7/20], Step [1750/2647], Loss: 0.52648, Disc Loss: 3.75950, Dur Loss: 1.09745, CE Loss: 0.06098, Norm Loss: 1.22259, F0 Loss: 2.83401, LM Loss: 2.02387, Gen Loss: 5.78084, Sty Loss: 0.11443, Diff Loss: 0.24710, DiscLM Loss: 0.00000, GenLM Loss: 1.00069 +INFO:2025-06-11 09:51:55,782: Epoch [7/20], Step [1800/2647], Loss: 0.53118, Disc Loss: 3.73378, Dur Loss: 0.84542, CE Loss: 0.04341, Norm Loss: 1.24001, F0 Loss: 3.14025, LM Loss: 2.15670, Gen Loss: 6.16900, Sty Loss: 0.11132, Diff Loss: 0.19574, DiscLM Loss: 0.00087, GenLM Loss: 0.99907 +INFO:2025-06-11 09:53:43,654: Epoch [7/20], Step [1850/2647], Loss: 0.53378, Disc Loss: 3.76958, Dur Loss: 0.83227, CE Loss: 0.04396, Norm Loss: 1.11704, F0 Loss: 2.48046, LM Loss: 2.19615, Gen Loss: 5.49111, Sty Loss: 0.08111, Diff Loss: 0.35420, DiscLM Loss: 0.00000, GenLM Loss: 0.98810 +INFO:2025-06-11 09:55:32,941: Epoch [7/20], Step [1900/2647], Loss: 0.52461, Disc Loss: 3.79048, Dur Loss: 0.96341, CE Loss: 0.05779, Norm Loss: 0.76949, F0 Loss: 2.71436, LM Loss: 2.10137, Gen Loss: 6.08659, Sty Loss: 0.12093, Diff Loss: 0.20245, DiscLM Loss: 0.00075, GenLM Loss: 0.99150 +INFO:2025-06-11 09:57:21,563: Epoch [7/20], Step [1950/2647], Loss: 0.52403, Disc Loss: 3.80519, Dur Loss: 1.22872, CE Loss: 0.07182, Norm Loss: 1.24356, F0 Loss: 3.13901, LM Loss: 2.11048, Gen Loss: 6.19882, Sty Loss: 0.12334, Diff Loss: 0.26893, DiscLM Loss: 0.00000, GenLM Loss: 1.00693 +INFO:2025-06-11 09:59:10,470: Epoch [7/20], Step [2000/2647], Loss: 0.52908, Disc Loss: 3.77999, Dur Loss: 0.92241, CE Loss: 0.04933, Norm Loss: 1.16554, F0 Loss: 2.53458, LM Loss: 2.05340, Gen Loss: 5.53483, Sty Loss: 0.09256, Diff Loss: 0.17428, DiscLM Loss: 0.00099, GenLM Loss: 1.00199 +INFO:2025-06-11 10:00:58,016: Epoch [7/20], Step [2050/2647], Loss: 0.53569, Disc Loss: 3.76563, Dur Loss: 0.84609, CE Loss: 0.04727, Norm Loss: 1.21041, F0 Loss: 3.71158, LM Loss: 2.19771, Gen Loss: 6.17212, Sty Loss: 0.12442, Diff Loss: 0.26626, DiscLM Loss: 0.00000, GenLM Loss: 0.99162 +INFO:2025-06-11 10:02:47,836: Epoch [7/20], Step [2100/2647], Loss: 0.53185, Disc Loss: 3.85946, Dur Loss: 0.71857, CE Loss: 0.04020, Norm Loss: 0.76183, F0 Loss: 2.97460, LM Loss: 2.21920, Gen Loss: 6.27501, Sty Loss: 0.12594, Diff Loss: 0.15355, DiscLM Loss: 0.00216, GenLM Loss: 0.98427 +INFO:2025-06-11 10:04:38,261: Epoch [7/20], Step [2150/2647], Loss: 0.52752, Disc Loss: 3.72685, Dur Loss: 1.08296, CE Loss: 0.06044, Norm Loss: 1.05219, F0 Loss: 2.78747, LM Loss: 2.19197, Gen Loss: 5.49318, Sty Loss: 0.11781, Diff Loss: 0.29362, DiscLM Loss: 0.00000, GenLM Loss: 1.00014 +INFO:2025-06-11 10:06:27,849: Epoch [7/20], Step [2200/2647], Loss: 0.51682, Disc Loss: 3.75142, Dur Loss: 0.95592, CE Loss: 0.05311, Norm Loss: 1.21637, F0 Loss: 3.08032, LM Loss: 2.22622, Gen Loss: 5.45300, Sty Loss: 0.08945, Diff Loss: 0.23803, DiscLM Loss: 0.00065, GenLM Loss: 0.99969 +INFO:2025-06-11 10:08:16,553: Epoch [7/20], Step [2250/2647], Loss: 0.53442, Disc Loss: 3.82881, Dur Loss: 0.93635, CE Loss: 0.05070, Norm Loss: 1.22357, F0 Loss: 3.05011, LM Loss: 2.19264, Gen Loss: 5.83861, Sty Loss: 0.09987, Diff Loss: 0.30967, DiscLM Loss: 0.00000, GenLM Loss: 0.99519 +INFO:2025-06-11 10:10:06,366: Epoch [7/20], Step [2300/2647], Loss: 0.52676, Disc Loss: 3.73377, Dur Loss: 0.98576, CE Loss: 0.04870, Norm Loss: 1.65941, F0 Loss: 3.20371, LM Loss: 2.22815, Gen Loss: 6.32417, Sty Loss: 0.13544, Diff Loss: 0.28011, DiscLM Loss: 0.00133, GenLM Loss: 1.00334 +INFO:2025-06-11 10:11:53,998: Epoch [7/20], Step [2350/2647], Loss: 0.52604, Disc Loss: 3.77574, Dur Loss: 0.72491, CE Loss: 0.03628, Norm Loss: 0.99246, F0 Loss: 2.74942, LM Loss: 2.09190, Gen Loss: 5.58038, Sty Loss: 0.09573, Diff Loss: 0.26259, DiscLM Loss: 0.00000, GenLM Loss: 1.00404 +INFO:2025-06-11 10:13:42,905: Epoch [7/20], Step [2400/2647], Loss: 0.52843, Disc Loss: 3.74014, Dur Loss: 0.94624, CE Loss: 0.04789, Norm Loss: 1.07024, F0 Loss: 3.97183, LM Loss: 2.13692, Gen Loss: 5.51731, Sty Loss: 0.11197, Diff Loss: 0.19640, DiscLM Loss: 0.00290, GenLM Loss: 1.00050 +INFO:2025-06-11 10:15:32,863: Epoch [7/20], Step [2450/2647], Loss: 0.53776, Disc Loss: 3.80308, Dur Loss: 1.07248, CE Loss: 0.05693, Norm Loss: 1.19328, F0 Loss: 2.93190, LM Loss: 2.27326, Gen Loss: 5.34020, Sty Loss: 0.09585, Diff Loss: 0.24703, DiscLM Loss: 0.00000, GenLM Loss: 0.99896 +INFO:2025-06-11 10:19:07,223: Epoch [7/20], Step [2550/2647], Loss: 1.05182, Disc Loss: 3.76704, Dur Loss: 0.95972, CE Loss: 0.05205, Norm Loss: 1.09085, F0 Loss: 3.44098, LM Loss: 2.17486, Gen Loss: 5.80333, Sty Loss: 0.10878, Diff Loss: 0.20696, DiscLM Loss: 0.00000, GenLM Loss: 0.98374 +INFO:2025-06-11 10:20:54,390: Epoch [7/20], Step [2600/2647], Loss: 0.52842, Disc Loss: 3.72566, Dur Loss: 1.13045, CE Loss: 0.05842, Norm Loss: 1.18470, F0 Loss: 3.02210, LM Loss: 2.37106, Gen Loss: 5.57565, Sty Loss: 0.11181, Diff Loss: 0.21855, DiscLM Loss: 0.00056, GenLM Loss: 1.00815 +INFO:2025-06-11 10:24:53,611: Validation loss: 0.549, Dur loss: 1.012, F0 loss: 3.360 + + + +INFO:2025-06-11 10:26:46,479: Epoch [8/20], Step [50/2647], Loss: 0.52895, Disc Loss: 3.83082, Dur Loss: 1.42409, CE Loss: 0.08100, Norm Loss: 1.31157, F0 Loss: 3.35943, LM Loss: 2.27588, Gen Loss: 5.79110, Sty Loss: 0.13984, Diff Loss: 0.23072, DiscLM Loss: 0.00000, GenLM Loss: 1.00980 +INFO:2025-06-11 10:28:34,609: Epoch [8/20], Step [100/2647], Loss: 0.53697, Disc Loss: 3.82314, Dur Loss: 0.76685, CE Loss: 0.04106, Norm Loss: 0.85047, F0 Loss: 2.42985, LM Loss: 2.17248, Gen Loss: 5.74191, Sty Loss: 0.09710, Diff Loss: 0.21070, DiscLM Loss: 0.00097, GenLM Loss: 0.99694 +INFO:2025-06-11 10:30:22,919: Epoch [8/20], Step [150/2647], Loss: 0.52444, Disc Loss: 3.65938, Dur Loss: 0.81104, CE Loss: 0.04450, Norm Loss: 0.94603, F0 Loss: 2.71824, LM Loss: 2.22094, Gen Loss: 6.08655, Sty Loss: 0.11245, Diff Loss: 0.22853, DiscLM Loss: 0.00000, GenLM Loss: 1.00016 +INFO:2025-06-11 10:32:11,150: Epoch [8/20], Step [200/2647], Loss: 0.52758, Disc Loss: 3.68259, Dur Loss: 1.03335, CE Loss: 0.05988, Norm Loss: 1.25218, F0 Loss: 3.23911, LM Loss: 2.23790, Gen Loss: 6.09470, Sty Loss: 0.11970, Diff Loss: 0.24096, DiscLM Loss: 0.00353, GenLM Loss: 0.99004 +INFO:2025-06-11 10:34:01,167: Epoch [8/20], Step [250/2647], Loss: 0.52584, Disc Loss: 3.78254, Dur Loss: 0.84550, CE Loss: 0.04222, Norm Loss: 1.10105, F0 Loss: 3.30420, LM Loss: 2.24761, Gen Loss: 5.69705, Sty Loss: 0.09858, Diff Loss: 0.22015, DiscLM Loss: 0.00000, GenLM Loss: 0.99435 +INFO:2025-06-11 10:35:51,202: Epoch [8/20], Step [300/2647], Loss: 0.52560, Disc Loss: 3.82372, Dur Loss: 0.78736, CE Loss: 0.04691, Norm Loss: 0.87923, F0 Loss: 2.56062, LM Loss: 2.14456, Gen Loss: 5.83039, Sty Loss: 0.11524, Diff Loss: 0.15204, DiscLM Loss: 0.00528, GenLM Loss: 1.01327 +INFO:2025-06-11 10:37:42,745: Epoch [8/20], Step [350/2647], Loss: 0.52531, Disc Loss: 3.86151, Dur Loss: 1.10847, CE Loss: 0.06166, Norm Loss: 1.01918, F0 Loss: 3.61445, LM Loss: 2.18279, Gen Loss: 6.06606, Sty Loss: 0.10980, Diff Loss: 0.32955, DiscLM Loss: 0.00000, GenLM Loss: 0.98925 +INFO:2025-06-11 10:39:34,487: Epoch [8/20], Step [400/2647], Loss: 0.51869, Disc Loss: 3.71240, Dur Loss: 0.93128, CE Loss: 0.05016, Norm Loss: 0.98534, F0 Loss: 4.50827, LM Loss: 2.19186, Gen Loss: 5.77258, Sty Loss: 0.12106, Diff Loss: 0.22933, DiscLM Loss: 0.00107, GenLM Loss: 1.00537 +INFO:2025-06-11 10:41:24,228: Epoch [8/20], Step [450/2647], Loss: 0.53182, Disc Loss: 3.76341, Dur Loss: 1.05109, CE Loss: 0.05864, Norm Loss: 1.06969, F0 Loss: 2.96860, LM Loss: 2.26382, Gen Loss: 5.53023, Sty Loss: 0.10308, Diff Loss: 0.31847, DiscLM Loss: 0.00000, GenLM Loss: 0.98581 +INFO:2025-06-11 10:43:13,662: Epoch [8/20], Step [500/2647], Loss: 0.52626, Disc Loss: 3.82739, Dur Loss: 0.94910, CE Loss: 0.05194, Norm Loss: 0.84135, F0 Loss: 3.43815, LM Loss: 2.18640, Gen Loss: 6.17925, Sty Loss: 0.12286, Diff Loss: 0.26947, DiscLM Loss: 0.00117, GenLM Loss: 0.99870 +INFO:2025-06-11 10:45:01,746: Epoch [8/20], Step [550/2647], Loss: 0.54493, Disc Loss: 3.83104, Dur Loss: 1.11025, CE Loss: 0.05673, Norm Loss: 0.98280, F0 Loss: 2.60478, LM Loss: 2.17791, Gen Loss: 5.46176, Sty Loss: 0.14113, Diff Loss: 0.23486, DiscLM Loss: 0.00000, GenLM Loss: 1.00076 +INFO:2025-06-11 10:46:52,433: Epoch [8/20], Step [600/2647], Loss: 0.53969, Disc Loss: 3.78320, Dur Loss: 1.11437, CE Loss: 0.06270, Norm Loss: 1.27460, F0 Loss: 3.13941, LM Loss: 2.34053, Gen Loss: 5.52052, Sty Loss: 0.11449, Diff Loss: 0.27206, DiscLM Loss: 0.00088, GenLM Loss: 0.98825 +INFO:2025-06-11 10:48:41,949: Epoch [8/20], Step [650/2647], Loss: 0.52874, Disc Loss: 3.80180, Dur Loss: 0.94579, CE Loss: 0.05486, Norm Loss: 1.03103, F0 Loss: 3.19439, LM Loss: 2.18403, Gen Loss: 5.66241, Sty Loss: 0.11188, Diff Loss: 0.30828, DiscLM Loss: 0.00000, GenLM Loss: 1.00708 +INFO:2025-06-11 10:50:32,607: Epoch [8/20], Step [700/2647], Loss: 0.52922, Disc Loss: 3.80417, Dur Loss: 1.01501, CE Loss: 0.05639, Norm Loss: 0.94582, F0 Loss: 3.02422, LM Loss: 2.14986, Gen Loss: 5.27952, Sty Loss: 0.11465, Diff Loss: 0.40946, DiscLM Loss: 0.00263, GenLM Loss: 0.99146 +INFO:2025-06-11 10:52:23,727: Epoch [8/20], Step [750/2647], Loss: 0.52777, Disc Loss: 3.76370, Dur Loss: 1.10122, CE Loss: 0.06206, Norm Loss: 1.28521, F0 Loss: 2.80173, LM Loss: 2.19742, Gen Loss: 6.04218, Sty Loss: 0.12590, Diff Loss: 0.20209, DiscLM Loss: 0.00000, GenLM Loss: 1.01747 +INFO:2025-06-11 10:54:11,787: Epoch [8/20], Step [800/2647], Loss: 0.53194, Disc Loss: 3.81265, Dur Loss: 0.88838, CE Loss: 0.05054, Norm Loss: 1.72027, F0 Loss: 3.60890, LM Loss: 2.24995, Gen Loss: 5.37311, Sty Loss: 0.09454, Diff Loss: 0.21764, DiscLM Loss: 0.00282, GenLM Loss: 0.99752 +INFO:2025-06-11 10:55:58,750: Epoch [8/20], Step [850/2647], Loss: 0.52911, Disc Loss: 3.84812, Dur Loss: 0.93656, CE Loss: 0.04939, Norm Loss: 1.21013, F0 Loss: 2.39601, LM Loss: 2.19948, Gen Loss: 5.56678, Sty Loss: 0.10666, Diff Loss: 0.21646, DiscLM Loss: 0.00000, GenLM Loss: 0.99908 +INFO:2025-06-11 10:57:46,770: Epoch [8/20], Step [900/2647], Loss: 0.52238, Disc Loss: 3.83334, Dur Loss: 0.98981, CE Loss: 0.05992, Norm Loss: 1.27861, F0 Loss: 3.57876, LM Loss: 2.28034, Gen Loss: 5.74585, Sty Loss: 0.14636, Diff Loss: 0.29015, DiscLM Loss: 0.00956, GenLM Loss: 0.98685 +INFO:2025-06-11 10:59:35,628: Epoch [8/20], Step [950/2647], Loss: 0.52415, Disc Loss: 3.77794, Dur Loss: 0.94225, CE Loss: 0.05728, Norm Loss: 0.95468, F0 Loss: 3.59573, LM Loss: 2.27260, Gen Loss: 6.60585, Sty Loss: 0.12368, Diff Loss: 0.24335, DiscLM Loss: 0.00000, GenLM Loss: 1.00556 +INFO:2025-06-11 11:01:23,189: Epoch [8/20], Step [1000/2647], Loss: 0.52894, Disc Loss: 3.69052, Dur Loss: 0.94646, CE Loss: 0.05999, Norm Loss: 0.87351, F0 Loss: 3.39958, LM Loss: 2.27249, Gen Loss: 7.32851, Sty Loss: 0.12461, Diff Loss: 0.25806, DiscLM Loss: 0.00851, GenLM Loss: 0.98969 +INFO:2025-06-11 11:03:15,318: Epoch [8/20], Step [1050/2647], Loss: 0.52021, Disc Loss: 3.76445, Dur Loss: 0.98442, CE Loss: 0.05510, Norm Loss: 1.05615, F0 Loss: 3.38875, LM Loss: 2.17522, Gen Loss: 6.28151, Sty Loss: 0.13084, Diff Loss: 0.30870, DiscLM Loss: 0.00000, GenLM Loss: 1.00242 +INFO:2025-06-11 11:05:04,591: Epoch [8/20], Step [1100/2647], Loss: 0.52929, Disc Loss: 3.80236, Dur Loss: 0.98540, CE Loss: 0.05229, Norm Loss: 0.96444, F0 Loss: 3.21723, LM Loss: 2.20009, Gen Loss: 6.09238, Sty Loss: 0.11990, Diff Loss: 0.22102, DiscLM Loss: 0.00220, GenLM Loss: 1.00143 +INFO:2025-06-11 11:06:54,688: Epoch [8/20], Step [1150/2647], Loss: 0.52572, Disc Loss: 3.82211, Dur Loss: 1.13277, CE Loss: 0.06427, Norm Loss: 1.00591, F0 Loss: 3.26995, LM Loss: 2.15489, Gen Loss: 5.75216, Sty Loss: 0.12800, Diff Loss: 0.23188, DiscLM Loss: 0.00000, GenLM Loss: 0.99052 +INFO:2025-06-11 11:08:46,581: Epoch [8/20], Step [1200/2647], Loss: 0.53723, Disc Loss: 3.74409, Dur Loss: 1.30274, CE Loss: 0.06491, Norm Loss: 1.13721, F0 Loss: 2.91879, LM Loss: 2.17171, Gen Loss: 5.30034, Sty Loss: 0.11857, Diff Loss: 0.28542, DiscLM Loss: 0.00136, GenLM Loss: 0.99877 +INFO:2025-06-11 11:10:34,680: Epoch [8/20], Step [1250/2647], Loss: 0.52922, Disc Loss: 3.72649, Dur Loss: 1.07023, CE Loss: 0.06258, Norm Loss: 1.00072, F0 Loss: 3.11682, LM Loss: 2.29232, Gen Loss: 6.53149, Sty Loss: 0.09353, Diff Loss: 0.20888, DiscLM Loss: 0.00000, GenLM Loss: 1.01830 +INFO:2025-06-11 11:12:22,576: Epoch [8/20], Step [1300/2647], Loss: 0.54346, Disc Loss: 3.82908, Dur Loss: 1.03327, CE Loss: 0.05437, Norm Loss: 1.06360, F0 Loss: 3.12020, LM Loss: 2.08960, Gen Loss: 5.36403, Sty Loss: 0.10129, Diff Loss: 0.27206, DiscLM Loss: 0.00079, GenLM Loss: 0.98891 +INFO:2025-06-11 11:14:08,635: Epoch [8/20], Step [1350/2647], Loss: 0.53864, Disc Loss: 3.78527, Dur Loss: 0.71302, CE Loss: 0.03790, Norm Loss: 0.87739, F0 Loss: 2.33831, LM Loss: 2.14035, Gen Loss: 5.76069, Sty Loss: 0.11187, Diff Loss: 0.33715, DiscLM Loss: 0.00000, GenLM Loss: 1.00983 +INFO:2025-06-11 11:15:57,073: Epoch [8/20], Step [1400/2647], Loss: 0.53526, Disc Loss: 3.80473, Dur Loss: 1.13310, CE Loss: 0.05767, Norm Loss: 1.21807, F0 Loss: 3.10207, LM Loss: 2.17208, Gen Loss: 5.23624, Sty Loss: 0.12679, Diff Loss: 0.21548, DiscLM Loss: 0.00096, GenLM Loss: 1.00578 +INFO:2025-06-11 11:17:43,720: Epoch [8/20], Step [1450/2647], Loss: 0.51697, Disc Loss: 3.80067, Dur Loss: 0.88467, CE Loss: 0.04745, Norm Loss: 0.95324, F0 Loss: 3.18024, LM Loss: 2.14904, Gen Loss: 5.64268, Sty Loss: 0.10005, Diff Loss: 0.20839, DiscLM Loss: 0.00000, GenLM Loss: 0.99796 +INFO:2025-06-11 11:19:31,987: Epoch [8/20], Step [1500/2647], Loss: 0.53218, Disc Loss: 3.70543, Dur Loss: 0.86111, CE Loss: 0.04265, Norm Loss: 1.25275, F0 Loss: 4.76625, LM Loss: 2.22111, Gen Loss: 6.22160, Sty Loss: 0.12495, Diff Loss: 0.31104, DiscLM Loss: 0.00061, GenLM Loss: 1.01437 +INFO:2025-06-11 11:21:22,730: Epoch [8/20], Step [1550/2647], Loss: 0.53798, Disc Loss: 3.72436, Dur Loss: 1.06087, CE Loss: 0.05897, Norm Loss: 1.41921, F0 Loss: 3.59401, LM Loss: 2.15778, Gen Loss: 6.37160, Sty Loss: 0.10329, Diff Loss: 0.29157, DiscLM Loss: 0.00000, GenLM Loss: 0.98781 +INFO:2025-06-11 11:23:10,633: Epoch [8/20], Step [1600/2647], Loss: 0.52860, Disc Loss: 3.81965, Dur Loss: 0.97999, CE Loss: 0.05286, Norm Loss: 1.19236, F0 Loss: 3.69457, LM Loss: 2.18442, Gen Loss: 5.74648, Sty Loss: 0.12578, Diff Loss: 0.18148, DiscLM Loss: 0.00165, GenLM Loss: 1.00377 +INFO:2025-06-11 11:24:58,433: Epoch [8/20], Step [1650/2647], Loss: 0.52091, Disc Loss: 3.80386, Dur Loss: 0.95332, CE Loss: 0.05022, Norm Loss: 1.05436, F0 Loss: 2.99888, LM Loss: 2.21205, Gen Loss: 5.20801, Sty Loss: 0.10892, Diff Loss: 0.28486, DiscLM Loss: 0.00000, GenLM Loss: 0.99558 +INFO:2025-06-11 11:26:48,598: Epoch [8/20], Step [1700/2647], Loss: 0.51962, Disc Loss: 3.65669, Dur Loss: 0.80876, CE Loss: 0.04199, Norm Loss: 0.92734, F0 Loss: 3.81188, LM Loss: 2.08391, Gen Loss: 6.64060, Sty Loss: 0.11812, Diff Loss: 0.29493, DiscLM Loss: 0.00047, GenLM Loss: 0.99758 +INFO:2025-06-11 11:28:38,706: Epoch [8/20], Step [1750/2647], Loss: 0.52963, Disc Loss: 3.83937, Dur Loss: 1.06808, CE Loss: 0.05791, Norm Loss: 0.97257, F0 Loss: 3.56695, LM Loss: 2.24752, Gen Loss: 5.52895, Sty Loss: 0.10464, Diff Loss: 0.34357, DiscLM Loss: 0.00000, GenLM Loss: 1.00399 +INFO:2025-06-11 11:30:26,474: Epoch [8/20], Step [1800/2647], Loss: 0.52571, Disc Loss: 3.72397, Dur Loss: 1.01032, CE Loss: 0.04830, Norm Loss: 1.14469, F0 Loss: 3.03847, LM Loss: 2.24553, Gen Loss: 6.15643, Sty Loss: 0.11232, Diff Loss: 0.19525, DiscLM Loss: 0.00064, GenLM Loss: 0.99579 +INFO:2025-06-11 11:32:18,242: Epoch [8/20], Step [1850/2647], Loss: 0.52259, Disc Loss: 3.76018, Dur Loss: 0.86085, CE Loss: 0.04501, Norm Loss: 1.12400, F0 Loss: 2.72979, LM Loss: 2.17553, Gen Loss: 5.99990, Sty Loss: 0.10972, Diff Loss: 0.25292, DiscLM Loss: 0.00000, GenLM Loss: 1.02484 +INFO:2025-06-11 11:34:05,877: Epoch [8/20], Step [1900/2647], Loss: 0.52562, Disc Loss: 3.76263, Dur Loss: 0.92239, CE Loss: 0.04777, Norm Loss: 1.02419, F0 Loss: 3.23948, LM Loss: 2.18688, Gen Loss: 5.76562, Sty Loss: 0.12967, Diff Loss: 0.15512, DiscLM Loss: 0.00287, GenLM Loss: 1.00042 +INFO:2025-06-11 11:35:54,869: Epoch [8/20], Step [1950/2647], Loss: 0.53099, Disc Loss: 3.78683, Dur Loss: 1.15849, CE Loss: 0.06481, Norm Loss: 1.05321, F0 Loss: 3.48271, LM Loss: 2.18188, Gen Loss: 5.67396, Sty Loss: 0.13890, Diff Loss: 0.27785, DiscLM Loss: 0.00000, GenLM Loss: 1.01643 +INFO:2025-06-11 11:37:44,932: Epoch [8/20], Step [2000/2647], Loss: 0.52688, Disc Loss: 3.74466, Dur Loss: 0.89192, CE Loss: 0.04788, Norm Loss: 0.84611, F0 Loss: 2.53172, LM Loss: 2.12953, Gen Loss: 5.82111, Sty Loss: 0.10301, Diff Loss: 0.17143, DiscLM Loss: 0.00224, GenLM Loss: 0.99385 +INFO:2025-06-11 11:39:35,103: Epoch [8/20], Step [2050/2647], Loss: 0.52917, Disc Loss: 3.70583, Dur Loss: 1.07364, CE Loss: 0.05977, Norm Loss: 0.82133, F0 Loss: 2.67944, LM Loss: 2.05838, Gen Loss: 6.29025, Sty Loss: 0.10967, Diff Loss: 0.20527, DiscLM Loss: 0.00000, GenLM Loss: 1.00314 +INFO:2025-06-11 11:41:23,149: Epoch [8/20], Step [2100/2647], Loss: 0.52533, Disc Loss: 3.79593, Dur Loss: 0.87676, CE Loss: 0.04914, Norm Loss: 1.14519, F0 Loss: 2.93386, LM Loss: 2.17572, Gen Loss: 5.73229, Sty Loss: 0.11200, Diff Loss: 0.24785, DiscLM Loss: 0.01585, GenLM Loss: 0.95795 +INFO:2025-06-11 11:43:10,070: Epoch [8/20], Step [2150/2647], Loss: 0.51476, Disc Loss: 3.67886, Dur Loss: 0.91579, CE Loss: 0.04804, Norm Loss: 1.16092, F0 Loss: 4.09003, LM Loss: 2.29600, Gen Loss: 6.19472, Sty Loss: 0.10910, Diff Loss: 0.33303, DiscLM Loss: 0.00000, GenLM Loss: 0.96449 +INFO:2025-06-11 11:44:57,461: Epoch [8/20], Step [2200/2647], Loss: 0.52484, Disc Loss: 3.80849, Dur Loss: 1.01674, CE Loss: 0.05460, Norm Loss: 1.14071, F0 Loss: 2.86087, LM Loss: 2.07594, Gen Loss: 6.26066, Sty Loss: 0.12490, Diff Loss: 0.41266, DiscLM Loss: 0.00191, GenLM Loss: 0.99738 +INFO:2025-06-11 11:46:44,321: Epoch [8/20], Step [2250/2647], Loss: 0.53001, Disc Loss: 3.75690, Dur Loss: 0.95144, CE Loss: 0.05458, Norm Loss: 1.14103, F0 Loss: 3.59194, LM Loss: 2.22057, Gen Loss: 5.98159, Sty Loss: 0.11234, Diff Loss: 0.22451, DiscLM Loss: 0.00000, GenLM Loss: 0.98272 +INFO:2025-06-11 11:48:33,555: Epoch [8/20], Step [2300/2647], Loss: 0.52875, Disc Loss: 3.77845, Dur Loss: 0.93113, CE Loss: 0.05430, Norm Loss: 1.17886, F0 Loss: 2.62447, LM Loss: 2.22921, Gen Loss: 5.90938, Sty Loss: 0.12259, Diff Loss: 0.21040, DiscLM Loss: 0.01400, GenLM Loss: 0.97385 +INFO:2025-06-11 11:50:26,562: Epoch [8/20], Step [2350/2647], Loss: 0.51766, Disc Loss: 3.79696, Dur Loss: 1.23369, CE Loss: 0.06905, Norm Loss: 1.08360, F0 Loss: 3.08257, LM Loss: 2.20586, Gen Loss: 5.91601, Sty Loss: 0.12704, Diff Loss: 0.23044, DiscLM Loss: 0.00000, GenLM Loss: 1.01903 +INFO:2025-06-11 11:52:15,048: Epoch [8/20], Step [2400/2647], Loss: 0.52031, Disc Loss: 3.74959, Dur Loss: 1.27368, CE Loss: 0.07308, Norm Loss: 1.20219, F0 Loss: 3.52482, LM Loss: 2.20528, Gen Loss: 5.37088, Sty Loss: 0.11483, Diff Loss: 0.25098, DiscLM Loss: 0.00109, GenLM Loss: 1.00368 +INFO:2025-06-11 11:53:59,805: Epoch [8/20], Step [2450/2647], Loss: 0.53341, Disc Loss: 3.71621, Dur Loss: 0.93432, CE Loss: 0.05012, Norm Loss: 0.96184, F0 Loss: 3.40406, LM Loss: 2.18020, Gen Loss: 5.68697, Sty Loss: 0.09749, Diff Loss: 0.23633, DiscLM Loss: 0.00000, GenLM Loss: 0.99846 +INFO:2025-06-11 11:55:48,803: Epoch [8/20], Step [2500/2647], Loss: 0.52860, Disc Loss: 3.76017, Dur Loss: 0.76232, CE Loss: 0.04062, Norm Loss: 0.76054, F0 Loss: 2.28357, LM Loss: 2.10124, Gen Loss: 6.07282, Sty Loss: 0.09120, Diff Loss: 0.28491, DiscLM Loss: 0.00083, GenLM Loss: 1.00923 +INFO:2025-06-11 11:57:38,063: Epoch [8/20], Step [2550/2647], Loss: 0.53850, Disc Loss: 3.95996, Dur Loss: 1.13372, CE Loss: 0.06883, Norm Loss: 1.35037, F0 Loss: 2.97741, LM Loss: 2.20283, Gen Loss: 5.97161, Sty Loss: 0.11835, Diff Loss: 0.36988, DiscLM Loss: 0.00000, GenLM Loss: 0.98959 +INFO:2025-06-11 11:59:26,581: Epoch [8/20], Step [2600/2647], Loss: 0.52769, Disc Loss: 3.79751, Dur Loss: 0.82849, CE Loss: 0.04363, Norm Loss: 0.96005, F0 Loss: 2.72072, LM Loss: 2.05661, Gen Loss: 5.85246, Sty Loss: 0.09279, Diff Loss: 0.23618, DiscLM Loss: 0.00113, GenLM Loss: 0.98254 +INFO:2025-06-11 12:03:29,103: Validation loss: 0.536, Dur loss: 1.004, F0 loss: 3.379 + + + +INFO:2025-06-11 12:05:20,161: Epoch [9/20], Step [50/2647], Loss: 0.52911, Disc Loss: 3.83181, Dur Loss: 1.32639, CE Loss: 0.07016, Norm Loss: 0.74175, F0 Loss: 3.46304, LM Loss: 2.18558, Gen Loss: 6.89716, Sty Loss: 0.11628, Diff Loss: 0.36618, DiscLM Loss: 0.00000, GenLM Loss: 1.00012 +INFO:2025-06-11 12:07:12,865: Epoch [9/20], Step [100/2647], Loss: 0.52296, Disc Loss: 3.77927, Dur Loss: 1.30871, CE Loss: 0.07086, Norm Loss: 1.30620, F0 Loss: 3.66128, LM Loss: 2.24188, Gen Loss: 5.29389, Sty Loss: 0.12469, Diff Loss: 0.28621, DiscLM Loss: 0.00093, GenLM Loss: 1.00525 +INFO:2025-06-11 12:09:02,437: Epoch [9/20], Step [150/2647], Loss: 0.52694, Disc Loss: 3.80380, Dur Loss: 1.06530, CE Loss: 0.06060, Norm Loss: 1.31263, F0 Loss: 3.81323, LM Loss: 2.13899, Gen Loss: 5.45759, Sty Loss: 0.11919, Diff Loss: 0.21519, DiscLM Loss: 0.00000, GenLM Loss: 0.98812 +INFO:2025-06-11 12:10:50,604: Epoch [9/20], Step [200/2647], Loss: 0.52674, Disc Loss: 3.82367, Dur Loss: 1.19521, CE Loss: 0.08171, Norm Loss: 1.48328, F0 Loss: 3.50732, LM Loss: 2.16539, Gen Loss: 5.35940, Sty Loss: 0.14776, Diff Loss: 0.17443, DiscLM Loss: 0.00112, GenLM Loss: 0.99793 +INFO:2025-06-11 12:12:40,225: Epoch [9/20], Step [250/2647], Loss: 0.52791, Disc Loss: 3.74063, Dur Loss: 0.93600, CE Loss: 0.05162, Norm Loss: 0.71600, F0 Loss: 2.24252, LM Loss: 2.20814, Gen Loss: 6.01471, Sty Loss: 0.12692, Diff Loss: 0.32761, DiscLM Loss: 0.00000, GenLM Loss: 1.00099 +INFO:2025-06-11 12:14:27,848: Epoch [9/20], Step [300/2647], Loss: 0.51878, Disc Loss: 3.76111, Dur Loss: 0.73207, CE Loss: 0.03945, Norm Loss: 0.99733, F0 Loss: 3.16955, LM Loss: 2.23577, Gen Loss: 7.59268, Sty Loss: 0.16089, Diff Loss: 0.51095, DiscLM Loss: 0.00276, GenLM Loss: 0.97755 +INFO:2025-06-11 12:16:17,894: Epoch [9/20], Step [350/2647], Loss: 0.53490, Disc Loss: 3.76313, Dur Loss: 0.81443, CE Loss: 0.04377, Norm Loss: 1.09384, F0 Loss: 4.08923, LM Loss: 2.12246, Gen Loss: 5.76952, Sty Loss: 0.10790, Diff Loss: 0.31740, DiscLM Loss: 0.00000, GenLM Loss: 1.00856 +INFO:2025-06-11 12:18:07,101: Epoch [9/20], Step [400/2647], Loss: 0.52329, Disc Loss: 3.70583, Dur Loss: 1.06560, CE Loss: 0.05765, Norm Loss: 1.00999, F0 Loss: 3.23379, LM Loss: 2.15855, Gen Loss: 6.15311, Sty Loss: 0.09160, Diff Loss: 0.18996, DiscLM Loss: 0.00071, GenLM Loss: 0.99982 +INFO:2025-06-11 12:19:55,959: Epoch [9/20], Step [450/2647], Loss: 0.53169, Disc Loss: 3.86265, Dur Loss: 0.86093, CE Loss: 0.04318, Norm Loss: 1.00336, F0 Loss: 3.09167, LM Loss: 2.15482, Gen Loss: 6.01100, Sty Loss: 0.10850, Diff Loss: 0.15590, DiscLM Loss: 0.00000, GenLM Loss: 0.99341 +INFO:2025-06-11 12:21:43,666: Epoch [9/20], Step [500/2647], Loss: 0.52625, Disc Loss: 3.74422, Dur Loss: 0.84051, CE Loss: 0.04661, Norm Loss: 0.94442, F0 Loss: 2.80668, LM Loss: 2.25907, Gen Loss: 6.37947, Sty Loss: 0.12766, Diff Loss: 0.24954, DiscLM Loss: 0.00042, GenLM Loss: 1.00813 +INFO:2025-06-11 12:23:31,796: Epoch [9/20], Step [550/2647], Loss: 0.53049, Disc Loss: 3.59038, Dur Loss: 0.95339, CE Loss: 0.05584, Norm Loss: 1.17174, F0 Loss: 3.63774, LM Loss: 2.10616, Gen Loss: 6.59901, Sty Loss: 0.12888, Diff Loss: 0.36910, DiscLM Loss: 0.00000, GenLM Loss: 0.99540 +INFO:2025-06-11 12:25:21,084: Epoch [9/20], Step [600/2647], Loss: 0.52011, Disc Loss: 3.77415, Dur Loss: 1.20151, CE Loss: 0.06765, Norm Loss: 1.51775, F0 Loss: 4.02223, LM Loss: 2.33203, Gen Loss: 5.69923, Sty Loss: 0.11395, Diff Loss: 0.27501, DiscLM Loss: 0.00061, GenLM Loss: 0.99714 +INFO:2025-06-11 12:27:09,663: Epoch [9/20], Step [650/2647], Loss: 0.52003, Disc Loss: 3.81128, Dur Loss: 0.82019, CE Loss: 0.04842, Norm Loss: 0.97007, F0 Loss: 3.42288, LM Loss: 2.24402, Gen Loss: 5.72253, Sty Loss: 0.09477, Diff Loss: 0.17622, DiscLM Loss: 0.00000, GenLM Loss: 0.99042 +INFO:2025-06-11 12:28:59,696: Epoch [9/20], Step [700/2647], Loss: 0.53263, Disc Loss: 3.83170, Dur Loss: 0.97125, CE Loss: 0.05270, Norm Loss: 1.04778, F0 Loss: 3.38900, LM Loss: 2.03851, Gen Loss: 5.69093, Sty Loss: 0.11200, Diff Loss: 0.32672, DiscLM Loss: 0.00105, GenLM Loss: 0.99157 +INFO:2025-06-11 12:30:50,655: Epoch [9/20], Step [750/2647], Loss: 0.52047, Disc Loss: 3.73067, Dur Loss: 1.34507, CE Loss: 0.08022, Norm Loss: 1.05112, F0 Loss: 3.16904, LM Loss: 2.29993, Gen Loss: 7.03241, Sty Loss: 0.16756, Diff Loss: 0.32265, DiscLM Loss: 0.00000, GenLM Loss: 1.03411 +INFO:2025-06-11 12:32:40,557: Epoch [9/20], Step [800/2647], Loss: 0.52877, Disc Loss: 3.74100, Dur Loss: 1.05685, CE Loss: 0.05753, Norm Loss: 1.07259, F0 Loss: 3.07100, LM Loss: 2.20256, Gen Loss: 6.07963, Sty Loss: 0.11846, Diff Loss: 0.27824, DiscLM Loss: 0.00109, GenLM Loss: 0.99419 +INFO:2025-06-11 12:34:31,914: Epoch [9/20], Step [850/2647], Loss: 0.52297, Disc Loss: 3.73364, Dur Loss: 0.90609, CE Loss: 0.04844, Norm Loss: 1.05883, F0 Loss: 3.27957, LM Loss: 2.16111, Gen Loss: 6.18116, Sty Loss: 0.13669, Diff Loss: 0.23035, DiscLM Loss: 0.00000, GenLM Loss: 1.00010 +INFO:2025-06-11 12:36:18,886: Epoch [9/20], Step [900/2647], Loss: 0.52945, Disc Loss: 3.72026, Dur Loss: 0.83542, CE Loss: 0.04083, Norm Loss: 0.87340, F0 Loss: 2.83074, LM Loss: 2.18859, Gen Loss: 6.00007, Sty Loss: 0.13853, Diff Loss: 0.34267, DiscLM Loss: 0.00215, GenLM Loss: 1.02089 +INFO:2025-06-11 12:38:09,059: Epoch [9/20], Step [950/2647], Loss: 0.52145, Disc Loss: 3.79461, Dur Loss: 0.90662, CE Loss: 0.05141, Norm Loss: 1.15429, F0 Loss: 2.64543, LM Loss: 2.29394, Gen Loss: 6.70608, Sty Loss: 0.14822, Diff Loss: 0.36085, DiscLM Loss: 0.00000, GenLM Loss: 0.98025 +INFO:2025-06-11 12:39:58,539: Epoch [9/20], Step [1000/2647], Loss: 0.54220, Disc Loss: 3.67223, Dur Loss: 1.12204, CE Loss: 0.06174, Norm Loss: 1.26753, F0 Loss: 3.16345, LM Loss: 2.19674, Gen Loss: 5.74358, Sty Loss: 0.09683, Diff Loss: 0.21663, DiscLM Loss: 0.00231, GenLM Loss: 1.00759 +INFO:2025-06-11 12:41:45,732: Epoch [9/20], Step [1050/2647], Loss: 0.53164, Disc Loss: 3.74208, Dur Loss: 0.86947, CE Loss: 0.04728, Norm Loss: 0.76955, F0 Loss: 2.86686, LM Loss: 2.13760, Gen Loss: 5.84819, Sty Loss: 0.12075, Diff Loss: 0.36709, DiscLM Loss: 0.00000, GenLM Loss: 0.98316 +INFO:2025-06-11 12:43:33,266: Epoch [9/20], Step [1100/2647], Loss: 0.52781, Disc Loss: 3.77581, Dur Loss: 0.76304, CE Loss: 0.03947, Norm Loss: 0.90048, F0 Loss: 3.04776, LM Loss: 2.15063, Gen Loss: 5.60579, Sty Loss: 0.11305, Diff Loss: 0.29727, DiscLM Loss: 0.05043, GenLM Loss: 0.93440 +INFO:2025-06-11 12:45:21,105: Epoch [9/20], Step [1150/2647], Loss: 0.53609, Disc Loss: 3.88202, Dur Loss: 1.01428, CE Loss: 0.05272, Norm Loss: 1.42251, F0 Loss: 2.86590, LM Loss: 2.24175, Gen Loss: 5.94089, Sty Loss: 0.11623, Diff Loss: 0.36608, DiscLM Loss: 0.00000, GenLM Loss: 0.99036 +INFO:2025-06-11 12:47:10,403: Epoch [9/20], Step [1200/2647], Loss: 0.53262, Disc Loss: 3.80371, Dur Loss: 1.12080, CE Loss: 0.06478, Norm Loss: 1.09304, F0 Loss: 2.69628, LM Loss: 2.23288, Gen Loss: 5.44033, Sty Loss: 0.10714, Diff Loss: 0.25623, DiscLM Loss: 0.01190, GenLM Loss: 0.95842 +INFO:2025-06-11 12:49:00,077: Epoch [9/20], Step [1250/2647], Loss: 0.52648, Disc Loss: 3.73476, Dur Loss: 0.96429, CE Loss: 0.05625, Norm Loss: 0.95635, F0 Loss: 3.22586, LM Loss: 2.19914, Gen Loss: 5.61376, Sty Loss: 0.11212, Diff Loss: 0.37230, DiscLM Loss: 0.00000, GenLM Loss: 1.01879 +INFO:2025-06-11 12:52:38,147: Epoch [9/20], Step [1350/2647], Loss: 1.05897, Disc Loss: 3.73795, Dur Loss: 0.83595, CE Loss: 0.04769, Norm Loss: 1.15697, F0 Loss: 3.65889, LM Loss: 2.21543, Gen Loss: 5.66985, Sty Loss: 0.13834, Diff Loss: 0.26705, DiscLM Loss: 0.00000, GenLM Loss: 1.07461 +INFO:2025-06-11 12:54:27,348: Epoch [9/20], Step [1400/2647], Loss: 0.53209, Disc Loss: 3.78085, Dur Loss: 0.72157, CE Loss: 0.04094, Norm Loss: 0.87983, F0 Loss: 2.59194, LM Loss: 2.22936, Gen Loss: 5.62492, Sty Loss: 0.10821, Diff Loss: 0.32149, DiscLM Loss: 0.00581, GenLM Loss: 1.00652 +INFO:2025-06-11 12:56:16,068: Epoch [9/20], Step [1450/2647], Loss: 0.53714, Disc Loss: 3.67134, Dur Loss: 0.98883, CE Loss: 0.05262, Norm Loss: 1.10746, F0 Loss: 3.35550, LM Loss: 2.16614, Gen Loss: 6.45942, Sty Loss: 0.11949, Diff Loss: 0.26386, DiscLM Loss: 0.00000, GenLM Loss: 0.95895 +INFO:2025-06-11 12:58:02,751: Epoch [9/20], Step [1500/2647], Loss: 0.52774, Disc Loss: 3.79738, Dur Loss: 1.08278, CE Loss: 0.06801, Norm Loss: 1.05918, F0 Loss: 3.14677, LM Loss: 2.33874, Gen Loss: 6.28927, Sty Loss: 0.16224, Diff Loss: 0.30673, DiscLM Loss: 0.02658, GenLM Loss: 0.99759 +INFO:2025-06-11 12:59:52,575: Epoch [9/20], Step [1550/2647], Loss: 0.52754, Disc Loss: 3.85107, Dur Loss: 0.81551, CE Loss: 0.04342, Norm Loss: 0.89075, F0 Loss: 2.94036, LM Loss: 2.13930, Gen Loss: 5.54411, Sty Loss: 0.09909, Diff Loss: 0.20804, DiscLM Loss: 0.00000, GenLM Loss: 0.99278 +INFO:2025-06-11 13:01:40,424: Epoch [9/20], Step [1600/2647], Loss: 0.52720, Disc Loss: 3.82614, Dur Loss: 1.01035, CE Loss: 0.05467, Norm Loss: 0.96744, F0 Loss: 4.13810, LM Loss: 2.09399, Gen Loss: 5.73624, Sty Loss: 0.14385, Diff Loss: 0.26461, DiscLM Loss: 0.00362, GenLM Loss: 0.94432 +INFO:2025-06-11 13:03:26,665: Epoch [9/20], Step [1650/2647], Loss: 0.52533, Disc Loss: 3.69507, Dur Loss: 0.70365, CE Loss: 0.03628, Norm Loss: 0.79121, F0 Loss: 3.04018, LM Loss: 2.11712, Gen Loss: 6.28542, Sty Loss: 0.11182, Diff Loss: 0.25237, DiscLM Loss: 0.00000, GenLM Loss: 0.96703 +INFO:2025-06-11 13:05:14,370: Epoch [9/20], Step [1700/2647], Loss: 0.52865, Disc Loss: 3.76510, Dur Loss: 1.04225, CE Loss: 0.05635, Norm Loss: 1.14445, F0 Loss: 2.89620, LM Loss: 2.44055, Gen Loss: 6.64951, Sty Loss: 0.14661, Diff Loss: 0.30632, DiscLM Loss: 0.01147, GenLM Loss: 0.99875 +INFO:2025-06-11 13:07:01,007: Epoch [9/20], Step [1750/2647], Loss: 0.53233, Disc Loss: 3.78112, Dur Loss: 0.88943, CE Loss: 0.04930, Norm Loss: 1.10057, F0 Loss: 3.24770, LM Loss: 2.35064, Gen Loss: 5.83504, Sty Loss: 0.20130, Diff Loss: 0.49878, DiscLM Loss: 0.00000, GenLM Loss: 0.98252 +INFO:2025-06-11 13:08:45,998: Epoch [9/20], Step [1800/2647], Loss: 0.53037, Disc Loss: 3.72860, Dur Loss: 0.86040, CE Loss: 0.04649, Norm Loss: 0.93370, F0 Loss: 2.57872, LM Loss: 2.21387, Gen Loss: 5.88364, Sty Loss: 0.11173, Diff Loss: 0.26918, DiscLM Loss: 0.00124, GenLM Loss: 0.95542 +INFO:2025-06-11 13:10:34,035: Epoch [9/20], Step [1850/2647], Loss: 0.53575, Disc Loss: 3.67746, Dur Loss: 1.00302, CE Loss: 0.05438, Norm Loss: 0.94508, F0 Loss: 2.60387, LM Loss: 2.07862, Gen Loss: 6.35928, Sty Loss: 0.11239, Diff Loss: 0.22886, DiscLM Loss: 0.00000, GenLM Loss: 1.00818 +INFO:2025-06-11 13:12:24,040: Epoch [9/20], Step [1900/2647], Loss: 0.52161, Disc Loss: 3.83966, Dur Loss: 0.99723, CE Loss: 0.05407, Norm Loss: 1.16022, F0 Loss: 2.84646, LM Loss: 2.18454, Gen Loss: 6.10271, Sty Loss: 0.10671, Diff Loss: 0.18454, DiscLM Loss: 0.00415, GenLM Loss: 1.03140 +INFO:2025-06-11 13:14:12,217: Epoch [9/20], Step [1950/2647], Loss: 0.51553, Disc Loss: 3.80020, Dur Loss: 1.03021, CE Loss: 0.05681, Norm Loss: 0.94484, F0 Loss: 3.10671, LM Loss: 2.13083, Gen Loss: 5.63932, Sty Loss: 0.11228, Diff Loss: 0.20391, DiscLM Loss: 0.00000, GenLM Loss: 0.96704 +INFO:2025-06-11 13:16:01,886: Epoch [9/20], Step [2000/2647], Loss: 0.53081, Disc Loss: 3.82111, Dur Loss: 0.91427, CE Loss: 0.04422, Norm Loss: 1.39399, F0 Loss: 2.80489, LM Loss: 2.25605, Gen Loss: 6.63038, Sty Loss: 0.18974, Diff Loss: 0.41729, DiscLM Loss: 0.00120, GenLM Loss: 0.97722 +INFO:2025-06-11 13:17:53,655: Epoch [9/20], Step [2050/2647], Loss: 0.52707, Disc Loss: 3.81507, Dur Loss: 0.99975, CE Loss: 0.05398, Norm Loss: 1.24741, F0 Loss: 3.87455, LM Loss: 2.16310, Gen Loss: 6.25404, Sty Loss: 0.11736, Diff Loss: 0.34090, DiscLM Loss: 0.00000, GenLM Loss: 0.97327 +INFO:2025-06-11 13:19:44,438: Epoch [9/20], Step [2100/2647], Loss: 0.53081, Disc Loss: 3.79040, Dur Loss: 0.97773, CE Loss: 0.05203, Norm Loss: 0.89671, F0 Loss: 2.67028, LM Loss: 2.07960, Gen Loss: 5.65264, Sty Loss: 0.11041, Diff Loss: 0.31415, DiscLM Loss: 0.00140, GenLM Loss: 1.02755 +INFO:2025-06-11 13:21:33,548: Epoch [9/20], Step [2150/2647], Loss: 0.52291, Disc Loss: 3.73875, Dur Loss: 0.83967, CE Loss: 0.04426, Norm Loss: 1.06390, F0 Loss: 2.71473, LM Loss: 2.14614, Gen Loss: 5.97567, Sty Loss: 0.11576, Diff Loss: 0.26637, DiscLM Loss: 0.00000, GenLM Loss: 0.99678 +INFO:2025-06-11 13:23:23,277: Epoch [9/20], Step [2200/2647], Loss: 0.53285, Disc Loss: 3.75952, Dur Loss: 0.78821, CE Loss: 0.04489, Norm Loss: 1.07420, F0 Loss: 3.26170, LM Loss: 2.21251, Gen Loss: 7.11360, Sty Loss: 0.15436, Diff Loss: 0.43940, DiscLM Loss: 0.00533, GenLM Loss: 0.98113 +INFO:2025-06-11 13:25:11,246: Epoch [9/20], Step [2250/2647], Loss: 0.53085, Disc Loss: 3.77748, Dur Loss: 1.15860, CE Loss: 0.06401, Norm Loss: 1.03289, F0 Loss: 3.64063, LM Loss: 2.16875, Gen Loss: 5.40736, Sty Loss: 0.13093, Diff Loss: 0.30944, DiscLM Loss: 0.00000, GenLM Loss: 1.00866 +INFO:2025-06-11 13:26:58,800: Epoch [9/20], Step [2300/2647], Loss: 0.53915, Disc Loss: 3.67986, Dur Loss: 0.86929, CE Loss: 0.04760, Norm Loss: 1.06371, F0 Loss: 4.06072, LM Loss: 2.26342, Gen Loss: 7.03274, Sty Loss: 0.15240, Diff Loss: 0.26422, DiscLM Loss: 0.00322, GenLM Loss: 1.01152 +INFO:2025-06-11 13:28:45,603: Epoch [9/20], Step [2350/2647], Loss: 0.55574, Disc Loss: 3.74579, Dur Loss: 1.14429, CE Loss: 0.06629, Norm Loss: 1.01558, F0 Loss: 3.32953, LM Loss: 2.26993, Gen Loss: 6.26807, Sty Loss: 0.14824, Diff Loss: 0.32309, DiscLM Loss: 0.00000, GenLM Loss: 0.98887 +INFO:2025-06-11 13:30:34,135: Epoch [9/20], Step [2400/2647], Loss: 0.53739, Disc Loss: 3.70057, Dur Loss: 0.91233, CE Loss: 0.05094, Norm Loss: 0.84887, F0 Loss: 2.27877, LM Loss: 2.25851, Gen Loss: 6.64269, Sty Loss: 0.17094, Diff Loss: 0.30924, DiscLM Loss: 0.00106, GenLM Loss: 0.99583 +INFO:2025-06-11 13:32:22,525: Epoch [9/20], Step [2450/2647], Loss: 0.54414, Disc Loss: 3.79604, Dur Loss: 0.86792, CE Loss: 0.04807, Norm Loss: 1.12886, F0 Loss: 3.12464, LM Loss: 2.11774, Gen Loss: 5.70613, Sty Loss: 0.12254, Diff Loss: 0.28788, DiscLM Loss: 0.00000, GenLM Loss: 0.99130 +INFO:2025-06-11 13:34:09,044: Epoch [9/20], Step [2500/2647], Loss: 0.52680, Disc Loss: 3.81289, Dur Loss: 1.14304, CE Loss: 0.05602, Norm Loss: 1.34825, F0 Loss: 3.63436, LM Loss: 2.38560, Gen Loss: 5.83267, Sty Loss: 0.11125, Diff Loss: 0.25806, DiscLM Loss: 0.00291, GenLM Loss: 0.98962 +INFO:2025-06-11 13:35:58,156: Epoch [9/20], Step [2550/2647], Loss: 0.53209, Disc Loss: 3.78680, Dur Loss: 1.07871, CE Loss: 0.06343, Norm Loss: 1.07377, F0 Loss: 2.82800, LM Loss: 2.18855, Gen Loss: 5.67456, Sty Loss: 0.12104, Diff Loss: 0.22633, DiscLM Loss: 0.00000, GenLM Loss: 0.96638 +INFO:2025-06-11 13:37:46,127: Epoch [9/20], Step [2600/2647], Loss: 0.53561, Disc Loss: 3.81835, Dur Loss: 1.09738, CE Loss: 0.05878, Norm Loss: 1.12078, F0 Loss: 2.36736, LM Loss: 2.20661, Gen Loss: 5.37058, Sty Loss: 0.12110, Diff Loss: 0.21349, DiscLM Loss: 0.00119, GenLM Loss: 1.00635 +INFO:2025-06-11 13:41:53,219: Validation loss: 0.544, Dur loss: 0.987, F0 loss: 3.405 + + + +INFO:2025-06-11 13:43:47,678: Epoch [10/20], Step [50/2647], Loss: 0.53243, Disc Loss: 3.77521, Dur Loss: 0.89580, CE Loss: 0.05116, Norm Loss: 0.91672, F0 Loss: 3.43917, LM Loss: 2.09117, Gen Loss: 5.71120, Sty Loss: 0.12639, Diff Loss: 0.19317, DiscLM Loss: 0.00000, GenLM Loss: 0.99241 +INFO:2025-06-11 13:45:36,748: Epoch [10/20], Step [100/2647], Loss: 0.52368, Disc Loss: 3.85066, Dur Loss: 0.97891, CE Loss: 0.05095, Norm Loss: 0.91867, F0 Loss: 2.51890, LM Loss: 2.02703, Gen Loss: 5.32963, Sty Loss: 0.10212, Diff Loss: 0.29426, DiscLM Loss: 0.04729, GenLM Loss: 0.92690 +INFO:2025-06-11 13:47:26,247: Epoch [10/20], Step [150/2647], Loss: 0.52638, Disc Loss: 3.81312, Dur Loss: 0.94623, CE Loss: 0.05465, Norm Loss: 1.13847, F0 Loss: 4.00868, LM Loss: 2.33179, Gen Loss: 5.92886, Sty Loss: 0.16816, Diff Loss: 0.24149, DiscLM Loss: 0.00000, GenLM Loss: 1.00479 +INFO:2025-06-11 13:49:14,156: Epoch [10/20], Step [200/2647], Loss: 0.52821, Disc Loss: 3.83271, Dur Loss: 0.82350, CE Loss: 0.04444, Norm Loss: 0.88954, F0 Loss: 3.09002, LM Loss: 2.07733, Gen Loss: 5.70111, Sty Loss: 0.10645, Diff Loss: 0.24212, DiscLM Loss: 0.00211, GenLM Loss: 0.99613 +INFO:2025-06-11 13:51:02,411: Epoch [10/20], Step [250/2647], Loss: 0.52176, Disc Loss: 3.63038, Dur Loss: 1.31099, CE Loss: 0.06878, Norm Loss: 1.50862, F0 Loss: 2.13607, LM Loss: 2.15224, Gen Loss: 6.43244, Sty Loss: 0.13407, Diff Loss: 0.26752, DiscLM Loss: 0.00000, GenLM Loss: 1.00196 +INFO:2025-06-11 13:52:50,374: Epoch [10/20], Step [300/2647], Loss: 0.52363, Disc Loss: 3.71783, Dur Loss: 1.08512, CE Loss: 0.05946, Norm Loss: 1.23042, F0 Loss: 2.93807, LM Loss: 2.13762, Gen Loss: 5.97199, Sty Loss: 0.10349, Diff Loss: 0.31400, DiscLM Loss: 0.01682, GenLM Loss: 1.00259 +INFO:2025-06-11 13:54:38,959: Epoch [10/20], Step [350/2647], Loss: 0.52827, Disc Loss: 3.73136, Dur Loss: 1.21186, CE Loss: 0.06302, Norm Loss: 1.33626, F0 Loss: 2.93758, LM Loss: 2.19108, Gen Loss: 6.01080, Sty Loss: 0.11201, Diff Loss: 0.24061, DiscLM Loss: 0.00000, GenLM Loss: 0.98892 +INFO:2025-06-11 13:56:28,650: Epoch [10/20], Step [400/2647], Loss: 0.52254, Disc Loss: 3.70257, Dur Loss: 0.89104, CE Loss: 0.04296, Norm Loss: 0.94827, F0 Loss: 3.80048, LM Loss: 2.12029, Gen Loss: 6.12635, Sty Loss: 0.10651, Diff Loss: 0.27236, DiscLM Loss: 0.00373, GenLM Loss: 1.00539 +INFO:2025-06-11 13:58:19,790: Epoch [10/20], Step [450/2647], Loss: 0.53101, Disc Loss: 3.77834, Dur Loss: 1.30836, CE Loss: 0.07983, Norm Loss: 1.19937, F0 Loss: 3.78322, LM Loss: 2.26717, Gen Loss: 5.97683, Sty Loss: 0.12862, Diff Loss: 0.23965, DiscLM Loss: 0.00000, GenLM Loss: 0.99841 +INFO:2025-06-11 14:00:08,061: Epoch [10/20], Step [500/2647], Loss: 0.55415, Disc Loss: 3.76726, Dur Loss: 0.93454, CE Loss: 0.05087, Norm Loss: 1.01111, F0 Loss: 2.97619, LM Loss: 2.02802, Gen Loss: 6.15582, Sty Loss: 0.14813, Diff Loss: 0.27870, DiscLM Loss: 0.00359, GenLM Loss: 0.98938 +INFO:2025-06-11 14:01:58,022: Epoch [10/20], Step [550/2647], Loss: 0.53932, Disc Loss: 3.79556, Dur Loss: 0.91859, CE Loss: 0.05248, Norm Loss: 0.84706, F0 Loss: 2.81421, LM Loss: 2.12380, Gen Loss: 5.53158, Sty Loss: 0.11794, Diff Loss: 0.33640, DiscLM Loss: 0.00000, GenLM Loss: 0.98432 +INFO:2025-06-11 14:03:43,871: Epoch [10/20], Step [600/2647], Loss: 0.52260, Disc Loss: 3.76653, Dur Loss: 1.10718, CE Loss: 0.05813, Norm Loss: 0.96488, F0 Loss: 3.38787, LM Loss: 2.10708, Gen Loss: 6.14433, Sty Loss: 0.11903, Diff Loss: 0.27606, DiscLM Loss: 0.00190, GenLM Loss: 1.01248 +INFO:2025-06-11 14:05:30,918: Epoch [10/20], Step [650/2647], Loss: 0.54040, Disc Loss: 3.75688, Dur Loss: 1.31700, CE Loss: 0.07937, Norm Loss: 1.40187, F0 Loss: 2.92972, LM Loss: 2.27861, Gen Loss: 6.21545, Sty Loss: 0.12446, Diff Loss: 0.39543, DiscLM Loss: 0.00000, GenLM Loss: 0.97804 +INFO:2025-06-11 14:07:22,918: Epoch [10/20], Step [700/2647], Loss: 0.54108, Disc Loss: 3.66789, Dur Loss: 0.80182, CE Loss: 0.04374, Norm Loss: 1.00914, F0 Loss: 2.64206, LM Loss: 2.21231, Gen Loss: 6.59751, Sty Loss: 0.12709, Diff Loss: 0.20822, DiscLM Loss: 0.00234, GenLM Loss: 0.97226 +INFO:2025-06-11 14:09:11,672: Epoch [10/20], Step [750/2647], Loss: 0.53131, Disc Loss: 3.78405, Dur Loss: 0.90691, CE Loss: 0.04992, Norm Loss: 1.00943, F0 Loss: 2.81485, LM Loss: 2.05972, Gen Loss: 5.39561, Sty Loss: 0.10965, Diff Loss: 0.20103, DiscLM Loss: 0.00000, GenLM Loss: 1.00473 +INFO:2025-06-11 14:10:58,665: Epoch [10/20], Step [800/2647], Loss: 0.52225, Disc Loss: 3.69894, Dur Loss: 0.65717, CE Loss: 0.03350, Norm Loss: 0.73499, F0 Loss: 2.38218, LM Loss: 2.10662, Gen Loss: 6.77530, Sty Loss: 0.12683, Diff Loss: 0.24137, DiscLM Loss: 0.00157, GenLM Loss: 1.01748 +INFO:2025-06-11 14:12:47,821: Epoch [10/20], Step [850/2647], Loss: 0.52803, Disc Loss: 3.73497, Dur Loss: 1.24401, CE Loss: 0.06988, Norm Loss: 1.06111, F0 Loss: 2.48508, LM Loss: 2.19786, Gen Loss: 5.88067, Sty Loss: 0.10655, Diff Loss: 0.26691, DiscLM Loss: 0.00000, GenLM Loss: 0.98896 +INFO:2025-06-11 14:14:34,751: Epoch [10/20], Step [900/2647], Loss: 0.53488, Disc Loss: 3.75160, Dur Loss: 0.83608, CE Loss: 0.04779, Norm Loss: 0.98570, F0 Loss: 3.76766, LM Loss: 2.17430, Gen Loss: 6.33411, Sty Loss: 0.14847, Diff Loss: 0.24000, DiscLM Loss: 0.00054, GenLM Loss: 0.99771 +INFO:2025-06-11 14:16:22,285: Epoch [10/20], Step [950/2647], Loss: 0.52049, Disc Loss: 3.72166, Dur Loss: 0.94203, CE Loss: 0.05116, Norm Loss: 0.92994, F0 Loss: 3.41632, LM Loss: 2.15693, Gen Loss: 5.90860, Sty Loss: 0.08693, Diff Loss: 0.22935, DiscLM Loss: 0.00000, GenLM Loss: 1.01746 +INFO:2025-06-11 14:18:11,171: Epoch [10/20], Step [1000/2647], Loss: 0.51525, Disc Loss: 3.86958, Dur Loss: 0.92498, CE Loss: 0.04640, Norm Loss: 0.79599, F0 Loss: 2.32585, LM Loss: 2.07668, Gen Loss: 5.29053, Sty Loss: 0.09959, Diff Loss: 0.17651, DiscLM Loss: 0.01320, GenLM Loss: 1.01859 +INFO:2025-06-11 14:20:00,299: Epoch [10/20], Step [1050/2647], Loss: 0.52727, Disc Loss: 3.82766, Dur Loss: 1.05415, CE Loss: 0.05871, Norm Loss: 0.95250, F0 Loss: 2.81724, LM Loss: 2.06613, Gen Loss: 5.51528, Sty Loss: 0.12487, Diff Loss: 0.28280, DiscLM Loss: 0.00000, GenLM Loss: 0.98177 +INFO:2025-06-11 14:21:50,453: Epoch [10/20], Step [1100/2647], Loss: 0.53379, Disc Loss: 3.79349, Dur Loss: 1.02568, CE Loss: 0.05883, Norm Loss: 1.14272, F0 Loss: 3.54217, LM Loss: 2.23776, Gen Loss: 5.64505, Sty Loss: 0.10911, Diff Loss: 0.26394, DiscLM Loss: 0.00280, GenLM Loss: 1.00260 +INFO:2025-06-11 14:23:38,404: Epoch [10/20], Step [1150/2647], Loss: 0.52757, Disc Loss: 3.70285, Dur Loss: 0.91790, CE Loss: 0.04645, Norm Loss: 1.28815, F0 Loss: 3.20095, LM Loss: 2.20457, Gen Loss: 5.91734, Sty Loss: 0.09799, Diff Loss: 0.34303, DiscLM Loss: 0.00000, GenLM Loss: 1.00964 +INFO:2025-06-11 14:25:27,980: Epoch [10/20], Step [1200/2647], Loss: 0.52724, Disc Loss: 3.82224, Dur Loss: 1.03180, CE Loss: 0.05843, Norm Loss: 1.11167, F0 Loss: 3.20703, LM Loss: 2.11212, Gen Loss: 5.85459, Sty Loss: 0.10336, Diff Loss: 0.16506, DiscLM Loss: 0.00162, GenLM Loss: 0.99479 +INFO:2025-06-11 14:27:16,509: Epoch [10/20], Step [1250/2647], Loss: 0.52105, Disc Loss: 3.75921, Dur Loss: 0.85375, CE Loss: 0.04538, Norm Loss: 1.12567, F0 Loss: 3.13426, LM Loss: 2.17779, Gen Loss: 5.73156, Sty Loss: 0.12068, Diff Loss: 0.23422, DiscLM Loss: 0.00000, GenLM Loss: 0.99831 +INFO:2025-06-11 14:29:05,930: Epoch [10/20], Step [1300/2647], Loss: 0.53351, Disc Loss: 3.94982, Dur Loss: 0.94396, CE Loss: 0.04680, Norm Loss: 1.10430, F0 Loss: 3.22661, LM Loss: 2.31568, Gen Loss: 5.84344, Sty Loss: 0.10128, Diff Loss: 0.20641, DiscLM Loss: 0.00054, GenLM Loss: 1.01758 +INFO:2025-06-11 14:30:54,723: Epoch [10/20], Step [1350/2647], Loss: 0.52820, Disc Loss: 3.67626, Dur Loss: 0.87963, CE Loss: 0.04676, Norm Loss: 0.89111, F0 Loss: 2.79600, LM Loss: 2.15412, Gen Loss: 6.24776, Sty Loss: 0.08985, Diff Loss: 0.27471, DiscLM Loss: 0.00000, GenLM Loss: 0.99025 +INFO:2025-06-11 14:32:43,463: Epoch [10/20], Step [1400/2647], Loss: 0.53345, Disc Loss: 3.68332, Dur Loss: 0.78514, CE Loss: 0.04050, Norm Loss: 0.93804, F0 Loss: 3.00363, LM Loss: 2.23330, Gen Loss: 6.48046, Sty Loss: 0.13310, Diff Loss: 0.33424, DiscLM Loss: 0.00222, GenLM Loss: 0.98544 +INFO:2025-06-11 14:34:31,025: Epoch [10/20], Step [1450/2647], Loss: 0.52477, Disc Loss: 3.76635, Dur Loss: 1.28523, CE Loss: 0.06951, Norm Loss: 1.09061, F0 Loss: 3.54206, LM Loss: 2.24530, Gen Loss: 5.94869, Sty Loss: 0.09957, Diff Loss: 0.30413, DiscLM Loss: 0.00000, GenLM Loss: 1.01026 +INFO:2025-06-11 14:36:20,054: Epoch [10/20], Step [1500/2647], Loss: 0.53076, Disc Loss: 3.95921, Dur Loss: 1.15946, CE Loss: 0.06050, Norm Loss: 1.52134, F0 Loss: 3.78455, LM Loss: 2.27623, Gen Loss: 6.85882, Sty Loss: 0.11973, Diff Loss: 0.31428, DiscLM Loss: 0.00454, GenLM Loss: 1.01777 +INFO:2025-06-11 14:38:10,350: Epoch [10/20], Step [1550/2647], Loss: 0.52696, Disc Loss: 3.70603, Dur Loss: 1.16301, CE Loss: 0.06764, Norm Loss: 1.21083, F0 Loss: 3.08367, LM Loss: 2.22900, Gen Loss: 6.23590, Sty Loss: 0.11937, Diff Loss: 0.28925, DiscLM Loss: 0.00000, GenLM Loss: 0.97901 +INFO:2025-06-11 14:40:00,254: Epoch [10/20], Step [1600/2647], Loss: 0.52612, Disc Loss: 3.83702, Dur Loss: 1.02916, CE Loss: 0.05820, Norm Loss: 0.89705, F0 Loss: 3.48513, LM Loss: 2.10389, Gen Loss: 5.37003, Sty Loss: 0.11607, Diff Loss: 0.30762, DiscLM Loss: 0.00747, GenLM Loss: 0.99859 +INFO:2025-06-11 14:41:49,219: Epoch [10/20], Step [1650/2647], Loss: 0.52795, Disc Loss: 3.77735, Dur Loss: 0.71515, CE Loss: 0.03587, Norm Loss: 0.64163, F0 Loss: 2.38367, LM Loss: 2.08702, Gen Loss: 6.46019, Sty Loss: 0.11933, Diff Loss: 0.28980, DiscLM Loss: 0.00000, GenLM Loss: 1.02787 +INFO:2025-06-11 14:43:36,699: Epoch [10/20], Step [1700/2647], Loss: 0.53817, Disc Loss: 3.65800, Dur Loss: 0.83299, CE Loss: 0.04420, Norm Loss: 0.86355, F0 Loss: 2.94063, LM Loss: 2.21808, Gen Loss: 6.46998, Sty Loss: 0.12238, Diff Loss: 0.26275, DiscLM Loss: 0.00387, GenLM Loss: 0.98790 +INFO:2025-06-11 14:45:26,928: Epoch [10/20], Step [1750/2647], Loss: 0.53445, Disc Loss: 3.73770, Dur Loss: 0.90302, CE Loss: 0.04810, Norm Loss: 1.22913, F0 Loss: 3.02445, LM Loss: 2.20198, Gen Loss: 5.97179, Sty Loss: 0.14360, Diff Loss: 0.39006, DiscLM Loss: 0.00000, GenLM Loss: 0.99235 +INFO:2025-06-11 14:47:14,511: Epoch [10/20], Step [1800/2647], Loss: 0.51371, Disc Loss: 3.69582, Dur Loss: 1.03117, CE Loss: 0.05794, Norm Loss: 1.17249, F0 Loss: 2.60876, LM Loss: 2.22532, Gen Loss: 5.91251, Sty Loss: 0.10294, Diff Loss: 0.18893, DiscLM Loss: 0.01767, GenLM Loss: 0.99820 +INFO:2025-06-11 14:49:02,525: Epoch [10/20], Step [1850/2647], Loss: 0.52380, Disc Loss: 3.88418, Dur Loss: 1.09118, CE Loss: 0.05697, Norm Loss: 1.01169, F0 Loss: 2.94328, LM Loss: 2.40554, Gen Loss: 5.36233, Sty Loss: 0.12611, Diff Loss: 0.18834, DiscLM Loss: 0.00000, GenLM Loss: 0.98995 +INFO:2025-06-11 14:50:51,954: Epoch [10/20], Step [1900/2647], Loss: 0.52584, Disc Loss: 3.69604, Dur Loss: 0.99452, CE Loss: 0.05587, Norm Loss: 1.21182, F0 Loss: 3.56855, LM Loss: 2.23234, Gen Loss: 7.10708, Sty Loss: 0.13925, Diff Loss: 0.33931, DiscLM Loss: 0.00562, GenLM Loss: 0.96370 +INFO:2025-06-11 14:52:43,617: Epoch [10/20], Step [1950/2647], Loss: 0.51799, Disc Loss: 3.82033, Dur Loss: 0.88800, CE Loss: 0.04392, Norm Loss: 1.27529, F0 Loss: 2.93823, LM Loss: 2.16372, Gen Loss: 5.04730, Sty Loss: 0.11671, Diff Loss: 0.22830, DiscLM Loss: 0.00000, GenLM Loss: 0.92450 +INFO:2025-06-11 14:54:30,443: Epoch [10/20], Step [2000/2647], Loss: 0.53489, Disc Loss: 3.69283, Dur Loss: 0.86222, CE Loss: 0.04954, Norm Loss: 1.08139, F0 Loss: 2.77376, LM Loss: 2.33331, Gen Loss: 6.22942, Sty Loss: 0.11226, Diff Loss: 0.33167, DiscLM Loss: 0.00157, GenLM Loss: 0.99195 +INFO:2025-06-11 14:56:18,118: Epoch [10/20], Step [2050/2647], Loss: 0.52869, Disc Loss: 3.69968, Dur Loss: 0.89860, CE Loss: 0.04975, Norm Loss: 0.88432, F0 Loss: 2.83304, LM Loss: 2.17673, Gen Loss: 5.74896, Sty Loss: 0.09182, Diff Loss: 0.17666, DiscLM Loss: 0.00000, GenLM Loss: 0.97345 +INFO:2025-06-11 14:58:07,031: Epoch [10/20], Step [2100/2647], Loss: 0.53293, Disc Loss: 3.76718, Dur Loss: 1.06577, CE Loss: 0.05675, Norm Loss: 1.23620, F0 Loss: 2.60812, LM Loss: 2.10005, Gen Loss: 6.37423, Sty Loss: 0.11604, Diff Loss: 0.21334, DiscLM Loss: 0.00715, GenLM Loss: 1.03929 +INFO:2025-06-11 14:59:57,011: Epoch [10/20], Step [2150/2647], Loss: 0.51390, Disc Loss: 3.58762, Dur Loss: 0.99240, CE Loss: 0.05584, Norm Loss: 1.08108, F0 Loss: 2.52349, LM Loss: 2.24512, Gen Loss: 7.19684, Sty Loss: 0.11570, Diff Loss: 0.18350, DiscLM Loss: 0.00000, GenLM Loss: 0.97007 +INFO:2025-06-11 15:01:45,566: Epoch [10/20], Step [2200/2647], Loss: 0.51889, Disc Loss: 3.76689, Dur Loss: 0.89230, CE Loss: 0.04587, Norm Loss: 1.02998, F0 Loss: 2.76603, LM Loss: 2.07702, Gen Loss: 5.88169, Sty Loss: 0.12865, Diff Loss: 0.36195, DiscLM Loss: 0.00054, GenLM Loss: 0.98302 +INFO:2025-06-11 15:03:38,160: Epoch [10/20], Step [2250/2647], Loss: 0.51521, Disc Loss: 3.87884, Dur Loss: 0.87148, CE Loss: 0.04488, Norm Loss: 1.13161, F0 Loss: 3.66550, LM Loss: 2.14648, Gen Loss: 6.20299, Sty Loss: 0.10995, Diff Loss: 0.19342, DiscLM Loss: 0.00000, GenLM Loss: 1.02246 +INFO:2025-06-11 15:05:25,558: Epoch [10/20], Step [2300/2647], Loss: 0.51576, Disc Loss: 3.78461, Dur Loss: 1.04176, CE Loss: 0.06670, Norm Loss: 1.26306, F0 Loss: 2.62376, LM Loss: 2.31561, Gen Loss: 5.45774, Sty Loss: 0.10013, Diff Loss: 0.35515, DiscLM Loss: 0.04739, GenLM Loss: 0.92602 +INFO:2025-06-11 15:09:00,750: Epoch [10/20], Step [2400/2647], Loss: 1.07618, Disc Loss: 3.66501, Dur Loss: 1.27710, CE Loss: 0.07407, Norm Loss: 1.67865, F0 Loss: 3.96697, LM Loss: 2.31815, Gen Loss: 7.09220, Sty Loss: 0.13545, Diff Loss: 0.27581, DiscLM Loss: 0.02569, GenLM Loss: 0.99818 +INFO:2025-06-11 15:10:48,518: Epoch [10/20], Step [2450/2647], Loss: 0.52274, Disc Loss: 3.72029, Dur Loss: 1.13460, CE Loss: 0.05911, Norm Loss: 1.12543, F0 Loss: 3.52352, LM Loss: 2.01679, Gen Loss: 6.29103, Sty Loss: 0.10756, Diff Loss: 0.16503, DiscLM Loss: 0.00000, GenLM Loss: 0.98710 +INFO:2025-06-11 15:12:39,037: Epoch [10/20], Step [2500/2647], Loss: 0.53142, Disc Loss: 3.77099, Dur Loss: 1.05714, CE Loss: 0.05600, Norm Loss: 1.25515, F0 Loss: 3.41673, LM Loss: 2.15202, Gen Loss: 6.36661, Sty Loss: 0.11606, Diff Loss: 0.29163, DiscLM Loss: 0.00645, GenLM Loss: 0.98705 +INFO:2025-06-11 15:14:29,378: Epoch [10/20], Step [2550/2647], Loss: 0.52895, Disc Loss: 3.75614, Dur Loss: 1.09950, CE Loss: 0.06286, Norm Loss: 1.01025, F0 Loss: 2.64045, LM Loss: 2.17723, Gen Loss: 5.85573, Sty Loss: 0.11273, Diff Loss: 0.27321, DiscLM Loss: 0.00000, GenLM Loss: 0.99231 +INFO:2025-06-11 15:16:21,423: Epoch [10/20], Step [2600/2647], Loss: 0.52523, Disc Loss: 3.74951, Dur Loss: 0.80199, CE Loss: 0.04238, Norm Loss: 1.05655, F0 Loss: 2.57024, LM Loss: 2.27273, Gen Loss: 6.49578, Sty Loss: 0.14933, Diff Loss: 0.32470, DiscLM Loss: 0.00225, GenLM Loss: 0.98849 +INFO:2025-06-11 15:20:27,014: Validation loss: 0.538, Dur loss: 0.995, F0 loss: 3.347 + + + +INFO:2025-06-11 15:22:18,659: Epoch [11/20], Step [50/2647], Loss: 0.51713, Disc Loss: 3.81481, Dur Loss: 0.99811, CE Loss: 0.05219, Norm Loss: 0.92400, F0 Loss: 3.14887, LM Loss: 2.06312, Gen Loss: 5.65958, Sty Loss: 0.11005, Diff Loss: 0.27182, DiscLM Loss: 0.00000, GenLM Loss: 1.00449 +INFO:2025-06-11 15:24:11,743: Epoch [11/20], Step [100/2647], Loss: 0.51961, Disc Loss: 3.81567, Dur Loss: 1.03545, CE Loss: 0.05812, Norm Loss: 0.93291, F0 Loss: 3.48249, LM Loss: 2.23547, Gen Loss: 6.33717, Sty Loss: 0.12736, Diff Loss: 0.23334, DiscLM Loss: 0.00148, GenLM Loss: 1.02285 +INFO:2025-06-11 15:26:00,830: Epoch [11/20], Step [150/2647], Loss: 0.52262, Disc Loss: 3.78448, Dur Loss: 1.22151, CE Loss: 0.06637, Norm Loss: 1.40865, F0 Loss: 4.23657, LM Loss: 2.26363, Gen Loss: 6.40089, Sty Loss: 0.11985, Diff Loss: 0.40310, DiscLM Loss: 0.00000, GenLM Loss: 0.97417 +INFO:2025-06-11 15:27:51,681: Epoch [11/20], Step [200/2647], Loss: 0.52653, Disc Loss: 3.74782, Dur Loss: 0.91698, CE Loss: 0.04976, Norm Loss: 1.03628, F0 Loss: 3.80911, LM Loss: 2.26556, Gen Loss: 6.42861, Sty Loss: 0.13798, Diff Loss: 0.27974, DiscLM Loss: 0.00161, GenLM Loss: 0.99459 +INFO:2025-06-11 15:29:41,349: Epoch [11/20], Step [250/2647], Loss: 0.52329, Disc Loss: 3.72818, Dur Loss: 0.91638, CE Loss: 0.04734, Norm Loss: 0.98529, F0 Loss: 3.21494, LM Loss: 2.12561, Gen Loss: 5.86278, Sty Loss: 0.11754, Diff Loss: 0.19561, DiscLM Loss: 0.00000, GenLM Loss: 1.01693 +INFO:2025-06-11 15:31:28,108: Epoch [11/20], Step [300/2647], Loss: 0.54508, Disc Loss: 3.72102, Dur Loss: 0.93531, CE Loss: 0.05285, Norm Loss: 1.15685, F0 Loss: 3.89844, LM Loss: 2.26224, Gen Loss: 6.44628, Sty Loss: 0.19626, Diff Loss: 0.38747, DiscLM Loss: 0.00114, GenLM Loss: 0.98152 +INFO:2025-06-11 15:33:17,228: Epoch [11/20], Step [350/2647], Loss: 0.53470, Disc Loss: 3.74646, Dur Loss: 1.23304, CE Loss: 0.05735, Norm Loss: 0.98357, F0 Loss: 3.04938, LM Loss: 2.11722, Gen Loss: 5.69645, Sty Loss: 0.12316, Diff Loss: 0.24071, DiscLM Loss: 0.00000, GenLM Loss: 0.99483 +INFO:2025-06-11 15:35:07,248: Epoch [11/20], Step [400/2647], Loss: 0.52892, Disc Loss: 3.73396, Dur Loss: 1.03547, CE Loss: 0.06179, Norm Loss: 0.95716, F0 Loss: 2.90792, LM Loss: 2.21948, Gen Loss: 6.13104, Sty Loss: 0.11151, Diff Loss: 0.24649, DiscLM Loss: 0.00085, GenLM Loss: 1.01780 +INFO:2025-06-11 15:36:57,862: Epoch [11/20], Step [450/2647], Loss: 0.53700, Disc Loss: 3.74615, Dur Loss: 1.00643, CE Loss: 0.05768, Norm Loss: 1.20660, F0 Loss: 4.16801, LM Loss: 2.34002, Gen Loss: 6.35522, Sty Loss: 0.12755, Diff Loss: 0.30651, DiscLM Loss: 0.00000, GenLM Loss: 0.98990 +INFO:2025-06-11 15:38:44,611: Epoch [11/20], Step [500/2647], Loss: 0.56428, Disc Loss: 3.63619, Dur Loss: 0.99749, CE Loss: 0.05639, Norm Loss: 1.13571, F0 Loss: 4.68205, LM Loss: 2.23766, Gen Loss: 6.26419, Sty Loss: 0.15618, Diff Loss: 0.49881, DiscLM Loss: 0.00088, GenLM Loss: 0.99309 +INFO:2025-06-11 15:40:34,764: Epoch [11/20], Step [550/2647], Loss: 0.53517, Disc Loss: 3.70535, Dur Loss: 0.85416, CE Loss: 0.04573, Norm Loss: 1.12777, F0 Loss: 3.77835, LM Loss: 2.13132, Gen Loss: 6.70767, Sty Loss: 0.12010, Diff Loss: 0.27824, DiscLM Loss: 0.00000, GenLM Loss: 1.00843 +INFO:2025-06-11 15:42:23,168: Epoch [11/20], Step [600/2647], Loss: 0.53877, Disc Loss: 3.80799, Dur Loss: 0.88307, CE Loss: 0.04613, Norm Loss: 0.84653, F0 Loss: 2.81744, LM Loss: 2.15066, Gen Loss: 5.57863, Sty Loss: 0.11127, Diff Loss: 0.36790, DiscLM Loss: 0.00072, GenLM Loss: 0.98971 +INFO:2025-06-11 15:44:11,370: Epoch [11/20], Step [650/2647], Loss: 0.53389, Disc Loss: 3.83375, Dur Loss: 0.92831, CE Loss: 0.05306, Norm Loss: 0.80243, F0 Loss: 3.38741, LM Loss: 2.10305, Gen Loss: 6.27039, Sty Loss: 0.20626, Diff Loss: 0.35138, DiscLM Loss: 0.00000, GenLM Loss: 1.00424 +INFO:2025-06-11 15:45:59,143: Epoch [11/20], Step [700/2647], Loss: 0.57049, Disc Loss: 3.70644, Dur Loss: 0.88982, CE Loss: 0.04425, Norm Loss: 0.96550, F0 Loss: 2.91543, LM Loss: 2.21502, Gen Loss: 6.30830, Sty Loss: 0.15388, Diff Loss: 0.34627, DiscLM Loss: 0.00094, GenLM Loss: 0.99220 +INFO:2025-06-11 15:47:46,950: Epoch [11/20], Step [750/2647], Loss: 0.54077, Disc Loss: 3.78815, Dur Loss: 0.95861, CE Loss: 0.05247, Norm Loss: 1.03499, F0 Loss: 2.98671, LM Loss: 2.19430, Gen Loss: 5.81758, Sty Loss: 0.20774, Diff Loss: 0.37576, DiscLM Loss: 0.00000, GenLM Loss: 0.99911 +INFO:2025-06-11 15:49:36,677: Epoch [11/20], Step [800/2647], Loss: 0.52982, Disc Loss: 3.68525, Dur Loss: 1.18027, CE Loss: 0.06653, Norm Loss: 1.21750, F0 Loss: 3.00512, LM Loss: 2.22816, Gen Loss: 5.77784, Sty Loss: 0.10653, Diff Loss: 0.17235, DiscLM Loss: 0.00108, GenLM Loss: 1.00551 +INFO:2025-06-11 15:51:25,798: Epoch [11/20], Step [850/2647], Loss: 0.52668, Disc Loss: 3.62656, Dur Loss: 0.80715, CE Loss: 0.04134, Norm Loss: 0.81528, F0 Loss: 2.76375, LM Loss: 2.23122, Gen Loss: 6.89269, Sty Loss: 0.27276, Diff Loss: 0.39991, DiscLM Loss: 0.00000, GenLM Loss: 1.00174 +INFO:2025-06-11 15:53:15,643: Epoch [11/20], Step [900/2647], Loss: 0.52899, Disc Loss: 3.83898, Dur Loss: 0.84631, CE Loss: 0.04449, Norm Loss: 0.80588, F0 Loss: 2.41857, LM Loss: 2.05376, Gen Loss: 5.81732, Sty Loss: 0.13016, Diff Loss: 0.25747, DiscLM Loss: 0.00080, GenLM Loss: 0.99273 +INFO:2025-06-11 15:55:01,401: Epoch [11/20], Step [950/2647], Loss: 0.52605, Disc Loss: 3.78682, Dur Loss: 1.09132, CE Loss: 0.06052, Norm Loss: 1.12068, F0 Loss: 2.84247, LM Loss: 2.20786, Gen Loss: 5.66925, Sty Loss: 0.10772, Diff Loss: 0.19273, DiscLM Loss: 0.00000, GenLM Loss: 0.99671 +INFO:2025-06-11 15:56:51,406: Epoch [11/20], Step [1000/2647], Loss: 0.52653, Disc Loss: 3.70671, Dur Loss: 0.73233, CE Loss: 0.03909, Norm Loss: 1.03375, F0 Loss: 3.30687, LM Loss: 2.05942, Gen Loss: 6.43840, Sty Loss: 0.13781, Diff Loss: 0.26025, DiscLM Loss: 0.00498, GenLM Loss: 0.98358 +INFO:2025-06-11 15:58:40,421: Epoch [11/20], Step [1050/2647], Loss: 0.52396, Disc Loss: 3.56850, Dur Loss: 1.06438, CE Loss: 0.05469, Norm Loss: 0.79623, F0 Loss: 2.50725, LM Loss: 2.47458, Gen Loss: 7.42983, Sty Loss: 0.88782, Diff Loss: 1.28521, DiscLM Loss: 0.00000, GenLM Loss: 1.00211 +INFO:2025-06-11 16:00:29,082: Epoch [11/20], Step [1100/2647], Loss: 0.57815, Disc Loss: 3.81426, Dur Loss: 0.83276, CE Loss: 0.04463, Norm Loss: 0.82367, F0 Loss: 2.89614, LM Loss: 2.24541, Gen Loss: 5.88015, Sty Loss: 0.17654, Diff Loss: 0.22161, DiscLM Loss: 0.00227, GenLM Loss: 0.98860 +INFO:2025-06-11 16:02:21,118: Epoch [11/20], Step [1150/2647], Loss: 0.53731, Disc Loss: 3.61205, Dur Loss: 1.14747, CE Loss: 0.06091, Norm Loss: 1.17897, F0 Loss: 3.53114, LM Loss: 2.21515, Gen Loss: 6.37409, Sty Loss: 0.18297, Diff Loss: 0.34530, DiscLM Loss: 0.00000, GenLM Loss: 1.00450 +INFO:2025-06-11 16:04:10,650: Epoch [11/20], Step [1200/2647], Loss: 0.54273, Disc Loss: 3.76885, Dur Loss: 1.17456, CE Loss: 0.06310, Norm Loss: 0.97186, F0 Loss: 2.97103, LM Loss: 2.04507, Gen Loss: 5.56695, Sty Loss: 0.10173, Diff Loss: 0.24944, DiscLM Loss: 0.00427, GenLM Loss: 0.99666 +INFO:2025-06-11 16:05:58,517: Epoch [11/20], Step [1250/2647], Loss: 0.53546, Disc Loss: 3.76174, Dur Loss: 1.12382, CE Loss: 0.07040, Norm Loss: 1.30060, F0 Loss: 3.30315, LM Loss: 2.28985, Gen Loss: 5.77836, Sty Loss: 0.10877, Diff Loss: 0.31868, DiscLM Loss: 0.00000, GenLM Loss: 0.99428 +INFO:2025-06-11 16:07:45,836: Epoch [11/20], Step [1300/2647], Loss: 0.53315, Disc Loss: 3.68174, Dur Loss: 1.13280, CE Loss: 0.06295, Norm Loss: 1.05942, F0 Loss: 4.19946, LM Loss: 2.20486, Gen Loss: 6.76393, Sty Loss: 0.14945, Diff Loss: 0.30915, DiscLM Loss: 0.00029, GenLM Loss: 1.00560 +INFO:2025-06-11 16:09:35,488: Epoch [11/20], Step [1350/2647], Loss: 0.52653, Disc Loss: 3.72809, Dur Loss: 0.98244, CE Loss: 0.05573, Norm Loss: 0.83716, F0 Loss: 3.03782, LM Loss: 2.23333, Gen Loss: 6.75345, Sty Loss: 0.13048, Diff Loss: 0.31899, DiscLM Loss: 0.00000, GenLM Loss: 0.99547 +INFO:2025-06-11 16:11:22,343: Epoch [11/20], Step [1400/2647], Loss: 0.52599, Disc Loss: 3.67971, Dur Loss: 0.80478, CE Loss: 0.04047, Norm Loss: 0.80458, F0 Loss: 3.02969, LM Loss: 2.17879, Gen Loss: 6.69750, Sty Loss: 0.14109, Diff Loss: 0.37901, DiscLM Loss: 0.00043, GenLM Loss: 1.00244 +INFO:2025-06-11 16:13:13,687: Epoch [11/20], Step [1450/2647], Loss: 0.52791, Disc Loss: 3.79059, Dur Loss: 0.96497, CE Loss: 0.05347, Norm Loss: 0.93186, F0 Loss: 2.43281, LM Loss: 2.18357, Gen Loss: 5.97887, Sty Loss: 0.11280, Diff Loss: 0.28360, DiscLM Loss: 0.00000, GenLM Loss: 1.00107 +INFO:2025-06-11 16:15:01,438: Epoch [11/20], Step [1500/2647], Loss: 0.52140, Disc Loss: 3.56692, Dur Loss: 0.93900, CE Loss: 0.04869, Norm Loss: 1.02715, F0 Loss: 3.42337, LM Loss: 2.11809, Gen Loss: 6.99684, Sty Loss: 0.12833, Diff Loss: 0.23247, DiscLM Loss: 0.00162, GenLM Loss: 0.99098 +INFO:2025-06-11 16:16:48,878: Epoch [11/20], Step [1550/2647], Loss: 0.53097, Disc Loss: 3.69355, Dur Loss: 0.90682, CE Loss: 0.04572, Norm Loss: 0.98004, F0 Loss: 2.41967, LM Loss: 2.09559, Gen Loss: 6.03577, Sty Loss: 0.10125, Diff Loss: 0.23568, DiscLM Loss: 0.00000, GenLM Loss: 1.01049 +INFO:2025-06-11 16:18:37,083: Epoch [11/20], Step [1600/2647], Loss: 0.52201, Disc Loss: 3.78309, Dur Loss: 0.94959, CE Loss: 0.05387, Norm Loss: 1.08819, F0 Loss: 3.43241, LM Loss: 2.19407, Gen Loss: 5.83513, Sty Loss: 0.10618, Diff Loss: 0.27632, DiscLM Loss: 0.00053, GenLM Loss: 1.00770 +INFO:2025-06-11 16:20:24,019: Epoch [11/20], Step [1650/2647], Loss: 0.52983, Disc Loss: 3.76692, Dur Loss: 1.10390, CE Loss: 0.06367, Norm Loss: 0.98128, F0 Loss: 3.31206, LM Loss: 2.14859, Gen Loss: 6.14628, Sty Loss: 0.14003, Diff Loss: 0.27382, DiscLM Loss: 0.00000, GenLM Loss: 0.98892 +INFO:2025-06-11 16:22:14,238: Epoch [11/20], Step [1700/2647], Loss: 0.56972, Disc Loss: 3.69428, Dur Loss: 1.01461, CE Loss: 0.05820, Norm Loss: 1.05689, F0 Loss: 3.30251, LM Loss: 2.18275, Gen Loss: 6.10679, Sty Loss: 0.18762, Diff Loss: 0.37356, DiscLM Loss: 0.00614, GenLM Loss: 1.00200 +INFO:2025-06-11 16:24:01,468: Epoch [11/20], Step [1750/2647], Loss: 0.55808, Disc Loss: 3.71418, Dur Loss: 0.72007, CE Loss: 0.03908, Norm Loss: 0.96088, F0 Loss: 2.69097, LM Loss: 2.38273, Gen Loss: 7.04031, Sty Loss: 0.47699, Diff Loss: 0.55552, DiscLM Loss: 0.00000, GenLM Loss: 1.00261 +INFO:2025-06-11 16:25:50,249: Epoch [11/20], Step [1800/2647], Loss: 0.53694, Disc Loss: 3.79859, Dur Loss: 0.95128, CE Loss: 0.05427, Norm Loss: 1.18134, F0 Loss: 3.92005, LM Loss: 2.17637, Gen Loss: 5.59037, Sty Loss: 0.14567, Diff Loss: 0.24418, DiscLM Loss: 0.00059, GenLM Loss: 1.01310 +INFO:2025-06-11 16:27:40,878: Epoch [11/20], Step [1850/2647], Loss: 0.56710, Disc Loss: 3.53264, Dur Loss: 0.99854, CE Loss: 0.05384, Norm Loss: 0.95417, F0 Loss: 3.08161, LM Loss: 2.40101, Gen Loss: 7.02104, Sty Loss: 0.30250, Diff Loss: 0.46497, DiscLM Loss: 0.00000, GenLM Loss: 0.95906 +INFO:2025-06-11 16:29:33,305: Epoch [11/20], Step [1900/2647], Loss: 0.55451, Disc Loss: 3.73771, Dur Loss: 1.03753, CE Loss: 0.05686, Norm Loss: 1.01102, F0 Loss: 2.86997, LM Loss: 2.18337, Gen Loss: 5.94349, Sty Loss: 0.15578, Diff Loss: 0.47505, DiscLM Loss: 0.00027, GenLM Loss: 0.99321 +INFO:2025-06-11 16:31:22,947: Epoch [11/20], Step [1950/2647], Loss: 0.54864, Disc Loss: 3.77077, Dur Loss: 0.97995, CE Loss: 0.05364, Norm Loss: 0.94783, F0 Loss: 3.35528, LM Loss: 2.30124, Gen Loss: 5.98901, Sty Loss: 0.20566, Diff Loss: 0.29762, DiscLM Loss: 0.00000, GenLM Loss: 1.01567 +INFO:2025-06-11 16:33:12,964: Epoch [11/20], Step [2000/2647], Loss: 0.54087, Disc Loss: 3.69250, Dur Loss: 0.96447, CE Loss: 0.05023, Norm Loss: 0.99708, F0 Loss: 3.65468, LM Loss: 2.18241, Gen Loss: 6.96128, Sty Loss: 0.18490, Diff Loss: 0.29895, DiscLM Loss: 0.00056, GenLM Loss: 0.97022 +INFO:2025-06-11 16:35:00,531: Epoch [11/20], Step [2050/2647], Loss: 0.52223, Disc Loss: 3.76167, Dur Loss: 1.00867, CE Loss: 0.05543, Norm Loss: 0.98677, F0 Loss: 2.95762, LM Loss: 2.21314, Gen Loss: 6.39048, Sty Loss: 0.13382, Diff Loss: 0.28624, DiscLM Loss: 0.00000, GenLM Loss: 1.00573 +INFO:2025-06-11 16:36:47,523: Epoch [11/20], Step [2100/2647], Loss: 0.53669, Disc Loss: 3.75542, Dur Loss: 0.88719, CE Loss: 0.04978, Norm Loss: 0.95403, F0 Loss: 3.33097, LM Loss: 2.22367, Gen Loss: 6.59041, Sty Loss: 0.16688, Diff Loss: 0.26328, DiscLM Loss: 0.00231, GenLM Loss: 1.00711 +INFO:2025-06-11 16:38:36,354: Epoch [11/20], Step [2150/2647], Loss: 0.53452, Disc Loss: 3.81947, Dur Loss: 0.89345, CE Loss: 0.04434, Norm Loss: 0.87342, F0 Loss: 3.19604, LM Loss: 2.12198, Gen Loss: 7.48328, Sty Loss: 0.18444, Diff Loss: 0.28275, DiscLM Loss: 0.00000, GenLM Loss: 0.99303 +INFO:2025-06-11 16:40:25,967: Epoch [11/20], Step [2200/2647], Loss: 0.52581, Disc Loss: 3.80766, Dur Loss: 0.75255, CE Loss: 0.03936, Norm Loss: 0.78422, F0 Loss: 2.80998, LM Loss: 2.19687, Gen Loss: 6.33129, Sty Loss: 0.17295, Diff Loss: 0.35190, DiscLM Loss: 0.00317, GenLM Loss: 1.00785 +INFO:2025-06-11 16:42:14,341: Epoch [11/20], Step [2250/2647], Loss: 0.53047, Disc Loss: 3.66552, Dur Loss: 0.92303, CE Loss: 0.06474, Norm Loss: 1.36390, F0 Loss: 3.79779, LM Loss: 2.22449, Gen Loss: 7.00212, Sty Loss: 0.14144, Diff Loss: 0.27173, DiscLM Loss: 0.00000, GenLM Loss: 0.99363 +INFO:2025-06-11 16:44:03,565: Epoch [11/20], Step [2300/2647], Loss: 0.52378, Disc Loss: 3.72230, Dur Loss: 0.94836, CE Loss: 0.05363, Norm Loss: 1.12532, F0 Loss: 2.94640, LM Loss: 2.15492, Gen Loss: 6.63085, Sty Loss: 0.10697, Diff Loss: 0.23829, DiscLM Loss: 0.00451, GenLM Loss: 0.98662 +INFO:2025-06-11 16:45:52,916: Epoch [11/20], Step [2350/2647], Loss: 0.52917, Disc Loss: 3.72045, Dur Loss: 1.09011, CE Loss: 0.05555, Norm Loss: 1.38123, F0 Loss: 2.40543, LM Loss: 2.24727, Gen Loss: 6.68350, Sty Loss: 0.13515, Diff Loss: 0.41470, DiscLM Loss: 0.00000, GenLM Loss: 0.97824 +INFO:2025-06-11 16:47:42,284: Epoch [11/20], Step [2400/2647], Loss: 0.52397, Disc Loss: 3.76845, Dur Loss: 0.92373, CE Loss: 0.04752, Norm Loss: 1.34304, F0 Loss: 3.88807, LM Loss: 2.22277, Gen Loss: 6.15415, Sty Loss: 0.10258, Diff Loss: 0.28305, DiscLM Loss: 0.00252, GenLM Loss: 0.97512 +INFO:2025-06-11 16:49:28,349: Epoch [11/20], Step [2450/2647], Loss: 0.52134, Disc Loss: 3.71224, Dur Loss: 1.20432, CE Loss: 0.05920, Norm Loss: 1.03435, F0 Loss: 3.29155, LM Loss: 2.12073, Gen Loss: 5.61864, Sty Loss: 0.10598, Diff Loss: 0.23031, DiscLM Loss: 0.00000, GenLM Loss: 1.02745 +INFO:2025-06-11 16:51:20,974: Epoch [11/20], Step [2500/2647], Loss: 0.51444, Disc Loss: 3.80276, Dur Loss: 0.82287, CE Loss: 0.04425, Norm Loss: 0.94038, F0 Loss: 2.41500, LM Loss: 2.14335, Gen Loss: 5.58410, Sty Loss: 0.09715, Diff Loss: 0.20559, DiscLM Loss: 0.00094, GenLM Loss: 1.01731 +INFO:2025-06-11 16:53:07,025: Epoch [11/20], Step [2550/2647], Loss: 0.52499, Disc Loss: 3.83589, Dur Loss: 0.84878, CE Loss: 0.04640, Norm Loss: 1.00307, F0 Loss: 2.69536, LM Loss: 2.12521, Gen Loss: 5.85293, Sty Loss: 0.09732, Diff Loss: 0.22414, DiscLM Loss: 0.00000, GenLM Loss: 0.97776 +INFO:2025-06-11 16:54:57,822: Epoch [11/20], Step [2600/2647], Loss: 0.51997, Disc Loss: 3.69473, Dur Loss: 1.16370, CE Loss: 0.06452, Norm Loss: 0.77840, F0 Loss: 2.77343, LM Loss: 2.13014, Gen Loss: 6.84996, Sty Loss: 0.11408, Diff Loss: 0.27258, DiscLM Loss: 0.00087, GenLM Loss: 1.00611 +INFO:2025-06-11 16:59:00,913: Validation loss: 0.537, Dur loss: 1.001, F0 loss: 3.433 + + + +INFO:2025-06-11 17:00:50,968: Epoch [12/20], Step [50/2647], Loss: 0.52183, Disc Loss: 3.67319, Dur Loss: 1.24689, CE Loss: 0.06298, Norm Loss: 1.46167, F0 Loss: 3.61017, LM Loss: 2.30810, Gen Loss: 6.70459, Sty Loss: 0.09924, Diff Loss: 0.30480, DiscLM Loss: 0.00000, GenLM Loss: 0.99028 +INFO:2025-06-11 17:02:36,052: Epoch [12/20], Step [100/2647], Loss: 0.51929, Disc Loss: 3.84032, Dur Loss: 0.73988, CE Loss: 0.03725, Norm Loss: 0.90121, F0 Loss: 3.39184, LM Loss: 2.20925, Gen Loss: 6.07048, Sty Loss: 0.12452, Diff Loss: 0.28019, DiscLM Loss: 0.01379, GenLM Loss: 0.99384 +INFO:2025-06-11 17:04:25,830: Epoch [12/20], Step [150/2647], Loss: 0.53167, Disc Loss: 3.66084, Dur Loss: 0.95799, CE Loss: 0.05277, Norm Loss: 0.93608, F0 Loss: 3.13039, LM Loss: 2.10066, Gen Loss: 6.13725, Sty Loss: 0.11814, Diff Loss: 0.28898, DiscLM Loss: 0.00000, GenLM Loss: 1.00167 +INFO:2025-06-11 17:06:16,506: Epoch [12/20], Step [200/2647], Loss: 0.52594, Disc Loss: 3.62057, Dur Loss: 0.99825, CE Loss: 0.05497, Norm Loss: 1.08949, F0 Loss: 3.17197, LM Loss: 2.19902, Gen Loss: 6.76931, Sty Loss: 0.11599, Diff Loss: 0.30210, DiscLM Loss: 0.01385, GenLM Loss: 0.94919 +INFO:2025-06-11 17:08:06,676: Epoch [12/20], Step [250/2647], Loss: 0.52812, Disc Loss: 3.74578, Dur Loss: 1.11567, CE Loss: 0.06226, Norm Loss: 0.87149, F0 Loss: 3.38506, LM Loss: 2.26449, Gen Loss: 6.73012, Sty Loss: 0.13002, Diff Loss: 0.24324, DiscLM Loss: 0.00000, GenLM Loss: 1.00902 +INFO:2025-06-11 17:09:56,765: Epoch [12/20], Step [300/2647], Loss: 0.51637, Disc Loss: 3.73174, Dur Loss: 0.93232, CE Loss: 0.04921, Norm Loss: 1.05553, F0 Loss: 2.51444, LM Loss: 2.09613, Gen Loss: 6.18632, Sty Loss: 0.10673, Diff Loss: 0.20467, DiscLM Loss: 0.00101, GenLM Loss: 0.96843 +INFO:2025-06-11 17:11:45,880: Epoch [12/20], Step [350/2647], Loss: 0.52475, Disc Loss: 3.77412, Dur Loss: 0.86104, CE Loss: 0.04836, Norm Loss: 0.88031, F0 Loss: 2.43473, LM Loss: 2.11919, Gen Loss: 5.75354, Sty Loss: 0.12454, Diff Loss: 0.26880, DiscLM Loss: 0.00000, GenLM Loss: 1.01656 +INFO:2025-06-11 17:13:33,955: Epoch [12/20], Step [400/2647], Loss: 0.52065, Disc Loss: 3.78652, Dur Loss: 1.10969, CE Loss: 0.06276, Norm Loss: 1.41167, F0 Loss: 3.81915, LM Loss: 2.30845, Gen Loss: 6.23169, Sty Loss: 0.13347, Diff Loss: 0.25675, DiscLM Loss: 0.00079, GenLM Loss: 1.00752 +INFO:2025-06-11 17:15:20,666: Epoch [12/20], Step [450/2647], Loss: 0.51549, Disc Loss: 3.73756, Dur Loss: 2.04189, CE Loss: 0.09540, Norm Loss: 1.48521, F0 Loss: 3.45860, LM Loss: 2.26962, Gen Loss: 5.99012, Sty Loss: 0.13271, Diff Loss: 0.24828, DiscLM Loss: 0.00000, GenLM Loss: 0.98169 +INFO:2025-06-11 17:18:59,749: Epoch [12/20], Step [550/2647], Loss: 1.03523, Disc Loss: 3.79653, Dur Loss: 0.99205, CE Loss: 0.05927, Norm Loss: 1.34409, F0 Loss: 2.92695, LM Loss: 2.17778, Gen Loss: 5.55244, Sty Loss: 0.13247, Diff Loss: 0.26606, DiscLM Loss: 0.00000, GenLM Loss: 1.00492 +INFO:2025-06-11 17:20:48,891: Epoch [12/20], Step [600/2647], Loss: 0.51088, Disc Loss: 3.75797, Dur Loss: 0.87717, CE Loss: 0.04930, Norm Loss: 1.05828, F0 Loss: 2.59795, LM Loss: 2.32587, Gen Loss: 5.69046, Sty Loss: 0.07559, Diff Loss: 0.25344, DiscLM Loss: 0.00234, GenLM Loss: 0.98557 +INFO:2025-06-11 17:22:37,864: Epoch [12/20], Step [650/2647], Loss: 0.52597, Disc Loss: 3.68322, Dur Loss: 0.98988, CE Loss: 0.05517, Norm Loss: 0.95099, F0 Loss: 3.72511, LM Loss: 2.22403, Gen Loss: 6.77663, Sty Loss: 0.11402, Diff Loss: 0.27088, DiscLM Loss: 0.00000, GenLM Loss: 1.00038 +INFO:2025-06-11 17:24:25,547: Epoch [12/20], Step [700/2647], Loss: 0.52130, Disc Loss: 3.62446, Dur Loss: 1.41666, CE Loss: 0.08272, Norm Loss: 0.98581, F0 Loss: 2.80942, LM Loss: 2.32816, Gen Loss: 7.09105, Sty Loss: 0.12707, Diff Loss: 0.28117, DiscLM Loss: 0.00078, GenLM Loss: 1.00303 +INFO:2025-06-11 17:26:15,485: Epoch [12/20], Step [750/2647], Loss: 0.51643, Disc Loss: 3.73966, Dur Loss: 0.98310, CE Loss: 0.05288, Norm Loss: 1.09268, F0 Loss: 3.56238, LM Loss: 2.26951, Gen Loss: 6.84534, Sty Loss: 0.13549, Diff Loss: 0.33683, DiscLM Loss: 0.00000, GenLM Loss: 0.99908 +INFO:2025-06-11 17:28:06,030: Epoch [12/20], Step [800/2647], Loss: 0.52217, Disc Loss: 3.73715, Dur Loss: 0.99895, CE Loss: 0.05747, Norm Loss: 1.12704, F0 Loss: 4.42933, LM Loss: 2.12645, Gen Loss: 6.65991, Sty Loss: 0.11388, Diff Loss: 0.35265, DiscLM Loss: 0.00158, GenLM Loss: 1.00138 +INFO:2025-06-11 17:29:56,754: Epoch [12/20], Step [850/2647], Loss: 0.53907, Disc Loss: 3.73279, Dur Loss: 1.23147, CE Loss: 0.05695, Norm Loss: 1.15590, F0 Loss: 3.32096, LM Loss: 2.17697, Gen Loss: 5.99868, Sty Loss: 0.10259, Diff Loss: 0.30060, DiscLM Loss: 0.00000, GenLM Loss: 1.00644 +INFO:2025-06-11 17:31:46,969: Epoch [12/20], Step [900/2647], Loss: 0.52595, Disc Loss: 3.73025, Dur Loss: 0.92545, CE Loss: 0.04741, Norm Loss: 1.18879, F0 Loss: 2.79152, LM Loss: 2.19606, Gen Loss: 6.28957, Sty Loss: 0.11868, Diff Loss: 0.32760, DiscLM Loss: 0.00040, GenLM Loss: 1.00820 +INFO:2025-06-11 17:33:36,110: Epoch [12/20], Step [950/2647], Loss: 0.53283, Disc Loss: 3.69336, Dur Loss: 0.89314, CE Loss: 0.04584, Norm Loss: 0.80165, F0 Loss: 2.58149, LM Loss: 2.11589, Gen Loss: 5.96977, Sty Loss: 0.10215, Diff Loss: 0.33583, DiscLM Loss: 0.00000, GenLM Loss: 0.99482 +INFO:2025-06-11 17:35:24,778: Epoch [12/20], Step [1000/2647], Loss: 0.52259, Disc Loss: 3.70645, Dur Loss: 0.95806, CE Loss: 0.05108, Norm Loss: 0.96093, F0 Loss: 2.84006, LM Loss: 2.12998, Gen Loss: 5.93899, Sty Loss: 0.09706, Diff Loss: 0.27160, DiscLM Loss: 0.00391, GenLM Loss: 0.99928 +INFO:2025-06-11 17:37:12,374: Epoch [12/20], Step [1050/2647], Loss: 0.51965, Disc Loss: 3.67477, Dur Loss: 0.90847, CE Loss: 0.04884, Norm Loss: 0.94526, F0 Loss: 2.71622, LM Loss: 2.18209, Gen Loss: 6.18975, Sty Loss: 0.10093, Diff Loss: 0.20832, DiscLM Loss: 0.00000, GenLM Loss: 0.99911 +INFO:2025-06-11 17:39:00,698: Epoch [12/20], Step [1100/2647], Loss: 0.53484, Disc Loss: 3.65136, Dur Loss: 0.99857, CE Loss: 0.05383, Norm Loss: 1.02258, F0 Loss: 2.85418, LM Loss: 2.20408, Gen Loss: 6.46505, Sty Loss: 0.11390, Diff Loss: 0.28234, DiscLM Loss: 0.00089, GenLM Loss: 0.99881 +INFO:2025-06-11 17:40:50,155: Epoch [12/20], Step [1150/2647], Loss: 0.53071, Disc Loss: 3.69514, Dur Loss: 0.83721, CE Loss: 0.04392, Norm Loss: 0.94407, F0 Loss: 3.40568, LM Loss: 2.18948, Gen Loss: 5.91559, Sty Loss: 0.09074, Diff Loss: 0.41498, DiscLM Loss: 0.00000, GenLM Loss: 1.00791 +INFO:2025-06-11 17:42:37,947: Epoch [12/20], Step [1200/2647], Loss: 0.51688, Disc Loss: 3.73397, Dur Loss: 0.94889, CE Loss: 0.05058, Norm Loss: 1.09823, F0 Loss: 2.48282, LM Loss: 2.15625, Gen Loss: 5.96233, Sty Loss: 0.11468, Diff Loss: 0.23100, DiscLM Loss: 0.01040, GenLM Loss: 0.96088 +INFO:2025-06-11 17:44:26,468: Epoch [12/20], Step [1250/2647], Loss: 0.52795, Disc Loss: 3.77700, Dur Loss: 0.90193, CE Loss: 0.04657, Norm Loss: 0.87129, F0 Loss: 2.95350, LM Loss: 2.17182, Gen Loss: 5.92017, Sty Loss: 0.10369, Diff Loss: 0.20366, DiscLM Loss: 0.00000, GenLM Loss: 0.95872 +INFO:2025-06-11 17:46:16,398: Epoch [12/20], Step [1300/2647], Loss: 0.53508, Disc Loss: 3.83153, Dur Loss: 0.89168, CE Loss: 0.04491, Norm Loss: 1.04989, F0 Loss: 2.67634, LM Loss: 2.20900, Gen Loss: 5.84728, Sty Loss: 0.09920, Diff Loss: 0.29562, DiscLM Loss: 0.00106, GenLM Loss: 0.95528 +INFO:2025-06-11 17:49:56,651: Epoch [12/20], Step [1400/2647], Loss: 1.05594, Disc Loss: 3.74018, Dur Loss: 0.82534, CE Loss: 0.04369, Norm Loss: 1.06593, F0 Loss: 2.82709, LM Loss: 2.26419, Gen Loss: 6.01623, Sty Loss: 0.16367, Diff Loss: 0.51007, DiscLM Loss: 0.01189, GenLM Loss: 1.20278 +INFO:2025-06-11 17:51:43,833: Epoch [12/20], Step [1450/2647], Loss: 0.54000, Disc Loss: 3.80152, Dur Loss: 0.99626, CE Loss: 0.05206, Norm Loss: 1.24619, F0 Loss: 3.08970, LM Loss: 2.20591, Gen Loss: 5.89575, Sty Loss: 0.12752, Diff Loss: 0.37713, DiscLM Loss: 0.00000, GenLM Loss: 0.92216 +INFO:2025-06-11 17:53:31,741: Epoch [12/20], Step [1500/2647], Loss: 0.53503, Disc Loss: 3.75839, Dur Loss: 0.96928, CE Loss: 0.05220, Norm Loss: 0.92169, F0 Loss: 2.75607, LM Loss: 2.14804, Gen Loss: 6.07310, Sty Loss: 0.11565, Diff Loss: 0.38198, DiscLM Loss: 0.00581, GenLM Loss: 0.87909 +INFO:2025-06-11 17:55:19,170: Epoch [12/20], Step [1550/2647], Loss: 0.52250, Disc Loss: 3.78216, Dur Loss: 0.87614, CE Loss: 0.04535, Norm Loss: 1.29442, F0 Loss: 3.58584, LM Loss: 2.26399, Gen Loss: 6.31635, Sty Loss: 0.13202, Diff Loss: 0.41112, DiscLM Loss: 0.00000, GenLM Loss: 1.01176 +INFO:2025-06-11 17:57:04,603: Epoch [12/20], Step [1600/2647], Loss: 0.51981, Disc Loss: 3.71145, Dur Loss: 1.23786, CE Loss: 0.07894, Norm Loss: 1.30345, F0 Loss: 3.14848, LM Loss: 2.33645, Gen Loss: 6.20620, Sty Loss: 0.13363, Diff Loss: 0.36996, DiscLM Loss: 0.01303, GenLM Loss: 1.05879 +INFO:2025-06-11 17:58:52,220: Epoch [12/20], Step [1650/2647], Loss: 0.52550, Disc Loss: 3.85024, Dur Loss: 0.93416, CE Loss: 0.04751, Norm Loss: 1.19705, F0 Loss: 2.79883, LM Loss: 2.22755, Gen Loss: 5.62396, Sty Loss: 0.11656, Diff Loss: 0.19741, DiscLM Loss: 0.00000, GenLM Loss: 0.96703 +INFO:2025-06-11 18:00:40,685: Epoch [12/20], Step [1700/2647], Loss: 0.51278, Disc Loss: 3.81316, Dur Loss: 0.83975, CE Loss: 0.04324, Norm Loss: 0.94870, F0 Loss: 2.77684, LM Loss: 2.12945, Gen Loss: 6.16310, Sty Loss: 0.11614, Diff Loss: 0.26683, DiscLM Loss: 0.03593, GenLM Loss: 0.86147 +INFO:2025-06-11 18:02:30,329: Epoch [12/20], Step [1750/2647], Loss: 0.51362, Disc Loss: 3.74871, Dur Loss: 0.89522, CE Loss: 0.04243, Norm Loss: 0.91347, F0 Loss: 2.62760, LM Loss: 2.17115, Gen Loss: 5.54312, Sty Loss: 0.10303, Diff Loss: 0.19540, DiscLM Loss: 0.00000, GenLM Loss: 0.91392 +INFO:2025-06-11 18:04:17,950: Epoch [12/20], Step [1800/2647], Loss: 0.51025, Disc Loss: 3.81463, Dur Loss: 1.07871, CE Loss: 0.06037, Norm Loss: 1.08090, F0 Loss: 3.08541, LM Loss: 2.19938, Gen Loss: 6.07789, Sty Loss: 0.11648, Diff Loss: 0.28297, DiscLM Loss: 0.00257, GenLM Loss: 1.04112 +INFO:2025-06-11 18:06:05,728: Epoch [12/20], Step [1850/2647], Loss: 0.52073, Disc Loss: 3.70083, Dur Loss: 0.91058, CE Loss: 0.05102, Norm Loss: 1.11521, F0 Loss: 3.61417, LM Loss: 2.24551, Gen Loss: 5.86667, Sty Loss: 0.10733, Diff Loss: 0.30148, DiscLM Loss: 0.00000, GenLM Loss: 0.99484 +INFO:2025-06-11 18:07:54,226: Epoch [12/20], Step [1900/2647], Loss: 0.51007, Disc Loss: 3.80482, Dur Loss: 0.80832, CE Loss: 0.04461, Norm Loss: 1.19462, F0 Loss: 3.16262, LM Loss: 2.07358, Gen Loss: 6.29988, Sty Loss: 0.11221, Diff Loss: 0.29103, DiscLM Loss: 0.00481, GenLM Loss: 0.96216 +INFO:2025-06-11 18:09:43,358: Epoch [12/20], Step [1950/2647], Loss: 0.52552, Disc Loss: 3.67067, Dur Loss: 0.97769, CE Loss: 0.05242, Norm Loss: 1.11921, F0 Loss: 2.81328, LM Loss: 2.11603, Gen Loss: 6.18359, Sty Loss: 0.10138, Diff Loss: 0.32390, DiscLM Loss: 0.00000, GenLM Loss: 0.98372 +INFO:2025-06-11 18:11:32,961: Epoch [12/20], Step [2000/2647], Loss: 0.51193, Disc Loss: 3.69295, Dur Loss: 0.99535, CE Loss: 0.05917, Norm Loss: 0.81698, F0 Loss: 2.73997, LM Loss: 2.20472, Gen Loss: 6.91174, Sty Loss: 0.11859, Diff Loss: 0.31473, DiscLM Loss: 0.00300, GenLM Loss: 1.02325 +INFO:2025-06-11 18:13:20,326: Epoch [12/20], Step [2050/2647], Loss: 0.51200, Disc Loss: 3.82092, Dur Loss: 0.84188, CE Loss: 0.04342, Norm Loss: 0.84152, F0 Loss: 2.29096, LM Loss: 2.25913, Gen Loss: 5.83914, Sty Loss: 0.13171, Diff Loss: 0.32112, DiscLM Loss: 0.00000, GenLM Loss: 1.03103 +INFO:2025-06-11 18:15:11,785: Epoch [12/20], Step [2100/2647], Loss: 0.52059, Disc Loss: 3.82686, Dur Loss: 0.77616, CE Loss: 0.04375, Norm Loss: 0.79957, F0 Loss: 2.42178, LM Loss: 2.02071, Gen Loss: 5.82214, Sty Loss: 0.12645, Diff Loss: 0.22847, DiscLM Loss: 0.00323, GenLM Loss: 1.00587 +INFO:2025-06-11 18:17:01,414: Epoch [12/20], Step [2150/2647], Loss: 0.52144, Disc Loss: 3.72464, Dur Loss: 1.13421, CE Loss: 0.06609, Norm Loss: 1.09116, F0 Loss: 2.87233, LM Loss: 2.20287, Gen Loss: 6.15435, Sty Loss: 0.10626, Diff Loss: 0.30079, DiscLM Loss: 0.00000, GenLM Loss: 0.96754 +INFO:2025-06-11 18:18:47,046: Epoch [12/20], Step [2200/2647], Loss: 0.51063, Disc Loss: 3.76107, Dur Loss: 0.83621, CE Loss: 0.04200, Norm Loss: 0.84377, F0 Loss: 2.75192, LM Loss: 2.12460, Gen Loss: 6.32867, Sty Loss: 0.07822, Diff Loss: 0.35195, DiscLM Loss: 0.00528, GenLM Loss: 0.98210 +INFO:2025-06-11 18:20:36,231: Epoch [12/20], Step [2250/2647], Loss: 0.52295, Disc Loss: 3.78753, Dur Loss: 0.92132, CE Loss: 0.04893, Norm Loss: 0.93336, F0 Loss: 2.57266, LM Loss: 2.10016, Gen Loss: 6.51191, Sty Loss: 0.10692, Diff Loss: 0.19357, DiscLM Loss: 0.00000, GenLM Loss: 1.01357 +INFO:2025-06-11 18:22:26,552: Epoch [12/20], Step [2300/2647], Loss: 0.52043, Disc Loss: 3.75486, Dur Loss: 1.03668, CE Loss: 0.05959, Norm Loss: 1.15224, F0 Loss: 2.76589, LM Loss: 2.22744, Gen Loss: 6.54124, Sty Loss: 0.12777, Diff Loss: 0.34295, DiscLM Loss: 0.00109, GenLM Loss: 1.00986 +INFO:2025-06-11 18:24:14,842: Epoch [12/20], Step [2350/2647], Loss: 0.51684, Disc Loss: 3.80050, Dur Loss: 0.78540, CE Loss: 0.04018, Norm Loss: 0.83771, F0 Loss: 2.58522, LM Loss: 2.09861, Gen Loss: 6.28770, Sty Loss: 0.10238, Diff Loss: 0.35927, DiscLM Loss: 0.00000, GenLM Loss: 0.99518 +INFO:2025-06-11 18:26:05,338: Epoch [12/20], Step [2400/2647], Loss: 0.51883, Disc Loss: 3.78840, Dur Loss: 0.95278, CE Loss: 0.05013, Norm Loss: 0.74518, F0 Loss: 2.52255, LM Loss: 2.19373, Gen Loss: 6.04486, Sty Loss: 0.10730, Diff Loss: 0.23173, DiscLM Loss: 0.00115, GenLM Loss: 0.99427 +INFO:2025-06-11 18:27:53,039: Epoch [12/20], Step [2450/2647], Loss: 0.51726, Disc Loss: 3.91304, Dur Loss: 0.99124, CE Loss: 0.05373, Norm Loss: 1.02251, F0 Loss: 2.43037, LM Loss: 2.17506, Gen Loss: 6.50561, Sty Loss: 0.12594, Diff Loss: 0.32674, DiscLM Loss: 0.00000, GenLM Loss: 1.00054 +INFO:2025-06-11 18:29:43,464: Epoch [12/20], Step [2500/2647], Loss: 0.52205, Disc Loss: 3.75898, Dur Loss: 1.00803, CE Loss: 0.06254, Norm Loss: 0.95992, F0 Loss: 3.36865, LM Loss: 2.17627, Gen Loss: 5.70309, Sty Loss: 0.12216, Diff Loss: 0.31390, DiscLM Loss: 0.00587, GenLM Loss: 1.00504 +INFO:2025-06-11 18:31:31,710: Epoch [12/20], Step [2550/2647], Loss: 0.51135, Disc Loss: 3.64915, Dur Loss: 0.88388, CE Loss: 0.04894, Norm Loss: 0.75852, F0 Loss: 3.74859, LM Loss: 2.10959, Gen Loss: 7.17331, Sty Loss: 0.11579, Diff Loss: 0.24679, DiscLM Loss: 0.00000, GenLM Loss: 0.99195 +INFO:2025-06-11 18:33:21,041: Epoch [12/20], Step [2600/2647], Loss: 0.51675, Disc Loss: 3.81408, Dur Loss: 1.00314, CE Loss: 0.05525, Norm Loss: 1.16903, F0 Loss: 2.83402, LM Loss: 2.13962, Gen Loss: 5.99870, Sty Loss: 0.08173, Diff Loss: 0.18767, DiscLM Loss: 0.00191, GenLM Loss: 1.00107 +INFO:2025-06-11 18:37:26,322: Validation loss: 0.538, Dur loss: 0.989, F0 loss: 3.243 + + + +INFO:2025-06-11 18:39:18,261: Epoch [13/20], Step [50/2647], Loss: 0.52345, Disc Loss: 3.71458, Dur Loss: 1.00484, CE Loss: 0.05379, Norm Loss: 0.98551, F0 Loss: 2.68780, LM Loss: 2.13450, Gen Loss: 6.12373, Sty Loss: 0.14488, Diff Loss: 0.30495, DiscLM Loss: 0.00000, GenLM Loss: 0.98847 +INFO:2025-06-11 18:41:05,543: Epoch [13/20], Step [100/2647], Loss: 0.51086, Disc Loss: 3.76694, Dur Loss: 0.97399, CE Loss: 0.05404, Norm Loss: 1.25347, F0 Loss: 2.78633, LM Loss: 2.24211, Gen Loss: 6.17653, Sty Loss: 0.10130, Diff Loss: 0.26030, DiscLM Loss: 0.00101, GenLM Loss: 0.99593 +INFO:2025-06-11 18:42:52,723: Epoch [13/20], Step [150/2647], Loss: 0.51422, Disc Loss: 3.73870, Dur Loss: 1.04605, CE Loss: 0.06043, Norm Loss: 0.81518, F0 Loss: 3.13902, LM Loss: 2.20253, Gen Loss: 6.57171, Sty Loss: 0.13041, Diff Loss: 0.35330, DiscLM Loss: 0.00000, GenLM Loss: 0.99605 +INFO:2025-06-11 18:44:43,256: Epoch [13/20], Step [200/2647], Loss: 0.51770, Disc Loss: 3.85131, Dur Loss: 0.85271, CE Loss: 0.04607, Norm Loss: 1.12947, F0 Loss: 2.74615, LM Loss: 2.11066, Gen Loss: 5.67371, Sty Loss: 0.10137, Diff Loss: 0.19139, DiscLM Loss: 0.00753, GenLM Loss: 0.98833 +INFO:2025-06-11 18:46:31,687: Epoch [13/20], Step [250/2647], Loss: 0.51865, Disc Loss: 3.69526, Dur Loss: 1.26713, CE Loss: 0.06164, Norm Loss: 1.18626, F0 Loss: 2.78506, LM Loss: 2.05754, Gen Loss: 6.10735, Sty Loss: 0.13125, Diff Loss: 0.22060, DiscLM Loss: 0.00000, GenLM Loss: 1.00487 +INFO:2025-06-11 18:48:20,730: Epoch [13/20], Step [300/2647], Loss: 0.51030, Disc Loss: 3.64689, Dur Loss: 1.13103, CE Loss: 0.07064, Norm Loss: 1.20296, F0 Loss: 3.07707, LM Loss: 2.17816, Gen Loss: 6.34592, Sty Loss: 0.11514, Diff Loss: 0.34015, DiscLM Loss: 0.00946, GenLM Loss: 0.98842 +INFO:2025-06-11 18:50:09,558: Epoch [13/20], Step [350/2647], Loss: 0.51412, Disc Loss: 3.83397, Dur Loss: 1.01360, CE Loss: 0.05960, Norm Loss: 0.92737, F0 Loss: 2.58758, LM Loss: 2.03606, Gen Loss: 6.53827, Sty Loss: 0.10977, Diff Loss: 0.24377, DiscLM Loss: 0.00000, GenLM Loss: 1.00069 +INFO:2025-06-11 18:52:00,365: Epoch [13/20], Step [400/2647], Loss: 0.52744, Disc Loss: 3.80972, Dur Loss: 1.02799, CE Loss: 0.05830, Norm Loss: 1.43756, F0 Loss: 3.19107, LM Loss: 2.18216, Gen Loss: 5.79894, Sty Loss: 0.10125, Diff Loss: 0.22267, DiscLM Loss: 0.00155, GenLM Loss: 0.99581 +INFO:2025-06-11 18:53:50,332: Epoch [13/20], Step [450/2647], Loss: 0.51798, Disc Loss: 3.65348, Dur Loss: 1.00541, CE Loss: 0.04944, Norm Loss: 1.02847, F0 Loss: 2.93152, LM Loss: 2.16390, Gen Loss: 6.87079, Sty Loss: 0.11764, Diff Loss: 0.27804, DiscLM Loss: 0.00000, GenLM Loss: 0.99520 +INFO:2025-06-11 18:55:40,605: Epoch [13/20], Step [500/2647], Loss: 0.51483, Disc Loss: 3.74631, Dur Loss: 0.82715, CE Loss: 0.04655, Norm Loss: 0.98747, F0 Loss: 3.26613, LM Loss: 2.17172, Gen Loss: 6.41145, Sty Loss: 0.10831, Diff Loss: 0.28784, DiscLM Loss: 0.00155, GenLM Loss: 0.99047 +INFO:2025-06-11 18:57:32,644: Epoch [13/20], Step [550/2647], Loss: 0.51420, Disc Loss: 3.79367, Dur Loss: 1.18130, CE Loss: 0.06802, Norm Loss: 0.90091, F0 Loss: 2.70026, LM Loss: 2.10615, Gen Loss: 5.88808, Sty Loss: 0.10178, Diff Loss: 0.29367, DiscLM Loss: 0.00000, GenLM Loss: 1.01516 +INFO:2025-06-11 18:59:24,335: Epoch [13/20], Step [600/2647], Loss: 0.52490, Disc Loss: 3.78087, Dur Loss: 0.82769, CE Loss: 0.04403, Norm Loss: 0.92621, F0 Loss: 2.99021, LM Loss: 2.05517, Gen Loss: 6.48239, Sty Loss: 0.10853, Diff Loss: 0.33542, DiscLM Loss: 0.00042, GenLM Loss: 1.01440 +INFO:2025-06-11 19:01:14,679: Epoch [13/20], Step [650/2647], Loss: 0.51954, Disc Loss: 3.72440, Dur Loss: 0.99600, CE Loss: 0.05412, Norm Loss: 1.20088, F0 Loss: 3.15595, LM Loss: 2.16078, Gen Loss: 7.27876, Sty Loss: 0.13415, Diff Loss: 0.29278, DiscLM Loss: 0.00000, GenLM Loss: 0.98586 +INFO:2025-06-11 19:03:04,264: Epoch [13/20], Step [700/2647], Loss: 0.50147, Disc Loss: 3.71886, Dur Loss: 0.73727, CE Loss: 0.03999, Norm Loss: 0.93464, F0 Loss: 2.56406, LM Loss: 2.22091, Gen Loss: 6.06678, Sty Loss: 0.10571, Diff Loss: 0.22168, DiscLM Loss: 0.00108, GenLM Loss: 0.98015 +INFO:2025-06-11 19:04:53,246: Epoch [13/20], Step [750/2647], Loss: 0.51194, Disc Loss: 3.75383, Dur Loss: 0.86247, CE Loss: 0.04664, Norm Loss: 0.99723, F0 Loss: 3.08351, LM Loss: 2.09051, Gen Loss: 5.63732, Sty Loss: 0.11371, Diff Loss: 0.19789, DiscLM Loss: 0.00000, GenLM Loss: 0.99128 +INFO:2025-06-11 19:06:41,049: Epoch [13/20], Step [800/2647], Loss: 0.51640, Disc Loss: 3.71377, Dur Loss: 1.06956, CE Loss: 0.05432, Norm Loss: 0.85341, F0 Loss: 3.31124, LM Loss: 2.17553, Gen Loss: 6.93797, Sty Loss: 0.15384, Diff Loss: 0.24917, DiscLM Loss: 0.00086, GenLM Loss: 1.00414 +INFO:2025-06-11 19:08:30,174: Epoch [13/20], Step [850/2647], Loss: 0.51228, Disc Loss: 3.82827, Dur Loss: 0.98636, CE Loss: 0.04953, Norm Loss: 1.06358, F0 Loss: 3.29878, LM Loss: 2.14329, Gen Loss: 6.29999, Sty Loss: 0.11087, Diff Loss: 0.26953, DiscLM Loss: 0.00000, GenLM Loss: 1.00573 +INFO:2025-06-11 19:10:16,319: Epoch [13/20], Step [900/2647], Loss: 0.52286, Disc Loss: 3.74022, Dur Loss: 1.34994, CE Loss: 0.07592, Norm Loss: 1.99557, F0 Loss: 3.41261, LM Loss: 2.27921, Gen Loss: 6.04049, Sty Loss: 0.11401, Diff Loss: 0.19853, DiscLM Loss: 0.00067, GenLM Loss: 0.99250 +INFO:2025-06-11 19:12:05,320: Epoch [13/20], Step [950/2647], Loss: 0.52144, Disc Loss: 3.78805, Dur Loss: 0.87302, CE Loss: 0.04531, Norm Loss: 1.04872, F0 Loss: 2.33635, LM Loss: 2.26156, Gen Loss: 6.57550, Sty Loss: 0.12634, Diff Loss: 0.24108, DiscLM Loss: 0.00000, GenLM Loss: 1.00421 +INFO:2025-06-11 19:13:55,672: Epoch [13/20], Step [1000/2647], Loss: 0.51620, Disc Loss: 3.73972, Dur Loss: 0.84553, CE Loss: 0.04522, Norm Loss: 0.87732, F0 Loss: 2.50358, LM Loss: 2.18668, Gen Loss: 7.24392, Sty Loss: 0.12360, Diff Loss: 0.30014, DiscLM Loss: 0.00076, GenLM Loss: 1.01261 +INFO:2025-06-11 19:15:45,125: Epoch [13/20], Step [1050/2647], Loss: 0.51893, Disc Loss: 3.75204, Dur Loss: 0.91727, CE Loss: 0.05085, Norm Loss: 0.80381, F0 Loss: 3.09813, LM Loss: 2.19423, Gen Loss: 6.20251, Sty Loss: 0.10084, Diff Loss: 0.28166, DiscLM Loss: 0.00000, GenLM Loss: 1.00246 +INFO:2025-06-11 19:17:34,033: Epoch [13/20], Step [1100/2647], Loss: 0.51753, Disc Loss: 3.64410, Dur Loss: 1.05004, CE Loss: 0.05836, Norm Loss: 0.91279, F0 Loss: 2.88298, LM Loss: 2.11075, Gen Loss: 6.20980, Sty Loss: 0.10334, Diff Loss: 0.28449, DiscLM Loss: 0.00272, GenLM Loss: 0.99575 +INFO:2025-06-11 19:19:22,966: Epoch [13/20], Step [1150/2647], Loss: 0.51437, Disc Loss: 3.75146, Dur Loss: 0.93111, CE Loss: 0.05530, Norm Loss: 1.18518, F0 Loss: 2.43523, LM Loss: 2.16546, Gen Loss: 6.37763, Sty Loss: 0.13162, Diff Loss: 0.32539, DiscLM Loss: 0.00000, GenLM Loss: 0.99510 +INFO:2025-06-11 19:21:12,470: Epoch [13/20], Step [1200/2647], Loss: 0.52141, Disc Loss: 3.72144, Dur Loss: 1.10462, CE Loss: 0.06261, Norm Loss: 1.05145, F0 Loss: 3.34900, LM Loss: 2.13712, Gen Loss: 6.48557, Sty Loss: 0.10754, Diff Loss: 0.28832, DiscLM Loss: 0.00713, GenLM Loss: 1.00583 +INFO:2025-06-11 19:23:00,854: Epoch [13/20], Step [1250/2647], Loss: 0.51530, Disc Loss: 3.71056, Dur Loss: 0.79778, CE Loss: 0.04289, Norm Loss: 0.90407, F0 Loss: 3.32165, LM Loss: 2.14752, Gen Loss: 6.65002, Sty Loss: 0.11495, Diff Loss: 0.25045, DiscLM Loss: 0.00000, GenLM Loss: 0.99100 +INFO:2025-06-11 19:24:51,339: Epoch [13/20], Step [1300/2647], Loss: 0.50467, Disc Loss: 3.65089, Dur Loss: 1.00932, CE Loss: 0.05268, Norm Loss: 1.03932, F0 Loss: 2.57716, LM Loss: 2.20927, Gen Loss: 6.65729, Sty Loss: 0.09720, Diff Loss: 0.25875, DiscLM Loss: 0.00082, GenLM Loss: 0.99158 +INFO:2025-06-11 19:26:39,130: Epoch [13/20], Step [1350/2647], Loss: 0.51965, Disc Loss: 3.79670, Dur Loss: 0.82982, CE Loss: 0.04358, Norm Loss: 0.87001, F0 Loss: 2.64377, LM Loss: 2.09772, Gen Loss: 5.98341, Sty Loss: 0.11569, Diff Loss: 0.21959, DiscLM Loss: 0.00000, GenLM Loss: 1.00398 +INFO:2025-06-11 19:28:30,293: Epoch [13/20], Step [1400/2647], Loss: 0.51882, Disc Loss: 3.66766, Dur Loss: 1.17753, CE Loss: 0.06167, Norm Loss: 1.16987, F0 Loss: 3.53042, LM Loss: 2.20966, Gen Loss: 5.98895, Sty Loss: 0.08914, Diff Loss: 0.23869, DiscLM Loss: 0.00085, GenLM Loss: 0.99809 +INFO:2025-06-11 19:30:19,079: Epoch [13/20], Step [1450/2647], Loss: 0.51036, Disc Loss: 3.76601, Dur Loss: 1.10221, CE Loss: 0.05510, Norm Loss: 1.28465, F0 Loss: 4.26801, LM Loss: 2.19657, Gen Loss: 5.65976, Sty Loss: 0.10856, Diff Loss: 0.27171, DiscLM Loss: 0.00000, GenLM Loss: 1.00382 +INFO:2025-06-11 19:32:07,747: Epoch [13/20], Step [1500/2647], Loss: 0.51425, Disc Loss: 3.75175, Dur Loss: 1.08338, CE Loss: 0.05789, Norm Loss: 1.31588, F0 Loss: 3.00211, LM Loss: 2.29547, Gen Loss: 6.71978, Sty Loss: 0.09426, Diff Loss: 0.21244, DiscLM Loss: 0.00044, GenLM Loss: 1.00707 +INFO:2025-06-11 19:33:54,262: Epoch [13/20], Step [1550/2647], Loss: 0.51624, Disc Loss: 3.72317, Dur Loss: 0.89914, CE Loss: 0.05359, Norm Loss: 0.85462, F0 Loss: 2.96089, LM Loss: 2.13700, Gen Loss: 6.56088, Sty Loss: 0.10413, Diff Loss: 0.29326, DiscLM Loss: 0.00000, GenLM Loss: 1.00006 +INFO:2025-06-11 19:35:41,893: Epoch [13/20], Step [1600/2647], Loss: 0.50645, Disc Loss: 3.76145, Dur Loss: 0.82335, CE Loss: 0.04337, Norm Loss: 1.00967, F0 Loss: 2.57134, LM Loss: 2.11041, Gen Loss: 5.61684, Sty Loss: 0.10832, Diff Loss: 0.34312, DiscLM Loss: 0.00053, GenLM Loss: 0.99163 +INFO:2025-06-11 19:37:28,436: Epoch [13/20], Step [1650/2647], Loss: 0.51729, Disc Loss: 3.71824, Dur Loss: 1.00300, CE Loss: 0.05334, Norm Loss: 0.97323, F0 Loss: 2.45834, LM Loss: 2.23808, Gen Loss: 6.20131, Sty Loss: 0.11338, Diff Loss: 0.34221, DiscLM Loss: 0.00000, GenLM Loss: 1.00286 +INFO:2025-06-11 19:39:19,357: Epoch [13/20], Step [1700/2647], Loss: 0.51787, Disc Loss: 3.71934, Dur Loss: 0.88312, CE Loss: 0.04494, Norm Loss: 0.85059, F0 Loss: 2.89971, LM Loss: 2.10472, Gen Loss: 6.78477, Sty Loss: 0.12656, Diff Loss: 0.37072, DiscLM Loss: 0.00163, GenLM Loss: 1.00731 +INFO:2025-06-11 19:41:09,272: Epoch [13/20], Step [1750/2647], Loss: 0.51241, Disc Loss: 3.72216, Dur Loss: 0.91673, CE Loss: 0.05272, Norm Loss: 1.35581, F0 Loss: 3.73751, LM Loss: 2.17953, Gen Loss: 6.38729, Sty Loss: 0.13090, Diff Loss: 0.26268, DiscLM Loss: 0.00000, GenLM Loss: 0.99882 +INFO:2025-06-11 19:42:57,430: Epoch [13/20], Step [1800/2647], Loss: 0.51510, Disc Loss: 3.86309, Dur Loss: 1.06140, CE Loss: 0.06101, Norm Loss: 1.11275, F0 Loss: 4.47999, LM Loss: 2.21338, Gen Loss: 6.03769, Sty Loss: 0.12024, Diff Loss: 0.21632, DiscLM Loss: 0.00070, GenLM Loss: 0.98666 +INFO:2025-06-11 19:44:48,238: Epoch [13/20], Step [1850/2647], Loss: 0.52273, Disc Loss: 3.77754, Dur Loss: 1.02940, CE Loss: 0.05681, Norm Loss: 0.96827, F0 Loss: 3.06999, LM Loss: 2.17562, Gen Loss: 6.50936, Sty Loss: 0.11364, Diff Loss: 0.23120, DiscLM Loss: 0.00000, GenLM Loss: 0.99267 +INFO:2025-06-11 19:46:36,547: Epoch [13/20], Step [1900/2647], Loss: 0.52293, Disc Loss: 3.83210, Dur Loss: 1.07391, CE Loss: 0.05791, Norm Loss: 1.20433, F0 Loss: 2.64112, LM Loss: 2.29419, Gen Loss: 6.07336, Sty Loss: 0.10941, Diff Loss: 0.28636, DiscLM Loss: 0.00079, GenLM Loss: 0.99349 +INFO:2025-06-11 19:48:26,730: Epoch [13/20], Step [1950/2647], Loss: 0.51004, Disc Loss: 3.67110, Dur Loss: 0.83338, CE Loss: 0.04798, Norm Loss: 1.02638, F0 Loss: 2.90863, LM Loss: 2.24412, Gen Loss: 6.74789, Sty Loss: 0.12941, Diff Loss: 0.40028, DiscLM Loss: 0.00000, GenLM Loss: 0.99705 +INFO:2025-06-11 19:50:14,777: Epoch [13/20], Step [2000/2647], Loss: 0.51334, Disc Loss: 3.76886, Dur Loss: 1.17780, CE Loss: 0.06288, Norm Loss: 1.08856, F0 Loss: 2.75687, LM Loss: 2.11259, Gen Loss: 5.74429, Sty Loss: 0.09535, Diff Loss: 0.16122, DiscLM Loss: 0.00233, GenLM Loss: 0.99425 +INFO:2025-06-11 19:52:04,729: Epoch [13/20], Step [2050/2647], Loss: 0.52487, Disc Loss: 3.69970, Dur Loss: 1.04649, CE Loss: 0.05698, Norm Loss: 1.23160, F0 Loss: 3.43804, LM Loss: 2.18784, Gen Loss: 6.98182, Sty Loss: 0.09899, Diff Loss: 0.25326, DiscLM Loss: 0.00000, GenLM Loss: 1.01663 +INFO:2025-06-11 19:53:53,484: Epoch [13/20], Step [2100/2647], Loss: 0.51783, Disc Loss: 3.73763, Dur Loss: 0.81350, CE Loss: 0.04228, Norm Loss: 1.02903, F0 Loss: 3.19173, LM Loss: 2.18354, Gen Loss: 6.49093, Sty Loss: 0.09883, Diff Loss: 0.33217, DiscLM Loss: 0.00125, GenLM Loss: 0.98434 +INFO:2025-06-11 19:55:42,772: Epoch [13/20], Step [2150/2647], Loss: 0.52747, Disc Loss: 3.64241, Dur Loss: 0.86387, CE Loss: 0.04460, Norm Loss: 0.89845, F0 Loss: 3.20525, LM Loss: 2.09705, Gen Loss: 6.62778, Sty Loss: 0.10030, Diff Loss: 0.22640, DiscLM Loss: 0.00000, GenLM Loss: 0.97701 +INFO:2025-06-11 19:57:33,332: Epoch [13/20], Step [2200/2647], Loss: 0.51932, Disc Loss: 3.76938, Dur Loss: 1.01372, CE Loss: 0.05162, Norm Loss: 0.68964, F0 Loss: 2.12756, LM Loss: 2.08066, Gen Loss: 6.29347, Sty Loss: 0.11136, Diff Loss: 0.31149, DiscLM Loss: 0.00234, GenLM Loss: 0.99968 +INFO:2025-06-11 19:59:23,815: Epoch [13/20], Step [2250/2647], Loss: 0.52152, Disc Loss: 3.70641, Dur Loss: 0.81146, CE Loss: 0.04067, Norm Loss: 0.89909, F0 Loss: 3.07016, LM Loss: 2.08968, Gen Loss: 6.42072, Sty Loss: 0.10280, Diff Loss: 0.30456, DiscLM Loss: 0.00000, GenLM Loss: 1.00712 +INFO:2025-06-11 20:01:15,322: Epoch [13/20], Step [2300/2647], Loss: 0.51207, Disc Loss: 3.71469, Dur Loss: 0.89795, CE Loss: 0.05078, Norm Loss: 0.86570, F0 Loss: 2.86099, LM Loss: 2.13131, Gen Loss: 6.34868, Sty Loss: 0.10500, Diff Loss: 0.26780, DiscLM Loss: 0.00071, GenLM Loss: 0.98843 +INFO:2025-06-11 20:03:05,243: Epoch [13/20], Step [2350/2647], Loss: 0.51611, Disc Loss: 3.77780, Dur Loss: 1.01001, CE Loss: 0.05466, Norm Loss: 1.33457, F0 Loss: 3.53870, LM Loss: 2.13687, Gen Loss: 5.97883, Sty Loss: 0.13095, Diff Loss: 0.37038, DiscLM Loss: 0.00000, GenLM Loss: 1.00326 +INFO:2025-06-11 20:04:52,997: Epoch [13/20], Step [2400/2647], Loss: 0.51782, Disc Loss: 3.71201, Dur Loss: 1.05179, CE Loss: 0.05521, Norm Loss: 1.17915, F0 Loss: 3.42898, LM Loss: 2.13641, Gen Loss: 6.01261, Sty Loss: 0.11318, Diff Loss: 0.27230, DiscLM Loss: 0.00412, GenLM Loss: 1.01958 +INFO:2025-06-11 20:06:38,816: Epoch [13/20], Step [2450/2647], Loss: 0.52452, Disc Loss: 3.85819, Dur Loss: 0.84975, CE Loss: 0.04431, Norm Loss: 0.96490, F0 Loss: 2.61791, LM Loss: 2.07820, Gen Loss: 5.81056, Sty Loss: 0.09292, Diff Loss: 0.20496, DiscLM Loss: 0.00000, GenLM Loss: 0.99043 +INFO:2025-06-11 20:08:27,656: Epoch [13/20], Step [2500/2647], Loss: 0.51822, Disc Loss: 3.80677, Dur Loss: 0.92513, CE Loss: 0.05145, Norm Loss: 0.82494, F0 Loss: 3.06872, LM Loss: 2.16151, Gen Loss: 6.49400, Sty Loss: 0.08999, Diff Loss: 0.18978, DiscLM Loss: 0.04157, GenLM Loss: 0.97431 +INFO:2025-06-11 20:10:17,011: Epoch [13/20], Step [2550/2647], Loss: 0.52214, Disc Loss: 3.85647, Dur Loss: 0.95785, CE Loss: 0.06202, Norm Loss: 1.28238, F0 Loss: 2.88777, LM Loss: 2.27343, Gen Loss: 6.54544, Sty Loss: 0.09564, Diff Loss: 0.28729, DiscLM Loss: 0.00000, GenLM Loss: 0.99118 +INFO:2025-06-11 20:12:05,746: Epoch [13/20], Step [2600/2647], Loss: 0.51218, Disc Loss: 3.63120, Dur Loss: 0.82729, CE Loss: 0.04317, Norm Loss: 0.78235, F0 Loss: 2.82965, LM Loss: 2.01792, Gen Loss: 7.01182, Sty Loss: 0.10449, Diff Loss: 0.23104, DiscLM Loss: 0.05483, GenLM Loss: 0.90484 +INFO:2025-06-11 20:16:09,317: Validation loss: 0.531, Dur loss: 0.985, F0 loss: 3.292 + + + +INFO:2025-06-11 20:17:59,655: Epoch [14/20], Step [50/2647], Loss: 0.50936, Disc Loss: 3.77848, Dur Loss: 0.93230, CE Loss: 0.04737, Norm Loss: 1.28929, F0 Loss: 4.11027, LM Loss: 2.26351, Gen Loss: 6.47830, Sty Loss: 0.11489, Diff Loss: 0.28803, DiscLM Loss: 0.00000, GenLM Loss: 0.92781 +INFO:2025-06-11 20:19:50,151: Epoch [14/20], Step [100/2647], Loss: 0.51033, Disc Loss: 3.75842, Dur Loss: 0.79831, CE Loss: 0.04021, Norm Loss: 1.04486, F0 Loss: 2.67573, LM Loss: 2.12281, Gen Loss: 5.95504, Sty Loss: 0.10671, Diff Loss: 0.29067, DiscLM Loss: 0.00501, GenLM Loss: 0.95058 +INFO:2025-06-11 20:21:41,882: Epoch [14/20], Step [150/2647], Loss: 0.51545, Disc Loss: 3.72865, Dur Loss: 1.03024, CE Loss: 0.05925, Norm Loss: 1.23192, F0 Loss: 3.14732, LM Loss: 2.20157, Gen Loss: 6.36697, Sty Loss: 0.09078, Diff Loss: 0.17852, DiscLM Loss: 0.00000, GenLM Loss: 1.05239 +INFO:2025-06-11 20:23:30,635: Epoch [14/20], Step [200/2647], Loss: 0.51752, Disc Loss: 3.79970, Dur Loss: 1.15058, CE Loss: 0.06640, Norm Loss: 0.96192, F0 Loss: 2.74976, LM Loss: 2.17900, Gen Loss: 6.48514, Sty Loss: 0.09844, Diff Loss: 0.32163, DiscLM Loss: 0.00127, GenLM Loss: 1.00772 +INFO:2025-06-11 20:25:22,069: Epoch [14/20], Step [250/2647], Loss: 0.50695, Disc Loss: 3.73406, Dur Loss: 0.97508, CE Loss: 0.05394, Norm Loss: 1.29061, F0 Loss: 2.84603, LM Loss: 2.22974, Gen Loss: 6.34321, Sty Loss: 0.13710, Diff Loss: 0.28731, DiscLM Loss: 0.00000, GenLM Loss: 0.96562 +INFO:2025-06-11 20:27:11,254: Epoch [14/20], Step [300/2647], Loss: 0.51351, Disc Loss: 3.72961, Dur Loss: 0.74401, CE Loss: 0.04053, Norm Loss: 1.02984, F0 Loss: 2.54439, LM Loss: 2.25228, Gen Loss: 6.80834, Sty Loss: 0.11871, Diff Loss: 0.20898, DiscLM Loss: 0.00208, GenLM Loss: 0.99502 +INFO:2025-06-11 20:29:00,766: Epoch [14/20], Step [350/2647], Loss: 0.52310, Disc Loss: 3.79191, Dur Loss: 1.02129, CE Loss: 0.05323, Norm Loss: 0.89807, F0 Loss: 3.39156, LM Loss: 2.09557, Gen Loss: 5.65930, Sty Loss: 0.11812, Diff Loss: 0.21498, DiscLM Loss: 0.00000, GenLM Loss: 1.01569 +INFO:2025-06-11 20:30:50,837: Epoch [14/20], Step [400/2647], Loss: 0.52516, Disc Loss: 3.74839, Dur Loss: 1.02792, CE Loss: 0.05385, Norm Loss: 1.03023, F0 Loss: 2.80819, LM Loss: 2.14363, Gen Loss: 6.21020, Sty Loss: 0.12040, Diff Loss: 0.41149, DiscLM Loss: 0.00055, GenLM Loss: 1.00027 +INFO:2025-06-11 20:32:40,203: Epoch [14/20], Step [450/2647], Loss: 0.51853, Disc Loss: 3.80016, Dur Loss: 1.01774, CE Loss: 0.05735, Norm Loss: 1.21325, F0 Loss: 3.47964, LM Loss: 2.17477, Gen Loss: 6.60010, Sty Loss: 0.10463, Diff Loss: 0.28524, DiscLM Loss: 0.00000, GenLM Loss: 0.90729 +INFO:2025-06-11 20:34:28,044: Epoch [14/20], Step [500/2647], Loss: 0.51225, Disc Loss: 3.81605, Dur Loss: 0.95060, CE Loss: 0.05101, Norm Loss: 1.13791, F0 Loss: 2.84423, LM Loss: 2.01218, Gen Loss: 6.13575, Sty Loss: 0.13439, Diff Loss: 0.25590, DiscLM Loss: 0.00339, GenLM Loss: 0.99728 +INFO:2025-06-11 20:36:20,067: Epoch [14/20], Step [550/2647], Loss: 0.51849, Disc Loss: 3.71579, Dur Loss: 1.39051, CE Loss: 0.07577, Norm Loss: 0.99083, F0 Loss: 3.62515, LM Loss: 2.25762, Gen Loss: 6.66777, Sty Loss: 0.13067, Diff Loss: 0.30243, DiscLM Loss: 0.00000, GenLM Loss: 0.97884 +INFO:2025-06-11 20:38:08,533: Epoch [14/20], Step [600/2647], Loss: 0.51229, Disc Loss: 3.70917, Dur Loss: 1.09791, CE Loss: 0.05963, Norm Loss: 1.16603, F0 Loss: 3.22417, LM Loss: 2.16000, Gen Loss: 6.63838, Sty Loss: 0.11433, Diff Loss: 0.33218, DiscLM Loss: 0.00199, GenLM Loss: 0.98854 +INFO:2025-06-11 20:39:57,972: Epoch [14/20], Step [650/2647], Loss: 0.51138, Disc Loss: 3.79361, Dur Loss: 0.97140, CE Loss: 0.04885, Norm Loss: 0.79159, F0 Loss: 2.84615, LM Loss: 2.04079, Gen Loss: 6.22556, Sty Loss: 0.09095, Diff Loss: 0.26317, DiscLM Loss: 0.00000, GenLM Loss: 0.98637 +INFO:2025-06-11 20:41:47,950: Epoch [14/20], Step [700/2647], Loss: 0.51302, Disc Loss: 3.83402, Dur Loss: 1.06340, CE Loss: 0.05591, Norm Loss: 1.24581, F0 Loss: 2.81096, LM Loss: 2.14920, Gen Loss: 6.06790, Sty Loss: 0.12375, Diff Loss: 0.37046, DiscLM Loss: 0.00066, GenLM Loss: 0.99431 +INFO:2025-06-11 20:43:36,251: Epoch [14/20], Step [750/2647], Loss: 0.52197, Disc Loss: 3.76508, Dur Loss: 0.99522, CE Loss: 0.05259, Norm Loss: 0.96561, F0 Loss: 3.01532, LM Loss: 2.11302, Gen Loss: 6.52322, Sty Loss: 0.12019, Diff Loss: 0.32872, DiscLM Loss: 0.00000, GenLM Loss: 1.00927 +INFO:2025-06-11 20:45:25,064: Epoch [14/20], Step [800/2647], Loss: 0.51137, Disc Loss: 3.68661, Dur Loss: 0.85720, CE Loss: 0.04673, Norm Loss: 0.86519, F0 Loss: 3.17590, LM Loss: 2.16091, Gen Loss: 6.82717, Sty Loss: 0.12087, Diff Loss: 0.24635, DiscLM Loss: 0.00255, GenLM Loss: 1.00728 +INFO:2025-06-11 20:47:14,253: Epoch [14/20], Step [850/2647], Loss: 0.51241, Disc Loss: 3.73232, Dur Loss: 1.11099, CE Loss: 0.05967, Norm Loss: 0.97624, F0 Loss: 3.38286, LM Loss: 2.08004, Gen Loss: 6.19668, Sty Loss: 0.11420, Diff Loss: 0.28718, DiscLM Loss: 0.00000, GenLM Loss: 0.99232 +INFO:2025-06-11 20:49:03,077: Epoch [14/20], Step [900/2647], Loss: 0.51527, Disc Loss: 3.83195, Dur Loss: 0.78199, CE Loss: 0.04356, Norm Loss: 0.88676, F0 Loss: 2.13514, LM Loss: 2.18374, Gen Loss: 5.80899, Sty Loss: 0.10990, Diff Loss: 0.22917, DiscLM Loss: 0.00048, GenLM Loss: 0.99473 +INFO:2025-06-11 20:50:51,375: Epoch [14/20], Step [950/2647], Loss: 0.51264, Disc Loss: 3.73809, Dur Loss: 0.88262, CE Loss: 0.05082, Norm Loss: 0.93100, F0 Loss: 2.96580, LM Loss: 2.02662, Gen Loss: 5.69160, Sty Loss: 0.09542, Diff Loss: 0.29046, DiscLM Loss: 0.00000, GenLM Loss: 1.00418 +INFO:2025-06-11 20:52:41,046: Epoch [14/20], Step [1000/2647], Loss: 0.52125, Disc Loss: 3.71884, Dur Loss: 0.91219, CE Loss: 0.04903, Norm Loss: 0.89837, F0 Loss: 2.68670, LM Loss: 2.11221, Gen Loss: 6.42926, Sty Loss: 0.10042, Diff Loss: 0.26314, DiscLM Loss: 0.00222, GenLM Loss: 0.98527 +INFO:2025-06-11 20:54:29,867: Epoch [14/20], Step [1050/2647], Loss: 0.51678, Disc Loss: 3.66530, Dur Loss: 0.88381, CE Loss: 0.04623, Norm Loss: 1.00183, F0 Loss: 3.62509, LM Loss: 2.07370, Gen Loss: 6.28993, Sty Loss: 0.10163, Diff Loss: 0.28014, DiscLM Loss: 0.00000, GenLM Loss: 0.97782 +INFO:2025-06-11 20:56:17,185: Epoch [14/20], Step [1100/2647], Loss: 0.51679, Disc Loss: 3.73835, Dur Loss: 0.83868, CE Loss: 0.04308, Norm Loss: 0.90095, F0 Loss: 3.40558, LM Loss: 2.17227, Gen Loss: 6.46909, Sty Loss: 0.10810, Diff Loss: 0.22624, DiscLM Loss: 0.00076, GenLM Loss: 0.98808 +INFO:2025-06-11 20:58:05,872: Epoch [14/20], Step [1150/2647], Loss: 0.50829, Disc Loss: 3.63165, Dur Loss: 0.83557, CE Loss: 0.04570, Norm Loss: 0.74760, F0 Loss: 3.23077, LM Loss: 2.12802, Gen Loss: 6.88201, Sty Loss: 0.10448, Diff Loss: 0.20623, DiscLM Loss: 0.00000, GenLM Loss: 0.99358 +INFO:2025-06-11 20:59:54,254: Epoch [14/20], Step [1200/2647], Loss: 0.52423, Disc Loss: 3.71995, Dur Loss: 0.88671, CE Loss: 0.04539, Norm Loss: 0.92892, F0 Loss: 2.78330, LM Loss: 2.18064, Gen Loss: 6.63753, Sty Loss: 0.09515, Diff Loss: 0.21811, DiscLM Loss: 0.00764, GenLM Loss: 0.97782 +INFO:2025-06-11 21:01:45,256: Epoch [14/20], Step [1250/2647], Loss: 0.50500, Disc Loss: 3.69033, Dur Loss: 0.84889, CE Loss: 0.04553, Norm Loss: 0.98940, F0 Loss: 2.88726, LM Loss: 2.12689, Gen Loss: 6.36942, Sty Loss: 0.10953, Diff Loss: 0.26824, DiscLM Loss: 0.00000, GenLM Loss: 1.01402 +INFO:2025-06-11 21:03:30,249: Epoch [14/20], Step [1300/2647], Loss: 0.51752, Disc Loss: 3.58667, Dur Loss: 0.84637, CE Loss: 0.04497, Norm Loss: 0.86654, F0 Loss: 3.22979, LM Loss: 2.20706, Gen Loss: 7.33110, Sty Loss: 0.13021, Diff Loss: 0.37916, DiscLM Loss: 0.00032, GenLM Loss: 0.99001 +INFO:2025-06-11 21:05:18,069: Epoch [14/20], Step [1350/2647], Loss: 0.51174, Disc Loss: 3.74099, Dur Loss: 1.00152, CE Loss: 0.05754, Norm Loss: 1.13583, F0 Loss: 3.05613, LM Loss: 2.14293, Gen Loss: 6.22585, Sty Loss: 0.08210, Diff Loss: 0.26480, DiscLM Loss: 0.00000, GenLM Loss: 0.98466 +INFO:2025-06-11 21:07:04,891: Epoch [14/20], Step [1400/2647], Loss: 0.51902, Disc Loss: 3.63978, Dur Loss: 0.82890, CE Loss: 0.04077, Norm Loss: 0.98479, F0 Loss: 3.59010, LM Loss: 2.10344, Gen Loss: 6.36888, Sty Loss: 0.09927, Diff Loss: 0.25370, DiscLM Loss: 0.00124, GenLM Loss: 1.00002 +INFO:2025-06-11 21:08:52,839: Epoch [14/20], Step [1450/2647], Loss: 0.50779, Disc Loss: 3.78649, Dur Loss: 0.76279, CE Loss: 0.03987, Norm Loss: 0.94612, F0 Loss: 2.68232, LM Loss: 2.09949, Gen Loss: 5.76443, Sty Loss: 0.10779, Diff Loss: 0.21402, DiscLM Loss: 0.00000, GenLM Loss: 1.00637 +INFO:2025-06-11 21:10:43,156: Epoch [14/20], Step [1500/2647], Loss: 0.51375, Disc Loss: 3.72926, Dur Loss: 0.83495, CE Loss: 0.04452, Norm Loss: 0.92087, F0 Loss: 2.91181, LM Loss: 2.14834, Gen Loss: 5.98568, Sty Loss: 0.10756, Diff Loss: 0.30029, DiscLM Loss: 0.00523, GenLM Loss: 0.96260 +INFO:2025-06-11 21:12:35,601: Epoch [14/20], Step [1550/2647], Loss: 0.52242, Disc Loss: 3.75456, Dur Loss: 0.90317, CE Loss: 0.04954, Norm Loss: 1.22716, F0 Loss: 3.34084, LM Loss: 2.22583, Gen Loss: 6.34058, Sty Loss: 0.12716, Diff Loss: 0.25147, DiscLM Loss: 0.00000, GenLM Loss: 1.01205 +INFO:2025-06-11 21:14:25,064: Epoch [14/20], Step [1600/2647], Loss: 0.51585, Disc Loss: 3.79369, Dur Loss: 1.06509, CE Loss: 0.05692, Norm Loss: 1.42551, F0 Loss: 3.66480, LM Loss: 2.19639, Gen Loss: 6.58044, Sty Loss: 0.09545, Diff Loss: 0.28039, DiscLM Loss: 0.00295, GenLM Loss: 1.00925 +INFO:2025-06-11 21:16:14,610: Epoch [14/20], Step [1650/2647], Loss: 0.51498, Disc Loss: 3.76772, Dur Loss: 1.01136, CE Loss: 0.05271, Norm Loss: 1.07427, F0 Loss: 2.43026, LM Loss: 2.09239, Gen Loss: 6.05201, Sty Loss: 0.09386, Diff Loss: 0.23971, DiscLM Loss: 0.00000, GenLM Loss: 0.97352 +INFO:2025-06-11 21:18:01,758: Epoch [14/20], Step [1700/2647], Loss: 0.52074, Disc Loss: 3.75412, Dur Loss: 0.88595, CE Loss: 0.04567, Norm Loss: 1.12246, F0 Loss: 3.13469, LM Loss: 2.12819, Gen Loss: 5.96668, Sty Loss: 0.11489, Diff Loss: 0.21738, DiscLM Loss: 0.00336, GenLM Loss: 0.98523 +INFO:2025-06-11 21:19:51,146: Epoch [14/20], Step [1750/2647], Loss: 0.51581, Disc Loss: 3.74151, Dur Loss: 0.83047, CE Loss: 0.04634, Norm Loss: 1.08705, F0 Loss: 3.24647, LM Loss: 2.17158, Gen Loss: 6.38366, Sty Loss: 0.12113, Diff Loss: 0.33755, DiscLM Loss: 0.00000, GenLM Loss: 1.00832 +INFO:2025-06-11 21:21:39,984: Epoch [14/20], Step [1800/2647], Loss: 0.51319, Disc Loss: 3.87173, Dur Loss: 0.93543, CE Loss: 0.04887, Norm Loss: 0.97254, F0 Loss: 3.22937, LM Loss: 2.18451, Gen Loss: 6.42584, Sty Loss: 0.14584, Diff Loss: 0.20122, DiscLM Loss: 0.00104, GenLM Loss: 1.00091 +INFO:2025-06-11 21:23:29,664: Epoch [14/20], Step [1850/2647], Loss: 0.51235, Disc Loss: 3.74567, Dur Loss: 0.96903, CE Loss: 0.05202, Norm Loss: 0.88879, F0 Loss: 2.65085, LM Loss: 2.05820, Gen Loss: 6.43961, Sty Loss: 0.09522, Diff Loss: 0.27636, DiscLM Loss: 0.00000, GenLM Loss: 0.99542 +INFO:2025-06-11 21:25:18,689: Epoch [14/20], Step [1900/2647], Loss: 0.50245, Disc Loss: 3.82295, Dur Loss: 0.88872, CE Loss: 0.04980, Norm Loss: 1.32183, F0 Loss: 2.92732, LM Loss: 2.13966, Gen Loss: 5.73975, Sty Loss: 0.11221, Diff Loss: 0.24389, DiscLM Loss: 0.00202, GenLM Loss: 1.00333 +INFO:2025-06-11 21:27:07,225: Epoch [14/20], Step [1950/2647], Loss: 0.51491, Disc Loss: 3.86929, Dur Loss: 0.84381, CE Loss: 0.04607, Norm Loss: 0.72070, F0 Loss: 3.09935, LM Loss: 2.16846, Gen Loss: 6.25058, Sty Loss: 0.11134, Diff Loss: 0.26617, DiscLM Loss: 0.00000, GenLM Loss: 0.99722 +INFO:2025-06-11 21:28:56,651: Epoch [14/20], Step [2000/2647], Loss: 0.51799, Disc Loss: 3.79313, Dur Loss: 1.32773, CE Loss: 0.06947, Norm Loss: 1.03590, F0 Loss: 3.33769, LM Loss: 2.05071, Gen Loss: 6.06810, Sty Loss: 0.11797, Diff Loss: 0.31985, DiscLM Loss: 0.00123, GenLM Loss: 1.00212 +INFO:2025-06-11 21:30:46,524: Epoch [14/20], Step [2050/2647], Loss: 0.51320, Disc Loss: 3.78102, Dur Loss: 0.73287, CE Loss: 0.03574, Norm Loss: 0.97383, F0 Loss: 2.76546, LM Loss: 2.14995, Gen Loss: 5.74116, Sty Loss: 0.11300, Diff Loss: 0.26219, DiscLM Loss: 0.00000, GenLM Loss: 0.99775 +INFO:2025-06-11 21:32:36,447: Epoch [14/20], Step [2100/2647], Loss: 0.52072, Disc Loss: 3.75039, Dur Loss: 0.83566, CE Loss: 0.04450, Norm Loss: 1.02623, F0 Loss: 3.09736, LM Loss: 2.10973, Gen Loss: 5.78009, Sty Loss: 0.11590, Diff Loss: 0.22921, DiscLM Loss: 0.00054, GenLM Loss: 1.00009 +INFO:2025-06-11 21:34:24,841: Epoch [14/20], Step [2150/2647], Loss: 0.51103, Disc Loss: 3.76632, Dur Loss: 0.74288, CE Loss: 0.03684, Norm Loss: 0.84857, F0 Loss: 2.96751, LM Loss: 2.16053, Gen Loss: 6.04558, Sty Loss: 0.11036, Diff Loss: 0.31290, DiscLM Loss: 0.00000, GenLM Loss: 1.00539 +INFO:2025-06-11 21:36:14,569: Epoch [14/20], Step [2200/2647], Loss: 0.51149, Disc Loss: 3.83818, Dur Loss: 0.74698, CE Loss: 0.03881, Norm Loss: 1.11262, F0 Loss: 3.24856, LM Loss: 2.19618, Gen Loss: 6.13497, Sty Loss: 0.09488, Diff Loss: 0.30690, DiscLM Loss: 0.00031, GenLM Loss: 1.00386 +INFO:2025-06-11 21:38:03,669: Epoch [14/20], Step [2250/2647], Loss: 0.51964, Disc Loss: 3.81449, Dur Loss: 0.82597, CE Loss: 0.04382, Norm Loss: 0.85225, F0 Loss: 2.96399, LM Loss: 2.14922, Gen Loss: 5.85931, Sty Loss: 0.11181, Diff Loss: 0.32276, DiscLM Loss: 0.00000, GenLM Loss: 0.99106 +INFO:2025-06-11 21:39:53,056: Epoch [14/20], Step [2300/2647], Loss: 0.50787, Disc Loss: 3.84439, Dur Loss: 1.00225, CE Loss: 0.05141, Norm Loss: 1.02434, F0 Loss: 2.74120, LM Loss: 2.14676, Gen Loss: 5.82271, Sty Loss: 0.08733, Diff Loss: 0.28571, DiscLM Loss: 0.00109, GenLM Loss: 0.98839 +INFO:2025-06-11 21:41:42,910: Epoch [14/20], Step [2350/2647], Loss: 0.50781, Disc Loss: 3.72657, Dur Loss: 0.92866, CE Loss: 0.04942, Norm Loss: 0.90175, F0 Loss: 4.11274, LM Loss: 2.13383, Gen Loss: 6.23394, Sty Loss: 0.11534, Diff Loss: 0.36582, DiscLM Loss: 0.00000, GenLM Loss: 0.99965 +INFO:2025-06-11 21:43:32,558: Epoch [14/20], Step [2400/2647], Loss: 0.52302, Disc Loss: 3.65844, Dur Loss: 1.16036, CE Loss: 0.06469, Norm Loss: 1.32683, F0 Loss: 3.20925, LM Loss: 2.19894, Gen Loss: 6.70688, Sty Loss: 0.12019, Diff Loss: 0.26801, DiscLM Loss: 0.00123, GenLM Loss: 0.99874 +INFO:2025-06-11 21:45:18,505: Epoch [14/20], Step [2450/2647], Loss: 0.52715, Disc Loss: 3.82931, Dur Loss: 0.90491, CE Loss: 0.04583, Norm Loss: 1.38949, F0 Loss: 3.34354, LM Loss: 2.25989, Gen Loss: 7.17880, Sty Loss: 0.12611, Diff Loss: 0.22003, DiscLM Loss: 0.00000, GenLM Loss: 1.01314 +INFO:2025-06-11 21:47:06,657: Epoch [14/20], Step [2500/2647], Loss: 0.52077, Disc Loss: 3.63923, Dur Loss: 1.11637, CE Loss: 0.05155, Norm Loss: 1.06464, F0 Loss: 3.57688, LM Loss: 2.26264, Gen Loss: 7.66099, Sty Loss: 0.10834, Diff Loss: 0.23423, DiscLM Loss: 0.00386, GenLM Loss: 0.95115 +INFO:2025-06-11 21:48:56,385: Epoch [14/20], Step [2550/2647], Loss: 0.51354, Disc Loss: 3.75394, Dur Loss: 0.66479, CE Loss: 0.03320, Norm Loss: 0.80245, F0 Loss: 2.72586, LM Loss: 2.16670, Gen Loss: 7.57326, Sty Loss: 0.12054, Diff Loss: 0.31621, DiscLM Loss: 0.00000, GenLM Loss: 0.97487 +INFO:2025-06-11 21:50:46,675: Epoch [14/20], Step [2600/2647], Loss: 0.50921, Disc Loss: 3.60249, Dur Loss: 0.95252, CE Loss: 0.04999, Norm Loss: 1.06507, F0 Loss: 2.65683, LM Loss: 2.15307, Gen Loss: 6.83682, Sty Loss: 0.12089, Diff Loss: 0.35288, DiscLM Loss: 0.00122, GenLM Loss: 1.04617 +INFO:2025-06-11 21:54:52,521: Validation loss: 0.547, Dur loss: 0.978, F0 loss: 3.250 + + + +INFO:2025-06-11 21:56:42,935: Epoch [15/20], Step [50/2647], Loss: 0.50732, Disc Loss: 3.69443, Dur Loss: 0.82750, CE Loss: 0.04623, Norm Loss: 1.05447, F0 Loss: 3.02611, LM Loss: 2.17656, Gen Loss: 6.38031, Sty Loss: 0.11796, Diff Loss: 0.20545, DiscLM Loss: 0.00000, GenLM Loss: 0.96015 +INFO:2025-06-11 21:58:34,043: Epoch [15/20], Step [100/2647], Loss: 0.51512, Disc Loss: 3.69942, Dur Loss: 0.86562, CE Loss: 0.04680, Norm Loss: 1.00187, F0 Loss: 2.29503, LM Loss: 2.07604, Gen Loss: 6.40987, Sty Loss: 0.09941, Diff Loss: 0.33107, DiscLM Loss: 0.00277, GenLM Loss: 0.93065 +INFO:2025-06-11 22:00:27,761: Epoch [15/20], Step [150/2647], Loss: 0.50778, Disc Loss: 3.67064, Dur Loss: 0.95605, CE Loss: 0.05091, Norm Loss: 1.18966, F0 Loss: 2.87127, LM Loss: 2.09357, Gen Loss: 6.33457, Sty Loss: 0.11963, Diff Loss: 0.35118, DiscLM Loss: 0.00000, GenLM Loss: 1.03267 +INFO:2025-06-11 22:02:18,462: Epoch [15/20], Step [200/2647], Loss: 0.51569, Disc Loss: 3.81682, Dur Loss: 0.98515, CE Loss: 0.05250, Norm Loss: 0.94314, F0 Loss: 3.33912, LM Loss: 2.15745, Gen Loss: 5.99765, Sty Loss: 0.12406, Diff Loss: 0.21164, DiscLM Loss: 0.00933, GenLM Loss: 1.04946 +INFO:2025-06-11 22:04:07,840: Epoch [15/20], Step [250/2647], Loss: 0.50705, Disc Loss: 3.82794, Dur Loss: 0.81117, CE Loss: 0.04085, Norm Loss: 1.06769, F0 Loss: 3.25738, LM Loss: 2.18560, Gen Loss: 5.94886, Sty Loss: 0.10526, Diff Loss: 0.28327, DiscLM Loss: 0.00000, GenLM Loss: 0.96471 +INFO:2025-06-11 22:05:57,664: Epoch [15/20], Step [300/2647], Loss: 0.50818, Disc Loss: 3.74520, Dur Loss: 1.02819, CE Loss: 0.05383, Norm Loss: 1.04314, F0 Loss: 2.81660, LM Loss: 2.16979, Gen Loss: 6.86461, Sty Loss: 0.12225, Diff Loss: 0.30150, DiscLM Loss: 0.00216, GenLM Loss: 0.93221 +INFO:2025-06-11 22:07:47,969: Epoch [15/20], Step [350/2647], Loss: 0.50981, Disc Loss: 3.73970, Dur Loss: 0.81079, CE Loss: 0.04315, Norm Loss: 0.95637, F0 Loss: 2.62625, LM Loss: 1.99294, Gen Loss: 6.31139, Sty Loss: 0.10807, Diff Loss: 0.25323, DiscLM Loss: 0.00000, GenLM Loss: 1.03002 +INFO:2025-06-11 22:09:40,635: Epoch [15/20], Step [400/2647], Loss: 0.51682, Disc Loss: 3.80967, Dur Loss: 1.12146, CE Loss: 0.06372, Norm Loss: 1.04636, F0 Loss: 2.24261, LM Loss: 2.11927, Gen Loss: 5.57585, Sty Loss: 0.11635, Diff Loss: 0.38738, DiscLM Loss: 0.01720, GenLM Loss: 1.04875 +INFO:2025-06-11 22:11:29,778: Epoch [15/20], Step [450/2647], Loss: 0.51989, Disc Loss: 3.77170, Dur Loss: 1.01092, CE Loss: 0.06103, Norm Loss: 0.95188, F0 Loss: 2.35078, LM Loss: 2.14425, Gen Loss: 6.22001, Sty Loss: 0.09852, Diff Loss: 0.27369, DiscLM Loss: 0.00000, GenLM Loss: 0.98378 +INFO:2025-06-11 22:13:18,451: Epoch [15/20], Step [500/2647], Loss: 0.52583, Disc Loss: 3.72110, Dur Loss: 0.85431, CE Loss: 0.04912, Norm Loss: 0.83613, F0 Loss: 2.58356, LM Loss: 2.15374, Gen Loss: 5.80706, Sty Loss: 0.10334, Diff Loss: 0.27930, DiscLM Loss: 0.02536, GenLM Loss: 0.90252 +INFO:2025-06-11 22:15:08,233: Epoch [15/20], Step [550/2647], Loss: 0.50238, Disc Loss: 3.94745, Dur Loss: 1.03715, CE Loss: 0.05378, Norm Loss: 1.05765, F0 Loss: 2.77876, LM Loss: 2.03419, Gen Loss: 6.20372, Sty Loss: 0.11983, Diff Loss: 0.25407, DiscLM Loss: 0.00000, GenLM Loss: 0.92355 +INFO:2025-06-11 22:16:58,146: Epoch [15/20], Step [600/2647], Loss: 0.51822, Disc Loss: 3.75979, Dur Loss: 1.00030, CE Loss: 0.05846, Norm Loss: 1.13234, F0 Loss: 3.88689, LM Loss: 2.26777, Gen Loss: 6.21262, Sty Loss: 0.11499, Diff Loss: 0.31045, DiscLM Loss: 0.01023, GenLM Loss: 1.01187 +INFO:2025-06-11 22:18:46,028: Epoch [15/20], Step [650/2647], Loss: 0.51963, Disc Loss: 3.67961, Dur Loss: 0.76297, CE Loss: 0.04214, Norm Loss: 0.99269, F0 Loss: 3.09099, LM Loss: 2.08331, Gen Loss: 6.74607, Sty Loss: 0.10369, Diff Loss: 0.21658, DiscLM Loss: 0.00000, GenLM Loss: 0.97468 +INFO:2025-06-11 22:20:35,144: Epoch [15/20], Step [700/2647], Loss: 0.51098, Disc Loss: 3.72524, Dur Loss: 0.90492, CE Loss: 0.05118, Norm Loss: 0.90452, F0 Loss: 2.26885, LM Loss: 2.14045, Gen Loss: 6.33561, Sty Loss: 0.10795, Diff Loss: 0.25721, DiscLM Loss: 0.00220, GenLM Loss: 0.98339 +INFO:2025-06-11 22:22:23,357: Epoch [15/20], Step [750/2647], Loss: 0.51190, Disc Loss: 3.81701, Dur Loss: 0.85968, CE Loss: 0.04569, Norm Loss: 0.76955, F0 Loss: 2.77818, LM Loss: 2.06976, Gen Loss: 6.38258, Sty Loss: 0.10728, Diff Loss: 0.21797, DiscLM Loss: 0.00000, GenLM Loss: 0.96355 +INFO:2025-06-11 22:24:12,592: Epoch [15/20], Step [800/2647], Loss: 0.52078, Disc Loss: 3.76002, Dur Loss: 0.99338, CE Loss: 0.05847, Norm Loss: 0.95975, F0 Loss: 2.51320, LM Loss: 2.12469, Gen Loss: 6.64015, Sty Loss: 0.12044, Diff Loss: 0.29905, DiscLM Loss: 0.02868, GenLM Loss: 0.97701 +INFO:2025-06-11 22:26:00,657: Epoch [15/20], Step [850/2647], Loss: 0.52083, Disc Loss: 3.78920, Dur Loss: 0.99975, CE Loss: 0.05659, Norm Loss: 1.15995, F0 Loss: 2.95273, LM Loss: 2.10967, Gen Loss: 5.61784, Sty Loss: 0.10890, Diff Loss: 0.26005, DiscLM Loss: 0.00000, GenLM Loss: 0.99136 +INFO:2025-06-11 22:27:49,666: Epoch [15/20], Step [900/2647], Loss: 0.53645, Disc Loss: 3.72965, Dur Loss: 0.93373, CE Loss: 0.04992, Norm Loss: 0.93228, F0 Loss: 2.89620, LM Loss: 2.16388, Gen Loss: 6.21336, Sty Loss: 0.13892, Diff Loss: 0.32966, DiscLM Loss: 0.00471, GenLM Loss: 0.99286 +INFO:2025-06-11 22:29:40,485: Epoch [15/20], Step [950/2647], Loss: 0.50463, Disc Loss: 3.73463, Dur Loss: 0.94777, CE Loss: 0.04653, Norm Loss: 0.92433, F0 Loss: 2.84822, LM Loss: 2.07343, Gen Loss: 5.81314, Sty Loss: 0.10236, Diff Loss: 0.26785, DiscLM Loss: 0.00000, GenLM Loss: 1.00113 +INFO:2025-06-11 22:31:28,538: Epoch [15/20], Step [1000/2647], Loss: 0.50894, Disc Loss: 3.70814, Dur Loss: 1.03408, CE Loss: 0.05758, Norm Loss: 0.92813, F0 Loss: 3.34438, LM Loss: 2.23138, Gen Loss: 7.06717, Sty Loss: 0.11447, Diff Loss: 0.17811, DiscLM Loss: 0.00155, GenLM Loss: 1.00691 +INFO:2025-06-11 22:33:16,965: Epoch [15/20], Step [1050/2647], Loss: 0.51257, Disc Loss: 3.73457, Dur Loss: 0.93818, CE Loss: 0.04813, Norm Loss: 0.85865, F0 Loss: 2.08767, LM Loss: 2.07752, Gen Loss: 5.99153, Sty Loss: 0.12720, Diff Loss: 0.35321, DiscLM Loss: 0.00000, GenLM Loss: 0.99835 +INFO:2025-06-11 22:35:07,937: Epoch [15/20], Step [1100/2647], Loss: 0.51484, Disc Loss: 3.73427, Dur Loss: 1.15444, CE Loss: 0.06519, Norm Loss: 0.98791, F0 Loss: 3.28013, LM Loss: 2.21350, Gen Loss: 6.19035, Sty Loss: 0.11559, Diff Loss: 0.22329, DiscLM Loss: 0.00789, GenLM Loss: 0.99477 +INFO:2025-06-11 22:36:58,198: Epoch [15/20], Step [1150/2647], Loss: 0.52536, Disc Loss: 3.76409, Dur Loss: 0.84768, CE Loss: 0.04341, Norm Loss: 0.93204, F0 Loss: 2.54490, LM Loss: 2.11076, Gen Loss: 6.01041, Sty Loss: 0.11546, Diff Loss: 0.26119, DiscLM Loss: 0.00000, GenLM Loss: 0.99366 +INFO:2025-06-11 22:38:49,328: Epoch [15/20], Step [1200/2647], Loss: 0.51547, Disc Loss: 3.61057, Dur Loss: 1.02747, CE Loss: 0.05603, Norm Loss: 1.17235, F0 Loss: 3.01095, LM Loss: 2.21935, Gen Loss: 7.20914, Sty Loss: 0.16769, Diff Loss: 0.35321, DiscLM Loss: 0.00668, GenLM Loss: 1.00318 +INFO:2025-06-11 22:40:37,232: Epoch [15/20], Step [1250/2647], Loss: 0.53040, Disc Loss: 3.69022, Dur Loss: 0.80743, CE Loss: 0.04883, Norm Loss: 0.93052, F0 Loss: 3.12101, LM Loss: 2.27555, Gen Loss: 6.42598, Sty Loss: 0.10519, Diff Loss: 0.41401, DiscLM Loss: 0.00000, GenLM Loss: 0.99544 +INFO:2025-06-11 22:42:26,663: Epoch [15/20], Step [1300/2647], Loss: 0.52229, Disc Loss: 3.81855, Dur Loss: 0.98717, CE Loss: 0.05096, Norm Loss: 0.86515, F0 Loss: 2.64012, LM Loss: 2.07211, Gen Loss: 6.12249, Sty Loss: 0.11880, Diff Loss: 0.33896, DiscLM Loss: 0.00350, GenLM Loss: 0.99777 +INFO:2025-06-11 22:44:16,303: Epoch [15/20], Step [1350/2647], Loss: 0.51314, Disc Loss: 3.74443, Dur Loss: 0.77881, CE Loss: 0.04081, Norm Loss: 0.83831, F0 Loss: 2.28386, LM Loss: 2.13947, Gen Loss: 6.16208, Sty Loss: 0.10126, Diff Loss: 0.23604, DiscLM Loss: 0.00000, GenLM Loss: 0.99890 +INFO:2025-06-11 22:46:04,222: Epoch [15/20], Step [1400/2647], Loss: 0.51299, Disc Loss: 3.74570, Dur Loss: 1.03734, CE Loss: 0.05855, Norm Loss: 0.78877, F0 Loss: 2.95834, LM Loss: 2.01377, Gen Loss: 6.48377, Sty Loss: 0.10062, Diff Loss: 0.29218, DiscLM Loss: 0.00356, GenLM Loss: 1.00550 +INFO:2025-06-11 22:47:52,663: Epoch [15/20], Step [1450/2647], Loss: 0.52465, Disc Loss: 3.76263, Dur Loss: 0.94031, CE Loss: 0.04954, Norm Loss: 1.30426, F0 Loss: 3.19372, LM Loss: 2.28592, Gen Loss: 6.14236, Sty Loss: 0.11436, Diff Loss: 0.28435, DiscLM Loss: 0.00000, GenLM Loss: 1.02400 +INFO:2025-06-11 22:49:42,809: Epoch [15/20], Step [1500/2647], Loss: 0.51728, Disc Loss: 3.63668, Dur Loss: 0.77838, CE Loss: 0.04146, Norm Loss: 0.79909, F0 Loss: 2.14939, LM Loss: 2.18429, Gen Loss: 6.50546, Sty Loss: 0.09063, Diff Loss: 0.25883, DiscLM Loss: 0.00288, GenLM Loss: 0.98510 +INFO:2025-06-11 22:51:31,750: Epoch [15/20], Step [1550/2647], Loss: 0.51259, Disc Loss: 3.68550, Dur Loss: 0.85157, CE Loss: 0.04898, Norm Loss: 1.07874, F0 Loss: 3.50362, LM Loss: 2.11026, Gen Loss: 6.75183, Sty Loss: 0.11458, Diff Loss: 0.18409, DiscLM Loss: 0.00000, GenLM Loss: 0.98995 +INFO:2025-06-11 22:53:19,067: Epoch [15/20], Step [1600/2647], Loss: 0.51379, Disc Loss: 3.72022, Dur Loss: 0.93615, CE Loss: 0.05096, Norm Loss: 0.98368, F0 Loss: 2.86528, LM Loss: 2.10392, Gen Loss: 6.21502, Sty Loss: 0.08830, Diff Loss: 0.28510, DiscLM Loss: 0.00355, GenLM Loss: 1.00847 +INFO:2025-06-11 22:55:07,864: Epoch [15/20], Step [1650/2647], Loss: 0.51840, Disc Loss: 3.69036, Dur Loss: 0.95480, CE Loss: 0.05596, Norm Loss: 1.15595, F0 Loss: 3.29952, LM Loss: 2.13109, Gen Loss: 6.55990, Sty Loss: 0.09936, Diff Loss: 0.31636, DiscLM Loss: 0.00000, GenLM Loss: 1.00466 +INFO:2025-06-11 22:56:53,323: Epoch [15/20], Step [1700/2647], Loss: 0.51854, Disc Loss: 3.75941, Dur Loss: 1.10620, CE Loss: 0.05401, Norm Loss: 1.01156, F0 Loss: 2.58701, LM Loss: 2.12373, Gen Loss: 6.34565, Sty Loss: 0.15244, Diff Loss: 0.44559, DiscLM Loss: 0.00341, GenLM Loss: 0.99804 +INFO:2025-06-11 22:58:45,383: Epoch [15/20], Step [1750/2647], Loss: 0.51620, Disc Loss: 3.70445, Dur Loss: 0.88877, CE Loss: 0.04478, Norm Loss: 0.94147, F0 Loss: 3.10541, LM Loss: 2.11598, Gen Loss: 6.85780, Sty Loss: 0.10267, Diff Loss: 0.32174, DiscLM Loss: 0.00000, GenLM Loss: 0.97679 +INFO:2025-06-11 23:00:35,743: Epoch [15/20], Step [1800/2647], Loss: 0.51051, Disc Loss: 3.67952, Dur Loss: 0.99206, CE Loss: 0.05692, Norm Loss: 0.82692, F0 Loss: 2.08994, LM Loss: 2.16078, Gen Loss: 6.70891, Sty Loss: 0.12992, Diff Loss: 0.30199, DiscLM Loss: 0.00061, GenLM Loss: 1.00383 +INFO:2025-06-11 23:02:27,983: Epoch [15/20], Step [1850/2647], Loss: 0.50621, Disc Loss: 3.78316, Dur Loss: 0.89265, CE Loss: 0.04973, Norm Loss: 1.05738, F0 Loss: 2.46336, LM Loss: 2.03652, Gen Loss: 6.10181, Sty Loss: 0.10602, Diff Loss: 0.29460, DiscLM Loss: 0.00000, GenLM Loss: 1.02059 +INFO:2025-06-11 23:04:18,817: Epoch [15/20], Step [1900/2647], Loss: 0.51578, Disc Loss: 3.68470, Dur Loss: 0.74561, CE Loss: 0.03837, Norm Loss: 1.07043, F0 Loss: 3.18597, LM Loss: 2.12449, Gen Loss: 6.55417, Sty Loss: 0.08785, Diff Loss: 0.22593, DiscLM Loss: 0.00053, GenLM Loss: 1.01713 +INFO:2025-06-11 23:06:09,204: Epoch [15/20], Step [1950/2647], Loss: 0.51398, Disc Loss: 3.83575, Dur Loss: 0.92386, CE Loss: 0.04871, Norm Loss: 0.82167, F0 Loss: 2.67501, LM Loss: 2.07410, Gen Loss: 6.97064, Sty Loss: 0.11899, Diff Loss: 0.28514, DiscLM Loss: 0.00000, GenLM Loss: 0.98550 +INFO:2025-06-11 23:07:58,172: Epoch [15/20], Step [2000/2647], Loss: 0.51131, Disc Loss: 3.72016, Dur Loss: 0.72770, CE Loss: 0.03905, Norm Loss: 0.95482, F0 Loss: 2.55742, LM Loss: 2.21747, Gen Loss: 6.36721, Sty Loss: 0.08957, Diff Loss: 0.22599, DiscLM Loss: 0.00680, GenLM Loss: 0.98880 +INFO:2025-06-11 23:09:47,822: Epoch [15/20], Step [2050/2647], Loss: 0.50860, Disc Loss: 3.74296, Dur Loss: 1.01084, CE Loss: 0.05482, Norm Loss: 1.20267, F0 Loss: 2.94848, LM Loss: 2.12823, Gen Loss: 6.02566, Sty Loss: 0.09778, Diff Loss: 0.32657, DiscLM Loss: 0.00000, GenLM Loss: 0.98914 +INFO:2025-06-11 23:11:36,643: Epoch [15/20], Step [2100/2647], Loss: 0.50610, Disc Loss: 3.72164, Dur Loss: 0.72043, CE Loss: 0.04433, Norm Loss: 1.45281, F0 Loss: 2.67837, LM Loss: 2.26544, Gen Loss: 7.03879, Sty Loss: 0.10436, Diff Loss: 0.24923, DiscLM Loss: 0.00088, GenLM Loss: 1.00579 +INFO:2025-06-11 23:13:25,453: Epoch [15/20], Step [2150/2647], Loss: 0.51609, Disc Loss: 3.68899, Dur Loss: 0.97833, CE Loss: 0.05976, Norm Loss: 0.96535, F0 Loss: 2.90374, LM Loss: 2.19521, Gen Loss: 6.90773, Sty Loss: 0.16644, Diff Loss: 0.33445, DiscLM Loss: 0.00000, GenLM Loss: 0.99153 +INFO:2025-06-11 23:15:17,568: Epoch [15/20], Step [2200/2647], Loss: 0.50979, Disc Loss: 3.69873, Dur Loss: 0.80827, CE Loss: 0.04406, Norm Loss: 1.09671, F0 Loss: 3.71666, LM Loss: 2.24132, Gen Loss: 6.63245, Sty Loss: 0.20112, Diff Loss: 0.49409, DiscLM Loss: 0.00301, GenLM Loss: 0.97855 +INFO:2025-06-11 23:17:04,559: Epoch [15/20], Step [2250/2647], Loss: 0.51340, Disc Loss: 3.72232, Dur Loss: 0.91063, CE Loss: 0.05462, Norm Loss: 0.96330, F0 Loss: 2.89004, LM Loss: 2.19744, Gen Loss: 6.55503, Sty Loss: 0.11019, Diff Loss: 0.39221, DiscLM Loss: 0.00000, GenLM Loss: 1.00840 +INFO:2025-06-11 23:18:52,945: Epoch [15/20], Step [2300/2647], Loss: 0.52000, Disc Loss: 3.68224, Dur Loss: 1.13075, CE Loss: 0.05206, Norm Loss: 1.00214, F0 Loss: 3.20208, LM Loss: 2.09630, Gen Loss: 6.68471, Sty Loss: 0.12451, Diff Loss: 0.23353, DiscLM Loss: 0.00087, GenLM Loss: 1.00923 +INFO:2025-06-11 23:20:39,575: Epoch [15/20], Step [2350/2647], Loss: 0.52296, Disc Loss: 3.76245, Dur Loss: 0.69324, CE Loss: 0.03329, Norm Loss: 0.77410, F0 Loss: 2.15943, LM Loss: 2.16160, Gen Loss: 6.80563, Sty Loss: 0.12968, Diff Loss: 0.33579, DiscLM Loss: 0.00000, GenLM Loss: 0.98741 +INFO:2025-06-11 23:22:30,476: Epoch [15/20], Step [2400/2647], Loss: 0.51344, Disc Loss: 3.71245, Dur Loss: 0.83261, CE Loss: 0.04413, Norm Loss: 0.89283, F0 Loss: 3.69722, LM Loss: 2.16513, Gen Loss: 6.20280, Sty Loss: 0.10636, Diff Loss: 0.34062, DiscLM Loss: 0.00058, GenLM Loss: 0.98987 +INFO:2025-06-11 23:24:17,903: Epoch [15/20], Step [2450/2647], Loss: 0.50704, Disc Loss: 3.66388, Dur Loss: 0.81251, CE Loss: 0.04321, Norm Loss: 0.91054, F0 Loss: 2.87979, LM Loss: 2.07839, Gen Loss: 6.35890, Sty Loss: 0.10483, Diff Loss: 0.27021, DiscLM Loss: 0.00000, GenLM Loss: 0.99620 +INFO:2025-06-11 23:26:06,383: Epoch [15/20], Step [2500/2647], Loss: 0.51257, Disc Loss: 3.79374, Dur Loss: 0.88805, CE Loss: 0.05108, Norm Loss: 0.96625, F0 Loss: 2.65359, LM Loss: 2.19942, Gen Loss: 6.92774, Sty Loss: 0.11569, Diff Loss: 0.23056, DiscLM Loss: 0.00060, GenLM Loss: 1.00026 +INFO:2025-06-11 23:27:56,535: Epoch [15/20], Step [2550/2647], Loss: 0.51606, Disc Loss: 3.78149, Dur Loss: 0.81217, CE Loss: 0.04550, Norm Loss: 0.98993, F0 Loss: 3.33384, LM Loss: 2.21757, Gen Loss: 6.57991, Sty Loss: 0.11876, Diff Loss: 0.23454, DiscLM Loss: 0.00000, GenLM Loss: 0.99519 +INFO:2025-06-11 23:29:44,783: Epoch [15/20], Step [2600/2647], Loss: 0.51697, Disc Loss: 3.69833, Dur Loss: 0.98687, CE Loss: 0.05339, Norm Loss: 0.89127, F0 Loss: 2.62820, LM Loss: 2.14018, Gen Loss: 5.83212, Sty Loss: 0.08451, Diff Loss: 0.24904, DiscLM Loss: 0.00059, GenLM Loss: 0.99465 +INFO:2025-06-11 23:33:51,298: Validation loss: 0.534, Dur loss: 1.006, F0 loss: 3.312 + + + +INFO:2025-06-11 23:35:43,043: Epoch [16/20], Step [50/2647], Loss: 0.51493, Disc Loss: 3.78993, Dur Loss: 1.12563, CE Loss: 0.06519, Norm Loss: 1.24103, F0 Loss: 3.43504, LM Loss: 2.19602, Gen Loss: 6.09565, Sty Loss: 0.09338, Diff Loss: 0.22853, DiscLM Loss: 0.00000, GenLM Loss: 1.00360 +INFO:2025-06-11 23:37:31,683: Epoch [16/20], Step [100/2647], Loss: 0.50592, Disc Loss: 3.77143, Dur Loss: 0.87443, CE Loss: 0.04626, Norm Loss: 0.83493, F0 Loss: 2.51332, LM Loss: 2.12126, Gen Loss: 6.01977, Sty Loss: 0.09337, Diff Loss: 0.24945, DiscLM Loss: 0.00029, GenLM Loss: 1.00203 +INFO:2025-06-11 23:39:21,770: Epoch [16/20], Step [150/2647], Loss: 0.51417, Disc Loss: 3.84841, Dur Loss: 0.98666, CE Loss: 0.05384, Norm Loss: 0.90344, F0 Loss: 2.44135, LM Loss: 2.05683, Gen Loss: 5.68590, Sty Loss: 0.06965, Diff Loss: 0.20490, DiscLM Loss: 0.00000, GenLM Loss: 0.97829 +INFO:2025-06-11 23:41:11,987: Epoch [16/20], Step [200/2647], Loss: 0.50848, Disc Loss: 3.69877, Dur Loss: 0.82246, CE Loss: 0.04654, Norm Loss: 1.02928, F0 Loss: 2.61281, LM Loss: 2.30025, Gen Loss: 6.56038, Sty Loss: 0.10535, Diff Loss: 0.22275, DiscLM Loss: 0.01699, GenLM Loss: 0.98387 +INFO:2025-06-11 23:43:03,265: Epoch [16/20], Step [250/2647], Loss: 0.51941, Disc Loss: 3.71046, Dur Loss: 0.78457, CE Loss: 0.04195, Norm Loss: 1.00404, F0 Loss: 2.95330, LM Loss: 2.23537, Gen Loss: 6.95319, Sty Loss: 0.10584, Diff Loss: 0.28728, DiscLM Loss: 0.00000, GenLM Loss: 1.01688 +INFO:2025-06-11 23:44:51,747: Epoch [16/20], Step [300/2647], Loss: 0.51871, Disc Loss: 3.73748, Dur Loss: 0.94766, CE Loss: 0.04929, Norm Loss: 1.13272, F0 Loss: 2.86504, LM Loss: 2.12163, Gen Loss: 6.72123, Sty Loss: 0.11864, Diff Loss: 0.26201, DiscLM Loss: 0.00307, GenLM Loss: 0.97431 +INFO:2025-06-11 23:46:39,375: Epoch [16/20], Step [350/2647], Loss: 0.51043, Disc Loss: 3.70593, Dur Loss: 0.98580, CE Loss: 0.04971, Norm Loss: 0.93896, F0 Loss: 2.87917, LM Loss: 2.10239, Gen Loss: 6.08899, Sty Loss: 0.08939, Diff Loss: 0.23293, DiscLM Loss: 0.00000, GenLM Loss: 0.99246 +INFO:2025-06-11 23:48:27,739: Epoch [16/20], Step [400/2647], Loss: 0.51217, Disc Loss: 3.73468, Dur Loss: 0.94486, CE Loss: 0.04926, Norm Loss: 1.19080, F0 Loss: 3.25731, LM Loss: 2.04675, Gen Loss: 6.30979, Sty Loss: 0.09022, Diff Loss: 0.23077, DiscLM Loss: 0.00954, GenLM Loss: 1.00458 +INFO:2025-06-11 23:50:15,476: Epoch [16/20], Step [450/2647], Loss: 0.51134, Disc Loss: 3.76374, Dur Loss: 0.92554, CE Loss: 0.05319, Norm Loss: 1.00345, F0 Loss: 3.01367, LM Loss: 2.16089, Gen Loss: 5.87786, Sty Loss: 0.09791, Diff Loss: 0.22209, DiscLM Loss: 0.00000, GenLM Loss: 0.99649 +INFO:2025-06-11 23:52:05,248: Epoch [16/20], Step [500/2647], Loss: 0.50175, Disc Loss: 3.73305, Dur Loss: 0.95285, CE Loss: 0.05015, Norm Loss: 0.97612, F0 Loss: 2.95431, LM Loss: 2.09318, Gen Loss: 6.27407, Sty Loss: 0.11628, Diff Loss: 0.18509, DiscLM Loss: 0.00156, GenLM Loss: 0.99897 +INFO:2025-06-11 23:53:57,124: Epoch [16/20], Step [550/2647], Loss: 0.49761, Disc Loss: 3.85405, Dur Loss: 1.01126, CE Loss: 0.05552, Norm Loss: 1.12237, F0 Loss: 3.39526, LM Loss: 2.07851, Gen Loss: 5.50972, Sty Loss: 0.08311, Diff Loss: 0.32201, DiscLM Loss: 0.00000, GenLM Loss: 1.00701 +INFO:2025-06-11 23:55:46,384: Epoch [16/20], Step [600/2647], Loss: 0.50399, Disc Loss: 3.77957, Dur Loss: 0.81413, CE Loss: 0.04710, Norm Loss: 0.79311, F0 Loss: 2.68119, LM Loss: 2.15017, Gen Loss: 6.37136, Sty Loss: 0.10058, Diff Loss: 0.30237, DiscLM Loss: 0.00139, GenLM Loss: 1.00018 +INFO:2025-06-11 23:57:36,409: Epoch [16/20], Step [650/2647], Loss: 0.51174, Disc Loss: 3.82159, Dur Loss: 0.87537, CE Loss: 0.04704, Norm Loss: 0.88562, F0 Loss: 2.74450, LM Loss: 2.13866, Gen Loss: 6.53672, Sty Loss: 0.09718, Diff Loss: 0.23819, DiscLM Loss: 0.00000, GenLM Loss: 0.98440 +INFO:2025-06-11 23:59:26,045: Epoch [16/20], Step [700/2647], Loss: 0.51263, Disc Loss: 3.70535, Dur Loss: 0.82250, CE Loss: 0.04110, Norm Loss: 1.05179, F0 Loss: 3.36710, LM Loss: 2.11642, Gen Loss: 6.30895, Sty Loss: 0.10391, Diff Loss: 0.25031, DiscLM Loss: 0.00070, GenLM Loss: 1.00545 +INFO:2025-06-12 00:01:14,039: Epoch [16/20], Step [750/2647], Loss: 0.51989, Disc Loss: 3.62456, Dur Loss: 0.78680, CE Loss: 0.04307, Norm Loss: 1.01445, F0 Loss: 2.23517, LM Loss: 2.24088, Gen Loss: 6.51610, Sty Loss: 0.09429, Diff Loss: 0.22907, DiscLM Loss: 0.00000, GenLM Loss: 0.99669 +INFO:2025-06-12 00:03:02,392: Epoch [16/20], Step [800/2647], Loss: 0.51291, Disc Loss: 3.79590, Dur Loss: 1.07250, CE Loss: 0.06330, Norm Loss: 0.86250, F0 Loss: 2.71754, LM Loss: 2.12848, Gen Loss: 6.43129, Sty Loss: 0.12154, Diff Loss: 0.32123, DiscLM Loss: 0.00929, GenLM Loss: 0.99331 +INFO:2025-06-12 00:04:51,515: Epoch [16/20], Step [850/2647], Loss: 0.49749, Disc Loss: 3.79089, Dur Loss: 0.82785, CE Loss: 0.04176, Norm Loss: 0.81602, F0 Loss: 2.53245, LM Loss: 2.09027, Gen Loss: 5.96031, Sty Loss: 0.08921, Diff Loss: 0.18269, DiscLM Loss: 0.00000, GenLM Loss: 1.00486 +INFO:2025-06-12 00:06:39,768: Epoch [16/20], Step [900/2647], Loss: 0.50745, Disc Loss: 3.74492, Dur Loss: 0.95222, CE Loss: 0.05115, Norm Loss: 1.15932, F0 Loss: 3.42034, LM Loss: 2.19814, Gen Loss: 6.81085, Sty Loss: 0.07395, Diff Loss: 0.24729, DiscLM Loss: 0.00256, GenLM Loss: 1.00588 +INFO:2025-06-12 00:08:28,162: Epoch [16/20], Step [950/2647], Loss: 0.52029, Disc Loss: 3.86732, Dur Loss: 1.21559, CE Loss: 0.05781, Norm Loss: 0.90898, F0 Loss: 2.53760, LM Loss: 2.19636, Gen Loss: 6.37320, Sty Loss: 0.09870, Diff Loss: 0.16795, DiscLM Loss: 0.00000, GenLM Loss: 0.99639 +INFO:2025-06-12 00:10:18,240: Epoch [16/20], Step [1000/2647], Loss: 0.50664, Disc Loss: 3.69273, Dur Loss: 0.72194, CE Loss: 0.03822, Norm Loss: 0.98334, F0 Loss: 3.15941, LM Loss: 2.04536, Gen Loss: 6.42227, Sty Loss: 0.09048, Diff Loss: 0.30433, DiscLM Loss: 0.00404, GenLM Loss: 0.98314 +INFO:2025-06-12 00:12:06,956: Epoch [16/20], Step [1050/2647], Loss: 0.51096, Disc Loss: 3.66954, Dur Loss: 0.94000, CE Loss: 0.05217, Norm Loss: 1.34247, F0 Loss: 3.76188, LM Loss: 2.28581, Gen Loss: 6.22797, Sty Loss: 0.10130, Diff Loss: 0.27088, DiscLM Loss: 0.00000, GenLM Loss: 1.00157 +INFO:2025-06-12 00:13:56,599: Epoch [16/20], Step [1100/2647], Loss: 0.51226, Disc Loss: 3.75038, Dur Loss: 0.86089, CE Loss: 0.04630, Norm Loss: 0.93733, F0 Loss: 2.47869, LM Loss: 2.05734, Gen Loss: 6.60830, Sty Loss: 0.10189, Diff Loss: 0.39395, DiscLM Loss: 0.00166, GenLM Loss: 1.00292 +INFO:2025-06-12 00:15:44,512: Epoch [16/20], Step [1150/2647], Loss: 0.51891, Disc Loss: 3.75900, Dur Loss: 0.91595, CE Loss: 0.04629, Norm Loss: 0.96960, F0 Loss: 2.62286, LM Loss: 2.20513, Gen Loss: 6.66725, Sty Loss: 0.10361, Diff Loss: 0.35212, DiscLM Loss: 0.00000, GenLM Loss: 0.99851 +INFO:2025-06-12 00:17:32,291: Epoch [16/20], Step [1200/2647], Loss: 0.49842, Disc Loss: 3.85413, Dur Loss: 0.90698, CE Loss: 0.04894, Norm Loss: 0.84666, F0 Loss: 2.84655, LM Loss: 2.01321, Gen Loss: 6.36212, Sty Loss: 0.10061, Diff Loss: 0.27229, DiscLM Loss: 0.00247, GenLM Loss: 0.99289 +INFO:2025-06-12 00:19:19,647: Epoch [16/20], Step [1250/2647], Loss: 0.50707, Disc Loss: 3.68949, Dur Loss: 0.74378, CE Loss: 0.03954, Norm Loss: 0.94276, F0 Loss: 2.91722, LM Loss: 2.14922, Gen Loss: 6.39954, Sty Loss: 0.08482, Diff Loss: 0.29938, DiscLM Loss: 0.00000, GenLM Loss: 1.00577 +INFO:2025-06-12 00:21:08,240: Epoch [16/20], Step [1300/2647], Loss: 0.50816, Disc Loss: 3.81243, Dur Loss: 0.90910, CE Loss: 0.05208, Norm Loss: 0.86284, F0 Loss: 2.28307, LM Loss: 2.08705, Gen Loss: 6.66090, Sty Loss: 0.11431, Diff Loss: 0.33529, DiscLM Loss: 0.00580, GenLM Loss: 0.99570 +INFO:2025-06-12 00:22:54,058: Epoch [16/20], Step [1350/2647], Loss: 0.51232, Disc Loss: 3.71591, Dur Loss: 0.96939, CE Loss: 0.04968, Norm Loss: 1.00862, F0 Loss: 3.31935, LM Loss: 2.17564, Gen Loss: 6.66757, Sty Loss: 0.10905, Diff Loss: 0.28418, DiscLM Loss: 0.00000, GenLM Loss: 1.00395 +INFO:2025-06-12 00:24:43,368: Epoch [16/20], Step [1400/2647], Loss: 0.50538, Disc Loss: 3.74754, Dur Loss: 0.95549, CE Loss: 0.05456, Norm Loss: 0.89249, F0 Loss: 2.62410, LM Loss: 2.16299, Gen Loss: 6.68374, Sty Loss: 0.12150, Diff Loss: 0.24169, DiscLM Loss: 0.00080, GenLM Loss: 1.00542 +INFO:2025-06-12 00:26:33,177: Epoch [16/20], Step [1450/2647], Loss: 0.52605, Disc Loss: 3.72071, Dur Loss: 0.90447, CE Loss: 0.05142, Norm Loss: 0.89060, F0 Loss: 3.13526, LM Loss: 2.21857, Gen Loss: 7.23047, Sty Loss: 0.11451, Diff Loss: 0.25222, DiscLM Loss: 0.00000, GenLM Loss: 1.00576 +INFO:2025-06-12 00:28:22,170: Epoch [16/20], Step [1500/2647], Loss: 0.50616, Disc Loss: 3.64737, Dur Loss: 0.97374, CE Loss: 0.04837, Norm Loss: 1.07713, F0 Loss: 2.56437, LM Loss: 2.17044, Gen Loss: 6.92191, Sty Loss: 0.11204, Diff Loss: 0.25383, DiscLM Loss: 0.00074, GenLM Loss: 1.00415 +INFO:2025-06-12 00:30:09,795: Epoch [16/20], Step [1550/2647], Loss: 0.50905, Disc Loss: 3.82629, Dur Loss: 0.90206, CE Loss: 0.04967, Norm Loss: 1.50868, F0 Loss: 3.10042, LM Loss: 2.24267, Gen Loss: 5.85967, Sty Loss: 0.10681, Diff Loss: 0.26946, DiscLM Loss: 0.00000, GenLM Loss: 0.99469 +INFO:2025-06-12 00:31:58,457: Epoch [16/20], Step [1600/2647], Loss: 0.51596, Disc Loss: 3.69046, Dur Loss: 0.98988, CE Loss: 0.05652, Norm Loss: 1.17236, F0 Loss: 3.02758, LM Loss: 2.27810, Gen Loss: 6.80791, Sty Loss: 0.09155, Diff Loss: 0.18078, DiscLM Loss: 0.00087, GenLM Loss: 1.00026 +INFO:2025-06-12 00:33:47,783: Epoch [16/20], Step [1650/2647], Loss: 0.50467, Disc Loss: 3.69012, Dur Loss: 1.00081, CE Loss: 0.05477, Norm Loss: 1.06611, F0 Loss: 2.75671, LM Loss: 2.17481, Gen Loss: 6.77080, Sty Loss: 0.09913, Diff Loss: 0.30282, DiscLM Loss: 0.00000, GenLM Loss: 1.00424 +INFO:2025-06-12 00:35:35,837: Epoch [16/20], Step [1700/2647], Loss: 0.51371, Disc Loss: 3.73889, Dur Loss: 1.22615, CE Loss: 0.06625, Norm Loss: 1.23806, F0 Loss: 3.32120, LM Loss: 2.22842, Gen Loss: 6.47539, Sty Loss: 0.10140, Diff Loss: 0.28024, DiscLM Loss: 0.00094, GenLM Loss: 0.99660 +INFO:2025-06-12 00:37:26,853: Epoch [16/20], Step [1750/2647], Loss: 0.51550, Disc Loss: 3.75629, Dur Loss: 0.82152, CE Loss: 0.04224, Norm Loss: 1.00250, F0 Loss: 3.70599, LM Loss: 2.19302, Gen Loss: 6.58219, Sty Loss: 0.11762, Diff Loss: 0.31800, DiscLM Loss: 0.00000, GenLM Loss: 0.99851 +INFO:2025-06-12 00:39:13,415: Epoch [16/20], Step [1800/2647], Loss: 0.51367, Disc Loss: 3.81352, Dur Loss: 0.84191, CE Loss: 0.04506, Norm Loss: 1.02451, F0 Loss: 3.24817, LM Loss: 2.10669, Gen Loss: 6.29929, Sty Loss: 0.10537, Diff Loss: 0.29032, DiscLM Loss: 0.00042, GenLM Loss: 1.00431 +INFO:2025-06-12 00:41:01,800: Epoch [16/20], Step [1850/2647], Loss: 0.52194, Disc Loss: 3.70576, Dur Loss: 0.94409, CE Loss: 0.05453, Norm Loss: 1.18354, F0 Loss: 3.27810, LM Loss: 2.20796, Gen Loss: 6.34319, Sty Loss: 0.10810, Diff Loss: 0.30415, DiscLM Loss: 0.00000, GenLM Loss: 0.99859 +INFO:2025-06-12 00:42:49,697: Epoch [16/20], Step [1900/2647], Loss: 0.51471, Disc Loss: 3.62125, Dur Loss: 1.33163, CE Loss: 0.06578, Norm Loss: 1.00795, F0 Loss: 2.90219, LM Loss: 2.16538, Gen Loss: 6.93944, Sty Loss: 0.12428, Diff Loss: 0.27322, DiscLM Loss: 0.00041, GenLM Loss: 0.99128 +INFO:2025-06-12 00:44:38,136: Epoch [16/20], Step [1950/2647], Loss: 0.51462, Disc Loss: 3.76461, Dur Loss: 1.24995, CE Loss: 0.06654, Norm Loss: 1.06404, F0 Loss: 3.63790, LM Loss: 2.24365, Gen Loss: 5.78886, Sty Loss: 0.12612, Diff Loss: 0.30394, DiscLM Loss: 0.00000, GenLM Loss: 0.99032 +INFO:2025-06-12 00:46:24,311: Epoch [16/20], Step [2000/2647], Loss: 0.51480, Disc Loss: 3.75359, Dur Loss: 0.75010, CE Loss: 0.03732, Norm Loss: 1.09081, F0 Loss: 3.14588, LM Loss: 2.25082, Gen Loss: 6.42551, Sty Loss: 0.11789, Diff Loss: 0.35975, DiscLM Loss: 0.00120, GenLM Loss: 1.00044 +INFO:2025-06-12 00:48:13,619: Epoch [16/20], Step [2050/2647], Loss: 0.50765, Disc Loss: 3.71481, Dur Loss: 0.88573, CE Loss: 0.04798, Norm Loss: 0.79843, F0 Loss: 2.54855, LM Loss: 2.05425, Gen Loss: 6.90509, Sty Loss: 0.09812, Diff Loss: 0.23747, DiscLM Loss: 0.00000, GenLM Loss: 0.98299 +INFO:2025-06-12 00:50:02,853: Epoch [16/20], Step [2100/2647], Loss: 0.50873, Disc Loss: 3.64105, Dur Loss: 1.06860, CE Loss: 0.05894, Norm Loss: 0.94194, F0 Loss: 2.99139, LM Loss: 2.03321, Gen Loss: 6.73912, Sty Loss: 0.09822, Diff Loss: 0.21797, DiscLM Loss: 0.00255, GenLM Loss: 1.01528 +INFO:2025-06-12 00:51:52,712: Epoch [16/20], Step [2150/2647], Loss: 0.50509, Disc Loss: 3.71943, Dur Loss: 1.22008, CE Loss: 0.07067, Norm Loss: 1.31394, F0 Loss: 3.42587, LM Loss: 2.33558, Gen Loss: 6.81695, Sty Loss: 0.11098, Diff Loss: 0.19329, DiscLM Loss: 0.00000, GenLM Loss: 0.97470 +INFO:2025-06-12 00:53:42,693: Epoch [16/20], Step [2200/2647], Loss: 0.51192, Disc Loss: 3.75251, Dur Loss: 1.12038, CE Loss: 0.06008, Norm Loss: 1.13269, F0 Loss: 3.38283, LM Loss: 2.09219, Gen Loss: 6.24014, Sty Loss: 0.11006, Diff Loss: 0.25837, DiscLM Loss: 0.00257, GenLM Loss: 0.99710 +INFO:2025-06-12 00:55:32,847: Epoch [16/20], Step [2250/2647], Loss: 0.52033, Disc Loss: 3.73831, Dur Loss: 0.94066, CE Loss: 0.04966, Norm Loss: 1.00090, F0 Loss: 2.82642, LM Loss: 2.18731, Gen Loss: 5.91472, Sty Loss: 0.10965, Diff Loss: 0.24652, DiscLM Loss: 0.00000, GenLM Loss: 1.00348 +INFO:2025-06-12 00:57:20,578: Epoch [16/20], Step [2300/2647], Loss: 0.52262, Disc Loss: 3.69579, Dur Loss: 1.02822, CE Loss: 0.06235, Norm Loss: 1.33336, F0 Loss: 3.41750, LM Loss: 2.18222, Gen Loss: 6.56247, Sty Loss: 0.10431, Diff Loss: 0.25634, DiscLM Loss: 0.00057, GenLM Loss: 0.99208 +INFO:2025-06-12 00:59:07,563: Epoch [16/20], Step [2350/2647], Loss: 0.51585, Disc Loss: 3.68274, Dur Loss: 0.97226, CE Loss: 0.04856, Norm Loss: 0.86141, F0 Loss: 2.57292, LM Loss: 2.14748, Gen Loss: 6.99263, Sty Loss: 0.09917, Diff Loss: 0.28451, DiscLM Loss: 0.00000, GenLM Loss: 1.01829 +INFO:2025-06-12 01:00:57,380: Epoch [16/20], Step [2400/2647], Loss: 0.51531, Disc Loss: 3.66655, Dur Loss: 0.89970, CE Loss: 0.05321, Norm Loss: 0.81971, F0 Loss: 2.86566, LM Loss: 2.14057, Gen Loss: 6.11192, Sty Loss: 0.09003, Diff Loss: 0.29748, DiscLM Loss: 0.01408, GenLM Loss: 0.96932 +INFO:2025-06-12 01:02:46,055: Epoch [16/20], Step [2450/2647], Loss: 0.50897, Disc Loss: 3.82092, Dur Loss: 1.03593, CE Loss: 0.05213, Norm Loss: 1.18503, F0 Loss: 2.99672, LM Loss: 2.12785, Gen Loss: 5.91967, Sty Loss: 0.11570, Diff Loss: 0.30246, DiscLM Loss: 0.00000, GenLM Loss: 0.99512 +INFO:2025-06-12 01:04:33,042: Epoch [16/20], Step [2500/2647], Loss: 0.51245, Disc Loss: 3.65627, Dur Loss: 1.05113, CE Loss: 0.04903, Norm Loss: 1.23614, F0 Loss: 3.09446, LM Loss: 2.20576, Gen Loss: 7.37715, Sty Loss: 0.10550, Diff Loss: 0.30270, DiscLM Loss: 0.00286, GenLM Loss: 0.98597 +INFO:2025-06-12 01:06:23,318: Epoch [16/20], Step [2550/2647], Loss: 0.50331, Disc Loss: 3.79642, Dur Loss: 0.92017, CE Loss: 0.04690, Norm Loss: 0.99073, F0 Loss: 2.75500, LM Loss: 2.10931, Gen Loss: 5.82668, Sty Loss: 0.10539, Diff Loss: 0.32993, DiscLM Loss: 0.00000, GenLM Loss: 1.00537 +INFO:2025-06-12 01:08:13,083: Epoch [16/20], Step [2600/2647], Loss: 0.50935, Disc Loss: 3.79878, Dur Loss: 0.82923, CE Loss: 0.04196, Norm Loss: 1.01818, F0 Loss: 2.73124, LM Loss: 2.04019, Gen Loss: 5.81141, Sty Loss: 0.10004, Diff Loss: 0.21687, DiscLM Loss: 0.00213, GenLM Loss: 1.01132 +INFO:2025-06-12 01:12:12,905: Validation loss: 0.536, Dur loss: 0.960, F0 loss: 3.295 + + + +INFO:2025-06-12 01:14:05,915: Epoch [17/20], Step [50/2647], Loss: 0.50984, Disc Loss: 3.73624, Dur Loss: 0.89910, CE Loss: 0.04836, Norm Loss: 0.88839, F0 Loss: 2.64986, LM Loss: 2.13131, Gen Loss: 6.38718, Sty Loss: 0.09231, Diff Loss: 0.28423, DiscLM Loss: 0.00000, GenLM Loss: 1.01166 +INFO:2025-06-12 01:15:54,341: Epoch [17/20], Step [100/2647], Loss: 0.50950, Disc Loss: 3.66753, Dur Loss: 1.01984, CE Loss: 0.05032, Norm Loss: 1.01385, F0 Loss: 3.15650, LM Loss: 2.12606, Gen Loss: 6.45031, Sty Loss: 0.10703, Diff Loss: 0.24534, DiscLM Loss: 0.00196, GenLM Loss: 1.00466 +INFO:2025-06-12 01:17:43,823: Epoch [17/20], Step [150/2647], Loss: 0.50887, Disc Loss: 3.74577, Dur Loss: 0.90225, CE Loss: 0.04778, Norm Loss: 0.80225, F0 Loss: 2.79039, LM Loss: 2.08971, Gen Loss: 6.46942, Sty Loss: 0.11004, Diff Loss: 0.31874, DiscLM Loss: 0.00000, GenLM Loss: 0.98507 +INFO:2025-06-12 01:19:32,601: Epoch [17/20], Step [200/2647], Loss: 0.50132, Disc Loss: 3.69851, Dur Loss: 0.94458, CE Loss: 0.05554, Norm Loss: 1.19935, F0 Loss: 2.70763, LM Loss: 2.26358, Gen Loss: 6.51885, Sty Loss: 0.08049, Diff Loss: 0.23993, DiscLM Loss: 0.00068, GenLM Loss: 0.99715 +INFO:2025-06-12 01:21:23,447: Epoch [17/20], Step [250/2647], Loss: 0.50283, Disc Loss: 3.73627, Dur Loss: 1.03491, CE Loss: 0.05377, Norm Loss: 1.08349, F0 Loss: 2.29525, LM Loss: 2.18966, Gen Loss: 6.14380, Sty Loss: 0.10321, Diff Loss: 0.27292, DiscLM Loss: 0.00000, GenLM Loss: 1.00231 +INFO:2025-06-12 01:23:13,735: Epoch [17/20], Step [300/2647], Loss: 0.50867, Disc Loss: 3.75492, Dur Loss: 0.88907, CE Loss: 0.04679, Norm Loss: 0.92207, F0 Loss: 2.22632, LM Loss: 2.05681, Gen Loss: 6.11267, Sty Loss: 0.09548, Diff Loss: 0.19309, DiscLM Loss: 0.00081, GenLM Loss: 0.99480 +INFO:2025-06-12 01:25:02,128: Epoch [17/20], Step [350/2647], Loss: 0.51646, Disc Loss: 3.75906, Dur Loss: 1.04736, CE Loss: 0.05440, Norm Loss: 1.04323, F0 Loss: 2.25657, LM Loss: 1.98872, Gen Loss: 6.25936, Sty Loss: 0.12822, Diff Loss: 0.30408, DiscLM Loss: 0.00000, GenLM Loss: 0.99667 +INFO:2025-06-12 01:26:49,084: Epoch [17/20], Step [400/2647], Loss: 0.51178, Disc Loss: 3.76904, Dur Loss: 1.50112, CE Loss: 0.07099, Norm Loss: 1.26828, F0 Loss: 3.29104, LM Loss: 2.27292, Gen Loss: 6.60125, Sty Loss: 0.10315, Diff Loss: 0.29567, DiscLM Loss: 0.00067, GenLM Loss: 0.99564 +INFO:2025-06-12 01:28:37,127: Epoch [17/20], Step [450/2647], Loss: 0.51256, Disc Loss: 3.69442, Dur Loss: 0.83299, CE Loss: 0.04058, Norm Loss: 0.83601, F0 Loss: 2.94516, LM Loss: 2.13248, Gen Loss: 6.93522, Sty Loss: 0.11145, Diff Loss: 0.33842, DiscLM Loss: 0.00000, GenLM Loss: 0.99486 +INFO:2025-06-12 01:30:26,709: Epoch [17/20], Step [500/2647], Loss: 0.50542, Disc Loss: 3.74600, Dur Loss: 0.95637, CE Loss: 0.04776, Norm Loss: 1.02622, F0 Loss: 3.18506, LM Loss: 2.20720, Gen Loss: 6.06392, Sty Loss: 0.10861, Diff Loss: 0.28325, DiscLM Loss: 0.00254, GenLM Loss: 1.00418 +INFO:2025-06-12 01:32:16,849: Epoch [17/20], Step [550/2647], Loss: 0.50895, Disc Loss: 3.74638, Dur Loss: 0.95635, CE Loss: 0.04882, Norm Loss: 1.29408, F0 Loss: 3.37996, LM Loss: 2.19263, Gen Loss: 6.50330, Sty Loss: 0.12590, Diff Loss: 0.26880, DiscLM Loss: 0.00000, GenLM Loss: 0.99525 +INFO:2025-06-12 01:34:05,853: Epoch [17/20], Step [600/2647], Loss: 0.51519, Disc Loss: 3.84961, Dur Loss: 0.98554, CE Loss: 0.05109, Norm Loss: 0.76880, F0 Loss: 2.31723, LM Loss: 2.13608, Gen Loss: 6.81770, Sty Loss: 0.12226, Diff Loss: 0.22904, DiscLM Loss: 0.00051, GenLM Loss: 0.99302 +INFO:2025-06-12 01:35:55,626: Epoch [17/20], Step [650/2647], Loss: 0.50611, Disc Loss: 3.57135, Dur Loss: 0.93256, CE Loss: 0.05152, Norm Loss: 0.95574, F0 Loss: 3.68171, LM Loss: 2.17459, Gen Loss: 7.59320, Sty Loss: 0.11946, Diff Loss: 0.26576, DiscLM Loss: 0.00000, GenLM Loss: 0.99204 +INFO:2025-06-12 01:37:41,871: Epoch [17/20], Step [700/2647], Loss: 0.51780, Disc Loss: 3.71755, Dur Loss: 0.86888, CE Loss: 0.04639, Norm Loss: 0.69745, F0 Loss: 2.42729, LM Loss: 2.13785, Gen Loss: 5.97522, Sty Loss: 0.09954, Diff Loss: 0.31842, DiscLM Loss: 0.00031, GenLM Loss: 0.99922 +INFO:2025-06-12 01:39:30,965: Epoch [17/20], Step [750/2647], Loss: 0.50923, Disc Loss: 3.70353, Dur Loss: 1.02248, CE Loss: 0.05809, Norm Loss: 1.01049, F0 Loss: 3.06125, LM Loss: 2.19598, Gen Loss: 7.25230, Sty Loss: 0.11144, Diff Loss: 0.23208, DiscLM Loss: 0.00000, GenLM Loss: 0.99625 +INFO:2025-06-12 01:41:18,525: Epoch [17/20], Step [800/2647], Loss: 0.51059, Disc Loss: 3.82143, Dur Loss: 0.90057, CE Loss: 0.04702, Norm Loss: 0.95186, F0 Loss: 2.83720, LM Loss: 2.10411, Gen Loss: 5.93472, Sty Loss: 0.09374, Diff Loss: 0.30988, DiscLM Loss: 0.00050, GenLM Loss: 1.00267 +INFO:2025-06-12 01:43:08,307: Epoch [17/20], Step [850/2647], Loss: 0.51036, Disc Loss: 3.78014, Dur Loss: 0.84776, CE Loss: 0.04218, Norm Loss: 0.95275, F0 Loss: 2.78830, LM Loss: 2.16739, Gen Loss: 6.22761, Sty Loss: 0.12256, Diff Loss: 0.32600, DiscLM Loss: 0.00000, GenLM Loss: 0.99708 +INFO:2025-06-12 01:45:00,821: Epoch [17/20], Step [900/2647], Loss: 0.51650, Disc Loss: 3.81036, Dur Loss: 0.88465, CE Loss: 0.04729, Norm Loss: 1.04479, F0 Loss: 2.80186, LM Loss: 2.12283, Gen Loss: 6.45403, Sty Loss: 0.11144, Diff Loss: 0.29614, DiscLM Loss: 0.00040, GenLM Loss: 0.99394 +INFO:2025-06-12 01:46:50,198: Epoch [17/20], Step [950/2647], Loss: 0.51204, Disc Loss: 3.80033, Dur Loss: 1.02102, CE Loss: 0.05640, Norm Loss: 1.33702, F0 Loss: 3.36946, LM Loss: 2.14872, Gen Loss: 5.53737, Sty Loss: 0.11461, Diff Loss: 0.22986, DiscLM Loss: 0.00000, GenLM Loss: 1.00613 +INFO:2025-06-12 01:48:38,357: Epoch [17/20], Step [1000/2647], Loss: 0.51077, Disc Loss: 3.70200, Dur Loss: 1.01259, CE Loss: 0.05532, Norm Loss: 1.29889, F0 Loss: 3.12851, LM Loss: 2.22517, Gen Loss: 6.68033, Sty Loss: 0.09755, Diff Loss: 0.19766, DiscLM Loss: 0.00030, GenLM Loss: 1.00287 +INFO:2025-06-12 01:50:27,037: Epoch [17/20], Step [1050/2647], Loss: 0.51220, Disc Loss: 3.85612, Dur Loss: 0.92736, CE Loss: 0.05219, Norm Loss: 0.97374, F0 Loss: 2.95011, LM Loss: 2.07901, Gen Loss: 6.02903, Sty Loss: 0.11269, Diff Loss: 0.33988, DiscLM Loss: 0.00000, GenLM Loss: 0.99561 +INFO:2025-06-12 01:52:16,268: Epoch [17/20], Step [1100/2647], Loss: 0.51091, Disc Loss: 3.69470, Dur Loss: 1.11619, CE Loss: 0.05886, Norm Loss: 1.06864, F0 Loss: 2.86407, LM Loss: 2.20072, Gen Loss: 6.33175, Sty Loss: 0.11761, Diff Loss: 0.31128, DiscLM Loss: 0.00033, GenLM Loss: 1.00522 +INFO:2025-06-12 01:54:03,230: Epoch [17/20], Step [1150/2647], Loss: 0.51347, Disc Loss: 3.74914, Dur Loss: 0.86652, CE Loss: 0.04528, Norm Loss: 0.77506, F0 Loss: 2.77935, LM Loss: 2.24199, Gen Loss: 7.00058, Sty Loss: 0.18202, Diff Loss: 0.40868, DiscLM Loss: 0.00000, GenLM Loss: 1.00298 +INFO:2025-06-12 01:55:53,033: Epoch [17/20], Step [1200/2647], Loss: 0.51023, Disc Loss: 3.73319, Dur Loss: 0.88893, CE Loss: 0.04726, Norm Loss: 1.12245, F0 Loss: 3.04706, LM Loss: 2.13667, Gen Loss: 6.76037, Sty Loss: 0.10083, Diff Loss: 0.26762, DiscLM Loss: 0.00655, GenLM Loss: 0.99299 +INFO:2025-06-12 01:57:41,221: Epoch [17/20], Step [1250/2647], Loss: 0.50423, Disc Loss: 3.62737, Dur Loss: 1.06455, CE Loss: 0.05821, Norm Loss: 0.88339, F0 Loss: 3.59779, LM Loss: 2.25473, Gen Loss: 7.49167, Sty Loss: 0.14212, Diff Loss: 0.28681, DiscLM Loss: 0.00000, GenLM Loss: 0.98257 +INFO:2025-06-12 01:59:29,871: Epoch [17/20], Step [1300/2647], Loss: 0.51460, Disc Loss: 3.69084, Dur Loss: 0.89821, CE Loss: 0.05140, Norm Loss: 1.04860, F0 Loss: 2.60251, LM Loss: 2.12457, Gen Loss: 6.15052, Sty Loss: 0.10072, Diff Loss: 0.29159, DiscLM Loss: 0.00680, GenLM Loss: 0.99553 +INFO:2025-06-12 02:01:19,941: Epoch [17/20], Step [1350/2647], Loss: 0.50788, Disc Loss: 3.59836, Dur Loss: 0.93835, CE Loss: 0.04832, Norm Loss: 0.90643, F0 Loss: 2.78236, LM Loss: 2.18145, Gen Loss: 6.72429, Sty Loss: 0.11948, Diff Loss: 0.33181, DiscLM Loss: 0.00000, GenLM Loss: 0.99194 +INFO:2025-06-12 02:03:10,171: Epoch [17/20], Step [1400/2647], Loss: 0.50605, Disc Loss: 3.76132, Dur Loss: 1.02775, CE Loss: 0.06418, Norm Loss: 0.89636, F0 Loss: 2.48573, LM Loss: 2.10035, Gen Loss: 6.61563, Sty Loss: 0.10736, Diff Loss: 0.27072, DiscLM Loss: 0.00579, GenLM Loss: 0.95218 +INFO:2025-06-12 02:05:00,469: Epoch [17/20], Step [1450/2647], Loss: 0.50954, Disc Loss: 3.78082, Dur Loss: 0.94739, CE Loss: 0.04998, Norm Loss: 1.26987, F0 Loss: 2.37723, LM Loss: 2.03932, Gen Loss: 5.75012, Sty Loss: 0.10972, Diff Loss: 0.27349, DiscLM Loss: 0.00000, GenLM Loss: 1.00816 +INFO:2025-06-12 02:06:50,657: Epoch [17/20], Step [1500/2647], Loss: 0.51157, Disc Loss: 3.79386, Dur Loss: 0.95591, CE Loss: 0.05042, Norm Loss: 0.88708, F0 Loss: 2.81942, LM Loss: 2.11191, Gen Loss: 6.18057, Sty Loss: 0.10508, Diff Loss: 0.22567, DiscLM Loss: 0.02722, GenLM Loss: 0.93434 +INFO:2025-06-12 02:08:39,684: Epoch [17/20], Step [1550/2647], Loss: 0.50591, Disc Loss: 3.79063, Dur Loss: 1.31781, CE Loss: 0.06880, Norm Loss: 1.07824, F0 Loss: 3.10584, LM Loss: 2.15326, Gen Loss: 5.67013, Sty Loss: 0.09864, Diff Loss: 0.25608, DiscLM Loss: 0.00000, GenLM Loss: 0.99615 +INFO:2025-06-12 02:10:28,640: Epoch [17/20], Step [1600/2647], Loss: 0.50517, Disc Loss: 3.78275, Dur Loss: 1.05773, CE Loss: 0.05905, Norm Loss: 0.74189, F0 Loss: 2.15674, LM Loss: 2.02539, Gen Loss: 6.69130, Sty Loss: 0.12256, Diff Loss: 0.37219, DiscLM Loss: 0.00225, GenLM Loss: 1.00758 +INFO:2025-06-12 02:12:17,479: Epoch [17/20], Step [1650/2647], Loss: 0.50895, Disc Loss: 3.73022, Dur Loss: 0.89783, CE Loss: 0.04788, Norm Loss: 0.77115, F0 Loss: 3.02381, LM Loss: 2.05054, Gen Loss: 7.16933, Sty Loss: 0.14684, Diff Loss: 0.39354, DiscLM Loss: 0.00000, GenLM Loss: 0.99218 +INFO:2025-06-12 02:14:03,498: Epoch [17/20], Step [1700/2647], Loss: 0.51173, Disc Loss: 3.81876, Dur Loss: 0.84151, CE Loss: 0.04697, Norm Loss: 0.94745, F0 Loss: 2.30642, LM Loss: 2.12739, Gen Loss: 6.09296, Sty Loss: 0.12853, Diff Loss: 0.23940, DiscLM Loss: 0.00208, GenLM Loss: 0.99249 +INFO:2025-06-12 02:15:49,157: Epoch [17/20], Step [1750/2647], Loss: 0.51953, Disc Loss: 3.77500, Dur Loss: 0.96359, CE Loss: 0.05741, Norm Loss: 0.96835, F0 Loss: 3.17491, LM Loss: 2.24872, Gen Loss: 6.57774, Sty Loss: 0.12983, Diff Loss: 0.31055, DiscLM Loss: 0.00000, GenLM Loss: 1.00837 +INFO:2025-06-12 02:17:37,833: Epoch [17/20], Step [1800/2647], Loss: 0.51551, Disc Loss: 3.73767, Dur Loss: 1.00818, CE Loss: 0.05048, Norm Loss: 1.05202, F0 Loss: 2.67205, LM Loss: 2.09515, Gen Loss: 6.32730, Sty Loss: 0.11609, Diff Loss: 0.26382, DiscLM Loss: 0.02566, GenLM Loss: 0.96087 +INFO:2025-06-12 02:19:27,078: Epoch [17/20], Step [1850/2647], Loss: 0.51023, Disc Loss: 3.65035, Dur Loss: 1.37250, CE Loss: 0.07107, Norm Loss: 1.01898, F0 Loss: 2.95602, LM Loss: 2.19285, Gen Loss: 6.12280, Sty Loss: 0.10055, Diff Loss: 0.35635, DiscLM Loss: 0.00000, GenLM Loss: 0.99536 +INFO:2025-06-12 02:21:17,557: Epoch [17/20], Step [1900/2647], Loss: 0.51209, Disc Loss: 3.68050, Dur Loss: 0.78122, CE Loss: 0.03970, Norm Loss: 0.85857, F0 Loss: 3.56104, LM Loss: 2.02599, Gen Loss: 6.20045, Sty Loss: 0.08324, Diff Loss: 0.33392, DiscLM Loss: 0.00277, GenLM Loss: 1.00161 +INFO:2025-06-12 02:23:06,134: Epoch [17/20], Step [1950/2647], Loss: 0.50947, Disc Loss: 3.78100, Dur Loss: 0.92640, CE Loss: 0.05232, Norm Loss: 1.12650, F0 Loss: 2.63802, LM Loss: 2.16817, Gen Loss: 6.20516, Sty Loss: 0.10918, Diff Loss: 0.27717, DiscLM Loss: 0.00000, GenLM Loss: 0.96086 +INFO:2025-06-12 02:24:53,343: Epoch [17/20], Step [2000/2647], Loss: 0.51373, Disc Loss: 3.78190, Dur Loss: 1.01646, CE Loss: 0.05952, Norm Loss: 0.93960, F0 Loss: 2.97104, LM Loss: 2.28344, Gen Loss: 6.63499, Sty Loss: 0.11071, Diff Loss: 0.29197, DiscLM Loss: 0.00563, GenLM Loss: 0.97314 +INFO:2025-06-12 02:26:42,538: Epoch [17/20], Step [2050/2647], Loss: 0.50129, Disc Loss: 3.73335, Dur Loss: 0.93813, CE Loss: 0.04881, Norm Loss: 1.03572, F0 Loss: 3.64037, LM Loss: 2.13972, Gen Loss: 6.98859, Sty Loss: 0.19064, Diff Loss: 0.39478, DiscLM Loss: 0.00000, GenLM Loss: 1.00856 +INFO:2025-06-12 02:28:31,104: Epoch [17/20], Step [2100/2647], Loss: 0.50571, Disc Loss: 3.83735, Dur Loss: 0.95514, CE Loss: 0.05075, Norm Loss: 0.74484, F0 Loss: 2.67487, LM Loss: 2.08428, Gen Loss: 5.65851, Sty Loss: 0.09859, Diff Loss: 0.21170, DiscLM Loss: 0.00163, GenLM Loss: 1.04082 +INFO:2025-06-12 02:30:18,515: Epoch [17/20], Step [2150/2647], Loss: 0.50976, Disc Loss: 3.76183, Dur Loss: 0.81720, CE Loss: 0.04575, Norm Loss: 1.23556, F0 Loss: 3.57589, LM Loss: 2.25340, Gen Loss: 6.66803, Sty Loss: 0.13207, Diff Loss: 0.30056, DiscLM Loss: 0.00000, GenLM Loss: 0.91617 +INFO:2025-06-12 02:32:08,013: Epoch [17/20], Step [2200/2647], Loss: 0.51398, Disc Loss: 3.70337, Dur Loss: 1.19296, CE Loss: 0.05812, Norm Loss: 0.85593, F0 Loss: 4.61472, LM Loss: 2.25737, Gen Loss: 6.95926, Sty Loss: 0.11626, Diff Loss: 0.21240, DiscLM Loss: 0.00182, GenLM Loss: 0.96974 +INFO:2025-06-12 02:33:54,966: Epoch [17/20], Step [2250/2647], Loss: 0.51735, Disc Loss: 3.71038, Dur Loss: 0.73538, CE Loss: 0.03637, Norm Loss: 0.93509, F0 Loss: 2.59100, LM Loss: 2.08214, Gen Loss: 6.39032, Sty Loss: 0.11424, Diff Loss: 0.33146, DiscLM Loss: 0.00000, GenLM Loss: 1.04324 +INFO:2025-06-12 02:35:44,748: Epoch [17/20], Step [2300/2647], Loss: 0.51758, Disc Loss: 3.74189, Dur Loss: 0.92476, CE Loss: 0.05187, Norm Loss: 0.72813, F0 Loss: 2.69518, LM Loss: 2.10561, Gen Loss: 6.39117, Sty Loss: 0.23141, Diff Loss: 0.44805, DiscLM Loss: 0.00364, GenLM Loss: 1.02973 +INFO:2025-06-12 02:37:32,162: Epoch [17/20], Step [2350/2647], Loss: 0.51540, Disc Loss: 3.59039, Dur Loss: 0.96586, CE Loss: 0.05133, Norm Loss: 0.87449, F0 Loss: 2.98185, LM Loss: 2.28508, Gen Loss: 7.11395, Sty Loss: 0.14688, Diff Loss: 0.26992, DiscLM Loss: 0.00000, GenLM Loss: 0.95604 +INFO:2025-06-12 02:39:17,966: Epoch [17/20], Step [2400/2647], Loss: 0.53214, Disc Loss: 3.69982, Dur Loss: 0.97594, CE Loss: 0.05343, Norm Loss: 0.93060, F0 Loss: 2.76695, LM Loss: 2.13445, Gen Loss: 6.39025, Sty Loss: 0.09872, Diff Loss: 0.27399, DiscLM Loss: 0.00183, GenLM Loss: 1.00039 +INFO:2025-06-12 02:41:05,140: Epoch [17/20], Step [2450/2647], Loss: 0.51972, Disc Loss: 3.68696, Dur Loss: 0.96170, CE Loss: 0.04824, Norm Loss: 1.17732, F0 Loss: 2.84667, LM Loss: 2.13912, Gen Loss: 6.35775, Sty Loss: 0.10979, Diff Loss: 0.33592, DiscLM Loss: 0.00000, GenLM Loss: 1.00506 +INFO:2025-06-12 02:42:53,482: Epoch [17/20], Step [2500/2647], Loss: 0.51674, Disc Loss: 3.68937, Dur Loss: 0.73133, CE Loss: 0.03848, Norm Loss: 0.81873, F0 Loss: 3.23141, LM Loss: 2.13254, Gen Loss: 7.24639, Sty Loss: 0.14551, Diff Loss: 0.31530, DiscLM Loss: 0.00107, GenLM Loss: 0.97310 +INFO:2025-06-12 02:44:42,920: Epoch [17/20], Step [2550/2647], Loss: 0.49554, Disc Loss: 3.72472, Dur Loss: 1.08261, CE Loss: 0.05335, Norm Loss: 0.83959, F0 Loss: 2.49059, LM Loss: 2.22569, Gen Loss: 6.70405, Sty Loss: 0.09360, Diff Loss: 0.22481, DiscLM Loss: 0.00000, GenLM Loss: 0.99795 +INFO:2025-06-12 02:46:33,840: Epoch [17/20], Step [2600/2647], Loss: 0.52214, Disc Loss: 3.81097, Dur Loss: 0.96504, CE Loss: 0.04564, Norm Loss: 1.15265, F0 Loss: 3.64287, LM Loss: 2.24046, Gen Loss: 5.82080, Sty Loss: 0.10637, Diff Loss: 0.22240, DiscLM Loss: 0.00116, GenLM Loss: 1.01311 +INFO:2025-06-12 02:50:38,619: Validation loss: 0.529, Dur loss: 0.968, F0 loss: 3.211 + + + +INFO:2025-06-12 02:52:30,055: Epoch [18/20], Step [50/2647], Loss: 0.51051, Disc Loss: 3.73320, Dur Loss: 0.96274, CE Loss: 0.05497, Norm Loss: 0.91959, F0 Loss: 3.07017, LM Loss: 2.20948, Gen Loss: 6.06614, Sty Loss: 0.11045, Diff Loss: 0.33736, DiscLM Loss: 0.00000, GenLM Loss: 0.98398 +INFO:2025-06-12 02:54:17,448: Epoch [18/20], Step [100/2647], Loss: 0.52276, Disc Loss: 3.65164, Dur Loss: 0.89647, CE Loss: 0.04922, Norm Loss: 0.75918, F0 Loss: 2.30585, LM Loss: 2.14879, Gen Loss: 8.03340, Sty Loss: 0.15195, Diff Loss: 0.43310, DiscLM Loss: 0.00059, GenLM Loss: 0.99916 +INFO:2025-06-12 02:56:06,681: Epoch [18/20], Step [150/2647], Loss: 0.50550, Disc Loss: 3.77631, Dur Loss: 0.89911, CE Loss: 0.04691, Norm Loss: 0.71981, F0 Loss: 2.10013, LM Loss: 2.04274, Gen Loss: 6.34502, Sty Loss: 0.09193, Diff Loss: 0.27463, DiscLM Loss: 0.00000, GenLM Loss: 1.01276 +INFO:2025-06-12 02:57:57,743: Epoch [18/20], Step [200/2647], Loss: 0.50161, Disc Loss: 3.72499, Dur Loss: 0.97274, CE Loss: 0.05777, Norm Loss: 1.07633, F0 Loss: 2.41945, LM Loss: 2.19687, Gen Loss: 6.79213, Sty Loss: 0.08699, Diff Loss: 0.24380, DiscLM Loss: 0.00492, GenLM Loss: 1.00095 +INFO:2025-06-12 02:59:47,690: Epoch [18/20], Step [250/2647], Loss: 0.51185, Disc Loss: 3.83471, Dur Loss: 0.78722, CE Loss: 0.03996, Norm Loss: 0.85784, F0 Loss: 2.91150, LM Loss: 2.15482, Gen Loss: 6.20388, Sty Loss: 0.09979, Diff Loss: 0.33158, DiscLM Loss: 0.00000, GenLM Loss: 0.99014 +INFO:2025-06-12 03:01:37,595: Epoch [18/20], Step [300/2647], Loss: 0.51158, Disc Loss: 3.76235, Dur Loss: 0.77123, CE Loss: 0.03941, Norm Loss: 0.85511, F0 Loss: 2.23855, LM Loss: 2.17666, Gen Loss: 6.46662, Sty Loss: 0.10329, Diff Loss: 0.28563, DiscLM Loss: 0.00040, GenLM Loss: 1.00803 +INFO:2025-06-12 03:03:27,276: Epoch [18/20], Step [350/2647], Loss: 0.51337, Disc Loss: 3.79245, Dur Loss: 1.18777, CE Loss: 0.06082, Norm Loss: 1.21649, F0 Loss: 2.50535, LM Loss: 2.00924, Gen Loss: 6.81282, Sty Loss: 0.10721, Diff Loss: 0.23785, DiscLM Loss: 0.00000, GenLM Loss: 1.00295 +INFO:2025-06-12 03:05:18,382: Epoch [18/20], Step [400/2647], Loss: 0.52498, Disc Loss: 3.73756, Dur Loss: 0.90844, CE Loss: 0.04741, Norm Loss: 1.02518, F0 Loss: 2.25958, LM Loss: 2.16132, Gen Loss: 6.19886, Sty Loss: 0.09306, Diff Loss: 0.28511, DiscLM Loss: 0.00049, GenLM Loss: 0.98986 +INFO:2025-06-12 03:07:06,343: Epoch [18/20], Step [450/2647], Loss: 0.51249, Disc Loss: 3.81084, Dur Loss: 1.01667, CE Loss: 0.05724, Norm Loss: 1.09877, F0 Loss: 2.60838, LM Loss: 2.15590, Gen Loss: 6.20372, Sty Loss: 0.08989, Diff Loss: 0.25455, DiscLM Loss: 0.00000, GenLM Loss: 1.00478 +INFO:2025-06-12 03:08:55,673: Epoch [18/20], Step [500/2647], Loss: 0.51924, Disc Loss: 3.70017, Dur Loss: 1.22768, CE Loss: 0.07342, Norm Loss: 1.54308, F0 Loss: 2.72705, LM Loss: 2.24615, Gen Loss: 7.07354, Sty Loss: 0.13755, Diff Loss: 0.33006, DiscLM Loss: 0.00059, GenLM Loss: 1.00990 +INFO:2025-06-12 03:10:45,268: Epoch [18/20], Step [550/2647], Loss: 0.51045, Disc Loss: 3.91572, Dur Loss: 1.30213, CE Loss: 0.07551, Norm Loss: 1.13258, F0 Loss: 3.86758, LM Loss: 2.26320, Gen Loss: 7.39655, Sty Loss: 0.12784, Diff Loss: 0.25731, DiscLM Loss: 0.00000, GenLM Loss: 0.99949 +INFO:2025-06-12 03:12:31,785: Epoch [18/20], Step [600/2647], Loss: 0.52614, Disc Loss: 3.82611, Dur Loss: 0.93577, CE Loss: 0.04528, Norm Loss: 0.71558, F0 Loss: 2.85161, LM Loss: 2.06839, Gen Loss: 6.25151, Sty Loss: 0.11741, Diff Loss: 0.30177, DiscLM Loss: 0.00081, GenLM Loss: 0.98758 +INFO:2025-06-12 03:14:18,515: Epoch [18/20], Step [650/2647], Loss: 0.51139, Disc Loss: 3.73810, Dur Loss: 0.81056, CE Loss: 0.04570, Norm Loss: 0.97564, F0 Loss: 2.65631, LM Loss: 2.15595, Gen Loss: 6.38373, Sty Loss: 0.10162, Diff Loss: 0.30356, DiscLM Loss: 0.00000, GenLM Loss: 1.00953 +INFO:2025-06-12 03:16:06,185: Epoch [18/20], Step [700/2647], Loss: 0.51250, Disc Loss: 3.60010, Dur Loss: 0.78684, CE Loss: 0.03891, Norm Loss: 0.83706, F0 Loss: 2.54285, LM Loss: 2.08215, Gen Loss: 6.64918, Sty Loss: 0.08760, Diff Loss: 0.37896, DiscLM Loss: 0.00094, GenLM Loss: 1.00675 +INFO:2025-06-12 03:17:54,251: Epoch [18/20], Step [750/2647], Loss: 0.50864, Disc Loss: 3.70712, Dur Loss: 0.86615, CE Loss: 0.04721, Norm Loss: 1.05832, F0 Loss: 3.12643, LM Loss: 2.10291, Gen Loss: 6.74822, Sty Loss: 0.08659, Diff Loss: 0.35475, DiscLM Loss: 0.00000, GenLM Loss: 0.99294 +INFO:2025-06-12 03:19:42,741: Epoch [18/20], Step [800/2647], Loss: 0.50726, Disc Loss: 3.74760, Dur Loss: 0.95216, CE Loss: 0.04626, Norm Loss: 0.86531, F0 Loss: 2.77802, LM Loss: 2.06750, Gen Loss: 6.20506, Sty Loss: 0.09326, Diff Loss: 0.26181, DiscLM Loss: 0.00110, GenLM Loss: 1.00024 +INFO:2025-06-12 03:21:30,683: Epoch [18/20], Step [850/2647], Loss: 0.51251, Disc Loss: 3.85982, Dur Loss: 0.93595, CE Loss: 0.05139, Norm Loss: 1.15052, F0 Loss: 3.78055, LM Loss: 2.26872, Gen Loss: 6.54530, Sty Loss: 0.12080, Diff Loss: 0.30523, DiscLM Loss: 0.00000, GenLM Loss: 1.00055 +INFO:2025-06-12 03:23:21,430: Epoch [18/20], Step [900/2647], Loss: 0.51564, Disc Loss: 3.73069, Dur Loss: 1.05289, CE Loss: 0.05557, Norm Loss: 0.94630, F0 Loss: 3.74544, LM Loss: 2.29209, Gen Loss: 6.38090, Sty Loss: 0.10554, Diff Loss: 0.27223, DiscLM Loss: 0.00044, GenLM Loss: 0.99147 +INFO:2025-06-12 03:25:10,314: Epoch [18/20], Step [950/2647], Loss: 0.51130, Disc Loss: 3.65259, Dur Loss: 1.02585, CE Loss: 0.05279, Norm Loss: 0.89290, F0 Loss: 3.49345, LM Loss: 2.08441, Gen Loss: 6.74868, Sty Loss: 0.10341, Diff Loss: 0.23994, DiscLM Loss: 0.00000, GenLM Loss: 0.99125 +INFO:2025-06-12 03:26:56,945: Epoch [18/20], Step [1000/2647], Loss: 0.51836, Disc Loss: 3.70773, Dur Loss: 1.12440, CE Loss: 0.06303, Norm Loss: 0.96504, F0 Loss: 3.18068, LM Loss: 2.17510, Gen Loss: 6.03543, Sty Loss: 0.11647, Diff Loss: 0.20383, DiscLM Loss: 0.00111, GenLM Loss: 1.00439 +INFO:2025-06-12 03:28:45,789: Epoch [18/20], Step [1050/2647], Loss: 0.51176, Disc Loss: 3.72813, Dur Loss: 0.82736, CE Loss: 0.04334, Norm Loss: 0.61183, F0 Loss: 2.50959, LM Loss: 2.08585, Gen Loss: 6.64595, Sty Loss: 0.08690, Diff Loss: 0.33090, DiscLM Loss: 0.00000, GenLM Loss: 0.98920 +INFO:2025-06-12 03:30:34,627: Epoch [18/20], Step [1100/2647], Loss: 0.52336, Disc Loss: 3.68560, Dur Loss: 0.73780, CE Loss: 0.03959, Norm Loss: 0.98535, F0 Loss: 2.65126, LM Loss: 2.15871, Gen Loss: 7.06236, Sty Loss: 0.12234, Diff Loss: 0.32426, DiscLM Loss: 0.00115, GenLM Loss: 1.00276 +INFO:2025-06-12 03:32:22,439: Epoch [18/20], Step [1150/2647], Loss: 0.51200, Disc Loss: 3.81630, Dur Loss: 0.89301, CE Loss: 0.04694, Norm Loss: 0.80686, F0 Loss: 2.81334, LM Loss: 2.14006, Gen Loss: 7.18504, Sty Loss: 0.11403, Diff Loss: 0.26041, DiscLM Loss: 0.00000, GenLM Loss: 0.98914 +INFO:2025-06-12 03:34:10,403: Epoch [18/20], Step [1200/2647], Loss: 0.50549, Disc Loss: 3.71378, Dur Loss: 0.91901, CE Loss: 0.05012, Norm Loss: 1.13238, F0 Loss: 2.29246, LM Loss: 2.05601, Gen Loss: 6.66451, Sty Loss: 0.12766, Diff Loss: 0.25438, DiscLM Loss: 0.00675, GenLM Loss: 0.97700 +INFO:2025-06-12 03:36:00,550: Epoch [18/20], Step [1250/2647], Loss: 0.50093, Disc Loss: 3.74818, Dur Loss: 0.88020, CE Loss: 0.04641, Norm Loss: 1.07545, F0 Loss: 3.17025, LM Loss: 2.19765, Gen Loss: 5.85360, Sty Loss: 0.12884, Diff Loss: 0.23630, DiscLM Loss: 0.00000, GenLM Loss: 1.00832 +INFO:2025-06-12 03:37:52,077: Epoch [18/20], Step [1300/2647], Loss: 0.52031, Disc Loss: 3.68713, Dur Loss: 1.08769, CE Loss: 0.06112, Norm Loss: 1.32050, F0 Loss: 3.64977, LM Loss: 2.26259, Gen Loss: 7.04869, Sty Loss: 0.12913, Diff Loss: 0.36942, DiscLM Loss: 0.00052, GenLM Loss: 0.99812 +INFO:2025-06-12 03:39:38,335: Epoch [18/20], Step [1350/2647], Loss: 0.52069, Disc Loss: 3.67533, Dur Loss: 0.78729, CE Loss: 0.04177, Norm Loss: 0.77369, F0 Loss: 2.39114, LM Loss: 2.08143, Gen Loss: 6.99353, Sty Loss: 0.14061, Diff Loss: 0.29407, DiscLM Loss: 0.00000, GenLM Loss: 0.99885 +INFO:2025-06-12 03:41:25,401: Epoch [18/20], Step [1400/2647], Loss: 0.50419, Disc Loss: 3.67339, Dur Loss: 0.90407, CE Loss: 0.04814, Norm Loss: 0.85361, F0 Loss: 2.21610, LM Loss: 2.09695, Gen Loss: 7.18783, Sty Loss: 0.09775, Diff Loss: 0.26177, DiscLM Loss: 0.00037, GenLM Loss: 1.00881 +INFO:2025-06-12 03:43:13,174: Epoch [18/20], Step [1450/2647], Loss: 0.52068, Disc Loss: 3.70664, Dur Loss: 1.09489, CE Loss: 0.06296, Norm Loss: 0.75137, F0 Loss: 3.20265, LM Loss: 2.14067, Gen Loss: 6.60586, Sty Loss: 0.12367, Diff Loss: 0.35171, DiscLM Loss: 0.00000, GenLM Loss: 0.99492 +INFO:2025-06-12 03:45:00,723: Epoch [18/20], Step [1500/2647], Loss: 0.50216, Disc Loss: 3.71484, Dur Loss: 0.85779, CE Loss: 0.03966, Norm Loss: 0.86528, F0 Loss: 2.64549, LM Loss: 2.03180, Gen Loss: 6.52711, Sty Loss: 0.09210, Diff Loss: 0.23322, DiscLM Loss: 0.00036, GenLM Loss: 0.99794 +INFO:2025-06-12 03:46:50,179: Epoch [18/20], Step [1550/2647], Loss: 0.50744, Disc Loss: 3.73301, Dur Loss: 1.00533, CE Loss: 0.04975, Norm Loss: 1.22432, F0 Loss: 2.89552, LM Loss: 2.34810, Gen Loss: 6.13733, Sty Loss: 0.11718, Diff Loss: 0.21445, DiscLM Loss: 0.00000, GenLM Loss: 0.99096 +INFO:2025-06-12 03:48:40,765: Epoch [18/20], Step [1600/2647], Loss: 0.51234, Disc Loss: 3.80223, Dur Loss: 0.77783, CE Loss: 0.04187, Norm Loss: 1.10562, F0 Loss: 3.31545, LM Loss: 2.20834, Gen Loss: 6.35038, Sty Loss: 0.10631, Diff Loss: 0.25080, DiscLM Loss: 0.00106, GenLM Loss: 0.99427 +INFO:2025-06-12 03:50:30,893: Epoch [18/20], Step [1650/2647], Loss: 0.50985, Disc Loss: 3.68222, Dur Loss: 1.11487, CE Loss: 0.06410, Norm Loss: 1.25775, F0 Loss: 2.97002, LM Loss: 2.22270, Gen Loss: 7.55921, Sty Loss: 0.10854, Diff Loss: 0.19625, DiscLM Loss: 0.00000, GenLM Loss: 1.01828 +INFO:2025-06-12 03:52:18,720: Epoch [18/20], Step [1700/2647], Loss: 0.51280, Disc Loss: 3.62282, Dur Loss: 0.85794, CE Loss: 0.04824, Norm Loss: 0.84966, F0 Loss: 3.03646, LM Loss: 2.15916, Gen Loss: 7.26140, Sty Loss: 0.12151, Diff Loss: 0.36288, DiscLM Loss: 0.00524, GenLM Loss: 0.98014 +INFO:2025-06-12 03:54:09,169: Epoch [18/20], Step [1750/2647], Loss: 0.50776, Disc Loss: 3.69609, Dur Loss: 0.88029, CE Loss: 0.04523, Norm Loss: 0.87718, F0 Loss: 2.83457, LM Loss: 2.03010, Gen Loss: 6.31088, Sty Loss: 0.10338, Diff Loss: 0.30375, DiscLM Loss: 0.00000, GenLM Loss: 0.99027 +INFO:2025-06-12 03:56:00,998: Epoch [18/20], Step [1800/2647], Loss: 0.51505, Disc Loss: 3.71584, Dur Loss: 0.95801, CE Loss: 0.05452, Norm Loss: 0.95005, F0 Loss: 2.77061, LM Loss: 2.11547, Gen Loss: 6.50632, Sty Loss: 0.09526, Diff Loss: 0.21218, DiscLM Loss: 0.00046, GenLM Loss: 1.00363 +INFO:2025-06-12 03:57:49,714: Epoch [18/20], Step [1850/2647], Loss: 0.51028, Disc Loss: 3.78714, Dur Loss: 0.79457, CE Loss: 0.04057, Norm Loss: 0.73714, F0 Loss: 3.19902, LM Loss: 2.10015, Gen Loss: 7.25479, Sty Loss: 0.09429, Diff Loss: 0.36452, DiscLM Loss: 0.00000, GenLM Loss: 0.99707 +INFO:2025-06-12 03:59:38,288: Epoch [18/20], Step [1900/2647], Loss: 0.52892, Disc Loss: 3.79455, Dur Loss: 0.90097, CE Loss: 0.04686, Norm Loss: 1.10090, F0 Loss: 2.87956, LM Loss: 2.09697, Gen Loss: 6.30213, Sty Loss: 0.09699, Diff Loss: 0.32377, DiscLM Loss: 0.00069, GenLM Loss: 1.00057 +INFO:2025-06-12 04:01:27,317: Epoch [18/20], Step [1950/2647], Loss: 0.51082, Disc Loss: 3.73887, Dur Loss: 0.90103, CE Loss: 0.04572, Norm Loss: 0.86642, F0 Loss: 2.30323, LM Loss: 2.05114, Gen Loss: 6.42158, Sty Loss: 0.09578, Diff Loss: 0.29481, DiscLM Loss: 0.00000, GenLM Loss: 0.99999 +INFO:2025-06-12 04:03:15,063: Epoch [18/20], Step [2000/2647], Loss: 0.51886, Disc Loss: 3.62770, Dur Loss: 0.95997, CE Loss: 0.05228, Norm Loss: 1.14458, F0 Loss: 3.24776, LM Loss: 2.26108, Gen Loss: 6.63195, Sty Loss: 0.21347, Diff Loss: 0.42126, DiscLM Loss: 0.00106, GenLM Loss: 0.99258 +INFO:2025-06-12 04:05:05,081: Epoch [18/20], Step [2050/2647], Loss: 0.52737, Disc Loss: 3.70401, Dur Loss: 0.92984, CE Loss: 0.04876, Norm Loss: 1.19721, F0 Loss: 3.71950, LM Loss: 2.23262, Gen Loss: 6.64766, Sty Loss: 0.11864, Diff Loss: 0.30932, DiscLM Loss: 0.00000, GenLM Loss: 0.99955 +INFO:2025-06-12 04:06:52,141: Epoch [18/20], Step [2100/2647], Loss: 0.51932, Disc Loss: 3.79248, Dur Loss: 1.03326, CE Loss: 0.05317, Norm Loss: 1.07260, F0 Loss: 2.95679, LM Loss: 2.14205, Gen Loss: 6.21589, Sty Loss: 0.10311, Diff Loss: 0.28278, DiscLM Loss: 0.00506, GenLM Loss: 0.98617 +INFO:2025-06-12 04:08:41,428: Epoch [18/20], Step [2150/2647], Loss: 0.50672, Disc Loss: 3.71346, Dur Loss: 0.77482, CE Loss: 0.03841, Norm Loss: 0.97380, F0 Loss: 2.65825, LM Loss: 2.08130, Gen Loss: 6.27503, Sty Loss: 0.17564, Diff Loss: 0.43252, DiscLM Loss: 0.00000, GenLM Loss: 0.98445 +INFO:2025-06-12 04:10:29,867: Epoch [18/20], Step [2200/2647], Loss: 0.51658, Disc Loss: 3.71804, Dur Loss: 1.09184, CE Loss: 0.05702, Norm Loss: 0.99719, F0 Loss: 3.35166, LM Loss: 2.23567, Gen Loss: 6.39616, Sty Loss: 0.10795, Diff Loss: 0.27281, DiscLM Loss: 0.00049, GenLM Loss: 0.99376 +INFO:2025-06-12 04:12:17,620: Epoch [18/20], Step [2250/2647], Loss: 0.51647, Disc Loss: 3.70844, Dur Loss: 0.89381, CE Loss: 0.04791, Norm Loss: 0.65427, F0 Loss: 2.48422, LM Loss: 2.17440, Gen Loss: 6.22562, Sty Loss: 0.13414, Diff Loss: 0.34116, DiscLM Loss: 0.00000, GenLM Loss: 1.00347 +INFO:2025-06-12 04:14:05,066: Epoch [18/20], Step [2300/2647], Loss: 0.51165, Disc Loss: 3.73375, Dur Loss: 0.89946, CE Loss: 0.05048, Norm Loss: 0.79361, F0 Loss: 2.35185, LM Loss: 2.20242, Gen Loss: 6.87476, Sty Loss: 0.08782, Diff Loss: 0.22286, DiscLM Loss: 0.00032, GenLM Loss: 1.00828 +INFO:2025-06-12 04:15:53,475: Epoch [18/20], Step [2350/2647], Loss: 0.51644, Disc Loss: 3.77359, Dur Loss: 0.87753, CE Loss: 0.04856, Norm Loss: 0.81465, F0 Loss: 2.63643, LM Loss: 2.11015, Gen Loss: 6.16488, Sty Loss: 0.07582, Diff Loss: 0.24133, DiscLM Loss: 0.00000, GenLM Loss: 0.99138 +INFO:2025-06-12 04:17:41,151: Epoch [18/20], Step [2400/2647], Loss: 0.52130, Disc Loss: 3.73788, Dur Loss: 0.88637, CE Loss: 0.04466, Norm Loss: 0.97914, F0 Loss: 3.89337, LM Loss: 2.12978, Gen Loss: 6.56553, Sty Loss: 0.08885, Diff Loss: 0.20498, DiscLM Loss: 0.00613, GenLM Loss: 1.00300 +INFO:2025-06-12 04:19:29,988: Epoch [18/20], Step [2450/2647], Loss: 0.51236, Disc Loss: 3.76198, Dur Loss: 0.84718, CE Loss: 0.04329, Norm Loss: 0.90187, F0 Loss: 3.48762, LM Loss: 2.07299, Gen Loss: 6.19451, Sty Loss: 0.10127, Diff Loss: 0.31109, DiscLM Loss: 0.00000, GenLM Loss: 1.00082 +INFO:2025-06-12 04:21:16,577: Epoch [18/20], Step [2500/2647], Loss: 0.50839, Disc Loss: 3.71522, Dur Loss: 0.95460, CE Loss: 0.04938, Norm Loss: 1.05538, F0 Loss: 2.78557, LM Loss: 2.07296, Gen Loss: 6.60763, Sty Loss: 0.11801, Diff Loss: 0.35339, DiscLM Loss: 0.00033, GenLM Loss: 1.00568 +INFO:2025-06-12 04:23:06,012: Epoch [18/20], Step [2550/2647], Loss: 0.50577, Disc Loss: 3.75891, Dur Loss: 0.99755, CE Loss: 0.05145, Norm Loss: 0.99530, F0 Loss: 2.43424, LM Loss: 2.09739, Gen Loss: 6.27699, Sty Loss: 0.10704, Diff Loss: 0.27155, DiscLM Loss: 0.00000, GenLM Loss: 0.98510 +INFO:2025-06-12 04:24:56,242: Epoch [18/20], Step [2600/2647], Loss: 0.51031, Disc Loss: 3.67023, Dur Loss: 1.21488, CE Loss: 0.06772, Norm Loss: 1.12786, F0 Loss: 2.53013, LM Loss: 2.09733, Gen Loss: 6.51526, Sty Loss: 0.11277, Diff Loss: 0.22038, DiscLM Loss: 0.00264, GenLM Loss: 1.00242 +INFO:2025-06-12 04:28:58,631: Validation loss: 0.530, Dur loss: 0.947, F0 loss: 3.223 + + + +INFO:2025-06-12 04:30:52,179: Epoch [19/20], Step [50/2647], Loss: 0.51145, Disc Loss: 3.73203, Dur Loss: 0.77079, CE Loss: 0.03980, Norm Loss: 0.83005, F0 Loss: 3.16592, LM Loss: 2.05908, Gen Loss: 5.85759, Sty Loss: 0.09740, Diff Loss: 0.27798, DiscLM Loss: 0.00000, GenLM Loss: 0.98709 +INFO:2025-06-12 04:32:38,731: Epoch [19/20], Step [100/2647], Loss: 0.50573, Disc Loss: 3.62815, Dur Loss: 0.95463, CE Loss: 0.04801, Norm Loss: 1.02305, F0 Loss: 3.07086, LM Loss: 2.21349, Gen Loss: 7.04055, Sty Loss: 0.08335, Diff Loss: 0.26859, DiscLM Loss: 0.00040, GenLM Loss: 1.01432 +INFO:2025-06-12 04:34:26,955: Epoch [19/20], Step [150/2647], Loss: 0.50246, Disc Loss: 3.77918, Dur Loss: 1.18013, CE Loss: 0.06373, Norm Loss: 1.50092, F0 Loss: 3.84459, LM Loss: 2.21737, Gen Loss: 6.46663, Sty Loss: 0.09677, Diff Loss: 0.30988, DiscLM Loss: 0.00000, GenLM Loss: 0.98620 +INFO:2025-06-12 04:36:15,650: Epoch [19/20], Step [200/2647], Loss: 0.50346, Disc Loss: 3.79402, Dur Loss: 0.93804, CE Loss: 0.04914, Norm Loss: 1.08862, F0 Loss: 3.33048, LM Loss: 2.26254, Gen Loss: 5.71882, Sty Loss: 0.09226, Diff Loss: 0.27893, DiscLM Loss: 0.00364, GenLM Loss: 0.99981 +INFO:2025-06-12 04:38:03,449: Epoch [19/20], Step [250/2647], Loss: 0.51057, Disc Loss: 3.82739, Dur Loss: 0.99137, CE Loss: 0.04963, Norm Loss: 0.89171, F0 Loss: 2.37251, LM Loss: 2.10606, Gen Loss: 5.78850, Sty Loss: 0.09361, Diff Loss: 0.26189, DiscLM Loss: 0.00000, GenLM Loss: 0.99562 +INFO:2025-06-12 04:39:51,282: Epoch [19/20], Step [300/2647], Loss: 0.51940, Disc Loss: 3.72196, Dur Loss: 0.97203, CE Loss: 0.04998, Norm Loss: 1.06134, F0 Loss: 3.17843, LM Loss: 2.15671, Gen Loss: 6.22776, Sty Loss: 0.11268, Diff Loss: 0.26854, DiscLM Loss: 0.00081, GenLM Loss: 0.99102 +INFO:2025-06-12 04:41:42,569: Epoch [19/20], Step [350/2647], Loss: 0.50592, Disc Loss: 3.77290, Dur Loss: 1.04785, CE Loss: 0.05264, Norm Loss: 0.99409, F0 Loss: 3.12431, LM Loss: 2.13388, Gen Loss: 6.70604, Sty Loss: 0.09927, Diff Loss: 0.37221, DiscLM Loss: 0.00000, GenLM Loss: 1.01926 +INFO:2025-06-12 04:43:32,759: Epoch [19/20], Step [400/2647], Loss: 0.51054, Disc Loss: 3.66136, Dur Loss: 1.19133, CE Loss: 0.06333, Norm Loss: 1.04854, F0 Loss: 2.55589, LM Loss: 2.02236, Gen Loss: 6.53441, Sty Loss: 0.10624, Diff Loss: 0.29094, DiscLM Loss: 0.00045, GenLM Loss: 1.00815 +INFO:2025-06-12 04:45:20,233: Epoch [19/20], Step [450/2647], Loss: 0.50333, Disc Loss: 3.66428, Dur Loss: 1.39780, CE Loss: 0.07678, Norm Loss: 1.08929, F0 Loss: 2.56294, LM Loss: 2.29889, Gen Loss: 6.84149, Sty Loss: 0.26174, Diff Loss: 0.35954, DiscLM Loss: 0.00000, GenLM Loss: 0.99773 +INFO:2025-06-12 04:47:08,964: Epoch [19/20], Step [500/2647], Loss: 0.51303, Disc Loss: 3.80693, Dur Loss: 1.14203, CE Loss: 0.06449, Norm Loss: 0.92628, F0 Loss: 2.60609, LM Loss: 2.06132, Gen Loss: 6.43628, Sty Loss: 0.08744, Diff Loss: 0.30390, DiscLM Loss: 0.01347, GenLM Loss: 1.00509 +INFO:2025-06-12 04:48:56,847: Epoch [19/20], Step [550/2647], Loss: 0.51237, Disc Loss: 3.66778, Dur Loss: 0.76626, CE Loss: 0.04010, Norm Loss: 0.92424, F0 Loss: 3.39610, LM Loss: 2.11739, Gen Loss: 6.53301, Sty Loss: 0.10672, Diff Loss: 0.23097, DiscLM Loss: 0.00000, GenLM Loss: 0.99734 +INFO:2025-06-12 04:50:45,163: Epoch [19/20], Step [600/2647], Loss: 0.51078, Disc Loss: 3.74542, Dur Loss: 1.06777, CE Loss: 0.05684, Norm Loss: 0.99847, F0 Loss: 3.08023, LM Loss: 2.14397, Gen Loss: 7.63768, Sty Loss: 0.09433, Diff Loss: 0.26873, DiscLM Loss: 0.00085, GenLM Loss: 1.00662 +INFO:2025-06-12 04:52:36,162: Epoch [19/20], Step [650/2647], Loss: 0.53116, Disc Loss: 3.82950, Dur Loss: 0.98874, CE Loss: 0.05628, Norm Loss: 1.01974, F0 Loss: 3.02903, LM Loss: 2.21292, Gen Loss: 6.93337, Sty Loss: 0.16785, Diff Loss: 0.44751, DiscLM Loss: 0.00000, GenLM Loss: 0.98517 +INFO:2025-06-12 04:54:27,178: Epoch [19/20], Step [700/2647], Loss: 0.51109, Disc Loss: 3.79943, Dur Loss: 0.97560, CE Loss: 0.05337, Norm Loss: 0.93804, F0 Loss: 1.85062, LM Loss: 2.12715, Gen Loss: 7.34709, Sty Loss: 0.12721, Diff Loss: 0.29906, DiscLM Loss: 0.00249, GenLM Loss: 0.99775 +INFO:2025-06-12 04:56:16,463: Epoch [19/20], Step [750/2647], Loss: 0.52277, Disc Loss: 3.80473, Dur Loss: 0.92034, CE Loss: 0.04761, Norm Loss: 0.90048, F0 Loss: 3.28343, LM Loss: 2.16329, Gen Loss: 5.93166, Sty Loss: 0.11311, Diff Loss: 0.27109, DiscLM Loss: 0.00000, GenLM Loss: 1.01376 +INFO:2025-06-12 04:58:05,437: Epoch [19/20], Step [800/2647], Loss: 0.50937, Disc Loss: 3.71170, Dur Loss: 1.17483, CE Loss: 0.05443, Norm Loss: 0.79696, F0 Loss: 3.18361, LM Loss: 2.16894, Gen Loss: 6.46687, Sty Loss: 0.08805, Diff Loss: 0.25807, DiscLM Loss: 0.00392, GenLM Loss: 1.02575 +INFO:2025-06-12 04:59:54,163: Epoch [19/20], Step [850/2647], Loss: 0.51204, Disc Loss: 3.76539, Dur Loss: 1.05957, CE Loss: 0.06334, Norm Loss: 0.84147, F0 Loss: 3.15822, LM Loss: 2.20798, Gen Loss: 6.26617, Sty Loss: 0.12536, Diff Loss: 0.27626, DiscLM Loss: 0.00000, GenLM Loss: 0.97728 +INFO:2025-06-12 05:01:44,073: Epoch [19/20], Step [900/2647], Loss: 0.51694, Disc Loss: 3.77817, Dur Loss: 0.88863, CE Loss: 0.04641, Norm Loss: 0.81953, F0 Loss: 2.76704, LM Loss: 1.99093, Gen Loss: 6.33646, Sty Loss: 0.08879, Diff Loss: 0.18963, DiscLM Loss: 0.00187, GenLM Loss: 1.00041 +INFO:2025-06-12 05:03:34,463: Epoch [19/20], Step [950/2647], Loss: 0.50284, Disc Loss: 3.67887, Dur Loss: 0.93986, CE Loss: 0.05132, Norm Loss: 0.82562, F0 Loss: 2.52824, LM Loss: 2.05552, Gen Loss: 6.63808, Sty Loss: 0.10306, Diff Loss: 0.26903, DiscLM Loss: 0.00000, GenLM Loss: 0.99255 +INFO:2025-06-12 05:05:23,398: Epoch [19/20], Step [1000/2647], Loss: 0.50408, Disc Loss: 3.83641, Dur Loss: 0.82571, CE Loss: 0.04199, Norm Loss: 0.88028, F0 Loss: 3.27680, LM Loss: 2.11390, Gen Loss: 6.44963, Sty Loss: 0.11071, Diff Loss: 0.33010, DiscLM Loss: 0.00037, GenLM Loss: 0.98550 +INFO:2025-06-12 05:07:14,150: Epoch [19/20], Step [1050/2647], Loss: 0.50363, Disc Loss: 3.65669, Dur Loss: 0.73503, CE Loss: 0.03995, Norm Loss: 0.85132, F0 Loss: 2.61610, LM Loss: 2.20481, Gen Loss: 6.72177, Sty Loss: 0.10672, Diff Loss: 0.31940, DiscLM Loss: 0.00000, GenLM Loss: 1.00787 +INFO:2025-06-12 05:09:02,942: Epoch [19/20], Step [1100/2647], Loss: 0.50159, Disc Loss: 3.84423, Dur Loss: 1.04219, CE Loss: 0.05297, Norm Loss: 1.02711, F0 Loss: 2.45452, LM Loss: 2.03010, Gen Loss: 5.64522, Sty Loss: 0.09734, Diff Loss: 0.23617, DiscLM Loss: 0.00080, GenLM Loss: 0.97969 +INFO:2025-06-12 05:10:47,240: Epoch [19/20], Step [1150/2647], Loss: 0.50799, Disc Loss: 3.77108, Dur Loss: 1.27323, CE Loss: 0.07048, Norm Loss: 1.24157, F0 Loss: 2.97496, LM Loss: 2.13232, Gen Loss: 5.62504, Sty Loss: 0.09931, Diff Loss: 0.23301, DiscLM Loss: 0.00000, GenLM Loss: 1.01504 +INFO:2025-06-12 05:12:37,814: Epoch [19/20], Step [1200/2647], Loss: 0.49692, Disc Loss: 3.79466, Dur Loss: 0.75909, CE Loss: 0.03935, Norm Loss: 1.03081, F0 Loss: 2.43827, LM Loss: 2.01154, Gen Loss: 6.14051, Sty Loss: 0.08482, Diff Loss: 0.20981, DiscLM Loss: 0.00088, GenLM Loss: 0.99751 +INFO:2025-06-12 05:14:28,178: Epoch [19/20], Step [1250/2647], Loss: 0.50907, Disc Loss: 3.57696, Dur Loss: 0.82416, CE Loss: 0.04326, Norm Loss: 1.49849, F0 Loss: 3.37232, LM Loss: 2.23115, Gen Loss: 7.61608, Sty Loss: 0.09113, Diff Loss: 0.31933, DiscLM Loss: 0.00000, GenLM Loss: 1.00452 +INFO:2025-06-12 05:16:17,677: Epoch [19/20], Step [1300/2647], Loss: 0.51564, Disc Loss: 3.70919, Dur Loss: 0.90977, CE Loss: 0.04739, Norm Loss: 0.88051, F0 Loss: 3.69468, LM Loss: 2.08139, Gen Loss: 6.26380, Sty Loss: 0.08769, Diff Loss: 0.26006, DiscLM Loss: 0.00053, GenLM Loss: 1.00828 +INFO:2025-06-12 05:18:07,090: Epoch [19/20], Step [1350/2647], Loss: 0.51449, Disc Loss: 3.67164, Dur Loss: 0.93032, CE Loss: 0.04921, Norm Loss: 0.84092, F0 Loss: 2.99532, LM Loss: 2.12310, Gen Loss: 6.68929, Sty Loss: 0.09873, Diff Loss: 0.39528, DiscLM Loss: 0.00000, GenLM Loss: 0.98412 +INFO:2025-06-12 05:19:57,930: Epoch [19/20], Step [1400/2647], Loss: 0.51317, Disc Loss: 3.65416, Dur Loss: 0.76424, CE Loss: 0.04152, Norm Loss: 0.89422, F0 Loss: 2.42864, LM Loss: 2.13582, Gen Loss: 6.78393, Sty Loss: 0.11571, Diff Loss: 0.32611, DiscLM Loss: 0.00284, GenLM Loss: 0.97149 +INFO:2025-06-12 05:21:50,094: Epoch [19/20], Step [1450/2647], Loss: 0.50613, Disc Loss: 3.64219, Dur Loss: 1.05420, CE Loss: 0.05567, Norm Loss: 0.89454, F0 Loss: 2.99405, LM Loss: 2.13614, Gen Loss: 7.03268, Sty Loss: 0.12125, Diff Loss: 0.23508, DiscLM Loss: 0.00000, GenLM Loss: 0.98513 +INFO:2025-06-12 05:23:37,038: Epoch [19/20], Step [1500/2647], Loss: 0.50271, Disc Loss: 3.66766, Dur Loss: 0.91716, CE Loss: 0.04754, Norm Loss: 0.92875, F0 Loss: 2.08423, LM Loss: 2.10148, Gen Loss: 6.28831, Sty Loss: 0.08237, Diff Loss: 0.25815, DiscLM Loss: 0.01061, GenLM Loss: 0.98953 +INFO:2025-06-12 05:25:27,265: Epoch [19/20], Step [1550/2647], Loss: 0.50327, Disc Loss: 3.81558, Dur Loss: 1.04270, CE Loss: 0.05843, Norm Loss: 0.81113, F0 Loss: 3.00298, LM Loss: 2.07420, Gen Loss: 6.57390, Sty Loss: 0.09340, Diff Loss: 0.23574, DiscLM Loss: 0.00000, GenLM Loss: 0.97927 +INFO:2025-06-12 05:27:16,035: Epoch [19/20], Step [1600/2647], Loss: 0.49620, Disc Loss: 3.76359, Dur Loss: 0.88459, CE Loss: 0.04265, Norm Loss: 0.95590, F0 Loss: 2.99161, LM Loss: 2.10955, Gen Loss: 7.36194, Sty Loss: 0.10931, Diff Loss: 0.24517, DiscLM Loss: 0.00269, GenLM Loss: 1.02237 +INFO:2025-06-12 05:29:04,916: Epoch [19/20], Step [1650/2647], Loss: 0.50022, Disc Loss: 3.76011, Dur Loss: 1.44520, CE Loss: 0.09217, Norm Loss: 1.18175, F0 Loss: 3.54837, LM Loss: 2.22538, Gen Loss: 6.58915, Sty Loss: 0.09580, Diff Loss: 0.33415, DiscLM Loss: 0.00000, GenLM Loss: 0.98322 +INFO:2025-06-12 05:30:54,073: Epoch [19/20], Step [1700/2647], Loss: 0.49958, Disc Loss: 3.75882, Dur Loss: 0.76462, CE Loss: 0.03928, Norm Loss: 0.78701, F0 Loss: 2.80466, LM Loss: 2.09034, Gen Loss: 6.65010, Sty Loss: 0.09664, Diff Loss: 0.28022, DiscLM Loss: 0.02079, GenLM Loss: 0.95081 +INFO:2025-06-12 05:32:45,165: Epoch [19/20], Step [1750/2647], Loss: 0.49948, Disc Loss: 3.78084, Dur Loss: 0.83411, CE Loss: 0.04506, Norm Loss: 0.78271, F0 Loss: 2.79323, LM Loss: 2.07920, Gen Loss: 6.92164, Sty Loss: 0.10312, Diff Loss: 0.28540, DiscLM Loss: 0.00000, GenLM Loss: 1.00718 +INFO:2025-06-12 05:34:33,617: Epoch [19/20], Step [1800/2647], Loss: 0.51084, Disc Loss: 3.63266, Dur Loss: 0.82999, CE Loss: 0.04258, Norm Loss: 1.08301, F0 Loss: 3.83270, LM Loss: 2.15167, Gen Loss: 7.19994, Sty Loss: 0.12075, Diff Loss: 0.31965, DiscLM Loss: 0.00628, GenLM Loss: 0.96012 +INFO:2025-06-12 05:36:20,562: Epoch [19/20], Step [1850/2647], Loss: 0.50486, Disc Loss: 3.64200, Dur Loss: 0.83043, CE Loss: 0.04233, Norm Loss: 0.93835, F0 Loss: 3.23190, LM Loss: 2.21742, Gen Loss: 7.10971, Sty Loss: 0.10756, Diff Loss: 0.25002, DiscLM Loss: 0.00000, GenLM Loss: 1.01225 +INFO:2025-06-12 05:38:07,883: Epoch [19/20], Step [1900/2647], Loss: 0.51196, Disc Loss: 3.66775, Dur Loss: 1.13661, CE Loss: 0.06407, Norm Loss: 1.00129, F0 Loss: 2.20360, LM Loss: 2.09540, Gen Loss: 7.44551, Sty Loss: 0.09411, Diff Loss: 0.35035, DiscLM Loss: 0.00282, GenLM Loss: 1.01635 +INFO:2025-06-12 05:39:56,479: Epoch [19/20], Step [1950/2647], Loss: 0.50931, Disc Loss: 3.78210, Dur Loss: 0.83807, CE Loss: 0.05443, Norm Loss: 0.97104, F0 Loss: 2.75234, LM Loss: 2.21861, Gen Loss: 6.64119, Sty Loss: 0.13202, Diff Loss: 0.28387, DiscLM Loss: 0.00000, GenLM Loss: 0.97725 +INFO:2025-06-12 05:41:45,358: Epoch [19/20], Step [2000/2647], Loss: 0.51458, Disc Loss: 3.67713, Dur Loss: 0.77982, CE Loss: 0.03984, Norm Loss: 1.00289, F0 Loss: 3.18925, LM Loss: 2.14923, Gen Loss: 6.20682, Sty Loss: 0.10677, Diff Loss: 0.21274, DiscLM Loss: 0.00113, GenLM Loss: 1.00379 +INFO:2025-06-12 05:43:33,758: Epoch [19/20], Step [2050/2647], Loss: 0.50509, Disc Loss: 3.75929, Dur Loss: 0.92393, CE Loss: 0.04403, Norm Loss: 0.80717, F0 Loss: 2.61615, LM Loss: 2.11093, Gen Loss: 6.84956, Sty Loss: 0.11880, Diff Loss: 0.22945, DiscLM Loss: 0.00000, GenLM Loss: 0.99432 +INFO:2025-06-12 05:45:19,895: Epoch [19/20], Step [2100/2647], Loss: 0.51824, Disc Loss: 3.80085, Dur Loss: 0.95039, CE Loss: 0.04665, Norm Loss: 0.99226, F0 Loss: 2.46738, LM Loss: 2.05780, Gen Loss: 5.74698, Sty Loss: 0.09029, Diff Loss: 0.30872, DiscLM Loss: 0.00092, GenLM Loss: 1.00431 +INFO:2025-06-12 05:47:09,542: Epoch [19/20], Step [2150/2647], Loss: 0.50464, Disc Loss: 3.66507, Dur Loss: 0.92305, CE Loss: 0.04603, Norm Loss: 1.00714, F0 Loss: 3.00874, LM Loss: 2.17680, Gen Loss: 6.79022, Sty Loss: 0.14959, Diff Loss: 0.35833, DiscLM Loss: 0.00000, GenLM Loss: 0.99537 +INFO:2025-06-12 05:48:59,840: Epoch [19/20], Step [2200/2647], Loss: 0.51614, Disc Loss: 3.75734, Dur Loss: 1.31694, CE Loss: 0.07078, Norm Loss: 1.24941, F0 Loss: 2.65531, LM Loss: 2.12550, Gen Loss: 6.14005, Sty Loss: 0.10833, Diff Loss: 0.19794, DiscLM Loss: 0.00278, GenLM Loss: 0.99909 +INFO:2025-06-12 05:50:50,231: Epoch [19/20], Step [2250/2647], Loss: 0.51140, Disc Loss: 3.82300, Dur Loss: 0.82524, CE Loss: 0.04081, Norm Loss: 0.90553, F0 Loss: 2.60019, LM Loss: 2.10201, Gen Loss: 6.18505, Sty Loss: 0.09450, Diff Loss: 0.24860, DiscLM Loss: 0.00000, GenLM Loss: 1.00049 +INFO:2025-06-12 05:52:40,709: Epoch [19/20], Step [2300/2647], Loss: 0.49549, Disc Loss: 3.72810, Dur Loss: 1.32065, CE Loss: 0.06369, Norm Loss: 1.39041, F0 Loss: 3.26597, LM Loss: 2.22050, Gen Loss: 5.71707, Sty Loss: 0.09776, Diff Loss: 0.22130, DiscLM Loss: 0.00073, GenLM Loss: 1.00234 +INFO:2025-06-12 05:54:29,535: Epoch [19/20], Step [2350/2647], Loss: 0.51247, Disc Loss: 3.73386, Dur Loss: 0.90536, CE Loss: 0.04797, Norm Loss: 0.89924, F0 Loss: 3.51529, LM Loss: 2.10408, Gen Loss: 6.23703, Sty Loss: 0.10203, Diff Loss: 0.25661, DiscLM Loss: 0.00000, GenLM Loss: 1.00025 +INFO:2025-06-12 05:56:18,657: Epoch [19/20], Step [2400/2647], Loss: 0.50158, Disc Loss: 3.71941, Dur Loss: 0.97890, CE Loss: 0.05357, Norm Loss: 0.93758, F0 Loss: 2.47247, LM Loss: 2.10605, Gen Loss: 6.74272, Sty Loss: 0.10349, Diff Loss: 0.29721, DiscLM Loss: 0.00070, GenLM Loss: 1.00173 +INFO:2025-06-12 05:58:08,086: Epoch [19/20], Step [2450/2647], Loss: 0.51719, Disc Loss: 3.78464, Dur Loss: 0.99051, CE Loss: 0.05558, Norm Loss: 1.12047, F0 Loss: 2.54311, LM Loss: 2.10177, Gen Loss: 5.85521, Sty Loss: 0.10293, Diff Loss: 0.25143, DiscLM Loss: 0.00000, GenLM Loss: 0.99002 +INFO:2025-06-12 05:59:54,850: Epoch [19/20], Step [2500/2647], Loss: 0.51636, Disc Loss: 3.81918, Dur Loss: 0.98838, CE Loss: 0.05341, Norm Loss: 0.94216, F0 Loss: 2.72591, LM Loss: 2.07947, Gen Loss: 6.02807, Sty Loss: 0.08156, Diff Loss: 0.30179, DiscLM Loss: 0.00041, GenLM Loss: 1.00454 +INFO:2025-06-12 06:01:44,249: Epoch [19/20], Step [2550/2647], Loss: 0.50680, Disc Loss: 3.75481, Dur Loss: 0.80627, CE Loss: 0.03910, Norm Loss: 0.99713, F0 Loss: 3.07970, LM Loss: 2.14512, Gen Loss: 6.37165, Sty Loss: 0.08302, Diff Loss: 0.29624, DiscLM Loss: 0.00000, GenLM Loss: 0.98649 +INFO:2025-06-12 06:03:35,350: Epoch [19/20], Step [2600/2647], Loss: 0.50553, Disc Loss: 3.62378, Dur Loss: 1.14465, CE Loss: 0.06727, Norm Loss: 1.15251, F0 Loss: 3.49035, LM Loss: 2.26790, Gen Loss: 6.91880, Sty Loss: 0.07768, Diff Loss: 0.23213, DiscLM Loss: 0.00274, GenLM Loss: 0.97967 +INFO:2025-06-12 06:07:36,615: Validation loss: 0.548, Dur loss: 0.954, F0 loss: 3.214 + + + +INFO:2025-06-12 06:09:25,190: Epoch [20/20], Step [50/2647], Loss: 0.51100, Disc Loss: 3.69412, Dur Loss: 0.88710, CE Loss: 0.04825, Norm Loss: 1.19608, F0 Loss: 2.66509, LM Loss: 2.14409, Gen Loss: 6.51191, Sty Loss: 0.08929, Diff Loss: 0.20616, DiscLM Loss: 0.00000, GenLM Loss: 0.99598 +INFO:2025-06-12 06:11:17,071: Epoch [20/20], Step [100/2647], Loss: 0.50396, Disc Loss: 3.71646, Dur Loss: 1.05318, CE Loss: 0.05925, Norm Loss: 1.21085, F0 Loss: 3.68178, LM Loss: 2.21367, Gen Loss: 6.62014, Sty Loss: 0.13286, Diff Loss: 0.20307, DiscLM Loss: 0.00291, GenLM Loss: 0.99233 +INFO:2025-06-12 06:13:07,650: Epoch [20/20], Step [150/2647], Loss: 0.51142, Disc Loss: 3.73300, Dur Loss: 0.80373, CE Loss: 0.04286, Norm Loss: 1.02170, F0 Loss: 2.93273, LM Loss: 2.11108, Gen Loss: 6.36150, Sty Loss: 0.09820, Diff Loss: 0.27615, DiscLM Loss: 0.00000, GenLM Loss: 1.00087 +INFO:2025-06-12 06:14:55,135: Epoch [20/20], Step [200/2647], Loss: 0.50611, Disc Loss: 3.75153, Dur Loss: 0.86464, CE Loss: 0.04654, Norm Loss: 0.96699, F0 Loss: 2.35337, LM Loss: 2.08993, Gen Loss: 7.25762, Sty Loss: 0.11174, Diff Loss: 0.25037, DiscLM Loss: 0.03207, GenLM Loss: 1.00059 +INFO:2025-06-12 06:16:44,778: Epoch [20/20], Step [250/2647], Loss: 0.50598, Disc Loss: 3.69117, Dur Loss: 0.87932, CE Loss: 0.04925, Norm Loss: 0.77184, F0 Loss: 2.63508, LM Loss: 2.07066, Gen Loss: 6.80729, Sty Loss: 0.08526, Diff Loss: 0.25443, DiscLM Loss: 0.00000, GenLM Loss: 1.01245 +INFO:2025-06-12 06:18:36,631: Epoch [20/20], Step [300/2647], Loss: 0.50768, Disc Loss: 3.73941, Dur Loss: 1.03968, CE Loss: 0.05163, Norm Loss: 1.15251, F0 Loss: 3.06194, LM Loss: 2.16587, Gen Loss: 6.71211, Sty Loss: 0.14126, Diff Loss: 0.32067, DiscLM Loss: 0.00273, GenLM Loss: 1.03317 +INFO:2025-06-12 06:20:27,292: Epoch [20/20], Step [350/2647], Loss: 0.50791, Disc Loss: 3.84679, Dur Loss: 0.93882, CE Loss: 0.05069, Norm Loss: 0.82173, F0 Loss: 3.04742, LM Loss: 2.34798, Gen Loss: 6.07823, Sty Loss: 0.11951, Diff Loss: 0.30375, DiscLM Loss: 0.00000, GenLM Loss: 0.95308 +INFO:2025-06-12 06:22:15,063: Epoch [20/20], Step [400/2647], Loss: 0.50914, Disc Loss: 3.68738, Dur Loss: 1.00036, CE Loss: 0.05502, Norm Loss: 0.97910, F0 Loss: 2.74178, LM Loss: 2.10094, Gen Loss: 7.21696, Sty Loss: 0.33048, Diff Loss: 0.54955, DiscLM Loss: 0.00244, GenLM Loss: 1.01119 +INFO:2025-06-12 06:24:06,214: Epoch [20/20], Step [450/2647], Loss: 0.51272, Disc Loss: 3.63793, Dur Loss: 1.05514, CE Loss: 0.05728, Norm Loss: 1.05818, F0 Loss: 2.73820, LM Loss: 2.06942, Gen Loss: 7.01114, Sty Loss: 0.13724, Diff Loss: 0.22606, DiscLM Loss: 0.00000, GenLM Loss: 0.98916 +INFO:2025-06-12 06:25:55,325: Epoch [20/20], Step [500/2647], Loss: 0.49993, Disc Loss: 3.81773, Dur Loss: 0.86260, CE Loss: 0.04397, Norm Loss: 0.94658, F0 Loss: 2.46833, LM Loss: 2.10175, Gen Loss: 6.01505, Sty Loss: 0.10955, Diff Loss: 0.26021, DiscLM Loss: 0.00123, GenLM Loss: 0.96341 +INFO:2025-06-12 06:27:43,185: Epoch [20/20], Step [550/2647], Loss: 0.50667, Disc Loss: 3.75728, Dur Loss: 0.80518, CE Loss: 0.04305, Norm Loss: 0.96605, F0 Loss: 2.88191, LM Loss: 2.15448, Gen Loss: 5.65782, Sty Loss: 0.09615, Diff Loss: 0.17058, DiscLM Loss: 0.00000, GenLM Loss: 1.02002 +INFO:2025-06-12 06:29:37,571: Epoch [20/20], Step [600/2647], Loss: 0.50630, Disc Loss: 3.84368, Dur Loss: 0.96221, CE Loss: 0.04988, Norm Loss: 0.91839, F0 Loss: 3.38176, LM Loss: 2.14046, Gen Loss: 6.06963, Sty Loss: 0.10012, Diff Loss: 0.18014, DiscLM Loss: 0.00426, GenLM Loss: 0.98988 +INFO:2025-06-12 06:31:25,954: Epoch [20/20], Step [650/2647], Loss: 0.50208, Disc Loss: 3.71256, Dur Loss: 0.80499, CE Loss: 0.04297, Norm Loss: 0.82453, F0 Loss: 2.62942, LM Loss: 2.10879, Gen Loss: 6.30727, Sty Loss: 0.09745, Diff Loss: 0.34296, DiscLM Loss: 0.00000, GenLM Loss: 0.99007 diff --git a/train_second.py b/train_second.py index fb1048dcffa1aada31f9918d426928abdfd4a21a..153ba9e471686cecaba902e1b614f1633e67dff6 100644 --- a/train_second.py +++ b/train_second.py @@ -31,6 +31,19 @@ from Modules.diffusion.sampler import DiffusionSampler, ADPM2Sampler, KarrasSche from optimizers import build_optimizer +def clip_to_bert(texts, mask, max_len: int = 510): + """ + Hard-clip batch to ≤ max_len tokens and return + (texts_clipped, **fresh full-width mask**, new_lengths). + """ + if texts.size(1) > max_len: + texts = texts[:, :max_len] + lengths = (texts != 0).sum(dim=1) # PAD id = 0 + seq_len = texts.size(1) + mask = torch.arange(seq_len, device=texts.device).unsqueeze(0) >= \ + lengths.unsqueeze(1) # shape [B, seq_len] + return texts, mask, lengths + # simple fix for dataparallel that allows access to class attributes class MyDataParallel(torch.nn.DataParallel): def __getattr__(self, name): @@ -262,10 +275,21 @@ def main(config_path): batch = [b.to(device) for b in batch[1:]] texts, input_lengths, ref_texts, ref_lengths, mels, mel_input_length, ref_mels = batch + # --------------- CLIP TEXTS *ONCE* ----------------- + text_mask = length_to_mask(input_lengths).to(texts.device) + texts, text_mask, input_lengths = clip_to_bert(texts, text_mask) + # ── drop rows that became all-PAD after clipping ─────────── + keep = (input_lengths > 0).nonzero(as_tuple=True)[0] + if keep.numel() != texts.size(0): + texts, text_mask, input_lengths = texts[keep], text_mask[keep], input_lengths[keep] + ref_texts, ref_lengths = ref_texts[keep], ref_lengths[keep] + mels, mel_input_length, ref_mels = mels[keep], mel_input_length[keep], ref_mels[keep] + waves = [waves[i] for i in keep.tolist()] + # ---------------------------------------------------- + with torch.no_grad(): mask = length_to_mask(mel_input_length // (2 ** n_down)).to(device) mel_mask = length_to_mask(mel_input_length).to(device) - text_mask = length_to_mask(input_lengths).to(texts.device) try: _, _, s2s_attn = model.text_aligner(mels, mask, texts) @@ -306,6 +330,46 @@ def main(config_path): gs = torch.stack(gs).squeeze() # global acoustic styles s_trg = torch.cat([gs, s_dur], dim=-1).detach() # ground truth for denoiser + # texts, input_lengths, ref_texts, ref_lengths, mels, mel_input_length, ref_mels = batch + + # # ────── PATCH: keep PL-BERT below 512 tokens ───────── + # MAX_BERT_LEN = 510 # leave room for [CLS] and [SEP] + # if texts.size(1) > MAX_BERT_LEN: # truncate batch-wise + # texts = texts[:, :MAX_BERT_LEN] + # seq_len = texts.size(1) # current padded width + # input_lengths = (texts != 0).sum(1) # 0 is PAD + # arange_row = torch.arange(seq_len, device=texts.device) # shape [L] + # text_mask = arange_row.unsqueeze(0) >= input_lengths.unsqueeze(1) + # # shape [B, L] + + # # keep only rows that still have at least one real token + # keep = (input_lengths > 0).nonzero(as_tuple=True)[0] + # if keep.numel() != texts.size(0): # a row was truncated to length 0 + # texts, text_mask, input_lengths = texts[keep], text_mask[keep], input_lengths[keep] + # ref_texts, ref_lengths = ref_texts[keep], ref_lengths[keep] + # mels, mel_input_length, ref_mels = mels[keep], mel_input_length[keep], ref_mels[keep] + # waves = [waves[i] for i in keep.tolist()] + + # # clip alignments to the *current* width (seq_len) + # s2s_attn_mono = s2s_attn_mono[:, :seq_len, :] + # d_gt = d_gt[:, :seq_len] + # # ───────────────────────────────────────────────────── + + # ------------------------------------------------------------- + # Now build *everything* that depends on token count + with torch.no_grad(): + t_en = model.text_encoder(texts, input_lengths, text_mask) + + _, _, s2s_attn = model.text_aligner(mels, mask, texts) + s2s_attn = s2s_attn.transpose(-1, -2)[..., 1:].transpose(-1, -2) + + mask_ST = mask_from_lens(s2s_attn, input_lengths, + mel_input_length // 2**n_down) + s2s_attn_mono = maximum_path(s2s_attn, mask_ST) + + asr = t_en @ s2s_attn_mono + d_gt = s2s_attn_mono.sum(dim=-1) + bert_dur = model.bert(texts, attention_mask=(~text_mask).int()) d_en = model.bert_encoder(bert_dur).transpose(-1, -2) @@ -477,6 +541,10 @@ def main(config_path): if use_ind: ref_lengths = input_lengths ref_texts = texts + + # ---- clip reference text exactly the same way ---- + ref_mask = length_to_mask(ref_lengths).to(ref_texts.device) + ref_texts, ref_mask, ref_lengths = clip_to_bert(ref_texts, ref_mask) slm_out = slmadv(i, y_rec_gt, @@ -575,23 +643,42 @@ def main(config_path): waves = batch[0] batch = [b.to(device) for b in batch[1:]] texts, input_lengths, ref_texts, ref_lengths, mels, mel_input_length, ref_mels = batch + + texts, text_mask, input_lengths = clip_to_bert(texts, text_mask) + keep = (input_lengths > 0).nonzero(as_tuple=True)[0] + if keep.numel() != texts.size(0): + texts, text_mask, input_lengths = texts[keep], text_mask[keep], input_lengths[keep] + ref_texts, ref_lengths = ref_texts[keep], ref_lengths[keep] + mels, mel_input_length, ref_mels = mels[keep], mel_input_length[keep], ref_mels[keep] + waves = [waves[i] for i in keep.tolist()] + with torch.no_grad(): - mask = length_to_mask(mel_input_length // (2 ** n_down)).to('cuda') - text_mask = length_to_mask(input_lengths).to(texts.device) + mask = length_to_mask(mel_input_length // (2 ** n_down)).to(texts.device) + # mask = length_to_mask(mel_input_length // (2 ** n_down)).to('cuda') - _, _, s2s_attn = model.text_aligner(mels, mask, texts) - s2s_attn = s2s_attn.transpose(-1, -2) - s2s_attn = s2s_attn[..., 1:] - s2s_attn = s2s_attn.transpose(-1, -2) + # _, _, s2s_attn = model.text_aligner(mels, mask, texts) + # s2s_attn = s2s_attn.transpose(-1, -2) + # s2s_attn = s2s_attn[..., 1:] + # s2s_attn = s2s_attn.transpose(-1, -2) + + # mask_ST = mask_from_lens(s2s_attn, input_lengths, mel_input_length // (2 ** n_down)) + # s2s_attn_mono = maximum_path(s2s_attn, mask_ST) - mask_ST = mask_from_lens(s2s_attn, input_lengths, mel_input_length // (2 ** n_down)) + # # encode + # t_en = model.text_encoder(texts, input_lengths, text_mask) + # asr = (t_en @ s2s_attn_mono) + + # d_gt = s2s_attn_mono.sum(axis=-1).detach() + + _, _, s2s_attn = model.text_aligner(mels, mask, texts) + s2s_attn = s2s_attn.transpose(-1, -2)[..., 1:].transpose(-1, -2) + mask_ST = mask_from_lens(s2s_attn, input_lengths, + mel_input_length // 2 ** n_down) s2s_attn_mono = maximum_path(s2s_attn, mask_ST) - # encode t_en = model.text_encoder(texts, input_lengths, text_mask) - asr = (t_en @ s2s_attn_mono) - - d_gt = s2s_attn_mono.sum(axis=-1).detach() + asr = t_en @ s2s_attn_mono + d_gt = s2s_attn_mono.sum(dim=-1).detach() ss = [] gs = [] @@ -789,4 +876,4 @@ def main(config_path): yaml.dump(config, outfile, default_flow_style=True) if __name__=="__main__": - main() + main() \ No newline at end of file