num_sent_processes: 4 [0] Creating shuffled datasets in /scratch/gz5hp/tfbs_region2vec_models/expr_universe_tile25k/50W10D-0.1000r/shuffled_datasets [1] Creating shuffled datasets in /scratch/gz5hp/tfbs_region2vec_models/expr_universe_tile25k/50W10D-0.1000r/shuffled_datasets [2] Creating shuffled datasets in /scratch/gz5hp/tfbs_region2vec_models/expr_universe_tile25k/50W10D-0.1000r/shuffled_datasets [3] Creating shuffled datasets in /scratch/gz5hp/tfbs_region2vec_models/expr_universe_tile25k/50W10D-0.1000r/shuffled_datasets {'num_shuffle': 100, 'embed_dim': 10, 'context_len': 50, 'nworkers': 6, 'save_freq': -1, 'save_dir': '/scratch/gz5hp/tfbs_region2vec_models/expr_universe_tile25k/50W10D-0.1000r', 'resume': '', 'train_alg': 'cbow', 'min_count': 5, 'neg_samples': 5, 'init_lr': 0.1, 'min_lr': 0.0001, 'lr_mode': 'linear', 'milestones': [], 'hier_softmax': False, 'update_vocab': 'once', 'seed': 0} Using cbow, negative sampling with 5 negative samples [03/12/24-18:09:54] Start training [03/12/24-18:09:54] Building vocabulary [03/12/24-18:10:09] Vocabulary size is 100053 [Shuffling 1] loss 4371840.5000 lr 0.1000 vocab_size 100053 (5.93s/10.12m) [Shuffling 2] loss 3961552.2500 lr 0.0990 vocab_size 100053 (6.19s/10.34m) [Shuffling 3] loss 3957232.2500 lr 0.0980 vocab_size 100053 (6.26s/10.46m) [Shuffling 4] loss 3943659.2500 lr 0.0970 vocab_size 100053 (6.29s/10.52m) [Shuffling 5] loss 3981849.2500 lr 0.0960 vocab_size 100053 (6.32s/10.57m) [Shuffling 6] loss 3895304.5000 lr 0.0950 vocab_size 100053 (6.31s/10.61m) [Shuffling 7] loss 3906431.0000 lr 0.0940 vocab_size 100053 (6.32s/10.63m) [Shuffling 8] loss 4039683.7500 lr 0.0930 vocab_size 100053 (6.27s/10.64m) [Shuffling 9] loss 3969390.5000 lr 0.0920 vocab_size 100053 (6.25s/10.64m) [Shuffling 10] loss 3727041.2500 lr 0.0910 vocab_size 100053 (6.26s/10.65m) [Shuffling 11] loss 3979958.7500 lr 0.0900 vocab_size 100053 (6.25s/10.65m) [Shuffling 12] loss 3921231.2500 lr 0.0890 vocab_size 100053 (6.26s/10.65m) [Shuffling 13] loss 3950111.0000 lr 0.0880 vocab_size 100053 (6.29s/10.66m) [Shuffling 14] loss 3910826.5000 lr 0.0870 vocab_size 100053 (6.27s/10.66m) [Shuffling 15] loss 3953330.2500 lr 0.0860 vocab_size 100053 (6.26s/10.66m) [Shuffling 16] loss 3940325.0000 lr 0.0850 vocab_size 100053 (6.31s/10.67m) [Shuffling 17] loss 3726625.0000 lr 0.0840 vocab_size 100053 (6.35s/10.68m) [Shuffling 18] loss 3810696.0000 lr 0.0830 vocab_size 100053 (6.30s/10.68m) [Shuffling 19] loss 3940788.0000 lr 0.0820 vocab_size 100053 (6.28s/10.68m) [Shuffling 20] loss 4018777.7500 lr 0.0810 vocab_size 100053 (6.32s/10.69m) [Shuffling 21] loss 3905921.7500 lr 0.0800 vocab_size 100053 (6.27s/10.69m) [Shuffling 22] loss 3909988.7500 lr 0.0790 vocab_size 100053 (6.30s/10.69m) [Shuffling 23] loss 3949095.0000 lr 0.0780 vocab_size 100053 (6.35s/10.70m) [Shuffling 24] loss 3917714.7500 lr 0.0770 vocab_size 100053 (6.22s/10.69m) [Shuffling 25] loss 3965227.0000 lr 0.0760 vocab_size 100053 (6.32s/10.70m) [Shuffling 26] loss 3905529.5000 lr 0.0750 vocab_size 100053 (6.29s/10.70m) [Shuffling 27] loss 3770560.2500 lr 0.0740 vocab_size 100053 (6.31s/10.70m) [Shuffling 28] loss 3800279.2500 lr 0.0730 vocab_size 100053 (6.25s/10.70m) [Shuffling 29] loss 3894136.0000 lr 0.0720 vocab_size 100053 (6.36s/10.70m) [Shuffling 30] loss 3788736.7500 lr 0.0710 vocab_size 100053 (6.30s/10.71m) [Shuffling 31] loss 3921935.0000 lr 0.0700 vocab_size 100053 (6.31s/10.71m) [Shuffling 32] loss 3972174.5000 lr 0.0690 vocab_size 100053 (6.28s/10.71m) [Shuffling 33] loss 3806870.2500 lr 0.0680 vocab_size 100053 (6.28s/10.71m) [Shuffling 34] loss 3890075.2500 lr 0.0670 vocab_size 100053 (6.35s/10.71m) [Shuffling 35] loss 3836115.5000 lr 0.0660 vocab_size 100053 (6.13s/10.70m) [Shuffling 36] loss 3765232.7500 lr 0.0650 vocab_size 100053 (6.16s/10.70m) [Shuffling 37] loss 3858733.7500 lr 0.0640 vocab_size 100053 (6.12s/10.69m) [Shuffling 38] loss 3845542.2500 lr 0.0630 vocab_size 100053 (6.16s/10.69m) [Shuffling 39] loss 3886882.5000 lr 0.0620 vocab_size 100053 (6.10s/10.68m) [Shuffling 40] loss 3702250.7500 lr 0.0610 vocab_size 100053 (6.17s/10.68m) [Shuffling 41] loss 3796411.5000 lr 0.0600 vocab_size 100053 (6.22s/10.67m) [Shuffling 42] loss 3908049.5000 lr 0.0590 vocab_size 100053 (6.29s/10.68m) [Shuffling 43] loss 3843461.5000 lr 0.0580 vocab_size 100053 (6.26s/10.68m) [Shuffling 44] loss 3802544.7500 lr 0.0570 vocab_size 100053 (6.26s/10.68m) [Shuffling 45] loss 3833951.5000 lr 0.0560 vocab_size 100053 (6.27s/10.68m) [Shuffling 46] loss 3768576.2500 lr 0.0550 vocab_size 100053 (6.27s/10.68m) [Shuffling 47] loss 3951148.2500 lr 0.0540 vocab_size 100053 (6.31s/10.68m) [Shuffling 48] loss 3791214.5000 lr 0.0530 vocab_size 100053 (6.34s/10.68m) [Shuffling 49] loss 3925194.7500 lr 0.0520 vocab_size 100053 (6.30s/10.68m) [Shuffling 50] loss 3797037.0000 lr 0.0510 vocab_size 100053 (6.28s/10.68m) [Shuffling 51] loss 3782341.2500 lr 0.0500 vocab_size 100053 (6.33s/10.69m) [Shuffling 52] loss 3790967.7500 lr 0.0491 vocab_size 100053 (6.24s/10.69m) [Shuffling 53] loss 3792357.7500 lr 0.0481 vocab_size 100053 (6.31s/10.69m) [Shuffling 54] loss 3804573.2500 lr 0.0471 vocab_size 100053 (6.33s/10.69m) [Shuffling 55] loss 3821758.2500 lr 0.0461 vocab_size 100053 (6.31s/10.69m) [Shuffling 56] loss 3848805.2500 lr 0.0451 vocab_size 100053 (6.24s/10.69m) [Shuffling 57] loss 3697114.7500 lr 0.0441 vocab_size 100053 (6.31s/10.69m) [Shuffling 58] loss 3789342.2500 lr 0.0431 vocab_size 100053 (6.20s/10.69m) [Shuffling 59] loss 3798432.7500 lr 0.0421 vocab_size 100053 (6.29s/10.69m) [Shuffling 60] loss 3839745.0000 lr 0.0411 vocab_size 100053 (6.23s/10.69m) [Shuffling 61] loss 3728444.7500 lr 0.0401 vocab_size 100053 (6.28s/10.69m) [Shuffling 62] loss 3783933.2500 lr 0.0391 vocab_size 100053 (6.31s/10.69m) [Shuffling 63] loss 3857920.2500 lr 0.0381 vocab_size 100053 (6.32s/10.69m) [Shuffling 64] loss 3732171.5000 lr 0.0371 vocab_size 100053 (6.18s/10.69m) [Shuffling 65] loss 3849039.7500 lr 0.0361 vocab_size 100053 (6.31s/10.69m) [Shuffling 66] loss 3661649.0000 lr 0.0351 vocab_size 100053 (6.33s/10.69m) [Shuffling 67] loss 3861618.7500 lr 0.0341 vocab_size 100053 (6.28s/10.69m) [Shuffling 68] loss 3765174.7500 lr 0.0331 vocab_size 100053 (6.33s/10.69m) [Shuffling 69] loss 3683450.2500 lr 0.0321 vocab_size 100053 (6.33s/10.69m) [Shuffling 70] loss 3649909.0000 lr 0.0311 vocab_size 100053 (6.36s/10.70m) [Shuffling 71] loss 3750059.5000 lr 0.0301 vocab_size 100053 (6.24s/10.70m) [Shuffling 72] loss 3707461.0000 lr 0.0291 vocab_size 100053 (6.32s/10.70m) [Shuffling 73] loss 3789411.7500 lr 0.0281 vocab_size 100053 (6.32s/10.70m) [Shuffling 74] loss 3797084.7500 lr 0.0271 vocab_size 100053 (6.32s/10.70m) [Shuffling 75] loss 3845826.0000 lr 0.0261 vocab_size 100053 (6.35s/10.70m) [Shuffling 76] loss 3703377.5000 lr 0.0251 vocab_size 100053 (6.28s/10.70m) [Shuffling 77] loss 3844592.2500 lr 0.0241 vocab_size 100053 (6.33s/10.70m) [Shuffling 78] loss 3804212.2500 lr 0.0231 vocab_size 100053 (6.30s/10.70m) [Shuffling 79] loss 3794253.7500 lr 0.0221 vocab_size 100053 (6.34s/10.70m) [Shuffling 80] loss 3816514.0000 lr 0.0211 vocab_size 100053 (6.29s/10.71m) [Shuffling 81] loss 3792219.0000 lr 0.0201 vocab_size 100053 (6.31s/10.71m) [Shuffling 82] loss 3690469.7500 lr 0.0191 vocab_size 100053 (6.31s/10.71m) [Shuffling 83] loss 3605237.2500 lr 0.0181 vocab_size 100053 (6.35s/10.71m) [Shuffling 84] loss 3722676.7500 lr 0.0171 vocab_size 100053 (6.29s/10.71m) [Shuffling 85] loss 3662822.5000 lr 0.0161 vocab_size 100053 (6.36s/10.71m) [Shuffling 86] loss 3657630.2500 lr 0.0151 vocab_size 100053 (6.24s/10.71m) [Shuffling 87] loss 3588744.5000 lr 0.0141 vocab_size 100053 (6.36s/10.71m) [Shuffling 88] loss 3690490.5000 lr 0.0131 vocab_size 100053 (6.26s/10.71m) [Shuffling 89] loss 3652875.7500 lr 0.0121 vocab_size 100053 (6.35s/10.71m) [Shuffling 90] loss 3641825.2500 lr 0.0111 vocab_size 100053 (6.16s/10.71m) [Shuffling 91] loss 3731320.2500 lr 0.0101 vocab_size 100053 (6.17s/10.71m) [Shuffling 92] loss 3700547.0000 lr 0.0091 vocab_size 100053 (6.32s/10.71m) [Shuffling 93] loss 3721811.2500 lr 0.0081 vocab_size 100053 (6.15s/10.71m) [Shuffling 94] loss 3722408.0000 lr 0.0071 vocab_size 100053 (6.21s/10.70m) [Shuffling 95] loss 3604954.5000 lr 0.0061 vocab_size 100053 (12.24s/10.81m) [Shuffling 96] loss 3559969.2500 lr 0.0051 vocab_size 100053 (6.30s/10.81m) [Shuffling 97] loss 3694484.7500 lr 0.0041 vocab_size 100053 (10.19s/10.87m) [Shuffling 98] loss 3724167.0000 lr 0.0031 vocab_size 100053 (6.07s/10.87m) [Shuffling 99] loss 3444920.7500 lr 0.0021 vocab_size 100053 (18.23s/11.07m) [Shuffling 100] loss 3696154.7500 lr 0.0011 vocab_size 100053 (17.15s/11.25m) [03/12/24-18:21:09] Training finished, training Time 11.25m