[2026-07-16 13:14:15] step 10050/1000000 | loss=3.8739 vocab=1.7239 token=2.3269 | 5.65 steps/s | VRAM=27.40 Go [2026-07-16 13:14:23] step 10100/1000000 | loss=3.8681 vocab=1.9317 token=1.9100 | 6.18 steps/s | VRAM=29.23 Go [2026-07-16 13:14:31] step 10150/1000000 | loss=3.8334 vocab=1.7620 token=1.2775 | 6.21 steps/s | VRAM=24.91 Go [2026-07-16 13:14:39] step 10200/1000000 | loss=3.8951 vocab=1.2236 token=2.1307 | 6.19 steps/s | VRAM=28.47 Go [2026-07-16 13:14:47] step 10250/1000000 | loss=3.9282 vocab=1.7882 token=2.0081 | 6.19 steps/s | VRAM=24.69 Go [2026-07-16 13:14:55] step 10300/1000000 | loss=3.8778 vocab=1.6420 token=2.1973 | 6.15 steps/s | VRAM=25.94 Go [2026-07-16 13:15:03] step 10350/1000000 | loss=4.0216 vocab=2.6756 token=2.3557 | 6.14 steps/s | VRAM=28.54 Go [2026-07-16 13:15:11] step 10400/1000000 | loss=3.8871 vocab=1.6342 token=1.8834 | 6.13 steps/s | VRAM=28.34 Go [2026-07-16 13:15:19] step 10450/1000000 | loss=3.8505 vocab=1.6026 token=1.8678 | 6.11 steps/s | VRAM=31.79 Go [2026-07-16 13:15:28] step 10500/1000000 | loss=3.8760 vocab=1.7592 token=1.8796 | 6.09 steps/s | VRAM=28.54 Go [2026-07-16 13:15:36] step 10550/1000000 | loss=3.8236 vocab=1.8689 token=2.1212 | 6.13 steps/s | VRAM=31.96 Go [2026-07-16 13:15:44] step 10600/1000000 | loss=3.9297 vocab=1.6670 token=1.7601 | 6.08 steps/s | VRAM=32.91 Go [2026-07-16 13:15:52] step 10650/1000000 | loss=3.8253 vocab=2.0230 token=2.1395 | 6.12 steps/s | VRAM=31.55 Go [2026-07-16 13:16:00] step 10700/1000000 | loss=3.7516 vocab=2.0719 token=2.1747 | 6.11 steps/s | VRAM=28.51 Go [2026-07-16 13:16:09] step 10750/1000000 | loss=3.8211 vocab=2.1703 token=2.1146 | 6.07 steps/s | VRAM=26.54 Go [2026-07-16 13:16:17] step 10800/1000000 | loss=3.7241 vocab=1.6038 token=1.7708 | 6.05 steps/s | VRAM=25.87 Go [2026-07-16 13:16:25] step 10850/1000000 | loss=3.7811 vocab=1.5625 token=1.9209 | 6.04 steps/s | VRAM=27.26 Go [2026-07-16 13:16:33] step 10900/1000000 | loss=3.7679 vocab=1.6722 token=2.1571 | 6.07 steps/s | VRAM=28.91 Go [2026-07-16 13:16:42] step 10950/1000000 | loss=3.8344 vocab=2.0745 token=2.1119 | 6.09 steps/s | VRAM=32.40 Go [2026-07-16 13:16:50] step 11000/1000000 | loss=3.7989 vocab=1.6713 token=1.9058 | 6.06 steps/s | VRAM=26.17 Go [2026-07-16 13:16:58] step 11050/1000000 | loss=3.7291 vocab=1.7958 token=1.9928 | 6.07 steps/s | VRAM=27.89 Go [2026-07-16 13:17:06] step 11100/1000000 | loss=3.8166 vocab=1.9557 token=1.9915 | 6.03 steps/s | VRAM=26.30 Go [2026-07-16 13:17:15] step 11150/1000000 | loss=3.7473 vocab=1.1555 token=1.8646 | 6.08 steps/s | VRAM=26.33 Go [2026-07-16 13:17:23] step 11200/1000000 | loss=3.7918 vocab=1.6894 token=2.1777 | 6.05 steps/s | VRAM=28.49 Go [2026-07-16 13:17:31] step 11250/1000000 | loss=3.7798 vocab=1.4382 token=2.2295 | 6.04 steps/s | VRAM=25.94 Go [2026-07-16 13:17:39] step 11300/1000000 | loss=3.7311 vocab=2.0292 token=1.8976 | 6.04 steps/s | VRAM=32.24 Go [2026-07-16 13:17:48] step 11350/1000000 | loss=3.7066 vocab=1.6495 token=1.9232 | 6.05 steps/s | VRAM=27.98 Go [2026-07-16 13:17:56] step 11400/1000000 | loss=3.6984 vocab=1.3906 token=1.8075 | 6.03 steps/s | VRAM=29.43 Go [2026-07-16 13:18:04] step 11450/1000000 | loss=3.7307 vocab=1.3202 token=2.0178 | 6.05 steps/s | VRAM=26.65 Go [2026-07-16 13:18:13] step 11500/1000000 | loss=3.8339 vocab=2.3844 token=1.9243 | 6.08 steps/s | VRAM=31.42 Go [2026-07-16 13:18:21] step 11550/1000000 | loss=3.6135 vocab=1.1015 token=1.9431 | 6.05 steps/s | VRAM=27.56 Go [2026-07-16 13:18:29] step 11600/1000000 | loss=3.7081 vocab=1.5618 token=2.2817 | 6.07 steps/s | VRAM=31.48 Go [2026-07-16 13:18:37] step 11650/1000000 | loss=3.7265 vocab=2.3474 token=2.1909 | 6.05 steps/s | VRAM=32.11 Go [2026-07-16 13:18:46] step 11700/1000000 | loss=3.7309 vocab=1.4281 token=2.0436 | 6.01 steps/s | VRAM=28.20 Go [2026-07-16 13:18:54] step 11750/1000000 | loss=3.7741 vocab=1.6148 token=1.7744 | 6.06 steps/s | VRAM=28.59 Go [2026-07-16 13:19:02] step 11800/1000000 | loss=3.6341 vocab=1.4528 token=2.1329 | 6.07 steps/s | VRAM=32.54 Go [2026-07-16 13:19:10] step 11850/1000000 | loss=3.6523 vocab=1.1635 token=2.1298 | 6.05 steps/s | VRAM=28.45 Go [2026-07-16 13:19:19] step 11900/1000000 | loss=3.6213 vocab=1.7239 token=1.5995 | 6.05 steps/s | VRAM=28.29 Go [2026-07-16 13:19:27] step 11950/1000000 | loss=3.5891 vocab=1.0419 token=1.7489 | 6.05 steps/s | VRAM=30.86 Go [2026-07-16 13:19:35] step 12000/1000000 | loss=3.7246 vocab=1.5289 token=1.8219 | 6.07 steps/s | VRAM=25.27 Go [2026-07-16 13:19:43] step 12050/1000000 | loss=3.6504 vocab=1.7099 token=1.5289 | 6.03 steps/s | VRAM=29.12 Go [2026-07-16 13:19:52] step 12100/1000000 | loss=3.6548 vocab=2.0487 token=2.2412 | 6.08 steps/s | VRAM=25.56 Go [2026-07-16 13:20:00] step 12150/1000000 | loss=3.7575 vocab=1.3653 token=1.8482 | 6.04 steps/s | VRAM=26.20 Go [2026-07-16 13:20:08] step 12200/1000000 | loss=3.6268 vocab=1.4738 token=1.8701 | 6.07 steps/s | VRAM=25.31 Go [2026-07-16 13:20:16] step 12250/1000000 | loss=3.7263 vocab=1.2757 token=2.1493 | 6.04 steps/s | VRAM=26.46 Go [2026-07-16 13:20:25] step 12300/1000000 | loss=3.6852 vocab=2.0068 token=1.7898 | 6.05 steps/s | VRAM=26.05 Go [2026-07-16 13:20:33] step 12350/1000000 | loss=3.6027 vocab=2.2706 token=2.0056 | 6.05 steps/s | VRAM=26.80 Go [2026-07-16 13:20:41] step 12400/1000000 | loss=3.5982 vocab=1.9058 token=1.9791 | 6.06 steps/s | VRAM=25.78 Go [2026-07-16 13:20:49] step 12450/1000000 | loss=3.4910 vocab=1.2847 token=1.8553 | 6.09 steps/s | VRAM=29.29 Go [2026-07-16 13:20:58] step 12500/1000000 | loss=3.6706 vocab=1.7749 token=2.0719 | 6.09 steps/s | VRAM=31.35 Go [2026-07-16 13:21:06] step 12550/1000000 | loss=3.5660 vocab=2.0661 token=1.8207 | 6.07 steps/s | VRAM=32.34 Go [2026-07-16 13:21:14] step 12600/1000000 | loss=3.5731 vocab=1.4163 token=1.5744 | 6.05 steps/s | VRAM=31.69 Go [2026-07-16 13:21:22] step 12650/1000000 | loss=3.5292 vocab=1.3156 token=1.6556 | 6.05 steps/s | VRAM=27.03 Go [2026-07-16 13:21:31] step 12700/1000000 | loss=3.5980 vocab=1.6908 token=2.1863 | 6.06 steps/s | VRAM=28.24 Go [2026-07-16 13:21:39] step 12750/1000000 | loss=3.5015 vocab=2.0064 token=1.8814 | 6.08 steps/s | VRAM=25.32 Go [2026-07-16 13:21:47] step 12800/1000000 | loss=3.5159 vocab=1.6461 token=2.1806 | 6.10 steps/s | VRAM=31.50 Go [2026-07-16 13:21:55] step 12850/1000000 | loss=3.5655 vocab=1.8533 token=1.7959 | 6.06 steps/s | VRAM=27.33 Go [2026-07-16 13:22:04] step 12900/1000000 | loss=3.4564 vocab=1.1577 token=1.7414 | 6.08 steps/s | VRAM=31.33 Go [2026-07-16 13:22:12] step 12950/1000000 | loss=3.5038 vocab=1.4553 token=1.8234 | 6.03 steps/s | VRAM=26.59 Go [2026-07-16 13:22:20] step 13000/1000000 | loss=3.5412 vocab=1.7704 token=2.1277 | 6.07 steps/s | VRAM=31.63 Go [2026-07-16 13:22:28] step 13050/1000000 | loss=3.6013 vocab=1.5972 token=1.9696 | 6.06 steps/s | VRAM=28.51 Go [2026-07-16 13:22:37] step 13100/1000000 | loss=3.5102 vocab=1.5341 token=1.5370 | 6.05 steps/s | VRAM=32.01 Go [2026-07-16 13:22:45] step 13150/1000000 | loss=3.5214 vocab=1.7149 token=1.8539 | 6.06 steps/s | VRAM=28.20 Go [2026-07-16 13:22:53] step 13200/1000000 | loss=3.4757 vocab=1.1822 token=1.5933 | 6.09 steps/s | VRAM=31.05 Go [2026-07-16 13:23:01] step 13250/1000000 | loss=3.5533 vocab=1.5801 token=1.9938 | 6.05 steps/s | VRAM=32.21 Go [2026-07-16 13:23:10] step 13300/1000000 | loss=3.6043 vocab=1.3672 token=1.9593 | 6.07 steps/s | VRAM=31.39 Go [2026-07-16 13:23:18] step 13350/1000000 | loss=3.5026 vocab=1.6949 token=1.9778 | 6.03 steps/s | VRAM=25.92 Go [2026-07-16 13:23:26] step 13400/1000000 | loss=3.4411 vocab=1.6656 token=1.9399 | 6.04 steps/s | VRAM=26.40 Go [2026-07-16 13:23:34] step 13450/1000000 | loss=3.4880 vocab=1.4465 token=1.7368 | 6.07 steps/s | VRAM=28.95 Go [2026-07-16 13:23:43] step 13500/1000000 | loss=3.4967 vocab=2.1710 token=1.7897 | 6.11 steps/s | VRAM=31.47 Go [2026-07-16 13:23:51] step 13550/1000000 | loss=3.6388 vocab=1.2918 token=1.5076 | 6.05 steps/s | VRAM=29.15 Go [2026-07-16 13:23:59] step 13600/1000000 | loss=3.4772 vocab=1.8389 token=1.8019 | 6.06 steps/s | VRAM=29.84 Go [2026-07-16 13:24:07] step 13650/1000000 | loss=3.5211 vocab=1.2228 token=1.7368 | 6.07 steps/s | VRAM=25.77 Go [2026-07-16 13:24:16] step 13700/1000000 | loss=3.5237 vocab=1.8125 token=1.9287 | 6.08 steps/s | VRAM=31.60 Go [2026-07-16 13:24:24] step 13750/1000000 | loss=3.4608 vocab=1.6799 token=2.1885 | 6.05 steps/s | VRAM=28.30 Go [2026-07-16 13:24:32] step 13800/1000000 | loss=3.5302 vocab=1.2252 token=1.5423 | 6.07 steps/s | VRAM=32.72 Go [2026-07-16 13:24:40] step 13850/1000000 | loss=3.4945 vocab=1.4872 token=1.6932 | 6.08 steps/s | VRAM=28.23 Go [2026-07-16 13:24:49] step 13900/1000000 | loss=3.5163 vocab=1.2317 token=1.8020 | 6.06 steps/s | VRAM=27.86 Go [2026-07-16 13:24:57] step 13950/1000000 | loss=3.5018 vocab=1.2810 token=1.4495 | 6.06 steps/s | VRAM=32.48 Go [2026-07-16 13:25:05] step 14000/1000000 | loss=3.3706 vocab=1.5247 token=1.8302 | 6.01 steps/s | VRAM=27.06 Go [2026-07-16 13:25:13] step 14050/1000000 | loss=3.4448 vocab=1.2280 token=1.7677 | 6.07 steps/s | VRAM=29.68 Go [2026-07-16 13:25:22] step 14100/1000000 | loss=3.5477 vocab=1.5798 token=1.8449 | 6.06 steps/s | VRAM=29.71 Go [2026-07-16 13:25:30] step 14150/1000000 | loss=3.3990 vocab=1.4638 token=1.7768 | 6.06 steps/s | VRAM=28.20 Go [2026-07-16 13:25:38] step 14200/1000000 | loss=3.5193 vocab=2.0078 token=1.9614 | 6.06 steps/s | VRAM=26.48 Go [2026-07-16 13:25:46] step 14250/1000000 | loss=3.5431 vocab=1.9855 token=1.9524 | 6.06 steps/s | VRAM=31.82 Go [2026-07-16 13:25:55] step 14300/1000000 | loss=3.4478 vocab=1.5832 token=1.9095 | 6.05 steps/s | VRAM=28.51 Go [2026-07-16 13:26:03] step 14350/1000000 | loss=3.4253 vocab=1.9695 token=1.8870 | 6.08 steps/s | VRAM=24.98 Go [2026-07-16 13:26:11] step 14400/1000000 | loss=3.5020 vocab=1.6930 token=1.6191 | 6.08 steps/s | VRAM=28.87 Go [2026-07-16 13:26:19] step 14450/1000000 | loss=3.4702 vocab=1.3892 token=1.6698 | 6.02 steps/s | VRAM=29.68 Go [2026-07-16 13:26:28] step 14500/1000000 | loss=3.4968 vocab=1.5200 token=1.9702 | 6.09 steps/s | VRAM=32.77 Go [2026-07-16 13:26:36] step 14550/1000000 | loss=3.4489 vocab=2.1107 token=1.4593 | 6.06 steps/s | VRAM=26.19 Go [2026-07-16 13:26:44] step 14600/1000000 | loss=3.5091 vocab=1.8104 token=1.8207 | 6.10 steps/s | VRAM=31.84 Go [2026-07-16 13:26:52] step 14650/1000000 | loss=3.5455 vocab=1.7622 token=1.9036 | 6.06 steps/s | VRAM=31.60 Go [2026-07-16 13:27:01] step 14700/1000000 | loss=3.4475 vocab=1.4747 token=1.9026 | 6.06 steps/s | VRAM=32.66 Go [2026-07-16 13:27:09] step 14750/1000000 | loss=3.4363 vocab=1.6731 token=1.8265 | 6.06 steps/s | VRAM=31.89 Go [2026-07-16 13:27:17] step 14800/1000000 | loss=3.3514 vocab=1.5178 token=1.8822 | 6.05 steps/s | VRAM=31.88 Go [2026-07-16 13:27:25] step 14850/1000000 | loss=3.3953 vocab=1.5458 token=1.8906 | 6.06 steps/s | VRAM=28.87 Go [2026-07-16 13:27:34] step 14900/1000000 | loss=3.3562 vocab=1.6461 token=1.9860 | 6.07 steps/s | VRAM=32.68 Go [2026-07-16 13:27:42] step 14950/1000000 | loss=3.4546 vocab=1.6309 token=1.9316 | 6.06 steps/s | VRAM=28.71 Go [2026-07-16 13:27:50] step 15000/1000000 | loss=3.3929 vocab=1.5374 token=2.1415 | 6.06 steps/s | VRAM=28.52 Go [2026-07-16 13:29:25] step 15050/1000000 | loss=3.4552 vocab=2.2542 token=1.9839 | 0.52 steps/s | VRAM=29.30 Go [2026-07-16 13:29:33] step 15100/1000000 | loss=3.4755 vocab=1.5358 token=1.9330 | 6.13 steps/s | VRAM=24.85 Go [2026-07-16 13:29:42] step 15150/1000000 | loss=3.3379 vocab=2.1379 token=1.9170 | 6.14 steps/s | VRAM=32.88 Go [2026-07-16 13:29:50] step 15200/1000000 | loss=3.3781 vocab=1.2795 token=1.6919 | 6.13 steps/s | VRAM=28.79 Go [2026-07-16 13:29:58] step 15250/1000000 | loss=3.3646 vocab=1.4624 token=2.2144 | 6.12 steps/s | VRAM=30.58 Go [2026-07-16 13:30:06] step 15300/1000000 | loss=3.4935 vocab=1.6495 token=1.2547 | 6.06 steps/s | VRAM=29.27 Go [2026-07-16 13:30:14] step 15350/1000000 | loss=3.4542 vocab=1.9404 token=1.8297 | 6.13 steps/s | VRAM=24.62 Go [2026-07-16 13:30:23] step 15400/1000000 | loss=3.3142 vocab=1.4720 token=1.6610 | 6.10 steps/s | VRAM=28.05 Go [2026-07-16 13:30:31] step 15450/1000000 | loss=3.3022 vocab=1.4505 token=1.6593 | 6.07 steps/s | VRAM=28.57 Go [2026-07-16 13:30:39] step 15500/1000000 | loss=3.4021 vocab=1.4326 token=1.8644 | 6.07 steps/s | VRAM=29.40 Go [2026-07-16 13:30:47] step 15550/1000000 | loss=3.2875 vocab=1.4938 token=2.0163 | 6.04 steps/s | VRAM=32.15 Go [2026-07-16 13:30:56] step 15600/1000000 | loss=3.4500 vocab=1.5720 token=1.5954 | 6.05 steps/s | VRAM=28.67 Go [2026-07-16 13:31:04] step 15650/1000000 | loss=3.4099 vocab=1.2948 token=2.1255 | 6.04 steps/s | VRAM=26.42 Go [2026-07-16 13:31:12] step 15700/1000000 | loss=3.3689 vocab=1.7104 token=1.9756 | 6.05 steps/s | VRAM=29.56 Go [2026-07-16 13:31:20] step 15750/1000000 | loss=3.4443 vocab=1.6018 token=1.9693 | 6.09 steps/s | VRAM=31.10 Go [2026-07-16 13:31:29] step 15800/1000000 | loss=3.2869 vocab=1.1517 token=2.0784 | 6.03 steps/s | VRAM=26.36 Go [2026-07-16 13:31:37] step 15850/1000000 | loss=3.3944 vocab=1.0527 token=1.7439 | 6.08 steps/s | VRAM=32.50 Go [2026-07-16 13:31:45] step 15900/1000000 | loss=3.4531 vocab=0.8431 token=1.2911 | 6.05 steps/s | VRAM=29.10 Go [2026-07-16 13:31:53] step 15950/1000000 | loss=3.4250 vocab=2.0311 token=1.8672 | 6.06 steps/s | VRAM=29.84 Go [2026-07-16 13:32:02] step 16000/1000000 | loss=3.3940 vocab=1.3661 token=1.5335 | 6.09 steps/s | VRAM=30.71 Go [2026-07-16 13:32:10] step 16050/1000000 | loss=3.3002 vocab=1.1642 token=1.7404 | 6.10 steps/s | VRAM=31.45 Go [2026-07-16 13:32:18] step 16100/1000000 | loss=3.3362 vocab=1.0531 token=1.4247 | 6.07 steps/s | VRAM=25.09 Go [2026-07-16 13:32:26] step 16150/1000000 | loss=3.2658 vocab=1.5283 token=1.6724 | 6.04 steps/s | VRAM=29.10 Go [2026-07-16 13:32:35] step 16200/1000000 | loss=3.4098 vocab=1.3256 token=1.8395 | 6.06 steps/s | VRAM=28.93 Go [2026-07-16 13:32:43] step 16250/1000000 | loss=3.2811 vocab=1.4859 token=1.6790 | 6.06 steps/s | VRAM=29.37 Go [2026-07-16 13:32:51] step 16300/1000000 | loss=3.3518 vocab=1.8447 token=2.0079 | 6.03 steps/s | VRAM=28.32 Go [2026-07-16 13:32:59] step 16350/1000000 | loss=3.3673 vocab=1.3095 token=1.7332 | 6.05 steps/s | VRAM=32.96 Go [2026-07-16 13:33:08] step 16400/1000000 | loss=3.3287 vocab=1.1124 token=1.7779 | 6.05 steps/s | VRAM=27.12 Go [2026-07-16 13:33:16] step 16450/1000000 | loss=3.4391 vocab=1.5113 token=2.0555 | 6.08 steps/s | VRAM=32.29 Go [2026-07-16 13:33:24] step 16500/1000000 | loss=3.3754 vocab=1.3893 token=1.8250 | 6.08 steps/s | VRAM=26.32 Go [2026-07-16 13:33:32] step 16550/1000000 | loss=3.2783 vocab=1.8261 token=1.9783 | 6.07 steps/s | VRAM=32.96 Go [2026-07-16 13:33:40] step 16600/1000000 | loss=3.3606 vocab=1.8394 token=1.6781 | 6.09 steps/s | VRAM=31.23 Go [2026-07-16 13:33:49] step 16650/1000000 | loss=3.3283 vocab=1.1527 token=1.9811 | 6.08 steps/s | VRAM=29.49 Go [2026-07-16 13:33:57] step 16700/1000000 | loss=3.3256 vocab=1.5758 token=1.9610 | 6.06 steps/s | VRAM=28.81 Go [2026-07-16 13:34:05] step 16750/1000000 | loss=3.2442 vocab=1.7105 token=1.6809 | 6.06 steps/s | VRAM=28.51 Go [2026-07-16 13:34:13] step 16800/1000000 | loss=3.3656 vocab=1.0128 token=1.7118 | 6.06 steps/s | VRAM=32.39 Go [2026-07-16 13:34:22] step 16850/1000000 | loss=3.2944 vocab=1.5475 token=1.6437 | 6.06 steps/s | VRAM=29.79 Go [2026-07-16 13:34:30] step 16900/1000000 | loss=3.4581 vocab=1.7179 token=1.6690 | 6.03 steps/s | VRAM=28.74 Go [2026-07-16 13:34:38] step 16950/1000000 | loss=3.3097 vocab=1.0794 token=1.9810 | 6.05 steps/s | VRAM=30.08 Go [2026-07-16 13:34:47] step 17000/1000000 | loss=3.3165 vocab=1.4162 token=1.8582 | 6.06 steps/s | VRAM=28.83 Go [2026-07-16 13:34:55] step 17050/1000000 | loss=3.2338 vocab=2.2261 token=2.0439 | 6.07 steps/s | VRAM=29.69 Go [2026-07-16 13:35:03] step 17100/1000000 | loss=3.2212 vocab=1.1779 token=1.6505 | 6.09 steps/s | VRAM=31.28 Go [2026-07-16 13:35:11] step 17150/1000000 | loss=3.2518 vocab=1.4289 token=1.8103 | 6.07 steps/s | VRAM=26.52 Go [2026-07-16 13:35:19] step 17200/1000000 | loss=3.3390 vocab=0.9055 token=1.4841 | 6.06 steps/s | VRAM=28.45 Go [2026-07-16 13:35:28] step 17250/1000000 | loss=3.3545 vocab=2.0166 token=2.1531 | 6.06 steps/s | VRAM=32.92 Go [2026-07-16 13:35:36] step 17300/1000000 | loss=3.2847 vocab=1.5599 token=2.2152 | 6.10 steps/s | VRAM=31.78 Go [2026-07-16 13:35:44] step 17350/1000000 | loss=3.2950 vocab=1.5167 token=2.1174 | 6.06 steps/s | VRAM=29.06 Go [2026-07-16 13:35:52] step 17400/1000000 | loss=3.2346 vocab=2.0571 token=1.8462 | 6.01 steps/s | VRAM=26.38 Go [2026-07-16 13:36:01] step 17450/1000000 | loss=3.2245 vocab=1.0204 token=1.6670 | 6.08 steps/s | VRAM=28.46 Go [2026-07-16 13:36:09] step 17500/1000000 | loss=3.2726 vocab=1.3302 token=1.7296 | 6.08 steps/s | VRAM=25.57 Go [2026-07-16 13:36:17] step 17550/1000000 | loss=3.1737 vocab=1.2514 token=1.6548 | 6.09 steps/s | VRAM=28.22 Go [2026-07-16 13:36:25] step 17600/1000000 | loss=3.3230 vocab=1.4542 token=1.6756 | 6.09 steps/s | VRAM=28.21 Go [2026-07-16 13:36:34] step 17650/1000000 | loss=3.1782 vocab=1.1670 token=1.8189 | 6.06 steps/s | VRAM=30.91 Go [2026-07-16 13:36:42] step 17700/1000000 | loss=3.3949 vocab=1.5185 token=2.3296 | 6.08 steps/s | VRAM=28.70 Go [2026-07-16 13:36:50] step 17750/1000000 | loss=3.1896 vocab=1.5056 token=1.7160 | 6.07 steps/s | VRAM=28.58 Go [2026-07-16 13:36:58] step 17800/1000000 | loss=3.1635 vocab=1.4237 token=2.0252 | 6.08 steps/s | VRAM=28.07 Go [2026-07-16 13:37:07] step 17850/1000000 | loss=3.1675 vocab=1.6449 token=1.8179 | 6.06 steps/s | VRAM=25.56 Go [2026-07-16 13:37:15] step 17900/1000000 | loss=3.3352 vocab=1.4909 token=1.9586 | 6.05 steps/s | VRAM=28.99 Go [2026-07-16 13:37:23] step 17950/1000000 | loss=3.2290 vocab=1.1355 token=1.7667 | 6.08 steps/s | VRAM=25.55 Go [2026-07-16 13:37:31] step 18000/1000000 | loss=3.3302 vocab=1.6732 token=2.0212 | 6.03 steps/s | VRAM=26.82 Go [2026-07-16 13:37:40] step 18050/1000000 | loss=3.1422 vocab=1.2693 token=1.8762 | 6.10 steps/s | VRAM=28.20 Go [2026-07-16 13:37:48] step 18100/1000000 | loss=3.2668 vocab=1.9062 token=1.7278 | 6.09 steps/s | VRAM=32.11 Go [2026-07-16 13:37:56] step 18150/1000000 | loss=3.1218 vocab=1.0251 token=1.6684 | 6.03 steps/s | VRAM=26.40 Go [2026-07-16 13:38:04] step 18200/1000000 | loss=3.2562 vocab=1.2566 token=2.1910 | 6.06 steps/s | VRAM=24.83 Go [2026-07-16 13:38:12] step 18250/1000000 | loss=3.3237 vocab=1.6520 token=1.9157 | 6.08 steps/s | VRAM=31.41 Go [2026-07-16 13:38:21] step 18300/1000000 | loss=3.1278 vocab=1.2809 token=1.6524 | 6.08 steps/s | VRAM=32.69 Go [2026-07-16 13:38:29] step 18350/1000000 | loss=3.3456 vocab=1.3939 token=1.9028 | 6.06 steps/s | VRAM=26.26 Go [2026-07-16 13:38:37] step 18400/1000000 | loss=3.2774 vocab=1.3403 token=1.6992 | 6.07 steps/s | VRAM=28.52 Go [2026-07-16 13:38:45] step 18450/1000000 | loss=3.2836 vocab=1.6797 token=2.1357 | 6.06 steps/s | VRAM=29.87 Go [2026-07-16 13:38:54] step 18500/1000000 | loss=3.2557 vocab=1.2393 token=1.9233 | 6.07 steps/s | VRAM=25.09 Go [2026-07-16 13:39:02] step 18550/1000000 | loss=3.2464 vocab=1.5721 token=1.7774 | 6.10 steps/s | VRAM=32.41 Go [2026-07-16 13:39:10] step 18600/1000000 | loss=3.4094 vocab=2.1771 token=2.4234 | 6.11 steps/s | VRAM=31.19 Go [2026-07-16 13:39:18] step 18650/1000000 | loss=3.1823 vocab=1.1661 token=1.7645 | 6.09 steps/s | VRAM=28.98 Go [2026-07-16 13:39:26] step 18700/1000000 | loss=3.1534 vocab=1.3172 token=1.6165 | 6.09 steps/s | VRAM=25.81 Go [2026-07-16 13:39:35] step 18750/1000000 | loss=3.1900 vocab=1.2487 token=1.8275 | 6.07 steps/s | VRAM=29.61 Go [2026-07-16 13:39:43] step 18800/1000000 | loss=3.2933 vocab=1.0319 token=1.7330 | 6.11 steps/s | VRAM=31.09 Go [2026-07-16 13:39:51] step 18850/1000000 | loss=3.2453 vocab=1.3143 token=1.9623 | 6.06 steps/s | VRAM=26.90 Go [2026-07-16 13:39:59] step 18900/1000000 | loss=3.1575 vocab=1.4243 token=1.7282 | 6.10 steps/s | VRAM=28.45 Go [2026-07-16 13:40:08] step 18950/1000000 | loss=3.2526 vocab=1.2329 token=2.0425 | 6.11 steps/s | VRAM=26.12 Go [2026-07-16 13:40:16] step 19000/1000000 | loss=3.1556 vocab=1.3633 token=1.5742 | 6.10 steps/s | VRAM=25.19 Go [2026-07-16 13:40:24] step 19050/1000000 | loss=3.1881 vocab=0.9993 token=1.9349 | 6.08 steps/s | VRAM=32.23 Go [2026-07-16 13:40:32] step 19100/1000000 | loss=3.1010 vocab=1.1277 token=1.6325 | 6.08 steps/s | VRAM=25.90 Go [2026-07-16 13:40:40] step 19150/1000000 | loss=3.3751 vocab=0.7025 token=1.5643 | 6.08 steps/s | VRAM=28.12 Go [2026-07-16 13:40:49] step 19200/1000000 | loss=3.0960 vocab=1.1892 token=1.6000 | 6.07 steps/s | VRAM=27.94 Go [2026-07-16 13:40:57] step 19250/1000000 | loss=3.1456 vocab=1.1529 token=1.8950 | 6.06 steps/s | VRAM=25.50 Go [2026-07-16 13:41:05] step 19300/1000000 | loss=3.2634 vocab=1.5473 token=2.0265 | 6.06 steps/s | VRAM=32.76 Go [2026-07-16 13:41:13] step 19350/1000000 | loss=3.2435 vocab=1.2101 token=1.9342 | 6.08 steps/s | VRAM=27.96 Go [2026-07-16 13:41:22] step 19400/1000000 | loss=3.2396 vocab=1.0117 token=1.8880 | 6.07 steps/s | VRAM=27.14 Go [2026-07-16 13:41:30] step 19450/1000000 | loss=3.1297 vocab=1.4123 token=1.6419 | 6.04 steps/s | VRAM=29.71 Go [2026-07-16 13:41:38] step 19500/1000000 | loss=3.1811 vocab=1.3341 token=1.8990 | 6.08 steps/s | VRAM=25.88 Go [2026-07-16 13:41:46] step 19550/1000000 | loss=3.1599 vocab=1.7748 token=1.7345 | 6.08 steps/s | VRAM=25.01 Go [2026-07-16 13:41:55] step 19600/1000000 | loss=3.3285 vocab=1.9843 token=1.5353 | 6.05 steps/s | VRAM=26.69 Go [2026-07-16 13:42:03] step 19650/1000000 | loss=3.0855 vocab=1.5508 token=2.1995 | 6.09 steps/s | VRAM=31.57 Go [2026-07-16 13:42:11] step 19700/1000000 | loss=3.1239 vocab=1.2235 token=1.7781 | 6.08 steps/s | VRAM=31.90 Go [2026-07-16 13:42:19] step 19750/1000000 | loss=3.1971 vocab=1.3878 token=1.7617 | 6.07 steps/s | VRAM=27.69 Go [2026-07-16 13:42:27] step 19800/1000000 | loss=3.2265 vocab=2.0193 token=2.0357 | 6.09 steps/s | VRAM=27.78 Go [2026-07-16 13:42:36] step 19850/1000000 | loss=3.1998 vocab=1.5677 token=1.5256 | 6.08 steps/s | VRAM=25.59 Go [2026-07-16 13:42:44] step 19900/1000000 | loss=3.1083 vocab=1.5843 token=1.5123 | 6.08 steps/s | VRAM=28.95 Go [2026-07-16 13:42:52] step 19950/1000000 | loss=3.0643 vocab=1.0883 token=2.0431 | 6.12 steps/s | VRAM=28.20 Go [2026-07-16 13:43:00] step 20000/1000000 | loss=3.1945 vocab=1.5023 token=1.7259 | 6.07 steps/s | VRAM=24.69 Go [2026-07-16 13:44:26] step 20050/1000000 | loss=3.2455 vocab=1.3425 token=1.9229 | 0.58 steps/s | VRAM=29.36 Go [2026-07-16 13:44:34] step 20100/1000000 | loss=3.2311 vocab=1.3833 token=1.9783 | 6.11 steps/s | VRAM=26.50 Go [2026-07-16 13:44:43] step 20150/1000000 | loss=3.2366 vocab=1.0358 token=1.9697 | 6.12 steps/s | VRAM=29.06 Go [2026-07-16 13:44:51] step 20200/1000000 | loss=3.1946 vocab=1.5736 token=1.9477 | 6.17 steps/s | VRAM=31.08 Go [2026-07-16 13:44:59] step 20250/1000000 | loss=3.2129 vocab=1.4786 token=2.1084 | 6.11 steps/s | VRAM=28.12 Go [2026-07-16 13:45:07] step 20300/1000000 | loss=3.0973 vocab=1.7824 token=2.0538 | 6.12 steps/s | VRAM=30.40 Go [2026-07-16 13:45:15] step 20350/1000000 | loss=3.1372 vocab=0.9605 token=1.6748 | 6.10 steps/s | VRAM=28.76 Go [2026-07-16 13:45:23] step 20400/1000000 | loss=3.0375 vocab=0.9342 token=1.5815 | 6.08 steps/s | VRAM=26.25 Go [2026-07-16 13:45:32] step 20450/1000000 | loss=3.1328 vocab=1.3649 token=1.8255 | 6.10 steps/s | VRAM=32.52 Go [2026-07-16 13:45:40] step 20500/1000000 | loss=3.1008 vocab=1.2425 token=1.4504 | 6.03 steps/s | VRAM=28.68 Go [2026-07-16 13:45:48] step 20550/1000000 | loss=3.1896 vocab=1.4082 token=1.8852 | 6.08 steps/s | VRAM=29.14 Go [2026-07-16 13:45:56] step 20600/1000000 | loss=3.1231 vocab=1.4459 token=1.5875 | 6.07 steps/s | VRAM=32.44 Go [2026-07-16 13:46:05] step 20650/1000000 | loss=3.0770 vocab=1.0808 token=1.7690 | 6.05 steps/s | VRAM=25.21 Go [2026-07-16 13:46:13] step 20700/1000000 | loss=3.0980 vocab=1.3628 token=1.9442 | 6.08 steps/s | VRAM=26.35 Go [2026-07-16 13:46:21] step 20750/1000000 | loss=3.1006 vocab=1.3216 token=1.9806 | 6.08 steps/s | VRAM=25.94 Go [2026-07-16 13:46:29] step 20800/1000000 | loss=3.0984 vocab=1.2814 token=2.0613 | 6.06 steps/s | VRAM=32.44 Go [2026-07-16 13:46:37] step 20850/1000000 | loss=3.1046 vocab=1.1552 token=1.9315 | 6.11 steps/s | VRAM=32.17 Go [2026-07-16 13:46:46] step 20900/1000000 | loss=2.9796 vocab=1.4748 token=1.8308 | 6.06 steps/s | VRAM=28.69 Go [2026-07-16 13:46:54] step 20950/1000000 | loss=3.0986 vocab=1.5001 token=1.8100 | 6.09 steps/s | VRAM=31.28 Go [2026-07-16 13:47:02] step 21000/1000000 | loss=3.1481 vocab=1.3697 token=1.5630 | 6.05 steps/s | VRAM=29.50 Go [2026-07-16 13:47:10] step 21050/1000000 | loss=3.1386 vocab=1.5932 token=2.1684 | 6.04 steps/s | VRAM=26.31 Go [2026-07-16 13:47:19] step 21100/1000000 | loss=3.1584 vocab=1.0546 token=1.6270 | 6.05 steps/s | VRAM=28.86 Go [2026-07-16 13:47:27] step 21150/1000000 | loss=3.1642 vocab=1.3901 token=1.7196 | 6.06 steps/s | VRAM=29.86 Go [2026-07-16 13:47:35] step 21200/1000000 | loss=3.1914 vocab=0.5846 token=1.4916 | 6.07 steps/s | VRAM=25.33 Go [2026-07-16 13:47:43] step 21250/1000000 | loss=3.1039 vocab=1.9147 token=1.7820 | 6.08 steps/s | VRAM=29.03 Go [2026-07-16 13:47:52] step 21300/1000000 | loss=3.1006 vocab=1.7804 token=2.0599 | 6.08 steps/s | VRAM=32.51 Go [2026-07-16 13:48:00] step 21350/1000000 | loss=3.0993 vocab=1.6326 token=1.9962 | 6.06 steps/s | VRAM=32.45 Go [2026-07-16 13:48:08] step 21400/1000000 | loss=3.0598 vocab=1.1362 token=1.5103 | 6.05 steps/s | VRAM=26.71 Go [2026-07-16 13:48:16] step 21450/1000000 | loss=3.1230 vocab=1.4084 token=1.7185 | 6.10 steps/s | VRAM=28.81 Go [2026-07-16 13:48:25] step 21500/1000000 | loss=3.1987 vocab=1.1804 token=1.8029 | 6.04 steps/s | VRAM=28.57 Go [2026-07-16 13:48:33] step 21550/1000000 | loss=3.1549 vocab=1.0846 token=1.8959 | 6.06 steps/s | VRAM=32.65 Go [2026-07-16 13:48:41] step 21600/1000000 | loss=3.0838 vocab=1.0966 token=1.8400 | 6.05 steps/s | VRAM=28.74 Go [2026-07-16 13:48:49] step 21650/1000000 | loss=3.0545 vocab=1.3310 token=1.7447 | 6.08 steps/s | VRAM=32.29 Go [2026-07-16 13:48:58] step 21700/1000000 | loss=3.1934 vocab=0.7256 token=1.4791 | 6.08 steps/s | VRAM=32.12 Go [2026-07-16 13:49:06] step 21750/1000000 | loss=3.1376 vocab=1.3412 token=1.5021 | 6.07 steps/s | VRAM=26.38 Go [2026-07-16 13:49:14] step 21800/1000000 | loss=3.1491 vocab=1.6570 token=1.5155 | 6.06 steps/s | VRAM=29.36 Go [2026-07-16 13:49:22] step 21850/1000000 | loss=3.2302 vocab=1.1587 token=1.7761 | 6.07 steps/s | VRAM=28.57 Go [2026-07-16 13:49:31] step 21900/1000000 | loss=3.1424 vocab=1.0514 token=1.8781 | 6.09 steps/s | VRAM=28.93 Go [2026-07-16 13:49:39] step 21950/1000000 | loss=3.1026 vocab=1.3346 token=1.7111 | 6.08 steps/s | VRAM=29.39 Go [2026-07-16 13:49:47] step 22000/1000000 | loss=3.0408 vocab=1.2508 token=1.5884 | 6.06 steps/s | VRAM=29.06 Go [2026-07-16 13:49:55] step 22050/1000000 | loss=3.1152 vocab=1.6156 token=1.9602 | 6.09 steps/s | VRAM=24.80 Go [2026-07-16 13:50:04] step 22100/1000000 | loss=2.9989 vocab=1.0564 token=1.7298 | 6.06 steps/s | VRAM=28.83 Go [2026-07-16 13:50:12] step 22150/1000000 | loss=3.0517 vocab=1.2250 token=1.8065 | 6.09 steps/s | VRAM=24.89 Go [2026-07-16 13:50:20] step 22200/1000000 | loss=3.1150 vocab=1.4511 token=2.1734 | 6.09 steps/s | VRAM=28.87 Go [2026-07-16 13:50:28] step 22250/1000000 | loss=3.0107 vocab=1.1324 token=1.6984 | 6.09 steps/s | VRAM=26.12 Go [2026-07-16 13:50:36] step 22300/1000000 | loss=3.0637 vocab=1.5151 token=1.7412 | 6.07 steps/s | VRAM=28.65 Go [2026-07-16 13:50:45] step 22350/1000000 | loss=3.0517 vocab=0.7277 token=1.3930 | 6.04 steps/s | VRAM=28.87 Go [2026-07-16 13:50:53] step 22400/1000000 | loss=3.0726 vocab=1.1561 token=1.4423 | 6.09 steps/s | VRAM=29.16 Go [2026-07-16 13:51:01] step 22450/1000000 | loss=3.1811 vocab=1.5002 token=1.9725 | 6.06 steps/s | VRAM=28.50 Go [2026-07-16 13:51:09] step 22500/1000000 | loss=3.1735 vocab=1.2494 token=2.0292 | 6.08 steps/s | VRAM=28.66 Go [2026-07-16 13:51:18] step 22550/1000000 | loss=3.1933 vocab=1.4277 token=1.6322 | 6.06 steps/s | VRAM=32.21 Go [2026-07-16 13:51:26] step 22600/1000000 | loss=3.0730 vocab=1.2467 token=1.6723 | 6.05 steps/s | VRAM=27.69 Go [2026-07-16 13:51:34] step 22650/1000000 | loss=3.1903 vocab=1.3207 token=2.0830 | 6.09 steps/s | VRAM=28.67 Go [2026-07-16 13:51:42] step 22700/1000000 | loss=3.0977 vocab=1.2146 token=1.6109 | 6.06 steps/s | VRAM=32.71 Go [2026-07-16 13:51:51] step 22750/1000000 | loss=3.1214 vocab=1.7986 token=2.0727 | 6.06 steps/s | VRAM=32.19 Go [2026-07-16 13:51:59] step 22800/1000000 | loss=2.9569 vocab=1.3906 token=1.7361 | 6.07 steps/s | VRAM=27.99 Go [2026-07-16 13:52:07] step 22850/1000000 | loss=2.9518 vocab=1.2045 token=1.6918 | 6.05 steps/s | VRAM=26.66 Go [2026-07-16 13:52:15] step 22900/1000000 | loss=3.1924 vocab=1.3093 token=1.6601 | 6.06 steps/s | VRAM=26.95 Go [2026-07-16 13:52:24] step 22950/1000000 | loss=3.0767 vocab=1.0779 token=1.7312 | 6.09 steps/s | VRAM=31.49 Go [2026-07-16 13:52:32] step 23000/1000000 | loss=3.0607 vocab=1.0981 token=1.6184 | 6.06 steps/s | VRAM=29.47 Go [2026-07-16 13:52:40] step 23050/1000000 | loss=3.2207 vocab=1.1523 token=1.5018 | 6.05 steps/s | VRAM=31.94 Go [2026-07-16 13:52:48] step 23100/1000000 | loss=3.0643 vocab=1.4155 token=1.9741 | 6.05 steps/s | VRAM=27.94 Go [2026-07-16 13:52:57] step 23150/1000000 | loss=3.0688 vocab=1.2830 token=1.7804 | 6.06 steps/s | VRAM=29.03 Go [2026-07-16 13:53:05] step 23200/1000000 | loss=3.0756 vocab=1.1241 token=1.8615 | 6.06 steps/s | VRAM=26.74 Go [2026-07-16 13:53:13] step 23250/1000000 | loss=3.0477 vocab=1.1491 token=1.5241 | 6.09 steps/s | VRAM=28.46 Go [2026-07-16 13:53:21] step 23300/1000000 | loss=2.9535 vocab=1.1481 token=1.6382 | 6.06 steps/s | VRAM=25.76 Go [2026-07-16 13:53:30] step 23350/1000000 | loss=3.1137 vocab=1.5984 token=1.8375 | 6.07 steps/s | VRAM=26.75 Go [2026-07-16 13:53:38] step 23400/1000000 | loss=3.0058 vocab=1.1744 token=1.5968 | 6.07 steps/s | VRAM=31.84 Go [2026-07-16 13:53:46] step 23450/1000000 | loss=3.0193 vocab=1.4850 token=2.0054 | 6.05 steps/s | VRAM=29.64 Go [2026-07-16 13:53:54] step 23500/1000000 | loss=3.0875 vocab=1.7557 token=1.6516 | 6.09 steps/s | VRAM=31.97 Go [2026-07-16 13:54:03] step 23550/1000000 | loss=3.0950 vocab=1.6815 token=1.8723 | 6.05 steps/s | VRAM=30.13 Go [2026-07-16 13:54:11] step 23600/1000000 | loss=3.1406 vocab=1.7796 token=1.9295 | 6.04 steps/s | VRAM=27.98 Go [2026-07-16 13:54:19] step 23650/1000000 | loss=3.0512 vocab=1.0153 token=1.2697 | 6.06 steps/s | VRAM=26.37 Go [2026-07-16 13:54:27] step 23700/1000000 | loss=3.1140 vocab=1.2112 token=1.7942 | 6.06 steps/s | VRAM=28.52 Go [2026-07-16 13:54:36] step 23750/1000000 | loss=3.0159 vocab=1.1659 token=1.6144 | 6.06 steps/s | VRAM=29.10 Go [2026-07-16 13:54:44] step 23800/1000000 | loss=2.9947 vocab=1.2984 token=1.6140 | 6.07 steps/s | VRAM=28.55 Go [2026-07-16 13:54:52] step 23850/1000000 | loss=3.1083 vocab=1.5957 token=1.9550 | 6.08 steps/s | VRAM=26.43 Go [2026-07-16 13:55:00] step 23900/1000000 | loss=3.0318 vocab=2.1004 token=2.0745 | 6.10 steps/s | VRAM=26.00 Go [2026-07-16 13:55:08] step 23950/1000000 | loss=3.0326 vocab=1.4521 token=2.1727 | 6.09 steps/s | VRAM=24.73 Go [2026-07-16 13:55:17] step 24000/1000000 | loss=3.0354 vocab=1.6759 token=1.9632 | 6.07 steps/s | VRAM=27.81 Go [2026-07-16 13:55:25] step 24050/1000000 | loss=2.8890 vocab=1.1809 token=1.7998 | 6.07 steps/s | VRAM=31.21 Go [2026-07-16 13:55:33] step 24100/1000000 | loss=3.0785 vocab=1.4994 token=1.9313 | 6.06 steps/s | VRAM=32.85 Go [2026-07-16 13:55:41] step 24150/1000000 | loss=3.0221 vocab=1.2556 token=1.6216 | 6.09 steps/s | VRAM=31.98 Go [2026-07-16 13:55:50] step 24200/1000000 | loss=3.0185 vocab=1.3110 token=1.7777 | 6.09 steps/s | VRAM=32.07 Go [2026-07-16 13:55:58] step 24250/1000000 | loss=3.0360 vocab=1.3873 token=1.7676 | 6.09 steps/s | VRAM=26.26 Go [2026-07-16 13:56:06] step 24300/1000000 | loss=3.0793 vocab=1.6482 token=1.9061 | 6.08 steps/s | VRAM=28.91 Go [2026-07-16 13:56:14] step 24350/1000000 | loss=3.1821 vocab=1.4521 token=2.0638 | 6.08 steps/s | VRAM=29.17 Go [2026-07-16 13:56:22] step 24400/1000000 | loss=2.9884 vocab=1.1384 token=1.7186 | 6.06 steps/s | VRAM=28.96 Go [2026-07-16 13:56:31] step 24450/1000000 | loss=3.1396 vocab=1.6711 token=1.7184 | 6.06 steps/s | VRAM=27.87 Go [2026-07-16 13:56:39] step 24500/1000000 | loss=2.9662 vocab=1.9881 token=1.9773 | 6.09 steps/s | VRAM=25.40 Go [2026-07-16 13:56:47] step 24550/1000000 | loss=2.9926 vocab=1.4198 token=1.9387 | 6.08 steps/s | VRAM=29.64 Go [2026-07-16 13:56:55] step 24600/1000000 | loss=2.9597 vocab=1.2274 token=1.6593 | 6.07 steps/s | VRAM=32.86 Go [2026-07-16 13:57:04] step 24650/1000000 | loss=2.9497 vocab=1.3364 token=1.6608 | 6.10 steps/s | VRAM=32.53 Go [2026-07-16 13:57:12] step 24700/1000000 | loss=3.1288 vocab=1.0727 token=1.7457 | 6.06 steps/s | VRAM=29.10 Go [2026-07-16 13:57:20] step 24750/1000000 | loss=3.0142 vocab=1.3945 token=1.6066 | 6.08 steps/s | VRAM=32.16 Go [2026-07-16 13:57:28] step 24800/1000000 | loss=2.9004 vocab=1.2680 token=1.8641 | 6.07 steps/s | VRAM=28.63 Go [2026-07-16 13:57:37] step 24850/1000000 | loss=2.9946 vocab=1.3344 token=1.7151 | 6.08 steps/s | VRAM=32.42 Go [2026-07-16 13:57:45] step 24900/1000000 | loss=2.9702 vocab=1.4906 token=1.7206 | 6.10 steps/s | VRAM=30.59 Go [2026-07-16 13:57:53] step 24950/1000000 | loss=2.9256 vocab=1.7221 token=1.8050 | 6.03 steps/s | VRAM=28.73 Go [2026-07-16 13:58:01] step 25000/1000000 | loss=3.0594 vocab=1.0389 token=1.9449 | 6.08 steps/s | VRAM=27.63 Go [2026-07-16 13:59:22] step 25050/1000000 | loss=2.9979 vocab=1.3002 token=1.3201 | 0.62 steps/s | VRAM=27.44 Go [2026-07-16 13:59:30] step 25100/1000000 | loss=3.0473 vocab=1.5148 token=1.6864 | 6.12 steps/s | VRAM=26.22 Go [2026-07-16 13:59:38] step 25150/1000000 | loss=2.9378 vocab=1.9780 token=1.9113 | 6.15 steps/s | VRAM=25.21 Go [2026-07-16 13:59:47] step 25200/1000000 | loss=2.9991 vocab=1.3898 token=1.5959 | 6.13 steps/s | VRAM=29.55 Go [2026-07-16 13:59:55] step 25250/1000000 | loss=3.0385 vocab=1.2542 token=1.6741 | 6.08 steps/s | VRAM=28.95 Go [2026-07-16 14:00:03] step 25300/1000000 | loss=2.9544 vocab=1.2888 token=1.8858 | 6.12 steps/s | VRAM=32.85 Go [2026-07-16 14:00:11] step 25350/1000000 | loss=2.9158 vocab=0.9302 token=1.6519 | 6.14 steps/s | VRAM=28.53 Go [2026-07-16 14:00:19] step 25400/1000000 | loss=3.1108 vocab=1.4442 token=1.9545 | 6.08 steps/s | VRAM=30.25 Go [2026-07-16 14:00:27] step 25450/1000000 | loss=3.0235 vocab=1.0768 token=1.5830 | 6.14 steps/s | VRAM=28.86 Go [2026-07-16 14:00:36] step 25500/1000000 | loss=2.9340 vocab=2.2214 token=1.8539 | 6.11 steps/s | VRAM=31.37 Go [2026-07-16 14:00:44] step 25550/1000000 | loss=2.9381 vocab=1.4959 token=1.9348 | 6.09 steps/s | VRAM=25.65 Go [2026-07-16 14:00:52] step 25600/1000000 | loss=3.0344 vocab=1.5424 token=1.6492 | 6.10 steps/s | VRAM=31.46 Go [2026-07-16 14:01:00] step 25650/1000000 | loss=3.0229 vocab=1.1826 token=1.9340 | 6.09 steps/s | VRAM=32.65 Go [2026-07-16 14:01:08] step 25700/1000000 | loss=2.9096 vocab=1.2034 token=1.5791 | 6.08 steps/s | VRAM=26.32 Go [2026-07-16 14:01:17] step 25750/1000000 | loss=3.0209 vocab=1.2564 token=2.0259 | 6.08 steps/s | VRAM=26.68 Go [2026-07-16 14:01:25] step 25800/1000000 | loss=2.9287 vocab=1.4585 token=1.5673 | 6.06 steps/s | VRAM=29.19 Go [2026-07-16 14:01:33] step 25850/1000000 | loss=3.0684 vocab=0.8219 token=1.4089 | 6.06 steps/s | VRAM=24.73 Go [2026-07-16 14:01:41] step 25900/1000000 | loss=2.9768 vocab=1.2960 token=1.9964 | 6.07 steps/s | VRAM=28.22 Go [2026-07-16 14:01:50] step 25950/1000000 | loss=3.0322 vocab=1.1128 token=1.7419 | 6.06 steps/s | VRAM=26.50 Go [2026-07-16 14:01:58] step 26000/1000000 | loss=3.0418 vocab=0.9729 token=1.6487 | 6.07 steps/s | VRAM=28.31 Go [2026-07-16 14:02:06] step 26050/1000000 | loss=3.1049 vocab=0.8590 token=1.7993 | 6.08 steps/s | VRAM=28.17 Go [2026-07-16 14:02:14] step 26100/1000000 | loss=2.9167 vocab=1.6264 token=1.7807 | 6.07 steps/s | VRAM=29.01 Go [2026-07-16 14:02:23] step 26150/1000000 | loss=3.0568 vocab=1.7965 token=1.9849 | 6.10 steps/s | VRAM=32.34 Go [2026-07-16 14:02:31] step 26200/1000000 | loss=3.1195 vocab=1.2126 token=1.6541 | 6.08 steps/s | VRAM=24.79 Go [2026-07-16 14:02:39] step 26250/1000000 | loss=3.0750 vocab=1.2687 token=2.0212 | 6.06 steps/s | VRAM=32.08 Go [2026-07-16 14:02:47] step 26300/1000000 | loss=2.9513 vocab=1.5222 token=2.0053 | 6.05 steps/s | VRAM=28.53 Go [2026-07-16 14:02:55] step 26350/1000000 | loss=2.9466 vocab=1.3802 token=1.4620 | 6.10 steps/s | VRAM=28.44 Go [2026-07-16 14:03:04] step 26400/1000000 | loss=3.0313 vocab=1.4096 token=1.5338 | 6.06 steps/s | VRAM=32.55 Go [2026-07-16 14:03:12] step 26450/1000000 | loss=3.0390 vocab=1.5837 token=1.4654 | 6.07 steps/s | VRAM=28.88 Go [2026-07-16 14:03:20] step 26500/1000000 | loss=2.9765 vocab=1.3030 token=1.8323 | 6.05 steps/s | VRAM=27.85 Go [2026-07-16 14:03:29] step 26550/1000000 | loss=3.0039 vocab=1.1364 token=1.4779 | 6.03 steps/s | VRAM=30.83 Go [2026-07-16 14:03:37] step 26600/1000000 | loss=3.0703 vocab=1.4342 token=1.6169 | 6.08 steps/s | VRAM=28.31 Go [2026-07-16 14:03:45] step 26650/1000000 | loss=3.0113 vocab=1.9463 token=1.7609 | 6.08 steps/s | VRAM=32.78 Go [2026-07-16 14:03:53] step 26700/1000000 | loss=3.0397 vocab=1.6969 token=1.5715 | 6.05 steps/s | VRAM=28.61 Go [2026-07-16 14:04:01] step 26750/1000000 | loss=3.0090 vocab=1.2531 token=1.7744 | 6.09 steps/s | VRAM=31.93 Go [2026-07-16 14:04:10] step 26800/1000000 | loss=3.0200 vocab=1.0505 token=2.0110 | 6.05 steps/s | VRAM=26.53 Go [2026-07-16 14:04:18] step 26850/1000000 | loss=2.8891 vocab=1.1847 token=1.9113 | 6.06 steps/s | VRAM=28.60 Go [2026-07-16 14:04:26] step 26900/1000000 | loss=2.9853 vocab=1.3813 token=1.7557 | 6.10 steps/s | VRAM=28.41 Go [2026-07-16 14:04:34] step 26950/1000000 | loss=3.0303 vocab=1.2491 token=1.6210 | 6.08 steps/s | VRAM=28.06 Go [2026-07-16 14:04:43] step 27000/1000000 | loss=3.0525 vocab=1.1884 token=1.4223 | 6.04 steps/s | VRAM=26.47 Go [2026-07-16 14:04:51] step 27050/1000000 | loss=3.0848 vocab=0.7769 token=1.8492 | 6.09 steps/s | VRAM=25.79 Go [2026-07-16 14:04:59] step 27100/1000000 | loss=2.9354 vocab=1.1135 token=1.5139 | 6.08 steps/s | VRAM=28.85 Go [2026-07-16 14:05:07] step 27150/1000000 | loss=2.9179 vocab=0.9376 token=1.6114 | 6.04 steps/s | VRAM=26.62 Go [2026-07-16 14:05:16] step 27200/1000000 | loss=2.9967 vocab=0.9439 token=1.8178 | 6.05 steps/s | VRAM=29.33 Go [2026-07-16 14:05:24] step 27250/1000000 | loss=2.9682 vocab=1.2387 token=1.7687 | 6.05 steps/s | VRAM=28.19 Go [2026-07-16 14:05:32] step 27300/1000000 | loss=2.9622 vocab=1.2206 token=1.9350 | 6.06 steps/s | VRAM=24.49 Go [2026-07-16 14:05:40] step 27350/1000000 | loss=3.0867 vocab=1.3542 token=1.7997 | 6.10 steps/s | VRAM=28.12 Go [2026-07-16 14:05:49] step 27400/1000000 | loss=2.9171 vocab=1.1340 token=1.8401 | 6.06 steps/s | VRAM=26.11 Go [2026-07-16 14:05:57] step 27450/1000000 | loss=2.9952 vocab=2.2723 token=1.9777 | 6.09 steps/s | VRAM=32.05 Go [2026-07-16 14:06:05] step 27500/1000000 | loss=2.9692 vocab=1.0094 token=1.6862 | 6.08 steps/s | VRAM=30.31 Go [2026-07-16 14:06:13] step 27550/1000000 | loss=2.8639 vocab=1.7346 token=1.9477 | 6.06 steps/s | VRAM=31.27 Go [2026-07-16 14:06:22] step 27600/1000000 | loss=2.9721 vocab=1.0441 token=1.7353 | 6.07 steps/s | VRAM=29.01 Go [2026-07-16 14:06:30] step 27650/1000000 | loss=2.9096 vocab=1.5178 token=1.8593 | 6.10 steps/s | VRAM=32.16 Go [2026-07-16 14:06:38] step 27700/1000000 | loss=2.9802 vocab=1.1746 token=1.6003 | 6.06 steps/s | VRAM=23.00 Go [2026-07-16 14:06:46] step 27750/1000000 | loss=2.8976 vocab=1.5831 token=1.5310 | 6.10 steps/s | VRAM=31.40 Go [2026-07-16 14:06:54] step 27800/1000000 | loss=2.9791 vocab=0.9152 token=1.6893 | 6.05 steps/s | VRAM=30.22 Go [2026-07-16 14:07:03] step 27850/1000000 | loss=2.8991 vocab=0.8800 token=1.8993 | 6.08 steps/s | VRAM=25.74 Go [2026-07-16 14:07:11] step 27900/1000000 | loss=2.9741 vocab=1.2143 token=1.5743 | 6.06 steps/s | VRAM=28.00 Go [2026-07-16 14:07:19] step 27950/1000000 | loss=2.9862 vocab=0.9734 token=1.6665 | 6.09 steps/s | VRAM=32.04 Go [2026-07-16 14:07:27] step 28000/1000000 | loss=2.8186 vocab=1.0480 token=2.1590 | 6.09 steps/s | VRAM=26.65 Go [2026-07-16 14:07:35] step 28050/1000000 | loss=2.9132 vocab=0.8028 token=1.3872 | 6.13 steps/s | VRAM=28.15 Go [2026-07-16 14:07:44] step 28100/1000000 | loss=2.8838 vocab=0.9595 token=1.5318 | 6.08 steps/s | VRAM=29.60 Go [2026-07-16 14:07:52] step 28150/1000000 | loss=2.9629 vocab=1.2806 token=1.8178 | 6.06 steps/s | VRAM=23.43 Go [2026-07-16 14:08:00] step 28200/1000000 | loss=2.8904 vocab=1.0917 token=1.7955 | 6.10 steps/s | VRAM=28.66 Go [2026-07-16 14:08:08] step 28250/1000000 | loss=2.9383 vocab=1.1666 token=1.5766 | 6.10 steps/s | VRAM=31.19 Go [2026-07-16 14:08:17] step 28300/1000000 | loss=2.9218 vocab=1.3515 token=1.4623 | 6.08 steps/s | VRAM=32.66 Go [2026-07-16 14:08:25] step 28350/1000000 | loss=3.0581 vocab=1.3153 token=1.8962 | 6.08 steps/s | VRAM=27.31 Go [2026-07-16 14:08:33] step 28400/1000000 | loss=2.9423 vocab=0.8765 token=1.8709 | 6.10 steps/s | VRAM=32.74 Go [2026-07-16 14:08:41] step 28450/1000000 | loss=2.9994 vocab=1.0780 token=1.4689 | 6.06 steps/s | VRAM=27.84 Go [2026-07-16 14:08:49] step 28500/1000000 | loss=2.9507 vocab=0.7147 token=1.4680 | 6.07 steps/s | VRAM=27.21 Go [2026-07-16 14:08:58] step 28550/1000000 | loss=2.9863 vocab=1.1102 token=2.1592 | 6.09 steps/s | VRAM=28.38 Go [2026-07-16 14:09:06] step 28600/1000000 | loss=2.9087 vocab=0.6624 token=1.5762 | 6.10 steps/s | VRAM=28.09 Go [2026-07-16 14:09:14] step 28650/1000000 | loss=3.0695 vocab=1.3016 token=1.7304 | 6.08 steps/s | VRAM=28.66 Go [2026-07-16 14:09:22] step 28700/1000000 | loss=2.8314 vocab=0.7886 token=1.5412 | 6.07 steps/s | VRAM=29.02 Go [2026-07-16 14:09:31] step 28750/1000000 | loss=3.0611 vocab=1.5471 token=1.6083 | 6.08 steps/s | VRAM=32.82 Go [2026-07-16 14:09:39] step 28800/1000000 | loss=2.8946 vocab=2.0823 token=2.2475 | 6.08 steps/s | VRAM=26.22 Go [2026-07-16 14:09:47] step 28850/1000000 | loss=2.9904 vocab=1.2505 token=1.9562 | 6.09 steps/s | VRAM=28.66 Go [2026-07-16 14:09:55] step 28900/1000000 | loss=2.9329 vocab=1.1590 token=2.0519 | 6.08 steps/s | VRAM=25.96 Go [2026-07-16 14:10:03] step 28950/1000000 | loss=3.0462 vocab=1.1172 token=1.9359 | 6.07 steps/s | VRAM=22.67 Go [2026-07-16 14:10:12] step 29000/1000000 | loss=2.9584 vocab=1.3646 token=1.8761 | 6.06 steps/s | VRAM=27.98 Go [2026-07-16 14:10:20] step 29050/1000000 | loss=2.8963 vocab=1.6789 token=1.8952 | 6.06 steps/s | VRAM=26.73 Go [2026-07-16 14:10:28] step 29100/1000000 | loss=2.8556 vocab=0.9038 token=1.6172 | 6.11 steps/s | VRAM=27.59 Go [2026-07-16 14:10:36] step 29150/1000000 | loss=2.7785 vocab=1.2721 token=1.6856 | 6.09 steps/s | VRAM=27.92 Go [2026-07-16 14:10:45] step 29200/1000000 | loss=2.8684 vocab=1.0434 token=1.3300 | 6.10 steps/s | VRAM=31.60 Go [2026-07-16 14:10:53] step 29250/1000000 | loss=2.8653 vocab=1.3876 token=1.8372 | 6.07 steps/s | VRAM=32.69 Go [2026-07-16 14:11:01] step 29300/1000000 | loss=2.9028 vocab=1.1061 token=1.8066 | 6.08 steps/s | VRAM=32.07 Go [2026-07-16 14:11:09] step 29350/1000000 | loss=2.9515 vocab=1.6591 token=1.7913 | 6.05 steps/s | VRAM=24.86 Go [2026-07-16 14:11:18] step 29400/1000000 | loss=2.8562 vocab=0.8514 token=1.2602 | 6.06 steps/s | VRAM=26.97 Go [2026-07-16 14:11:26] step 29450/1000000 | loss=2.9215 vocab=1.2225 token=1.6609 | 6.09 steps/s | VRAM=28.94 Go [2026-07-16 14:11:34] step 29500/1000000 | loss=2.8911 vocab=1.2633 token=1.4298 | 6.06 steps/s | VRAM=25.49 Go [2026-07-16 14:11:42] step 29550/1000000 | loss=3.0396 vocab=1.6864 token=1.8990 | 6.10 steps/s | VRAM=32.77 Go [2026-07-16 14:11:50] step 29600/1000000 | loss=2.8643 vocab=0.9863 token=1.2650 | 6.04 steps/s | VRAM=29.12 Go [2026-07-16 14:11:59] step 29650/1000000 | loss=2.8982 vocab=1.1088 token=1.3502 | 6.08 steps/s | VRAM=26.15 Go [2026-07-16 14:12:07] step 29700/1000000 | loss=2.9110 vocab=1.0000 token=1.8527 | 6.05 steps/s | VRAM=28.80 Go [2026-07-16 14:12:15] step 29750/1000000 | loss=2.9299 vocab=1.1617 token=1.5648 | 6.08 steps/s | VRAM=31.99 Go [2026-07-16 14:12:23] step 29800/1000000 | loss=2.9018 vocab=0.9607 token=1.4781 | 6.10 steps/s | VRAM=30.77 Go [2026-07-16 14:12:32] step 29850/1000000 | loss=2.9018 vocab=0.6982 token=1.7083 | 6.05 steps/s | VRAM=28.23 Go [2026-07-16 14:12:40] step 29900/1000000 | loss=2.7670 vocab=0.6966 token=1.4957 | 6.07 steps/s | VRAM=29.29 Go [2026-07-16 14:12:48] step 29950/1000000 | loss=2.8985 vocab=1.0302 token=1.6806 | 6.05 steps/s | VRAM=29.53 Go [2026-07-16 14:12:56] step 30000/1000000 | loss=2.9381 vocab=1.3018 token=1.6013 | 6.07 steps/s | VRAM=28.75 Go [2026-07-16 14:14:29] step 30050/1000000 | loss=2.9273 vocab=0.6591 token=1.7012 | 0.54 steps/s | VRAM=26.70 Go [2026-07-16 14:14:37] step 30100/1000000 | loss=2.9896 vocab=1.5112 token=1.9762 | 6.14 steps/s | VRAM=32.48 Go [2026-07-16 14:14:45] step 30150/1000000 | loss=2.9989 vocab=1.6078 token=1.9233 | 6.13 steps/s | VRAM=26.02 Go [2026-07-16 14:14:54] step 30200/1000000 | loss=2.8746 vocab=0.8898 token=1.6425 | 6.14 steps/s | VRAM=32.15 Go [2026-07-16 14:15:02] step 30250/1000000 | loss=3.0180 vocab=1.0902 token=1.5275 | 6.11 steps/s | VRAM=28.78 Go [2026-07-16 14:15:10] step 30300/1000000 | loss=2.9465 vocab=1.2643 token=1.4417 | 6.10 steps/s | VRAM=32.01 Go [2026-07-16 14:15:18] step 30350/1000000 | loss=2.8185 vocab=0.9634 token=1.3799 | 6.10 steps/s | VRAM=32.66 Go [2026-07-16 14:15:26] step 30400/1000000 | loss=2.9116 vocab=1.2891 token=2.0811 | 6.11 steps/s | VRAM=32.17 Go [2026-07-16 14:15:35] step 30450/1000000 | loss=2.7820 vocab=0.9380 token=1.2723 | 6.09 steps/s | VRAM=27.97 Go [2026-07-16 14:15:43] step 30500/1000000 | loss=2.9332 vocab=0.8718 token=1.3737 | 6.08 steps/s | VRAM=28.13 Go [2026-07-16 14:15:51] step 30550/1000000 | loss=2.8640 vocab=0.9403 token=1.8009 | 6.05 steps/s | VRAM=26.92 Go [2026-07-16 14:15:59] step 30600/1000000 | loss=2.9243 vocab=1.9607 token=1.6871 | 6.06 steps/s | VRAM=28.35 Go [2026-07-16 14:16:08] step 30650/1000000 | loss=2.9090 vocab=0.9877 token=1.6632 | 6.07 steps/s | VRAM=32.91 Go [2026-07-16 14:16:16] step 30700/1000000 | loss=3.0472 vocab=1.1577 token=1.9937 | 6.06 steps/s | VRAM=28.43 Go [2026-07-16 14:16:24] step 30750/1000000 | loss=2.8448 vocab=1.2797 token=1.4975 | 6.08 steps/s | VRAM=28.63 Go [2026-07-16 14:16:32] step 30800/1000000 | loss=2.8263 vocab=0.9361 token=1.2904 | 6.08 steps/s | VRAM=31.61 Go [2026-07-16 14:16:40] step 30850/1000000 | loss=2.9502 vocab=2.1490 token=1.8199 | 6.06 steps/s | VRAM=26.74 Go [2026-07-16 14:16:49] step 30900/1000000 | loss=2.8735 vocab=1.2316 token=2.0424 | 6.04 steps/s | VRAM=29.10 Go [2026-07-16 14:16:57] step 30950/1000000 | loss=2.9440 vocab=1.2797 token=1.8056 | 6.04 steps/s | VRAM=29.33 Go [2026-07-16 14:17:05] step 31000/1000000 | loss=2.8445 vocab=1.2104 token=1.5392 | 6.08 steps/s | VRAM=25.22 Go [2026-07-16 14:17:13] step 31050/1000000 | loss=2.9318 vocab=1.1467 token=1.4070 | 6.07 steps/s | VRAM=24.95 Go [2026-07-16 14:17:22] step 31100/1000000 | loss=2.9250 vocab=0.9859 token=1.5841 | 6.08 steps/s | VRAM=31.67 Go [2026-07-16 14:17:30] step 31150/1000000 | loss=2.8878 vocab=1.1283 token=2.0119 | 6.05 steps/s | VRAM=29.45 Go [2026-07-16 14:17:38] step 31200/1000000 | loss=2.8808 vocab=1.0882 token=1.7492 | 6.04 steps/s | VRAM=30.62 Go [2026-07-16 14:17:46] step 31250/1000000 | loss=2.8523 vocab=1.0207 token=1.4878 | 6.07 steps/s | VRAM=25.70 Go [2026-07-16 14:17:55] step 31300/1000000 | loss=2.8724 vocab=1.4852 token=1.5938 | 6.07 steps/s | VRAM=26.44 Go [2026-07-16 14:18:03] step 31350/1000000 | loss=3.0868 vocab=0.7344 token=1.4479 | 6.07 steps/s | VRAM=29.53 Go [2026-07-16 14:18:11] step 31400/1000000 | loss=2.8794 vocab=1.2553 token=1.6808 | 6.09 steps/s | VRAM=30.90 Go [2026-07-16 14:18:19] step 31450/1000000 | loss=2.9242 vocab=1.0139 token=1.6514 | 6.05 steps/s | VRAM=32.80 Go [2026-07-16 14:18:28] step 31500/1000000 | loss=2.9361 vocab=0.7826 token=1.7830 | 6.04 steps/s | VRAM=29.07 Go [2026-07-16 14:18:36] step 31550/1000000 | loss=2.8636 vocab=1.1072 token=1.8090 | 6.06 steps/s | VRAM=28.11 Go [2026-07-16 14:18:44] step 31600/1000000 | loss=2.9284 vocab=1.4244 token=1.7699 | 6.08 steps/s | VRAM=29.07 Go [2026-07-16 14:18:52] step 31650/1000000 | loss=2.9116 vocab=1.4628 token=1.9594 | 6.09 steps/s | VRAM=27.96 Go [2026-07-16 14:19:01] step 31700/1000000 | loss=3.0047 vocab=1.4200 token=1.8062 | 6.06 steps/s | VRAM=29.79 Go [2026-07-16 14:19:09] step 31750/1000000 | loss=2.9376 vocab=1.3495 token=1.4475 | 6.09 steps/s | VRAM=28.54 Go [2026-07-16 14:19:17] step 31800/1000000 | loss=2.8911 vocab=1.0147 token=1.4064 | 6.06 steps/s | VRAM=28.43 Go [2026-07-16 14:19:25] step 31850/1000000 | loss=2.8294 vocab=1.5926 token=1.7659 | 6.08 steps/s | VRAM=32.95 Go [2026-07-16 14:19:34] step 31900/1000000 | loss=2.8147 vocab=1.0834 token=1.3689 | 6.07 steps/s | VRAM=29.46 Go [2026-07-16 14:19:42] step 31950/1000000 | loss=2.7756 vocab=1.3180 token=1.6077 | 6.07 steps/s | VRAM=31.69 Go [2026-07-16 14:19:50] step 32000/1000000 | loss=2.7967 vocab=0.9708 token=1.4747 | 6.07 steps/s | VRAM=28.62 Go [2026-07-16 14:19:58] step 32050/1000000 | loss=2.8133 vocab=0.7337 token=1.6165 | 6.06 steps/s | VRAM=28.37 Go [2026-07-16 14:20:07] step 32100/1000000 | loss=2.8280 vocab=0.8891 token=1.5560 | 6.07 steps/s | VRAM=29.78 Go [2026-07-16 14:20:15] step 32150/1000000 | loss=2.9507 vocab=1.3850 token=1.8069 | 6.09 steps/s | VRAM=32.13 Go [2026-07-16 14:20:23] step 32200/1000000 | loss=2.9013 vocab=0.9985 token=1.4056 | 6.08 steps/s | VRAM=31.53 Go [2026-07-16 14:20:31] step 32250/1000000 | loss=2.8659 vocab=2.0136 token=1.9480 | 6.10 steps/s | VRAM=31.87 Go [2026-07-16 14:20:39] step 32300/1000000 | loss=2.7660 vocab=1.1304 token=1.8820 | 6.07 steps/s | VRAM=29.01 Go [2026-07-16 14:20:48] step 32350/1000000 | loss=2.9159 vocab=1.4871 token=1.9490 | 6.08 steps/s | VRAM=25.66 Go [2026-07-16 14:20:56] step 32400/1000000 | loss=2.8546 vocab=1.3534 token=1.5292 | 6.07 steps/s | VRAM=28.95 Go [2026-07-16 14:21:04] step 32450/1000000 | loss=2.8062 vocab=0.9528 token=1.2584 | 6.09 steps/s | VRAM=27.94 Go [2026-07-16 14:21:12] step 32500/1000000 | loss=2.8154 vocab=1.2123 token=1.8008 | 6.08 steps/s | VRAM=26.40 Go [2026-07-16 14:21:21] step 32550/1000000 | loss=2.9534 vocab=1.5287 token=2.0934 | 6.09 steps/s | VRAM=25.94 Go [2026-07-16 14:21:29] step 32600/1000000 | loss=2.9132 vocab=1.5060 token=1.7083 | 6.10 steps/s | VRAM=30.78 Go [2026-07-16 14:21:37] step 32650/1000000 | loss=2.9103 vocab=0.6114 token=1.4000 | 6.07 steps/s | VRAM=27.69 Go [2026-07-16 14:21:45] step 32700/1000000 | loss=2.9014 vocab=1.2301 token=1.8036 | 6.07 steps/s | VRAM=26.41 Go [2026-07-16 14:21:53] step 32750/1000000 | loss=2.8039 vocab=1.0814 token=1.6357 | 6.08 steps/s | VRAM=31.78 Go [2026-07-16 14:22:02] step 32800/1000000 | loss=2.8401 vocab=1.0241 token=1.5317 | 6.08 steps/s | VRAM=28.35 Go [2026-07-16 14:22:10] step 32850/1000000 | loss=2.8109 vocab=1.1034 token=1.5049 | 6.07 steps/s | VRAM=28.66 Go [2026-07-16 14:22:18] step 32900/1000000 | loss=2.7848 vocab=1.1571 token=1.6564 | 6.07 steps/s | VRAM=31.46 Go [2026-07-16 14:22:26] step 32950/1000000 | loss=2.9050 vocab=1.0203 token=1.5097 | 6.08 steps/s | VRAM=29.42 Go [2026-07-16 14:22:35] step 33000/1000000 | loss=2.8448 vocab=1.1623 token=1.5849 | 6.06 steps/s | VRAM=26.37 Go [2026-07-16 14:22:43] step 33050/1000000 | loss=2.9443 vocab=1.2326 token=1.8927 | 6.09 steps/s | VRAM=28.51 Go [2026-07-16 14:22:51] step 33100/1000000 | loss=2.9325 vocab=1.0410 token=1.6608 | 6.07 steps/s | VRAM=28.14 Go [2026-07-16 14:22:59] step 33150/1000000 | loss=2.8087 vocab=0.6605 token=1.5363 | 6.07 steps/s | VRAM=28.99 Go [2026-07-16 14:23:08] step 33200/1000000 | loss=2.8312 vocab=0.9091 token=1.2230 | 6.08 steps/s | VRAM=28.10 Go [2026-07-16 14:23:16] step 33250/1000000 | loss=2.9043 vocab=1.3992 token=1.6766 | 6.06 steps/s | VRAM=28.94 Go [2026-07-16 14:23:24] step 33300/1000000 | loss=2.9264 vocab=1.0179 token=1.4585 | 6.05 steps/s | VRAM=29.13 Go [2026-07-16 14:23:32] step 33350/1000000 | loss=2.7916 vocab=1.0503 token=1.6644 | 6.03 steps/s | VRAM=26.32 Go [2026-07-16 14:23:41] step 33400/1000000 | loss=2.9211 vocab=1.2697 token=1.8097 | 6.08 steps/s | VRAM=29.75 Go [2026-07-16 14:23:49] step 33450/1000000 | loss=2.8876 vocab=1.2824 token=1.8104 | 6.07 steps/s | VRAM=25.94 Go [2026-07-16 14:23:57] step 33500/1000000 | loss=2.9248 vocab=1.6811 token=1.7090 | 6.07 steps/s | VRAM=28.88 Go [2026-07-16 14:24:05] step 33550/1000000 | loss=2.8751 vocab=1.4966 token=1.8833 | 6.10 steps/s | VRAM=28.08 Go [2026-07-16 14:24:13] step 33600/1000000 | loss=2.8712 vocab=1.6024 token=1.8146 | 6.08 steps/s | VRAM=29.77 Go [2026-07-16 14:24:22] step 33650/1000000 | loss=2.9516 vocab=1.1407 token=1.7643 | 6.06 steps/s | VRAM=25.53 Go [2026-07-16 14:24:30] step 33700/1000000 | loss=2.8653 vocab=1.4415 token=1.9362 | 6.08 steps/s | VRAM=26.02 Go [2026-07-16 14:24:38] step 33750/1000000 | loss=2.9105 vocab=1.1836 token=1.6561 | 6.08 steps/s | VRAM=28.97 Go [2026-07-16 14:24:46] step 33800/1000000 | loss=2.8049 vocab=1.3646 token=1.7499 | 6.07 steps/s | VRAM=31.76 Go [2026-07-16 14:24:55] step 33850/1000000 | loss=2.8927 vocab=1.3501 token=1.7473 | 6.10 steps/s | VRAM=24.95 Go [2026-07-16 14:25:03] step 33900/1000000 | loss=2.8445 vocab=1.2115 token=2.0260 | 6.06 steps/s | VRAM=26.67 Go [2026-07-16 14:25:11] step 33950/1000000 | loss=2.8194 vocab=1.2810 token=1.4940 | 6.07 steps/s | VRAM=26.14 Go [2026-07-16 14:25:19] step 34000/1000000 | loss=2.9126 vocab=1.6474 token=2.0437 | 6.10 steps/s | VRAM=29.24 Go [2026-07-16 14:25:28] step 34050/1000000 | loss=2.8775 vocab=0.8272 token=1.7087 | 6.09 steps/s | VRAM=30.49 Go [2026-07-16 14:25:36] step 34100/1000000 | loss=3.0218 vocab=1.3407 token=1.7072 | 6.05 steps/s | VRAM=30.98 Go [2026-07-16 14:25:44] step 34150/1000000 | loss=2.9250 vocab=0.9758 token=1.4714 | 6.07 steps/s | VRAM=32.33 Go [2026-07-16 14:25:52] step 34200/1000000 | loss=2.9317 vocab=1.2919 token=1.5885 | 6.07 steps/s | VRAM=26.33 Go [2026-07-16 14:26:00] step 34250/1000000 | loss=2.8425 vocab=0.9097 token=1.8250 | 6.08 steps/s | VRAM=29.49 Go [2026-07-16 14:26:09] step 34300/1000000 | loss=2.8732 vocab=1.4901 token=1.9776 | 6.09 steps/s | VRAM=24.74 Go [2026-07-16 14:26:17] step 34350/1000000 | loss=2.7841 vocab=1.2413 token=1.4973 | 6.08 steps/s | VRAM=29.32 Go [2026-07-16 14:26:25] step 34400/1000000 | loss=2.8375 vocab=1.7468 token=1.8222 | 6.09 steps/s | VRAM=28.97 Go [2026-07-16 14:26:33] step 34450/1000000 | loss=2.8735 vocab=0.8889 token=1.3473 | 6.11 steps/s | VRAM=29.20 Go [2026-07-16 14:26:42] step 34500/1000000 | loss=2.7970 vocab=1.1347 token=1.6426 | 6.08 steps/s | VRAM=29.61 Go [2026-07-16 14:26:50] step 34550/1000000 | loss=2.8640 vocab=1.0186 token=1.9487 | 6.10 steps/s | VRAM=28.29 Go [2026-07-16 14:26:58] step 34600/1000000 | loss=2.8016 vocab=0.6595 token=1.2145 | 6.08 steps/s | VRAM=29.14 Go [2026-07-16 14:27:06] step 34650/1000000 | loss=2.6974 vocab=1.2224 token=1.6901 | 6.09 steps/s | VRAM=28.08 Go [2026-07-16 14:27:14] step 34700/1000000 | loss=2.8205 vocab=1.1909 token=1.1633 | 6.08 steps/s | VRAM=25.64 Go [2026-07-16 14:27:23] step 34750/1000000 | loss=2.8030 vocab=1.5375 token=1.6443 | 6.09 steps/s | VRAM=32.79 Go [2026-07-16 14:27:31] step 34800/1000000 | loss=2.8234 vocab=0.8885 token=1.6512 | 6.03 steps/s | VRAM=26.12 Go [2026-07-16 14:27:39] step 34850/1000000 | loss=2.9530 vocab=1.1755 token=1.8621 | 6.12 steps/s | VRAM=24.35 Go [2026-07-16 14:27:47] step 34900/1000000 | loss=2.8447 vocab=1.0513 token=1.9251 | 6.05 steps/s | VRAM=29.02 Go [2026-07-16 14:27:56] step 34950/1000000 | loss=2.9014 vocab=0.9748 token=1.5523 | 6.07 steps/s | VRAM=29.45 Go [2026-07-16 14:28:04] step 35000/1000000 | loss=2.9066 vocab=1.1640 token=1.6268 | 6.10 steps/s | VRAM=31.12 Go [2026-07-16 14:29:32] step 35050/1000000 | loss=2.8518 vocab=0.9473 token=2.0141 | 0.57 steps/s | VRAM=26.34 Go [2026-07-16 14:29:40] step 35100/1000000 | loss=2.8335 vocab=1.5574 token=1.6215 | 6.12 steps/s | VRAM=27.15 Go [2026-07-16 14:29:48] step 35150/1000000 | loss=2.8467 vocab=1.0384 token=0.9041 | 6.17 steps/s | VRAM=28.56 Go [2026-07-16 14:29:56] step 35200/1000000 | loss=2.8726 vocab=0.9145 token=1.6347 | 6.11 steps/s | VRAM=26.69 Go [2026-07-16 14:30:04] step 35250/1000000 | loss=2.8661 vocab=1.5360 token=1.6457 | 6.13 steps/s | VRAM=31.73 Go [2026-07-16 14:30:12] step 35300/1000000 | loss=2.8613 vocab=1.3462 token=1.6361 | 6.09 steps/s | VRAM=26.67 Go [2026-07-16 14:30:21] step 35350/1000000 | loss=2.8531 vocab=1.2669 token=1.7901 | 6.11 steps/s | VRAM=28.35 Go [2026-07-16 14:30:29] step 35400/1000000 | loss=2.8210 vocab=0.9689 token=1.7858 | 6.09 steps/s | VRAM=29.53 Go [2026-07-16 14:30:37] step 35450/1000000 | loss=2.8921 vocab=1.2459 token=1.7175 | 6.10 steps/s | VRAM=28.12 Go [2026-07-16 14:30:45] step 35500/1000000 | loss=2.7958 vocab=1.6215 token=1.7121 | 6.07 steps/s | VRAM=28.41 Go [2026-07-16 14:30:54] step 35550/1000000 | loss=2.8962 vocab=1.1439 token=1.7440 | 6.08 steps/s | VRAM=26.06 Go [2026-07-16 14:31:02] step 35600/1000000 | loss=2.8272 vocab=1.5094 token=1.9268 | 6.07 steps/s | VRAM=26.76 Go [2026-07-16 14:31:10] step 35650/1000000 | loss=2.7639 vocab=1.0715 token=1.7870 | 6.07 steps/s | VRAM=25.60 Go [2026-07-16 14:31:18] step 35700/1000000 | loss=2.8093 vocab=0.9299 token=1.5570 | 6.08 steps/s | VRAM=27.86 Go [2026-07-16 14:31:26] step 35750/1000000 | loss=2.8532 vocab=1.0805 token=1.6024 | 6.09 steps/s | VRAM=32.80 Go [2026-07-16 14:31:35] step 35800/1000000 | loss=2.8347 vocab=0.7481 token=1.6337 | 6.08 steps/s | VRAM=32.48 Go [2026-07-16 14:31:43] step 35850/1000000 | loss=2.7523 vocab=0.7456 token=1.4769 | 6.08 steps/s | VRAM=32.11 Go [2026-07-16 14:31:51] step 35900/1000000 | loss=2.8988 vocab=2.4879 token=1.5600 | 6.08 steps/s | VRAM=32.66 Go [2026-07-16 14:31:59] step 35950/1000000 | loss=2.8510 vocab=1.5391 token=1.7508 | 6.09 steps/s | VRAM=32.04 Go [2026-07-16 14:32:08] step 36000/1000000 | loss=2.7161 vocab=0.6738 token=1.3084 | 6.08 steps/s | VRAM=32.71 Go [2026-07-16 14:32:16] step 36050/1000000 | loss=2.6769 vocab=1.2838 token=1.6433 | 6.06 steps/s | VRAM=28.33 Go [2026-07-16 14:32:24] step 36100/1000000 | loss=2.8463 vocab=0.9004 token=1.6998 | 6.06 steps/s | VRAM=28.89 Go [2026-07-16 14:32:32] step 36150/1000000 | loss=2.8361 vocab=0.9847 token=1.5812 | 6.08 steps/s | VRAM=28.51 Go [2026-07-16 14:32:40] step 36200/1000000 | loss=2.8083 vocab=0.9525 token=1.6541 | 6.09 steps/s | VRAM=25.66 Go [2026-07-16 14:32:49] step 36250/1000000 | loss=2.8326 vocab=1.3677 token=1.8355 | 6.07 steps/s | VRAM=32.62 Go [2026-07-16 14:32:57] step 36300/1000000 | loss=2.8124 vocab=0.9196 token=1.4956 | 6.07 steps/s | VRAM=24.44 Go [2026-07-16 14:33:05] step 36350/1000000 | loss=2.7377 vocab=0.7957 token=1.1646 | 6.07 steps/s | VRAM=30.03 Go [2026-07-16 14:33:13] step 36400/1000000 | loss=2.8214 vocab=1.2018 token=1.8582 | 6.09 steps/s | VRAM=28.20 Go [2026-07-16 14:33:22] step 36450/1000000 | loss=2.8989 vocab=1.2888 token=1.8720 | 6.05 steps/s | VRAM=28.99 Go [2026-07-16 14:33:30] step 36500/1000000 | loss=2.8347 vocab=1.8819 token=1.9105 | 6.08 steps/s | VRAM=31.60 Go [2026-07-16 14:33:38] step 36550/1000000 | loss=2.8733 vocab=1.4524 token=1.8584 | 6.07 steps/s | VRAM=30.79 Go [2026-07-16 14:33:46] step 36600/1000000 | loss=2.7297 vocab=1.4440 token=1.7968 | 6.08 steps/s | VRAM=26.86 Go [2026-07-16 14:33:55] step 36650/1000000 | loss=2.8874 vocab=0.6896 token=1.3184 | 6.06 steps/s | VRAM=28.74 Go [2026-07-16 14:34:03] step 36700/1000000 | loss=2.8003 vocab=1.1136 token=1.5194 | 6.04 steps/s | VRAM=26.03 Go [2026-07-16 14:34:11] step 36750/1000000 | loss=2.7282 vocab=1.0568 token=1.6254 | 6.08 steps/s | VRAM=31.80 Go [2026-07-16 14:34:19] step 36800/1000000 | loss=2.7692 vocab=1.5016 token=1.8040 | 6.03 steps/s | VRAM=28.16 Go [2026-07-16 14:34:28] step 36850/1000000 | loss=2.8357 vocab=1.0531 token=1.8897 | 6.08 steps/s | VRAM=26.16 Go [2026-07-16 14:34:36] step 36900/1000000 | loss=2.8915 vocab=1.0633 token=1.6251 | 6.08 steps/s | VRAM=29.54 Go [2026-07-16 14:34:44] step 36950/1000000 | loss=2.8190 vocab=1.2317 token=1.7595 | 6.09 steps/s | VRAM=32.53 Go [2026-07-16 14:34:52] step 37000/1000000 | loss=2.7488 vocab=1.7480 token=1.5509 | 6.08 steps/s | VRAM=28.81 Go [2026-07-16 14:35:01] step 37050/1000000 | loss=2.8757 vocab=1.5712 token=1.8197 | 6.07 steps/s | VRAM=29.67 Go [2026-07-16 14:35:09] step 37100/1000000 | loss=2.7544 vocab=0.9965 token=1.3855 | 6.06 steps/s | VRAM=32.37 Go [2026-07-16 14:35:17] step 37150/1000000 | loss=2.8564 vocab=1.1536 token=1.8431 | 6.05 steps/s | VRAM=24.52 Go [2026-07-16 14:35:25] step 37200/1000000 | loss=2.7594 vocab=1.0330 token=1.6532 | 6.11 steps/s | VRAM=28.35 Go [2026-07-16 14:35:33] step 37250/1000000 | loss=2.6887 vocab=1.1462 token=1.6104 | 6.10 steps/s | VRAM=29.54 Go [2026-07-16 14:35:42] step 37300/1000000 | loss=2.9052 vocab=1.0633 token=1.5203 | 6.07 steps/s | VRAM=31.82 Go [2026-07-16 14:35:50] step 37350/1000000 | loss=2.8011 vocab=1.0285 token=1.7130 | 6.05 steps/s | VRAM=26.87 Go [2026-07-16 14:35:58] step 37400/1000000 | loss=2.8413 vocab=0.8294 token=1.3322 | 6.09 steps/s | VRAM=31.65 Go [2026-07-16 14:36:06] step 37450/1000000 | loss=2.8806 vocab=1.0935 token=1.9495 | 6.06 steps/s | VRAM=26.49 Go [2026-07-16 14:36:15] step 37500/1000000 | loss=2.8188 vocab=0.8481 token=1.0823 | 6.10 steps/s | VRAM=25.72 Go [2026-07-16 14:36:23] step 37550/1000000 | loss=2.8979 vocab=1.2648 token=1.5623 | 6.09 steps/s | VRAM=31.01 Go [2026-07-16 14:36:31] step 37600/1000000 | loss=2.7093 vocab=1.0838 token=1.6109 | 6.08 steps/s | VRAM=31.87 Go [2026-07-16 14:36:39] step 37650/1000000 | loss=2.7593 vocab=1.0984 token=1.8385 | 6.09 steps/s | VRAM=32.01 Go [2026-07-16 14:36:47] step 37700/1000000 | loss=2.8168 vocab=1.1916 token=1.4782 | 6.06 steps/s | VRAM=26.82 Go [2026-07-16 14:36:56] step 37750/1000000 | loss=2.8058 vocab=1.3099 token=1.8425 | 6.05 steps/s | VRAM=26.21 Go [2026-07-16 14:37:04] step 37800/1000000 | loss=2.8733 vocab=1.2384 token=1.7635 | 6.07 steps/s | VRAM=28.68 Go [2026-07-16 14:37:12] step 37850/1000000 | loss=2.8641 vocab=1.3319 token=1.6519 | 6.08 steps/s | VRAM=28.67 Go [2026-07-16 14:37:20] step 37900/1000000 | loss=2.6897 vocab=1.2677 token=1.9476 | 6.09 steps/s | VRAM=24.94 Go [2026-07-16 14:37:29] step 37950/1000000 | loss=2.8541 vocab=1.0095 token=1.6501 | 6.08 steps/s | VRAM=32.58 Go [2026-07-16 14:37:37] step 38000/1000000 | loss=2.8762 vocab=0.9221 token=1.5737 | 6.08 steps/s | VRAM=28.97 Go [2026-07-16 14:37:45] step 38050/1000000 | loss=2.8076 vocab=0.9600 token=1.6180 | 6.04 steps/s | VRAM=26.15 Go [2026-07-16 14:37:53] step 38100/1000000 | loss=2.8117 vocab=0.8904 token=1.5625 | 6.08 steps/s | VRAM=25.95 Go [2026-07-16 14:38:02] step 38150/1000000 | loss=2.7526 vocab=1.6142 token=1.3312 | 6.09 steps/s | VRAM=28.19 Go [2026-07-16 14:38:10] step 38200/1000000 | loss=2.7405 vocab=0.9543 token=1.6417 | 6.07 steps/s | VRAM=26.18 Go [2026-07-16 14:38:18] step 38250/1000000 | loss=2.8017 vocab=1.0369 token=1.7555 | 6.06 steps/s | VRAM=29.62 Go [2026-07-16 14:38:26] step 38300/1000000 | loss=2.7896 vocab=0.6381 token=1.6883 | 6.06 steps/s | VRAM=28.97 Go [2026-07-16 14:38:35] step 38350/1000000 | loss=2.7674 vocab=1.2942 token=1.8206 | 6.07 steps/s | VRAM=32.64 Go [2026-07-16 14:38:43] step 38400/1000000 | loss=2.8264 vocab=1.3874 token=1.5772 | 6.08 steps/s | VRAM=26.18 Go [2026-07-16 14:38:51] step 38450/1000000 | loss=2.8848 vocab=1.2216 token=1.9058 | 6.06 steps/s | VRAM=32.91 Go [2026-07-16 14:38:59] step 38500/1000000 | loss=2.7459 vocab=0.9925 token=1.8333 | 6.08 steps/s | VRAM=29.59 Go [2026-07-16 14:39:07] step 38550/1000000 | loss=2.7119 vocab=0.9719 token=1.4411 | 6.08 steps/s | VRAM=28.41 Go [2026-07-16 14:39:16] step 38600/1000000 | loss=2.7930 vocab=1.2806 token=1.8223 | 6.10 steps/s | VRAM=25.89 Go [2026-07-16 14:39:24] step 38650/1000000 | loss=2.7448 vocab=0.8741 token=1.6103 | 6.07 steps/s | VRAM=27.68 Go [2026-07-16 14:39:32] step 38700/1000000 | loss=2.8518 vocab=1.0330 token=1.7505 | 6.07 steps/s | VRAM=25.99 Go [2026-07-16 14:39:40] step 38750/1000000 | loss=2.7366 vocab=0.7424 token=1.7991 | 6.10 steps/s | VRAM=25.72 Go [2026-07-16 14:39:49] step 38800/1000000 | loss=2.8066 vocab=0.9284 token=1.5209 | 6.09 steps/s | VRAM=24.98 Go [2026-07-16 14:39:57] step 38850/1000000 | loss=2.8895 vocab=0.6773 token=1.8185 | 6.10 steps/s | VRAM=28.42 Go [2026-07-16 14:40:05] step 38900/1000000 | loss=2.7448 vocab=1.3258 token=1.2427 | 6.08 steps/s | VRAM=24.63 Go [2026-07-16 14:40:13] step 38950/1000000 | loss=2.8764 vocab=1.3461 token=1.3124 | 6.10 steps/s | VRAM=29.42 Go [2026-07-16 14:40:21] step 39000/1000000 | loss=2.8747 vocab=1.0319 token=1.9667 | 6.06 steps/s | VRAM=25.90 Go [2026-07-16 14:40:30] step 39050/1000000 | loss=2.8355 vocab=1.4772 token=1.7628 | 6.10 steps/s | VRAM=25.90 Go [2026-07-16 14:40:38] step 39100/1000000 | loss=2.8822 vocab=0.9778 token=1.5434 | 6.07 steps/s | VRAM=30.18 Go [2026-07-16 14:40:46] step 39150/1000000 | loss=2.7602 vocab=0.9143 token=1.8620 | 6.07 steps/s | VRAM=28.42 Go [2026-07-16 14:40:54] step 39200/1000000 | loss=2.6903 vocab=1.1326 token=1.4763 | 6.06 steps/s | VRAM=29.86 Go [2026-07-16 14:41:03] step 39250/1000000 | loss=2.7623 vocab=1.2987 token=1.6339 | 6.06 steps/s | VRAM=28.37 Go [2026-07-16 14:41:11] step 39300/1000000 | loss=2.7284 vocab=1.0429 token=1.7477 | 6.08 steps/s | VRAM=28.78 Go [2026-07-16 14:41:19] step 39350/1000000 | loss=2.7879 vocab=1.2290 token=1.9023 | 6.08 steps/s | VRAM=32.21 Go [2026-07-16 14:41:27] step 39400/1000000 | loss=2.8660 vocab=1.3344 token=1.6822 | 6.04 steps/s | VRAM=29.51 Go [2026-07-16 14:41:36] step 39450/1000000 | loss=2.7003 vocab=1.2937 token=1.5806 | 6.09 steps/s | VRAM=32.09 Go [2026-07-16 14:41:44] step 39500/1000000 | loss=2.7795 vocab=1.3402 token=1.8478 | 6.09 steps/s | VRAM=28.16 Go [2026-07-16 14:41:52] step 39550/1000000 | loss=2.7715 vocab=1.2500 token=1.5027 | 6.09 steps/s | VRAM=25.79 Go [2026-07-16 14:42:00] step 39600/1000000 | loss=2.8171 vocab=1.0008 token=1.4528 | 6.09 steps/s | VRAM=28.50 Go [2026-07-16 14:42:08] step 39650/1000000 | loss=2.8052 vocab=0.4939 token=1.3772 | 6.09 steps/s | VRAM=25.68 Go [2026-07-16 14:42:17] step 39700/1000000 | loss=2.7788 vocab=1.0499 token=1.3884 | 6.08 steps/s | VRAM=31.33 Go [2026-07-16 14:42:25] step 39750/1000000 | loss=2.6920 vocab=1.2768 token=1.5157 | 6.06 steps/s | VRAM=28.79 Go [2026-07-16 14:42:33] step 39800/1000000 | loss=2.7709 vocab=0.9833 token=1.8241 | 6.09 steps/s | VRAM=32.57 Go [2026-07-16 14:42:41] step 39850/1000000 | loss=2.7207 vocab=1.1452 token=1.5781 | 6.08 steps/s | VRAM=27.88 Go [2026-07-16 14:42:50] step 39900/1000000 | loss=2.8811 vocab=1.3034 token=1.7122 | 6.06 steps/s | VRAM=27.96 Go [2026-07-16 14:42:58] step 39950/1000000 | loss=2.7484 vocab=1.0987 token=1.6038 | 6.10 steps/s | VRAM=30.35 Go [2026-07-16 14:43:06] step 40000/1000000 | loss=2.7819 vocab=1.4021 token=1.5289 | 6.10 steps/s | VRAM=28.09 Go [2026-07-16 14:44:44] step 40050/1000000 | loss=2.7870 vocab=1.0277 token=1.7012 | 0.51 steps/s | VRAM=29.39 Go [2026-07-16 14:44:52] step 40100/1000000 | loss=2.8332 vocab=1.6216 token=2.0557 | 6.13 steps/s | VRAM=29.82 Go [2026-07-16 14:45:00] step 40150/1000000 | loss=2.6899 vocab=0.8339 token=1.7054 | 6.15 steps/s | VRAM=28.09 Go [2026-07-16 14:45:08] step 40200/1000000 | loss=2.8133 vocab=0.9980 token=1.8037 | 6.14 steps/s | VRAM=28.26 Go [2026-07-16 14:45:17] step 40250/1000000 | loss=2.7772 vocab=1.0098 token=1.6001 | 6.15 steps/s | VRAM=27.54 Go [2026-07-16 14:45:25] step 40300/1000000 | loss=2.7770 vocab=0.8968 token=1.3266 | 6.12 steps/s | VRAM=32.40 Go [2026-07-16 14:45:33] step 40350/1000000 | loss=2.7319 vocab=1.0565 token=1.3791 | 6.08 steps/s | VRAM=23.75 Go [2026-07-16 14:45:41] step 40400/1000000 | loss=2.7413 vocab=0.8745 token=1.5212 | 6.12 steps/s | VRAM=28.71 Go [2026-07-16 14:45:49] step 40450/1000000 | loss=2.7994 vocab=1.2827 token=1.5200 | 6.08 steps/s | VRAM=29.44 Go [2026-07-16 14:45:58] step 40500/1000000 | loss=2.8167 vocab=1.1444 token=1.9615 | 6.08 steps/s | VRAM=28.27 Go [2026-07-16 14:46:06] step 40550/1000000 | loss=2.8005 vocab=1.5813 token=1.7096 | 6.11 steps/s | VRAM=31.30 Go [2026-07-16 14:46:14] step 40600/1000000 | loss=2.7257 vocab=1.2027 token=1.8763 | 6.05 steps/s | VRAM=26.99 Go [2026-07-16 14:46:22] step 40650/1000000 | loss=2.7753 vocab=1.2682 token=1.6456 | 6.08 steps/s | VRAM=28.59 Go [2026-07-16 14:46:31] step 40700/1000000 | loss=2.8700 vocab=1.3593 token=1.9284 | 6.08 steps/s | VRAM=31.51 Go [2026-07-16 14:46:39] step 40750/1000000 | loss=2.8841 vocab=1.1945 token=1.8386 | 6.04 steps/s | VRAM=27.08 Go [2026-07-16 14:46:47] step 40800/1000000 | loss=2.7734 vocab=1.1109 token=1.7535 | 6.09 steps/s | VRAM=29.22 Go [2026-07-16 14:46:55] step 40850/1000000 | loss=2.8317 vocab=0.8133 token=1.7618 | 6.03 steps/s | VRAM=29.35 Go [2026-07-16 14:47:04] step 40900/1000000 | loss=2.6940 vocab=0.7257 token=1.7731 | 6.08 steps/s | VRAM=27.71 Go [2026-07-16 14:47:12] step 40950/1000000 | loss=2.8229 vocab=1.2716 token=1.7087 | 6.07 steps/s | VRAM=25.70 Go [2026-07-16 14:47:20] step 41000/1000000 | loss=2.8244 vocab=1.1944 token=1.6893 | 6.06 steps/s | VRAM=28.42 Go [2026-07-16 14:47:28] step 41050/1000000 | loss=2.8231 vocab=0.8856 token=1.7094 | 6.06 steps/s | VRAM=25.09 Go [2026-07-16 14:47:37] step 41100/1000000 | loss=2.6457 vocab=1.1771 token=1.6643 | 6.04 steps/s | VRAM=28.57 Go [2026-07-16 14:47:45] step 41150/1000000 | loss=2.8531 vocab=1.8931 token=1.8386 | 6.07 steps/s | VRAM=32.45 Go [2026-07-16 14:47:53] step 41200/1000000 | loss=2.8167 vocab=0.8126 token=1.6493 | 6.06 steps/s | VRAM=31.85 Go [2026-07-16 14:48:01] step 41250/1000000 | loss=2.7793 vocab=0.8780 token=1.2229 | 6.06 steps/s | VRAM=25.51 Go [2026-07-16 14:48:10] step 41300/1000000 | loss=2.7517 vocab=1.4085 token=1.6486 | 6.05 steps/s | VRAM=29.24 Go [2026-07-16 14:48:18] step 41350/1000000 | loss=2.7517 vocab=1.0707 token=1.6311 | 6.08 steps/s | VRAM=32.04 Go [2026-07-16 14:48:26] step 41400/1000000 | loss=2.7179 vocab=1.1750 token=1.5558 | 6.04 steps/s | VRAM=29.29 Go [2026-07-16 14:48:34] step 41450/1000000 | loss=2.7727 vocab=1.1565 token=1.3788 | 6.04 steps/s | VRAM=26.28 Go [2026-07-16 14:48:43] step 41500/1000000 | loss=2.7505 vocab=0.8422 token=1.4889 | 6.08 steps/s | VRAM=28.31 Go [2026-07-16 14:48:51] step 41550/1000000 | loss=2.7685 vocab=1.0396 token=1.6363 | 6.05 steps/s | VRAM=29.25 Go [2026-07-16 14:48:59] step 41600/1000000 | loss=2.6522 vocab=0.6602 token=1.4990 | 6.06 steps/s | VRAM=32.54 Go [2026-07-16 14:49:07] step 41650/1000000 | loss=2.6534 vocab=1.1187 token=1.5645 | 6.04 steps/s | VRAM=29.81 Go [2026-07-16 14:49:16] step 41700/1000000 | loss=2.8227 vocab=1.0704 token=1.4378 | 6.04 steps/s | VRAM=31.63 Go [2026-07-16 14:49:24] step 41750/1000000 | loss=2.7750 vocab=0.7588 token=1.6791 | 6.05 steps/s | VRAM=31.44 Go [2026-07-16 14:49:32] step 41800/1000000 | loss=2.7916 vocab=1.0355 token=1.5460 | 6.08 steps/s | VRAM=25.88 Go [2026-07-16 14:49:40] step 41850/1000000 | loss=2.8213 vocab=1.1584 token=1.7601 | 6.06 steps/s | VRAM=24.75 Go [2026-07-16 14:49:49] step 41900/1000000 | loss=2.8724 vocab=1.6662 token=1.9932 | 6.05 steps/s | VRAM=28.88 Go [2026-07-16 14:49:57] step 41950/1000000 | loss=2.7050 vocab=0.7458 token=1.6338 | 6.04 steps/s | VRAM=29.11 Go [2026-07-16 14:50:05] step 42000/1000000 | loss=2.7387 vocab=1.2654 token=1.6605 | 6.07 steps/s | VRAM=29.03 Go [2026-07-16 14:50:13] step 42050/1000000 | loss=2.7781 vocab=0.8164 token=1.9013 | 6.06 steps/s | VRAM=28.18 Go [2026-07-16 14:50:22] step 42100/1000000 | loss=2.7808 vocab=1.1123 token=1.7610 | 6.06 steps/s | VRAM=29.27 Go [2026-07-16 14:50:30] step 42150/1000000 | loss=2.7716 vocab=0.9859 token=1.6613 | 6.07 steps/s | VRAM=32.46 Go [2026-07-16 14:50:38] step 42200/1000000 | loss=2.7663 vocab=0.7307 token=1.3980 | 6.06 steps/s | VRAM=28.20 Go [2026-07-16 14:50:46] step 42250/1000000 | loss=2.8196 vocab=0.9406 token=1.4288 | 6.06 steps/s | VRAM=26.49 Go [2026-07-16 14:50:55] step 42300/1000000 | loss=2.7488 vocab=1.6307 token=1.6865 | 6.09 steps/s | VRAM=25.20 Go [2026-07-16 14:51:03] step 42350/1000000 | loss=2.7566 vocab=0.8734 token=1.5566 | 6.11 steps/s | VRAM=31.74 Go [2026-07-16 14:51:11] step 42400/1000000 | loss=2.7749 vocab=1.5646 token=1.7870 | 6.04 steps/s | VRAM=28.33 Go [2026-07-16 14:51:19] step 42450/1000000 | loss=2.7063 vocab=0.8951 token=1.4499 | 6.07 steps/s | VRAM=26.57 Go [2026-07-16 14:51:27] step 42500/1000000 | loss=2.6667 vocab=0.7531 token=1.7562 | 6.10 steps/s | VRAM=31.74 Go [2026-07-16 14:51:36] step 42550/1000000 | loss=2.7422 vocab=1.7850 token=1.8896 | 6.08 steps/s | VRAM=28.98 Go [2026-07-16 14:51:44] step 42600/1000000 | loss=2.7758 vocab=1.0496 token=1.8034 | 6.10 steps/s | VRAM=31.60 Go [2026-07-16 14:51:52] step 42650/1000000 | loss=2.7120 vocab=1.2440 token=1.7906 | 6.07 steps/s | VRAM=28.79 Go [2026-07-16 14:52:00] step 42700/1000000 | loss=2.6997 vocab=1.0257 token=1.7845 | 6.09 steps/s | VRAM=32.13 Go [2026-07-16 14:52:09] step 42750/1000000 | loss=2.7867 vocab=1.1591 token=2.1638 | 6.06 steps/s | VRAM=26.26 Go [2026-07-16 14:52:17] step 42800/1000000 | loss=2.7425 vocab=1.6699 token=1.7469 | 6.08 steps/s | VRAM=24.27 Go [2026-07-16 14:52:25] step 42850/1000000 | loss=2.7168 vocab=0.9549 token=1.1666 | 6.06 steps/s | VRAM=28.43 Go [2026-07-16 14:52:33] step 42900/1000000 | loss=2.7921 vocab=1.1129 token=1.6236 | 6.06 steps/s | VRAM=28.56 Go [2026-07-16 14:52:42] step 42950/1000000 | loss=2.7970 vocab=0.8234 token=1.5022 | 6.07 steps/s | VRAM=32.45 Go [2026-07-16 14:52:50] step 43000/1000000 | loss=2.8549 vocab=1.4202 token=2.1470 | 6.09 steps/s | VRAM=25.33 Go [2026-07-16 14:52:58] step 43050/1000000 | loss=2.7782 vocab=1.0932 token=1.4106 | 6.04 steps/s | VRAM=26.10 Go [2026-07-16 14:53:06] step 43100/1000000 | loss=2.7514 vocab=0.7096 token=1.5312 | 6.09 steps/s | VRAM=32.26 Go [2026-07-16 14:53:14] step 43150/1000000 | loss=2.7291 vocab=0.8588 token=1.6237 | 6.06 steps/s | VRAM=26.59 Go [2026-07-16 14:53:23] step 43200/1000000 | loss=2.8061 vocab=1.2323 token=1.5827 | 6.09 steps/s | VRAM=32.49 Go [2026-07-16 14:53:31] step 43250/1000000 | loss=2.7589 vocab=0.8479 token=1.3379 | 6.07 steps/s | VRAM=27.86 Go [2026-07-16 14:53:39] step 43300/1000000 | loss=2.7527 vocab=0.9880 token=1.3157 | 6.10 steps/s | VRAM=29.25 Go [2026-07-16 14:53:47] step 43350/1000000 | loss=2.7911 vocab=1.2724 token=1.1875 | 6.09 steps/s | VRAM=25.34 Go [2026-07-16 14:53:56] step 43400/1000000 | loss=2.7344 vocab=1.1411 token=1.6194 | 6.08 steps/s | VRAM=25.89 Go [2026-07-16 14:54:04] step 43450/1000000 | loss=2.6721 vocab=1.0120 token=1.7046 | 6.07 steps/s | VRAM=28.49 Go [2026-07-16 14:54:12] step 43500/1000000 | loss=2.7546 vocab=0.9392 token=1.7798 | 6.07 steps/s | VRAM=30.25 Go [2026-07-16 14:54:20] step 43550/1000000 | loss=2.7065 vocab=0.9873 token=1.6548 | 6.07 steps/s | VRAM=29.04 Go [2026-07-16 14:54:28] step 43600/1000000 | loss=2.6812 vocab=1.1935 token=1.8561 | 6.09 steps/s | VRAM=29.00 Go [2026-07-16 14:54:37] step 43650/1000000 | loss=2.6632 vocab=1.3663 token=1.6163 | 6.08 steps/s | VRAM=32.13 Go [2026-07-16 14:54:45] step 43700/1000000 | loss=2.7419 vocab=1.4427 token=1.7781 | 6.10 steps/s | VRAM=24.57 Go [2026-07-16 14:54:53] step 43750/1000000 | loss=2.8004 vocab=2.0006 token=1.8761 | 6.09 steps/s | VRAM=28.46 Go [2026-07-16 14:55:01] step 43800/1000000 | loss=2.7720 vocab=1.1785 token=1.5829 | 6.10 steps/s | VRAM=29.38 Go [2026-07-16 14:55:09] step 43850/1000000 | loss=2.6975 vocab=0.9591 token=1.5378 | 6.12 steps/s | VRAM=31.32 Go [2026-07-16 14:55:18] step 43900/1000000 | loss=2.7988 vocab=1.0209 token=1.5929 | 6.07 steps/s | VRAM=32.85 Go [2026-07-16 14:55:26] step 43950/1000000 | loss=2.7207 vocab=1.4435 token=1.4521 | 6.06 steps/s | VRAM=29.02 Go [2026-07-16 14:55:34] step 44000/1000000 | loss=2.7431 vocab=0.8480 token=1.3745 | 6.07 steps/s | VRAM=26.42 Go [2026-07-16 14:55:42] step 44050/1000000 | loss=2.6282 vocab=1.3531 token=1.4745 | 6.11 steps/s | VRAM=31.72 Go [2026-07-16 14:55:51] step 44100/1000000 | loss=2.8207 vocab=1.4919 token=1.9680 | 6.11 steps/s | VRAM=31.67 Go [2026-07-16 14:55:59] step 44150/1000000 | loss=2.6922 vocab=1.0264 token=1.8858 | 6.06 steps/s | VRAM=25.86 Go [2026-07-16 14:56:07] step 44200/1000000 | loss=2.7259 vocab=1.4073 token=1.5732 | 6.10 steps/s | VRAM=25.39 Go [2026-07-16 14:56:15] step 44250/1000000 | loss=2.6780 vocab=0.7947 token=1.6318 | 6.10 steps/s | VRAM=31.68 Go [2026-07-16 14:56:23] step 44300/1000000 | loss=2.7008 vocab=0.7793 token=1.3646 | 6.07 steps/s | VRAM=29.29 Go [2026-07-16 14:56:32] step 44350/1000000 | loss=2.7419 vocab=1.0149 token=1.8667 | 6.09 steps/s | VRAM=28.55 Go [2026-07-16 14:56:40] step 44400/1000000 | loss=2.6979 vocab=0.8122 token=1.3300 | 6.08 steps/s | VRAM=28.50 Go [2026-07-16 14:56:48] step 44450/1000000 | loss=2.7684 vocab=1.1623 token=1.5533 | 6.10 steps/s | VRAM=32.79 Go [2026-07-16 14:56:56] step 44500/1000000 | loss=2.6509 vocab=0.8525 token=1.5553 | 6.11 steps/s | VRAM=27.78 Go [2026-07-16 14:57:04] step 44550/1000000 | loss=2.6954 vocab=1.2064 token=1.7886 | 6.09 steps/s | VRAM=31.81 Go [2026-07-16 14:57:13] step 44600/1000000 | loss=2.6685 vocab=0.8845 token=1.7901 | 6.08 steps/s | VRAM=29.99 Go [2026-07-16 14:57:21] step 44650/1000000 | loss=2.6911 vocab=1.0688 token=1.5848 | 6.09 steps/s | VRAM=31.90 Go [2026-07-16 14:57:29] step 44700/1000000 | loss=2.6927 vocab=0.8374 token=1.2652 | 6.10 steps/s | VRAM=31.05 Go [2026-07-16 14:57:37] step 44750/1000000 | loss=2.7414 vocab=0.8539 token=1.5455 | 6.09 steps/s | VRAM=27.76 Go [2026-07-16 14:57:46] step 44800/1000000 | loss=2.7767 vocab=1.2873 token=1.4553 | 6.09 steps/s | VRAM=32.88 Go [2026-07-16 14:57:54] step 44850/1000000 | loss=2.7306 vocab=1.1263 token=1.8248 | 6.09 steps/s | VRAM=29.07 Go [2026-07-16 14:58:02] step 44900/1000000 | loss=2.7312 vocab=0.7972 token=1.5948 | 6.09 steps/s | VRAM=24.89 Go [2026-07-16 14:58:10] step 44950/1000000 | loss=2.6637 vocab=1.2910 token=1.4039 | 6.06 steps/s | VRAM=28.81 Go [2026-07-16 14:58:18] step 45000/1000000 | loss=2.6950 vocab=1.2556 token=1.8418 | 6.08 steps/s | VRAM=28.72 Go [2026-07-16 14:59:46] step 45050/1000000 | loss=2.7033 vocab=0.7626 token=1.7227 | 0.57 steps/s | VRAM=28.43 Go [2026-07-16 14:59:54] step 45100/1000000 | loss=2.6832 vocab=0.8660 token=1.8705 | 6.16 steps/s | VRAM=24.71 Go [2026-07-16 15:00:02] step 45150/1000000 | loss=2.7174 vocab=0.8616 token=1.7434 | 6.15 steps/s | VRAM=32.18 Go [2026-07-16 15:00:10] step 45200/1000000 | loss=2.6935 vocab=1.1717 token=1.4654 | 6.11 steps/s | VRAM=26.03 Go [2026-07-16 15:00:19] step 45250/1000000 | loss=2.7677 vocab=1.2863 token=1.9005 | 6.14 steps/s | VRAM=29.05 Go [2026-07-16 15:00:27] step 45300/1000000 | loss=2.7387 vocab=0.9566 token=1.6501 | 6.16 steps/s | VRAM=32.05 Go [2026-07-16 15:00:35] step 45350/1000000 | loss=2.6590 vocab=1.0596 token=1.3116 | 6.11 steps/s | VRAM=26.04 Go [2026-07-16 15:00:43] step 45400/1000000 | loss=2.6805 vocab=0.9805 token=1.5134 | 6.13 steps/s | VRAM=28.73 Go [2026-07-16 15:00:51] step 45450/1000000 | loss=2.8120 vocab=1.1013 token=1.7508 | 6.10 steps/s | VRAM=32.23 Go [2026-07-16 15:00:59] step 45500/1000000 | loss=2.8113 vocab=1.1986 token=1.8063 | 6.09 steps/s | VRAM=32.43 Go [2026-07-16 15:01:08] step 45550/1000000 | loss=2.6507 vocab=1.0833 token=1.8942 | 6.10 steps/s | VRAM=32.25 Go [2026-07-16 15:01:16] step 45600/1000000 | loss=2.6757 vocab=0.6885 token=1.0878 | 6.09 steps/s | VRAM=31.49 Go [2026-07-16 15:01:24] step 45650/1000000 | loss=2.7730 vocab=0.9753 token=1.6706 | 6.10 steps/s | VRAM=28.80 Go [2026-07-16 15:01:32] step 45700/1000000 | loss=2.7478 vocab=1.2670 token=1.7686 | 6.10 steps/s | VRAM=32.51 Go [2026-07-16 15:01:40] step 45750/1000000 | loss=2.7405 vocab=1.0198 token=1.9466 | 6.10 steps/s | VRAM=29.13 Go [2026-07-16 15:01:49] step 45800/1000000 | loss=2.7640 vocab=1.1054 token=1.4857 | 6.07 steps/s | VRAM=32.39 Go [2026-07-16 15:01:57] step 45850/1000000 | loss=2.7327 vocab=1.7583 token=1.6332 | 6.11 steps/s | VRAM=31.20 Go [2026-07-16 15:02:05] step 45900/1000000 | loss=2.7197 vocab=0.9732 token=1.6737 | 6.05 steps/s | VRAM=28.19 Go [2026-07-16 15:02:13] step 45950/1000000 | loss=2.7794 vocab=1.2752 token=1.8094 | 6.08 steps/s | VRAM=25.14 Go [2026-07-16 15:02:22] step 46000/1000000 | loss=2.6851 vocab=2.1057 token=2.1603 | 6.09 steps/s | VRAM=32.15 Go [2026-07-16 15:02:30] step 46050/1000000 | loss=2.6934 vocab=1.6491 token=1.7189 | 6.07 steps/s | VRAM=28.93 Go [2026-07-16 15:02:38] step 46100/1000000 | loss=2.7218 vocab=1.1171 token=1.7656 | 6.06 steps/s | VRAM=28.31 Go [2026-07-16 15:02:46] step 46150/1000000 | loss=2.7802 vocab=1.2127 token=1.6128 | 6.08 steps/s | VRAM=24.11 Go [2026-07-16 15:02:55] step 46200/1000000 | loss=2.7565 vocab=0.7396 token=1.3653 | 6.09 steps/s | VRAM=28.27 Go [2026-07-16 15:03:03] step 46250/1000000 | loss=2.6639 vocab=1.0787 token=1.9606 | 6.07 steps/s | VRAM=31.07 Go [2026-07-16 15:03:11] step 46300/1000000 | loss=2.6256 vocab=1.3751 token=1.9708 | 6.10 steps/s | VRAM=30.78 Go [2026-07-16 15:03:19] step 46350/1000000 | loss=2.7483 vocab=1.2824 token=1.2615 | 6.08 steps/s | VRAM=28.18 Go [2026-07-16 15:03:27] step 46400/1000000 | loss=2.7462 vocab=1.3159 token=1.6118 | 6.10 steps/s | VRAM=27.30 Go [2026-07-16 15:03:36] step 46450/1000000 | loss=2.6825 vocab=1.1669 token=1.8130 | 6.11 steps/s | VRAM=31.24 Go [2026-07-16 15:03:44] step 46500/1000000 | loss=2.6729 vocab=1.3518 token=2.2081 | 6.09 steps/s | VRAM=31.15 Go [2026-07-16 15:03:52] step 46550/1000000 | loss=2.7532 vocab=1.1784 token=1.5526 | 6.10 steps/s | VRAM=27.25 Go [2026-07-16 15:04:00] step 46600/1000000 | loss=2.7121 vocab=1.3438 token=1.5437 | 6.09 steps/s | VRAM=24.26 Go [2026-07-16 15:04:08] step 46650/1000000 | loss=2.6585 vocab=1.0454 token=1.6065 | 6.10 steps/s | VRAM=27.62 Go [2026-07-16 15:04:17] step 46700/1000000 | loss=2.6736 vocab=0.9199 token=1.3679 | 6.09 steps/s | VRAM=27.96 Go [2026-07-16 15:04:25] step 46750/1000000 | loss=2.7671 vocab=0.7971 token=1.5882 | 6.07 steps/s | VRAM=28.25 Go [2026-07-16 15:04:33] step 46800/1000000 | loss=2.6945 vocab=0.7883 token=1.5410 | 6.10 steps/s | VRAM=27.93 Go [2026-07-16 15:04:41] step 46850/1000000 | loss=2.8519 vocab=0.8837 token=1.5384 | 6.08 steps/s | VRAM=28.80 Go [2026-07-16 15:04:49] step 46900/1000000 | loss=2.7774 vocab=1.1623 token=2.0151 | 6.08 steps/s | VRAM=28.56 Go [2026-07-16 15:04:58] step 46950/1000000 | loss=2.6743 vocab=0.6921 token=1.5090 | 6.09 steps/s | VRAM=29.11 Go [2026-07-16 15:05:06] step 47000/1000000 | loss=2.6377 vocab=1.0096 token=1.8025 | 6.03 steps/s | VRAM=28.43 Go [2026-07-16 15:05:14] step 47050/1000000 | loss=2.6981 vocab=1.3167 token=1.7684 | 6.10 steps/s | VRAM=31.21 Go [2026-07-16 15:05:22] step 47100/1000000 | loss=2.6912 vocab=1.0169 token=1.4870 | 6.10 steps/s | VRAM=28.33 Go [2026-07-16 15:05:31] step 47150/1000000 | loss=2.6940 vocab=1.2298 token=1.4332 | 6.09 steps/s | VRAM=30.41 Go [2026-07-16 15:05:39] step 47200/1000000 | loss=2.7657 vocab=1.4875 token=1.8891 | 6.09 steps/s | VRAM=28.52 Go [2026-07-16 15:05:47] step 47250/1000000 | loss=2.6686 vocab=1.2701 token=1.7128 | 6.09 steps/s | VRAM=29.64 Go [2026-07-16 15:05:55] step 47300/1000000 | loss=2.7551 vocab=1.2270 token=1.4195 | 6.13 steps/s | VRAM=31.93 Go [2026-07-16 15:06:03] step 47350/1000000 | loss=2.5954 vocab=1.3002 token=1.8120 | 6.07 steps/s | VRAM=28.21 Go [2026-07-16 15:06:12] step 47400/1000000 | loss=2.6817 vocab=1.2553 token=1.3979 | 6.09 steps/s | VRAM=26.30 Go [2026-07-16 15:06:20] step 47450/1000000 | loss=2.7882 vocab=0.6736 token=1.1232 | 6.07 steps/s | VRAM=26.46 Go [2026-07-16 15:06:28] step 47500/1000000 | loss=2.7427 vocab=0.9340 token=1.4148 | 6.09 steps/s | VRAM=31.50 Go [2026-07-16 15:06:36] step 47550/1000000 | loss=2.6787 vocab=1.2373 token=1.4045 | 6.07 steps/s | VRAM=26.08 Go [2026-07-16 15:06:45] step 47600/1000000 | loss=2.7332 vocab=1.0308 token=1.6756 | 6.07 steps/s | VRAM=28.92 Go [2026-07-16 15:06:53] step 47650/1000000 | loss=2.7886 vocab=1.2535 token=1.8136 | 6.10 steps/s | VRAM=27.96 Go [2026-07-16 15:07:01] step 47700/1000000 | loss=2.6854 vocab=1.3300 token=1.6119 | 6.10 steps/s | VRAM=32.68 Go [2026-07-16 15:07:09] step 47750/1000000 | loss=2.8476 vocab=0.9179 token=1.8564 | 6.09 steps/s | VRAM=28.13 Go [2026-07-16 15:07:17] step 47800/1000000 | loss=2.6878 vocab=1.2415 token=1.9041 | 6.11 steps/s | VRAM=32.46 Go [2026-07-16 15:07:25] step 47850/1000000 | loss=2.7606 vocab=1.2209 token=1.5621 | 6.11 steps/s | VRAM=31.76 Go [2026-07-16 15:07:34] step 47900/1000000 | loss=2.7729 vocab=0.6254 token=1.3917 | 6.11 steps/s | VRAM=32.52 Go [2026-07-16 15:07:42] step 47950/1000000 | loss=2.7427 vocab=0.6700 token=1.4050 | 6.10 steps/s | VRAM=29.25 Go [2026-07-16 15:07:50] step 48000/1000000 | loss=2.6489 vocab=0.8917 token=1.6986 | 6.10 steps/s | VRAM=27.69 Go [2026-07-16 15:07:58] step 48050/1000000 | loss=2.7400 vocab=0.9810 token=1.7101 | 6.13 steps/s | VRAM=27.79 Go [2026-07-16 15:08:06] step 48100/1000000 | loss=2.7738 vocab=1.0920 token=1.8126 | 6.08 steps/s | VRAM=26.29 Go [2026-07-16 15:08:15] step 48150/1000000 | loss=2.7408 vocab=0.9376 token=1.2988 | 6.08 steps/s | VRAM=27.97 Go [2026-07-16 15:08:23] step 48200/1000000 | loss=2.7635 vocab=0.6003 token=1.4942 | 6.13 steps/s | VRAM=31.91 Go [2026-07-16 15:08:31] step 48250/1000000 | loss=2.7938 vocab=0.9671 token=1.6977 | 6.12 steps/s | VRAM=29.09 Go [2026-07-16 15:08:39] step 48300/1000000 | loss=2.5986 vocab=0.8627 token=1.4923 | 6.11 steps/s | VRAM=28.75 Go [2026-07-16 15:08:47] step 48350/1000000 | loss=2.7065 vocab=1.0064 token=1.5355 | 6.07 steps/s | VRAM=30.18 Go [2026-07-16 15:08:56] step 48400/1000000 | loss=2.6321 vocab=1.3546 token=1.5307 | 6.11 steps/s | VRAM=31.82 Go [2026-07-16 15:09:04] step 48450/1000000 | loss=2.7323 vocab=1.2504 token=1.7882 | 6.09 steps/s | VRAM=29.72 Go [2026-07-16 15:09:12] step 48500/1000000 | loss=2.7534 vocab=0.9807 token=1.7534 | 6.09 steps/s | VRAM=28.28 Go [2026-07-16 15:09:20] step 48550/1000000 | loss=2.6564 vocab=1.0257 token=1.7049 | 6.09 steps/s | VRAM=28.78 Go [2026-07-16 15:09:28] step 48600/1000000 | loss=2.6160 vocab=1.3078 token=1.6263 | 6.08 steps/s | VRAM=29.01 Go [2026-07-16 15:09:37] step 48650/1000000 | loss=2.7841 vocab=0.8144 token=1.3738 | 6.10 steps/s | VRAM=25.52 Go [2026-07-16 15:09:45] step 48700/1000000 | loss=2.6415 vocab=1.6512 token=1.9274 | 6.05 steps/s | VRAM=29.06 Go [2026-07-16 15:09:53] step 48750/1000000 | loss=2.6367 vocab=1.3027 token=1.7459 | 6.09 steps/s | VRAM=30.03 Go [2026-07-16 15:10:01] step 48800/1000000 | loss=2.7000 vocab=1.0842 token=2.0068 | 6.10 steps/s | VRAM=24.78 Go [2026-07-16 15:10:09] step 48850/1000000 | loss=2.7415 vocab=1.4237 token=1.9873 | 6.13 steps/s | VRAM=30.81 Go [2026-07-16 15:10:18] step 48900/1000000 | loss=2.7346 vocab=1.2059 token=1.8186 | 6.09 steps/s | VRAM=28.89 Go [2026-07-16 15:10:26] step 48950/1000000 | loss=2.7366 vocab=0.8661 token=1.2875 | 6.11 steps/s | VRAM=31.50 Go [2026-07-16 15:10:34] step 49000/1000000 | loss=2.8153 vocab=1.4330 token=1.8052 | 6.10 steps/s | VRAM=29.26 Go [2026-07-16 15:10:42] step 49050/1000000 | loss=2.5545 vocab=1.0795 token=1.9317 | 6.08 steps/s | VRAM=31.85 Go [2026-07-16 15:10:51] step 49100/1000000 | loss=2.6470 vocab=1.0077 token=1.5194 | 6.08 steps/s | VRAM=25.74 Go [2026-07-16 15:10:59] step 49150/1000000 | loss=2.6826 vocab=0.9798 token=1.7130 | 6.11 steps/s | VRAM=25.36 Go [2026-07-16 15:11:07] step 49200/1000000 | loss=2.6668 vocab=1.2064 token=1.5360 | 6.11 steps/s | VRAM=31.91 Go [2026-07-16 15:11:15] step 49250/1000000 | loss=2.6478 vocab=1.0316 token=1.5472 | 6.07 steps/s | VRAM=29.21 Go [2026-07-16 15:11:23] step 49300/1000000 | loss=2.7462 vocab=0.6562 token=1.1903 | 6.08 steps/s | VRAM=26.32 Go [2026-07-16 15:11:32] step 49350/1000000 | loss=2.8296 vocab=1.0780 token=1.5527 | 6.10 steps/s | VRAM=29.46 Go [2026-07-16 15:11:40] step 49400/1000000 | loss=2.7394 vocab=0.9850 token=2.1648 | 6.11 steps/s | VRAM=32.40 Go [2026-07-16 15:11:48] step 49450/1000000 | loss=2.6985 vocab=1.1913 token=1.7664 | 6.10 steps/s | VRAM=28.48 Go [2026-07-16 15:11:56] step 49500/1000000 | loss=2.7053 vocab=1.1991 token=1.5770 | 6.09 steps/s | VRAM=28.91 Go [2026-07-16 15:12:04] step 49550/1000000 | loss=2.7288 vocab=1.4662 token=1.7882 | 6.07 steps/s | VRAM=29.25 Go [2026-07-16 15:12:13] step 49600/1000000 | loss=2.6596 vocab=1.0705 token=1.3213 | 6.11 steps/s | VRAM=29.12 Go [2026-07-16 15:12:21] step 49650/1000000 | loss=2.6973 vocab=0.9660 token=1.7182 | 6.12 steps/s | VRAM=29.02 Go [2026-07-16 15:12:29] step 49700/1000000 | loss=2.7667 vocab=1.2056 token=1.8117 | 6.11 steps/s | VRAM=32.56 Go [2026-07-16 15:12:37] step 49750/1000000 | loss=2.7525 vocab=1.2631 token=1.7192 | 6.10 steps/s | VRAM=29.02 Go [2026-07-16 15:12:45] step 49800/1000000 | loss=2.5456 vocab=0.9459 token=1.8213 | 6.11 steps/s | VRAM=28.78 Go [2026-07-16 15:12:54] step 49850/1000000 | loss=2.7042 vocab=0.4311 token=1.0816 | 6.07 steps/s | VRAM=26.46 Go [2026-07-16 15:13:02] step 49900/1000000 | loss=2.5996 vocab=0.9437 token=1.8553 | 6.13 steps/s | VRAM=31.09 Go [2026-07-16 15:13:10] step 49950/1000000 | loss=2.7427 vocab=1.2111 token=1.4384 | 6.07 steps/s | VRAM=31.69 Go [2026-07-16 15:13:18] step 50000/1000000 | loss=2.7306 vocab=1.1594 token=2.0558 | 6.09 steps/s | VRAM=27.09 Go [2026-07-16 15:14:41] step 50050/1000000 | loss=2.7107 vocab=1.0425 token=1.5464 | 0.60 steps/s | VRAM=28.21 Go [2026-07-16 15:14:49] step 50100/1000000 | loss=2.7792 vocab=1.0721 token=1.4941 | 6.18 steps/s | VRAM=31.98 Go [2026-07-16 15:14:57] step 50150/1000000 | loss=2.7650 vocab=1.7694 token=1.6995 | 6.15 steps/s | VRAM=26.31 Go [2026-07-16 15:15:06] step 50200/1000000 | loss=2.6256 vocab=0.8742 token=1.5176 | 6.14 steps/s | VRAM=29.12 Go [2026-07-16 15:15:14] step 50250/1000000 | loss=2.7551 vocab=1.1368 token=1.7679 | 6.13 steps/s | VRAM=30.30 Go [2026-07-16 15:15:22] step 50300/1000000 | loss=2.6524 vocab=1.0538 token=1.8291 | 6.13 steps/s | VRAM=27.96 Go [2026-07-16 15:15:30] step 50350/1000000 | loss=2.6566 vocab=0.8311 token=1.5366 | 6.07 steps/s | VRAM=24.20 Go [2026-07-16 15:15:38] step 50400/1000000 | loss=2.6017 vocab=1.1430 token=1.2892 | 6.11 steps/s | VRAM=27.66 Go [2026-07-16 15:15:46] step 50450/1000000 | loss=2.7929 vocab=1.2205 token=1.6726 | 6.11 steps/s | VRAM=29.43 Go [2026-07-16 15:15:55] step 50500/1000000 | loss=2.6920 vocab=1.1504 token=1.3754 | 6.11 steps/s | VRAM=31.52 Go [2026-07-16 15:16:03] step 50550/1000000 | loss=2.6600 vocab=1.3321 token=1.5845 | 6.11 steps/s | VRAM=31.28 Go [2026-07-16 15:16:11] step 50600/1000000 | loss=2.6901 vocab=1.0358 token=1.6243 | 6.08 steps/s | VRAM=29.63 Go [2026-07-16 15:16:19] step 50650/1000000 | loss=2.6923 vocab=0.5847 token=1.4245 | 6.07 steps/s | VRAM=25.76 Go [2026-07-16 15:16:28] step 50700/1000000 | loss=2.7274 vocab=1.1044 token=1.7400 | 6.04 steps/s | VRAM=28.05 Go [2026-07-16 15:16:36] step 50750/1000000 | loss=2.7285 vocab=0.7022 token=1.6817 | 6.07 steps/s | VRAM=32.45 Go [2026-07-16 15:16:44] step 50800/1000000 | loss=2.7436 vocab=0.8088 token=1.2754 | 6.09 steps/s | VRAM=31.81 Go [2026-07-16 15:16:52] step 50850/1000000 | loss=2.7209 vocab=1.0835 token=1.8069 | 6.06 steps/s | VRAM=29.07 Go [2026-07-16 15:17:00] step 50900/1000000 | loss=2.6419 vocab=0.8787 token=1.4829 | 6.09 steps/s | VRAM=26.32 Go [2026-07-16 15:17:09] step 50950/1000000 | loss=2.6936 vocab=0.9384 token=1.4978 | 6.08 steps/s | VRAM=29.03 Go [2026-07-16 15:17:17] step 51000/1000000 | loss=2.7553 vocab=1.3143 token=1.5480 | 6.09 steps/s | VRAM=30.35 Go [2026-07-16 15:17:25] step 51050/1000000 | loss=2.7460 vocab=1.0668 token=1.8098 | 6.05 steps/s | VRAM=29.92 Go [2026-07-16 15:17:33] step 51100/1000000 | loss=2.6844 vocab=1.3329 token=1.8769 | 6.08 steps/s | VRAM=28.40 Go [2026-07-16 15:17:42] step 51150/1000000 | loss=2.6168 vocab=1.2657 token=1.8273 | 6.08 steps/s | VRAM=32.19 Go [2026-07-16 15:17:50] step 51200/1000000 | loss=2.6824 vocab=0.7894 token=1.2813 | 6.05 steps/s | VRAM=26.64 Go [2026-07-16 15:17:58] step 51250/1000000 | loss=2.6949 vocab=0.8056 token=1.6571 | 6.06 steps/s | VRAM=26.48 Go [2026-07-16 15:18:06] step 51300/1000000 | loss=2.6244 vocab=1.2800 token=1.3265 | 6.05 steps/s | VRAM=29.09 Go [2026-07-16 15:18:15] step 51350/1000000 | loss=2.6447 vocab=1.7096 token=1.4538 | 6.08 steps/s | VRAM=28.68 Go [2026-07-16 15:18:23] step 51400/1000000 | loss=2.5669 vocab=0.8169 token=1.1112 | 6.07 steps/s | VRAM=29.05 Go [2026-07-16 15:18:31] step 51450/1000000 | loss=2.7717 vocab=1.4070 token=2.0248 | 6.04 steps/s | VRAM=29.53 Go [2026-07-16 15:18:39] step 51500/1000000 | loss=2.6946 vocab=0.8411 token=1.6349 | 6.08 steps/s | VRAM=32.55 Go [2026-07-16 15:18:48] step 51550/1000000 | loss=2.6622 vocab=0.6248 token=1.6764 | 6.06 steps/s | VRAM=26.05 Go [2026-07-16 15:18:56] step 51600/1000000 | loss=2.6221 vocab=1.0183 token=1.6980 | 6.05 steps/s | VRAM=30.19 Go [2026-07-16 15:19:04] step 51650/1000000 | loss=2.6337 vocab=1.1724 token=1.8193 | 6.09 steps/s | VRAM=28.55 Go [2026-07-16 15:19:12] step 51700/1000000 | loss=2.6328 vocab=0.9283 token=1.7899 | 6.10 steps/s | VRAM=25.90 Go [2026-07-16 15:19:20] step 51750/1000000 | loss=2.6426 vocab=0.9630 token=1.6618 | 6.08 steps/s | VRAM=26.49 Go [2026-07-16 15:19:29] step 51800/1000000 | loss=2.6551 vocab=1.8979 token=1.6487 | 6.07 steps/s | VRAM=28.82 Go [2026-07-16 15:19:37] step 51850/1000000 | loss=2.6783 vocab=1.3413 token=1.8430 | 6.09 steps/s | VRAM=28.28 Go [2026-07-16 15:19:45] step 51900/1000000 | loss=2.7205 vocab=1.1295 token=1.7757 | 6.08 steps/s | VRAM=32.86 Go [2026-07-16 15:19:53] step 51950/1000000 | loss=2.6456 vocab=1.3031 token=2.0383 | 6.08 steps/s | VRAM=32.23 Go [2026-07-16 15:20:02] step 52000/1000000 | loss=2.6547 vocab=1.1763 token=1.5891 | 6.11 steps/s | VRAM=28.39 Go [2026-07-16 15:20:10] step 52050/1000000 | loss=2.6646 vocab=0.6989 token=1.0577 | 6.08 steps/s | VRAM=29.75 Go [2026-07-16 15:20:18] step 52100/1000000 | loss=2.7282 vocab=1.6626 token=2.1814 | 6.09 steps/s | VRAM=28.40 Go [2026-07-16 15:20:26] step 52150/1000000 | loss=2.6228 vocab=1.3279 token=1.6790 | 6.05 steps/s | VRAM=28.89 Go [2026-07-16 15:20:35] step 52200/1000000 | loss=2.7275 vocab=1.2929 token=1.7912 | 6.05 steps/s | VRAM=26.71 Go [2026-07-16 15:20:43] step 52250/1000000 | loss=2.6837 vocab=0.7479 token=1.3113 | 6.08 steps/s | VRAM=28.97 Go [2026-07-16 15:20:51] step 52300/1000000 | loss=2.6980 vocab=0.9343 token=1.1640 | 6.07 steps/s | VRAM=28.52 Go [2026-07-16 15:20:59] step 52350/1000000 | loss=2.7174 vocab=1.3456 token=1.8534 | 6.09 steps/s | VRAM=31.52 Go [2026-07-16 15:21:07] step 52400/1000000 | loss=2.5343 vocab=0.8398 token=1.5747 | 6.09 steps/s | VRAM=29.25 Go [2026-07-16 15:21:16] step 52450/1000000 | loss=2.6707 vocab=1.5522 token=1.8477 | 6.08 steps/s | VRAM=32.16 Go [2026-07-16 15:21:24] step 52500/1000000 | loss=2.6093 vocab=1.7162 token=1.5787 | 6.06 steps/s | VRAM=32.75 Go [2026-07-16 15:21:32] step 52550/1000000 | loss=2.8082 vocab=1.0904 token=1.2679 | 6.09 steps/s | VRAM=28.41 Go [2026-07-16 15:21:40] step 52600/1000000 | loss=2.5844 vocab=1.5426 token=1.5902 | 6.06 steps/s | VRAM=26.50 Go [2026-07-16 15:21:49] step 52650/1000000 | loss=2.5640 vocab=1.3315 token=1.1385 | 6.11 steps/s | VRAM=32.87 Go [2026-07-16 15:21:57] step 52700/1000000 | loss=2.7682 vocab=0.9555 token=1.4127 | 6.08 steps/s | VRAM=28.28 Go [2026-07-16 15:22:05] step 52750/1000000 | loss=2.6780 vocab=0.8745 token=1.6493 | 6.10 steps/s | VRAM=31.65 Go [2026-07-16 15:22:13] step 52800/1000000 | loss=2.7009 vocab=1.0474 token=1.6104 | 6.07 steps/s | VRAM=25.95 Go [2026-07-16 15:22:21] step 52850/1000000 | loss=2.6701 vocab=0.7707 token=0.8810 | 6.08 steps/s | VRAM=25.51 Go [2026-07-16 15:22:30] step 52900/1000000 | loss=2.6640 vocab=1.3937 token=1.8978 | 6.08 steps/s | VRAM=28.43 Go [2026-07-16 15:22:38] step 52950/1000000 | loss=2.7199 vocab=1.1950 token=1.5742 | 6.08 steps/s | VRAM=32.39 Go [2026-07-16 15:22:46] step 53000/1000000 | loss=2.6894 vocab=0.9872 token=1.3472 | 6.06 steps/s | VRAM=28.15 Go [2026-07-16 15:22:54] step 53050/1000000 | loss=2.5838 vocab=1.0826 token=1.7106 | 6.11 steps/s | VRAM=29.33 Go [2026-07-16 15:23:03] step 53100/1000000 | loss=2.6406 vocab=0.7939 token=1.5030 | 6.06 steps/s | VRAM=28.59 Go [2026-07-16 15:23:11] step 53150/1000000 | loss=2.7348 vocab=0.8815 token=1.3043 | 6.08 steps/s | VRAM=29.36 Go [2026-07-16 15:23:19] step 53200/1000000 | loss=2.5815 vocab=1.6920 token=1.8306 | 6.08 steps/s | VRAM=29.66 Go [2026-07-16 15:23:27] step 53250/1000000 | loss=2.6548 vocab=1.1112 token=1.7627 | 6.09 steps/s | VRAM=28.56 Go [2026-07-16 15:23:35] step 53300/1000000 | loss=2.6212 vocab=0.8442 token=1.0079 | 6.10 steps/s | VRAM=32.72 Go [2026-07-16 15:23:44] step 53350/1000000 | loss=2.6162 vocab=1.1771 token=1.7813 | 6.07 steps/s | VRAM=32.03 Go [2026-07-16 15:23:52] step 53400/1000000 | loss=2.6819 vocab=1.8342 token=1.9711 | 6.09 steps/s | VRAM=26.44 Go [2026-07-16 15:24:00] step 53450/1000000 | loss=2.7135 vocab=1.1129 token=1.7413 | 6.10 steps/s | VRAM=32.15 Go [2026-07-16 15:24:08] step 53500/1000000 | loss=2.6819 vocab=1.6525 token=1.8864 | 6.08 steps/s | VRAM=28.25 Go [2026-07-16 15:24:17] step 53550/1000000 | loss=2.6694 vocab=0.6733 token=1.2440 | 6.09 steps/s | VRAM=29.30 Go [2026-07-16 15:24:25] step 53600/1000000 | loss=2.7170 vocab=1.0867 token=1.5489 | 6.10 steps/s | VRAM=28.15 Go [2026-07-16 15:24:33] step 53650/1000000 | loss=2.7178 vocab=1.1522 token=1.5092 | 6.10 steps/s | VRAM=28.62 Go [2026-07-16 15:24:41] step 53700/1000000 | loss=2.6758 vocab=0.9442 token=1.3156 | 6.10 steps/s | VRAM=30.86 Go [2026-07-16 15:24:49] step 53750/1000000 | loss=2.7240 vocab=0.9210 token=1.7864 | 6.04 steps/s | VRAM=27.92 Go [2026-07-16 15:24:58] step 53800/1000000 | loss=2.8175 vocab=0.8206 token=1.7280 | 6.09 steps/s | VRAM=28.31 Go [2026-07-16 15:25:06] step 53850/1000000 | loss=2.7849 vocab=1.1387 token=1.6526 | 6.09 steps/s | VRAM=25.72 Go [2026-07-16 15:25:14] step 53900/1000000 | loss=2.5362 vocab=0.7095 token=1.5393 | 6.07 steps/s | VRAM=28.89 Go [2026-07-16 15:25:22] step 53950/1000000 | loss=2.7410 vocab=1.5559 token=1.6763 | 6.08 steps/s | VRAM=32.68 Go [2026-07-16 15:25:30] step 54000/1000000 | loss=2.6476 vocab=0.8871 token=1.8059 | 6.12 steps/s | VRAM=28.22 Go [2026-07-16 15:25:39] step 54050/1000000 | loss=2.6526 vocab=1.1736 token=1.7030 | 6.07 steps/s | VRAM=29.54 Go [2026-07-16 15:25:47] step 54100/1000000 | loss=2.6444 vocab=1.0574 token=1.4718 | 6.10 steps/s | VRAM=31.89 Go [2026-07-16 15:25:55] step 54150/1000000 | loss=2.6592 vocab=2.0777 token=1.7650 | 6.08 steps/s | VRAM=30.05 Go [2026-07-16 15:26:03] step 54200/1000000 | loss=2.6437 vocab=0.8110 token=1.4674 | 6.07 steps/s | VRAM=29.02 Go [2026-07-16 15:26:12] step 54250/1000000 | loss=2.5754 vocab=1.1305 token=1.5398 | 6.10 steps/s | VRAM=32.95 Go [2026-07-16 15:26:20] step 54300/1000000 | loss=2.7452 vocab=0.9122 token=1.5916 | 6.07 steps/s | VRAM=29.00 Go [2026-07-16 15:26:28] step 54350/1000000 | loss=2.6721 vocab=1.1238 token=1.5194 | 6.09 steps/s | VRAM=25.17 Go [2026-07-16 15:26:36] step 54400/1000000 | loss=2.6675 vocab=0.8290 token=1.6287 | 6.10 steps/s | VRAM=29.52 Go [2026-07-16 15:26:44] step 54450/1000000 | loss=2.7121 vocab=1.2785 token=1.8667 | 6.09 steps/s | VRAM=28.44 Go [2026-07-16 15:26:53] step 54500/1000000 | loss=2.6960 vocab=0.9240 token=1.6324 | 6.08 steps/s | VRAM=32.61 Go [2026-07-16 15:27:01] step 54550/1000000 | loss=2.6796 vocab=1.1494 token=1.4364 | 6.07 steps/s | VRAM=25.65 Go [2026-07-16 15:27:09] step 54600/1000000 | loss=2.6649 vocab=1.1846 token=1.5214 | 6.07 steps/s | VRAM=29.37 Go [2026-07-16 15:27:17] step 54650/1000000 | loss=2.6697 vocab=1.9277 token=2.1539 | 6.07 steps/s | VRAM=26.40 Go [2026-07-16 15:27:26] step 54700/1000000 | loss=2.6457 vocab=1.5997 token=1.6237 | 6.08 steps/s | VRAM=28.84 Go [2026-07-16 15:27:34] step 54750/1000000 | loss=2.6671 vocab=0.9621 token=1.6971 | 6.08 steps/s | VRAM=28.54 Go [2026-07-16 15:27:42] step 54800/1000000 | loss=2.6108 vocab=1.1446 token=1.7171 | 6.09 steps/s | VRAM=26.11 Go [2026-07-16 15:27:50] step 54850/1000000 | loss=2.6313 vocab=1.0931 token=1.4133 | 6.11 steps/s | VRAM=28.28 Go [2026-07-16 15:27:58] step 54900/1000000 | loss=2.5733 vocab=0.6871 token=1.0481 | 6.09 steps/s | VRAM=28.63 Go [2026-07-16 15:28:07] step 54950/1000000 | loss=2.7174 vocab=1.3220 token=1.6593 | 6.07 steps/s | VRAM=25.74 Go [2026-07-16 15:28:15] step 55000/1000000 | loss=2.7060 vocab=0.7157 token=1.5937 | 6.11 steps/s | VRAM=29.30 Go [2026-07-16 15:29:44] step 55050/1000000 | loss=2.7189 vocab=1.3553 token=1.6113 | 0.56 steps/s | VRAM=28.18 Go [2026-07-16 15:29:53] step 55100/1000000 | loss=2.5792 vocab=0.5307 token=1.5611 | 6.13 steps/s | VRAM=26.38 Go [2026-07-16 15:30:01] step 55150/1000000 | loss=2.6826 vocab=1.0684 token=1.3207 | 6.16 steps/s | VRAM=29.43 Go [2026-07-16 15:30:09] step 55200/1000000 | loss=2.6632 vocab=0.7146 token=1.2367 | 6.10 steps/s | VRAM=26.44 Go [2026-07-16 15:30:17] step 55250/1000000 | loss=2.6358 vocab=0.8344 token=1.7912 | 6.11 steps/s | VRAM=28.75 Go [2026-07-16 15:30:25] step 55300/1000000 | loss=2.5791 vocab=0.6976 token=1.2108 | 6.13 steps/s | VRAM=27.75 Go [2026-07-16 15:30:33] step 55350/1000000 | loss=2.7174 vocab=1.0298 token=1.9256 | 6.09 steps/s | VRAM=26.46 Go [2026-07-16 15:30:42] step 55400/1000000 | loss=2.7144 vocab=0.9641 token=1.4323 | 6.13 steps/s | VRAM=32.29 Go [2026-07-16 15:30:50] step 55450/1000000 | loss=2.6432 vocab=1.0471 token=1.9454 | 6.11 steps/s | VRAM=24.98 Go [2026-07-16 15:30:58] step 55500/1000000 | loss=2.6902 vocab=1.5933 token=1.3888 | 6.10 steps/s | VRAM=30.15 Go [2026-07-16 15:31:06] step 55550/1000000 | loss=2.6798 vocab=1.0197 token=1.3406 | 6.08 steps/s | VRAM=28.92 Go [2026-07-16 15:31:14] step 55600/1000000 | loss=2.7149 vocab=1.0945 token=1.5666 | 6.07 steps/s | VRAM=28.30 Go [2026-07-16 15:31:23] step 55650/1000000 | loss=2.6418 vocab=0.9685 token=1.6622 | 6.09 steps/s | VRAM=32.47 Go [2026-07-16 15:31:31] step 55700/1000000 | loss=2.6890 vocab=1.0972 token=1.6642 | 6.06 steps/s | VRAM=26.30 Go [2026-07-16 15:31:39] step 55750/1000000 | loss=2.7073 vocab=1.1195 token=1.6422 | 6.06 steps/s | VRAM=26.24 Go [2026-07-16 15:31:47] step 55800/1000000 | loss=2.6885 vocab=1.0389 token=1.7455 | 6.10 steps/s | VRAM=29.86 Go [2026-07-16 15:31:55] step 55850/1000000 | loss=2.6337 vocab=1.4210 token=1.9170 | 6.11 steps/s | VRAM=28.36 Go [2026-07-16 15:32:04] step 55900/1000000 | loss=2.6505 vocab=1.1538 token=1.6658 | 6.08 steps/s | VRAM=29.56 Go [2026-07-16 15:32:12] step 55950/1000000 | loss=2.6174 vocab=0.7786 token=1.9194 | 6.08 steps/s | VRAM=30.59 Go [2026-07-16 15:32:20] step 56000/1000000 | loss=2.6639 vocab=1.6014 token=1.5529 | 6.10 steps/s | VRAM=31.68 Go [2026-07-16 15:32:28] step 56050/1000000 | loss=2.7066 vocab=0.9798 token=1.5401 | 6.07 steps/s | VRAM=32.25 Go [2026-07-16 15:32:37] step 56100/1000000 | loss=2.5655 vocab=1.5416 token=1.7786 | 6.06 steps/s | VRAM=24.27 Go [2026-07-16 15:32:45] step 56150/1000000 | loss=2.5502 vocab=1.1298 token=1.3745 | 6.08 steps/s | VRAM=27.63 Go [2026-07-16 15:32:53] step 56200/1000000 | loss=2.6091 vocab=1.2446 token=1.7488 | 6.09 steps/s | VRAM=29.37 Go [2026-07-16 15:33:01] step 56250/1000000 | loss=2.6939 vocab=1.5190 token=1.5868 | 6.05 steps/s | VRAM=28.17 Go [2026-07-16 15:33:10] step 56300/1000000 | loss=2.7581 vocab=1.2892 token=2.0284 | 6.08 steps/s | VRAM=29.94 Go [2026-07-16 15:33:18] step 56350/1000000 | loss=2.6557 vocab=1.1479 token=1.8156 | 6.06 steps/s | VRAM=28.52 Go [2026-07-16 15:33:26] step 56400/1000000 | loss=2.7338 vocab=1.1946 token=1.4085 | 6.06 steps/s | VRAM=29.78 Go [2026-07-16 15:33:34] step 56450/1000000 | loss=2.7482 vocab=1.0368 token=2.1529 | 6.09 steps/s | VRAM=29.07 Go [2026-07-16 15:33:42] step 56500/1000000 | loss=2.6574 vocab=1.1801 token=1.6277 | 6.08 steps/s | VRAM=28.73 Go [2026-07-16 15:33:51] step 56550/1000000 | loss=2.5911 vocab=1.0801 token=1.6441 | 6.09 steps/s | VRAM=28.63 Go [2026-07-16 15:33:59] step 56600/1000000 | loss=2.5685 vocab=0.9972 token=1.4696 | 6.08 steps/s | VRAM=27.12 Go [2026-07-16 15:34:07] step 56650/1000000 | loss=2.7304 vocab=1.1080 token=1.9930 | 6.09 steps/s | VRAM=28.33 Go [2026-07-16 15:34:15] step 56700/1000000 | loss=2.6927 vocab=1.2436 token=1.6204 | 6.06 steps/s | VRAM=31.56 Go [2026-07-16 15:34:24] step 56750/1000000 | loss=2.5803 vocab=1.5093 token=1.6193 | 6.10 steps/s | VRAM=30.32 Go [2026-07-16 15:34:32] step 56800/1000000 | loss=2.5622 vocab=1.1524 token=1.4064 | 6.09 steps/s | VRAM=31.23 Go [2026-07-16 15:34:40] step 56850/1000000 | loss=2.6048 vocab=1.1122 token=1.5493 | 6.08 steps/s | VRAM=26.45 Go [2026-07-16 15:34:48] step 56900/1000000 | loss=2.5863 vocab=1.2578 token=1.5629 | 6.12 steps/s | VRAM=27.93 Go [2026-07-16 15:34:56] step 56950/1000000 | loss=2.6158 vocab=1.1454 token=1.7390 | 6.07 steps/s | VRAM=29.69 Go [2026-07-16 15:35:05] step 57000/1000000 | loss=2.5278 vocab=1.2109 token=1.8070 | 6.08 steps/s | VRAM=28.86 Go [2026-07-16 15:35:13] step 57050/1000000 | loss=2.6101 vocab=0.7946 token=1.4984 | 6.08 steps/s | VRAM=29.27 Go [2026-07-16 15:35:21] step 57100/1000000 | loss=2.6683 vocab=0.7643 token=1.6279 | 6.10 steps/s | VRAM=29.42 Go [2026-07-16 15:35:29] step 57150/1000000 | loss=2.6784 vocab=1.1687 token=1.4116 | 6.09 steps/s | VRAM=27.99 Go [2026-07-16 15:35:37] step 57200/1000000 | loss=2.6699 vocab=1.2403 token=1.5338 | 6.09 steps/s | VRAM=28.93 Go [2026-07-16 15:35:46] step 57250/1000000 | loss=2.5532 vocab=1.3175 token=1.4627 | 6.09 steps/s | VRAM=29.11 Go [2026-07-16 15:35:54] step 57300/1000000 | loss=2.5693 vocab=0.8808 token=1.3834 | 6.11 steps/s | VRAM=27.71 Go [2026-07-16 15:36:02] step 57350/1000000 | loss=2.5236 vocab=0.9560 token=1.6016 | 6.12 steps/s | VRAM=28.42 Go [2026-07-16 15:36:10] step 57400/1000000 | loss=2.5975 vocab=1.3133 token=1.8215 | 6.07 steps/s | VRAM=27.20 Go [2026-07-16 15:36:18] step 57450/1000000 | loss=2.6254 vocab=1.0945 token=1.7849 | 6.10 steps/s | VRAM=31.61 Go [2026-07-16 15:36:27] step 57500/1000000 | loss=2.6068 vocab=1.0936 token=1.7409 | 6.10 steps/s | VRAM=27.97 Go [2026-07-16 15:36:35] step 57550/1000000 | loss=2.6561 vocab=0.8373 token=1.4420 | 6.08 steps/s | VRAM=31.79 Go [2026-07-16 15:36:43] step 57600/1000000 | loss=2.6961 vocab=0.8493 token=1.7009 | 6.06 steps/s | VRAM=29.34 Go [2026-07-16 15:36:51] step 57650/1000000 | loss=2.6843 vocab=0.9613 token=1.8147 | 6.10 steps/s | VRAM=28.84 Go [2026-07-16 15:37:00] step 57700/1000000 | loss=2.6580 vocab=1.0626 token=1.5933 | 6.09 steps/s | VRAM=28.57 Go [2026-07-16 15:37:08] step 57750/1000000 | loss=2.7291 vocab=0.9236 token=1.5073 | 6.08 steps/s | VRAM=29.80 Go [2026-07-16 15:37:16] step 57800/1000000 | loss=2.6752 vocab=0.7733 token=1.6470 | 6.10 steps/s | VRAM=29.58 Go [2026-07-16 15:37:24] step 57850/1000000 | loss=2.6802 vocab=0.7042 token=1.5196 | 6.08 steps/s | VRAM=28.36 Go [2026-07-16 15:37:32] step 57900/1000000 | loss=2.7398 vocab=1.1844 token=1.6255 | 6.10 steps/s | VRAM=28.59 Go [2026-07-16 15:37:41] step 57950/1000000 | loss=2.6779 vocab=1.6207 token=1.6755 | 6.10 steps/s | VRAM=29.14 Go [2026-07-16 15:37:49] step 58000/1000000 | loss=2.5992 vocab=1.2302 token=1.8504 | 6.07 steps/s | VRAM=26.43 Go [2026-07-16 15:37:57] step 58050/1000000 | loss=2.7211 vocab=1.4973 token=1.5852 | 6.08 steps/s | VRAM=26.19 Go [2026-07-16 15:38:05] step 58100/1000000 | loss=2.5491 vocab=1.1045 token=1.5908 | 6.11 steps/s | VRAM=28.69 Go [2026-07-16 15:38:13] step 58150/1000000 | loss=2.5841 vocab=0.4037 token=1.1718 | 6.11 steps/s | VRAM=33.00 Go [2026-07-16 15:38:22] step 58200/1000000 | loss=2.6239 vocab=1.1766 token=1.4726 | 6.09 steps/s | VRAM=32.43 Go [2026-07-16 15:38:30] step 58250/1000000 | loss=2.6819 vocab=1.1438 token=1.7774 | 6.08 steps/s | VRAM=29.12 Go [2026-07-16 15:38:38] step 58300/1000000 | loss=2.6074 vocab=0.9855 token=1.1121 | 6.09 steps/s | VRAM=26.05 Go [2026-07-16 15:38:46] step 58350/1000000 | loss=2.6622 vocab=1.4099 token=1.6871 | 6.13 steps/s | VRAM=32.65 Go [2026-07-16 15:38:54] step 58400/1000000 | loss=2.7024 vocab=1.0752 token=1.7357 | 6.09 steps/s | VRAM=29.16 Go [2026-07-16 15:39:03] step 58450/1000000 | loss=2.6890 vocab=0.9239 token=1.6915 | 6.08 steps/s | VRAM=25.20 Go [2026-07-16 15:39:11] step 58500/1000000 | loss=2.6868 vocab=0.9926 token=1.5481 | 6.11 steps/s | VRAM=28.07 Go [2026-07-16 15:39:19] step 58550/1000000 | loss=2.5470 vocab=0.8237 token=1.7114 | 6.11 steps/s | VRAM=29.06 Go [2026-07-16 15:39:27] step 58600/1000000 | loss=2.6689 vocab=1.1832 token=1.9826 | 6.05 steps/s | VRAM=27.29 Go [2026-07-16 15:39:35] step 58650/1000000 | loss=2.6582 vocab=1.0866 token=1.4641 | 6.10 steps/s | VRAM=28.90 Go [2026-07-16 15:39:44] step 58700/1000000 | loss=2.6991 vocab=1.2905 token=1.6065 | 6.10 steps/s | VRAM=29.25 Go [2026-07-16 15:39:52] step 58750/1000000 | loss=2.5737 vocab=1.5724 token=1.5640 | 6.06 steps/s | VRAM=27.00 Go [2026-07-16 15:40:00] step 58800/1000000 | loss=2.6357 vocab=1.1743 token=1.7060 | 6.07 steps/s | VRAM=29.40 Go [2026-07-16 15:40:08] step 58850/1000000 | loss=2.6401 vocab=1.1634 token=1.5401 | 6.05 steps/s | VRAM=29.41 Go [2026-07-16 15:40:17] step 58900/1000000 | loss=2.6047 vocab=0.9825 token=1.4729 | 6.10 steps/s | VRAM=32.54 Go [2026-07-16 15:40:25] step 58950/1000000 | loss=2.6303 vocab=0.9758 token=1.8199 | 6.07 steps/s | VRAM=25.11 Go [2026-07-16 15:40:33] step 59000/1000000 | loss=2.6443 vocab=0.5812 token=1.6228 | 6.10 steps/s | VRAM=28.06 Go [2026-07-16 15:40:41] step 59050/1000000 | loss=2.7283 vocab=1.0183 token=1.5300 | 6.09 steps/s | VRAM=29.32 Go [2026-07-16 15:40:49] step 59100/1000000 | loss=2.6595 vocab=1.1886 token=1.6305 | 6.06 steps/s | VRAM=27.97 Go [2026-07-16 15:40:58] step 59150/1000000 | loss=2.5230 vocab=1.3386 token=1.8168 | 6.09 steps/s | VRAM=31.92 Go [2026-07-16 15:41:06] step 59200/1000000 | loss=2.6661 vocab=0.9352 token=1.4737 | 6.09 steps/s | VRAM=28.84 Go [2026-07-16 15:41:14] step 59250/1000000 | loss=2.7240 vocab=0.8918 token=1.4929 | 6.06 steps/s | VRAM=27.21 Go [2026-07-16 15:41:22] step 59300/1000000 | loss=2.6122 vocab=1.0343 token=1.7348 | 6.08 steps/s | VRAM=28.71 Go [2026-07-16 15:41:31] step 59350/1000000 | loss=2.6365 vocab=0.5763 token=1.3393 | 6.09 steps/s | VRAM=30.15 Go [2026-07-16 15:41:39] step 59400/1000000 | loss=2.5939 vocab=1.1525 token=1.2277 | 6.09 steps/s | VRAM=30.99 Go [2026-07-16 15:41:47] step 59450/1000000 | loss=2.6052 vocab=1.2944 token=1.2355 | 6.08 steps/s | VRAM=28.48 Go [2026-07-16 15:41:55] step 59500/1000000 | loss=2.6283 vocab=0.6471 token=1.2768 | 6.11 steps/s | VRAM=26.21 Go [2026-07-16 15:42:03] step 59550/1000000 | loss=2.7602 vocab=1.3240 token=1.7720 | 6.10 steps/s | VRAM=25.01 Go [2026-07-16 15:42:12] step 59600/1000000 | loss=2.5836 vocab=0.8987 token=1.3378 | 6.09 steps/s | VRAM=29.62 Go [2026-07-16 15:42:20] step 59650/1000000 | loss=2.6642 vocab=1.0251 token=1.2947 | 6.10 steps/s | VRAM=29.19 Go [2026-07-16 15:42:28] step 59700/1000000 | loss=2.4945 vocab=1.0796 token=1.6283 | 6.07 steps/s | VRAM=26.25 Go [2026-07-16 15:42:36] step 59750/1000000 | loss=2.5649 vocab=1.0631 token=1.4304 | 6.07 steps/s | VRAM=29.52 Go [2026-07-16 15:42:44] step 59800/1000000 | loss=2.5527 vocab=1.2219 token=1.8293 | 6.12 steps/s | VRAM=28.44 Go [2026-07-16 15:42:53] step 59850/1000000 | loss=2.5869 vocab=0.7950 token=1.2508 | 6.07 steps/s | VRAM=26.16 Go [2026-07-16 15:43:01] step 59900/1000000 | loss=2.7299 vocab=1.0482 token=1.4273 | 6.08 steps/s | VRAM=28.93 Go [2026-07-16 15:43:09] step 59950/1000000 | loss=2.6269 vocab=0.7788 token=1.5339 | 6.09 steps/s | VRAM=28.48 Go [2026-07-16 15:43:17] step 60000/1000000 | loss=2.7341 vocab=0.7758 token=1.6690 | 6.11 steps/s | VRAM=32.70 Go [2026-07-16 15:44:48] step 60050/1000000 | loss=2.6180 vocab=1.0175 token=1.2305 | 0.55 steps/s | VRAM=28.33 Go [2026-07-16 15:44:56] step 60100/1000000 | loss=2.6198 vocab=1.1407 token=1.4017 | 6.16 steps/s | VRAM=28.37 Go [2026-07-16 15:45:04] step 60150/1000000 | loss=2.6042 vocab=1.1274 token=1.6097 | 6.16 steps/s | VRAM=31.28 Go [2026-07-16 15:45:13] step 60200/1000000 | loss=2.6067 vocab=1.1095 token=1.7000 | 6.15 steps/s | VRAM=26.74 Go [2026-07-16 15:45:21] step 60250/1000000 | loss=2.6000 vocab=1.6087 token=1.6125 | 6.13 steps/s | VRAM=29.02 Go [2026-07-16 15:45:29] step 60300/1000000 | loss=2.6748 vocab=1.1109 token=1.7620 | 6.16 steps/s | VRAM=31.59 Go [2026-07-16 15:45:37] step 60350/1000000 | loss=2.6794 vocab=1.0976 token=1.5171 | 6.11 steps/s | VRAM=29.50 Go [2026-07-16 15:45:45] step 60400/1000000 | loss=2.6826 vocab=0.7727 token=1.3011 | 6.12 steps/s | VRAM=31.73 Go [2026-07-16 15:45:53] step 60450/1000000 | loss=2.6950 vocab=0.8341 token=1.4001 | 6.11 steps/s | VRAM=31.97 Go [2026-07-16 15:46:02] step 60500/1000000 | loss=2.7111 vocab=1.8323 token=1.6930 | 6.11 steps/s | VRAM=26.49 Go [2026-07-16 15:46:10] step 60550/1000000 | loss=2.6858 vocab=1.1809 token=1.7503 | 6.09 steps/s | VRAM=29.00 Go [2026-07-16 15:46:18] step 60600/1000000 | loss=2.5611 vocab=0.7724 token=1.3538 | 6.10 steps/s | VRAM=32.08 Go [2026-07-16 15:46:26] step 60650/1000000 | loss=2.6176 vocab=1.2497 token=1.8101 | 6.05 steps/s | VRAM=26.61 Go [2026-07-16 15:46:35] step 60700/1000000 | loss=2.5526 vocab=1.0723 token=1.8848 | 6.02 steps/s | VRAM=26.74 Go [2026-07-16 15:46:43] step 60750/1000000 | loss=2.5859 vocab=0.9501 token=1.2549 | 6.06 steps/s | VRAM=28.87 Go [2026-07-16 15:46:51] step 60800/1000000 | loss=2.6079 vocab=1.1732 token=1.2590 | 6.09 steps/s | VRAM=31.00 Go [2026-07-16 15:46:59] step 60850/1000000 | loss=2.7268 vocab=1.1457 token=1.8634 | 6.08 steps/s | VRAM=26.64 Go [2026-07-16 15:47:07] step 60900/1000000 | loss=2.6112 vocab=1.1327 token=1.4650 | 6.09 steps/s | VRAM=31.01 Go [2026-07-16 15:47:16] step 60950/1000000 | loss=2.6058 vocab=1.3082 token=1.8927 | 6.06 steps/s | VRAM=26.35 Go [2026-07-16 15:47:24] step 61000/1000000 | loss=2.5608 vocab=0.8987 token=1.6211 | 6.08 steps/s | VRAM=32.54 Go [2026-07-16 15:47:32] step 61050/1000000 | loss=2.6238 vocab=0.7252 token=1.4871 | 6.10 steps/s | VRAM=32.81 Go [2026-07-16 15:47:40] step 61100/1000000 | loss=2.5947 vocab=0.9781 token=0.9105 | 6.07 steps/s | VRAM=29.72 Go [2026-07-16 15:47:49] step 61150/1000000 | loss=2.6690 vocab=0.8730 token=1.8799 | 6.08 steps/s | VRAM=27.46 Go [2026-07-16 15:47:57] step 61200/1000000 | loss=2.5962 vocab=1.5404 token=1.9083 | 6.11 steps/s | VRAM=32.34 Go [2026-07-16 15:48:05] step 61250/1000000 | loss=2.6476 vocab=1.0536 token=1.5690 | 6.09 steps/s | VRAM=26.10 Go [2026-07-16 15:48:13] step 61300/1000000 | loss=2.6871 vocab=1.1680 token=1.5738 | 6.09 steps/s | VRAM=30.91 Go [2026-07-16 15:48:21] step 61350/1000000 | loss=2.6015 vocab=1.4954 token=1.8401 | 6.06 steps/s | VRAM=32.18 Go [2026-07-16 15:48:30] step 61400/1000000 | loss=2.6062 vocab=1.1249 token=1.4058 | 6.09 steps/s | VRAM=31.22 Go [2026-07-16 15:48:38] step 61450/1000000 | loss=2.7486 vocab=1.2645 token=1.5869 | 6.07 steps/s | VRAM=26.95 Go [2026-07-16 15:48:46] step 61500/1000000 | loss=2.7356 vocab=1.0519 token=1.9585 | 6.06 steps/s | VRAM=25.73 Go [2026-07-16 15:48:54] step 61550/1000000 | loss=2.7456 vocab=1.0475 token=1.4781 | 6.06 steps/s | VRAM=30.13 Go [2026-07-16 15:49:03] step 61600/1000000 | loss=2.7203 vocab=1.2163 token=1.5125 | 6.10 steps/s | VRAM=32.11 Go [2026-07-16 15:49:11] step 61650/1000000 | loss=2.5493 vocab=1.0882 token=1.5446 | 6.08 steps/s | VRAM=29.58 Go [2026-07-16 15:49:19] step 61700/1000000 | loss=2.6402 vocab=1.0806 token=1.6849 | 6.06 steps/s | VRAM=28.87 Go [2026-07-16 15:49:27] step 61750/1000000 | loss=2.5906 vocab=1.0158 token=1.6986 | 6.10 steps/s | VRAM=27.93 Go [2026-07-16 15:49:35] step 61800/1000000 | loss=2.7008 vocab=1.2252 token=2.1082 | 6.08 steps/s | VRAM=28.02 Go [2026-07-16 15:49:44] step 61850/1000000 | loss=2.5946 vocab=1.3508 token=1.5999 | 6.05 steps/s | VRAM=25.84 Go [2026-07-16 15:49:52] step 61900/1000000 | loss=2.6035 vocab=1.0874 token=1.6767 | 6.04 steps/s | VRAM=29.86 Go [2026-07-16 15:50:00] step 61950/1000000 | loss=2.6547 vocab=0.8453 token=1.5947 | 6.08 steps/s | VRAM=32.65 Go [2026-07-16 15:50:08] step 62000/1000000 | loss=2.6113 vocab=1.1774 token=1.5227 | 6.07 steps/s | VRAM=28.38 Go [2026-07-16 15:50:17] step 62050/1000000 | loss=2.5999 vocab=1.0861 token=1.5641 | 6.07 steps/s | VRAM=28.39 Go [2026-07-16 15:50:25] step 62100/1000000 | loss=2.6126 vocab=1.1049 token=1.5350 | 6.09 steps/s | VRAM=28.22 Go [2026-07-16 15:50:33] step 62150/1000000 | loss=2.6346 vocab=0.6294 token=1.4300 | 6.08 steps/s | VRAM=32.14 Go [2026-07-16 15:50:41] step 62200/1000000 | loss=2.6738 vocab=1.0006 token=1.3198 | 6.07 steps/s | VRAM=28.63 Go [2026-07-16 15:50:50] step 62250/1000000 | loss=2.6640 vocab=0.8787 token=1.9745 | 6.08 steps/s | VRAM=28.92 Go [2026-07-16 15:50:58] step 62300/1000000 | loss=2.6903 vocab=0.6776 token=1.3696 | 6.06 steps/s | VRAM=29.04 Go [2026-07-16 15:51:06] step 62350/1000000 | loss=2.6531 vocab=0.9884 token=1.6482 | 6.11 steps/s | VRAM=31.49 Go [2026-07-16 15:51:14] step 62400/1000000 | loss=2.6176 vocab=1.1075 token=1.3823 | 6.10 steps/s | VRAM=31.20 Go [2026-07-16 15:51:22] step 62450/1000000 | loss=2.6193 vocab=1.3697 token=1.2703 | 6.08 steps/s | VRAM=29.58 Go [2026-07-16 15:51:31] step 62500/1000000 | loss=2.6636 vocab=0.9347 token=1.8835 | 6.05 steps/s | VRAM=28.85 Go [2026-07-16 15:51:39] step 62550/1000000 | loss=2.5863 vocab=0.7140 token=1.6172 | 6.11 steps/s | VRAM=32.58 Go [2026-07-16 15:51:47] step 62600/1000000 | loss=2.5883 vocab=1.4701 token=1.6951 | 6.09 steps/s | VRAM=29.02 Go [2026-07-16 15:51:55] step 62650/1000000 | loss=2.6389 vocab=0.7141 token=1.1296 | 6.09 steps/s | VRAM=25.39 Go [2026-07-16 15:52:04] step 62700/1000000 | loss=2.6946 vocab=0.9785 token=1.8781 | 6.08 steps/s | VRAM=29.24 Go [2026-07-16 15:52:12] step 62750/1000000 | loss=2.7057 vocab=1.4685 token=1.7052 | 6.10 steps/s | VRAM=32.42 Go [2026-07-16 15:52:20] step 62800/1000000 | loss=2.6228 vocab=0.8528 token=1.6193 | 6.05 steps/s | VRAM=26.20 Go [2026-07-16 15:52:28] step 62850/1000000 | loss=2.6268 vocab=1.2193 token=1.3750 | 6.10 steps/s | VRAM=28.00 Go [2026-07-16 15:52:36] step 62900/1000000 | loss=2.6820 vocab=1.1153 token=1.9759 | 6.11 steps/s | VRAM=28.79 Go [2026-07-16 15:52:45] step 62950/1000000 | loss=2.4780 vocab=0.8593 token=1.2629 | 6.04 steps/s | VRAM=26.38 Go [2026-07-16 15:52:53] step 63000/1000000 | loss=2.7355 vocab=1.3461 token=1.3550 | 6.08 steps/s | VRAM=27.55 Go [2026-07-16 15:53:01] step 63050/1000000 | loss=2.6348 vocab=1.0424 token=1.4769 | 6.12 steps/s | VRAM=32.15 Go [2026-07-16 15:53:09] step 63100/1000000 | loss=2.6463 vocab=1.3785 token=1.8761 | 6.09 steps/s | VRAM=25.06 Go [2026-07-16 15:53:17] step 63150/1000000 | loss=2.6581 vocab=0.8091 token=1.5853 | 6.10 steps/s | VRAM=29.98 Go [2026-07-16 15:53:26] step 63200/1000000 | loss=2.6775 vocab=1.1312 token=1.3722 | 6.09 steps/s | VRAM=31.99 Go [2026-07-16 15:53:34] step 63250/1000000 | loss=2.5702 vocab=0.7646 token=1.8262 | 6.12 steps/s | VRAM=32.63 Go [2026-07-16 15:53:42] step 63300/1000000 | loss=2.6675 vocab=1.8242 token=1.5863 | 6.07 steps/s | VRAM=30.02 Go [2026-07-16 15:53:50] step 63350/1000000 | loss=2.5915 vocab=0.5725 token=1.2260 | 6.10 steps/s | VRAM=24.62 Go [2026-07-16 15:53:59] step 63400/1000000 | loss=2.6696 vocab=0.8378 token=1.6844 | 6.08 steps/s | VRAM=25.56 Go [2026-07-16 15:54:07] step 63450/1000000 | loss=2.5525 vocab=1.1857 token=1.5337 | 6.06 steps/s | VRAM=29.26 Go [2026-07-16 15:54:15] step 63500/1000000 | loss=2.5659 vocab=0.8223 token=1.4238 | 6.09 steps/s | VRAM=29.02 Go [2026-07-16 15:54:23] step 63550/1000000 | loss=2.6130 vocab=1.0674 token=2.2070 | 6.07 steps/s | VRAM=30.38 Go [2026-07-16 15:54:31] step 63600/1000000 | loss=2.5813 vocab=0.9741 token=1.2283 | 6.10 steps/s | VRAM=29.60 Go [2026-07-16 15:54:40] step 63650/1000000 | loss=2.6357 vocab=0.7789 token=1.5638 | 6.09 steps/s | VRAM=31.39 Go [2026-07-16 15:54:48] step 63700/1000000 | loss=2.5154 vocab=0.9834 token=1.3036 | 6.10 steps/s | VRAM=28.32 Go [2026-07-16 15:54:56] step 63750/1000000 | loss=2.5205 vocab=0.8734 token=1.2658 | 6.08 steps/s | VRAM=31.89 Go [2026-07-16 15:55:04] step 63800/1000000 | loss=2.7852 vocab=0.8339 token=1.3080 | 6.09 steps/s | VRAM=28.76 Go [2026-07-16 15:55:12] step 63850/1000000 | loss=2.5866 vocab=0.6126 token=1.1979 | 6.09 steps/s | VRAM=25.96 Go [2026-07-16 15:55:21] step 63900/1000000 | loss=2.4916 vocab=1.0425 token=1.2015 | 6.09 steps/s | VRAM=25.89 Go [2026-07-16 15:55:29] step 63950/1000000 | loss=2.6705 vocab=1.0020 token=1.3448 | 6.10 steps/s | VRAM=26.50 Go [2026-07-16 15:55:37] step 64000/1000000 | loss=2.6809 vocab=0.7842 token=1.5115 | 6.09 steps/s | VRAM=25.97 Go [2026-07-16 15:55:45] step 64050/1000000 | loss=2.6546 vocab=0.5568 token=1.4738 | 6.10 steps/s | VRAM=25.09 Go [2026-07-16 15:55:54] step 64100/1000000 | loss=2.5069 vocab=1.0973 token=1.4788 | 6.08 steps/s | VRAM=25.70 Go [2026-07-16 15:56:02] step 64150/1000000 | loss=2.5603 vocab=0.7527 token=1.5064 | 6.08 steps/s | VRAM=31.93 Go [2026-07-16 15:56:10] step 64200/1000000 | loss=2.7029 vocab=1.1606 token=1.6742 | 6.05 steps/s | VRAM=28.61 Go [2026-07-16 15:56:18] step 64250/1000000 | loss=2.5080 vocab=0.9905 token=1.3780 | 6.12 steps/s | VRAM=32.07 Go [2026-07-16 15:56:26] step 64300/1000000 | loss=2.6707 vocab=1.1935 token=1.4832 | 6.10 steps/s | VRAM=24.83 Go [2026-07-16 15:56:35] step 64350/1000000 | loss=2.5566 vocab=1.0095 token=1.8677 | 6.08 steps/s | VRAM=32.54 Go [2026-07-16 15:56:43] step 64400/1000000 | loss=2.6740 vocab=1.2487 token=1.8731 | 6.08 steps/s | VRAM=27.98 Go [2026-07-16 15:56:51] step 64450/1000000 | loss=2.6262 vocab=1.1121 token=1.3702 | 6.08 steps/s | VRAM=29.06 Go [2026-07-16 15:56:59] step 64500/1000000 | loss=2.6769 vocab=0.8910 token=1.4995 | 6.10 steps/s | VRAM=28.18 Go [2026-07-16 15:57:07] step 64550/1000000 | loss=2.6926 vocab=0.9011 token=1.4861 | 6.06 steps/s | VRAM=26.52 Go [2026-07-16 15:57:16] step 64600/1000000 | loss=2.7122 vocab=1.3749 token=1.5118 | 6.11 steps/s | VRAM=28.60 Go [2026-07-16 15:57:24] step 64650/1000000 | loss=2.6350 vocab=1.2047 token=1.6293 | 6.10 steps/s | VRAM=25.99 Go [2026-07-16 15:57:32] step 64700/1000000 | loss=2.6482 vocab=1.4584 token=1.6077 | 6.09 steps/s | VRAM=27.99 Go [2026-07-16 15:57:40] step 64750/1000000 | loss=2.5985 vocab=0.8941 token=1.5400 | 6.10 steps/s | VRAM=28.47 Go [2026-07-16 15:57:48] step 64800/1000000 | loss=2.6564 vocab=1.1002 token=1.5536 | 6.07 steps/s | VRAM=26.65 Go [2026-07-16 15:57:57] step 64850/1000000 | loss=2.5959 vocab=0.9906 token=1.7709 | 6.09 steps/s | VRAM=28.91 Go [2026-07-16 15:58:05] step 64900/1000000 | loss=2.7000 vocab=1.1118 token=1.3851 | 6.06 steps/s | VRAM=31.67 Go [2026-07-16 15:58:13] step 64950/1000000 | loss=2.6691 vocab=1.2957 token=1.8429 | 6.08 steps/s | VRAM=27.80 Go [2026-07-16 15:58:21] step 65000/1000000 | loss=2.6223 vocab=1.5770 token=1.8805 | 6.08 steps/s | VRAM=32.28 Go [2026-07-16 15:59:49] step 65050/1000000 | loss=2.6558 vocab=0.8093 token=1.3388 | 0.57 steps/s | VRAM=28.11 Go [2026-07-16 15:59:57] step 65100/1000000 | loss=2.6907 vocab=0.9236 token=1.4178 | 6.13 steps/s | VRAM=28.10 Go [2026-07-16 16:00:06] step 65150/1000000 | loss=2.5726 vocab=1.3585 token=1.5720 | 6.11 steps/s | VRAM=27.27 Go [2026-07-16 16:00:14] step 65200/1000000 | loss=2.6116 vocab=1.0059 token=1.7317 | 6.17 steps/s | VRAM=28.62 Go [2026-07-16 16:00:22] step 65250/1000000 | loss=2.6004 vocab=1.0005 token=1.7528 | 6.12 steps/s | VRAM=28.69 Go [2026-07-16 16:00:30] step 65300/1000000 | loss=2.5794 vocab=0.6970 token=1.5773 | 6.11 steps/s | VRAM=26.53 Go [2026-07-16 16:00:38] step 65350/1000000 | loss=2.6448 vocab=1.0300 token=1.7453 | 6.11 steps/s | VRAM=23.21 Go [2026-07-16 16:00:46] step 65400/1000000 | loss=2.6440 vocab=0.7831 token=1.3137 | 6.10 steps/s | VRAM=28.92 Go [2026-07-16 16:00:55] step 65450/1000000 | loss=2.5725 vocab=0.8326 token=1.3982 | 6.12 steps/s | VRAM=29.93 Go [2026-07-16 16:01:03] step 65500/1000000 | loss=2.6289 vocab=0.8523 token=1.4901 | 6.14 steps/s | VRAM=30.68 Go [2026-07-16 16:01:11] step 65550/1000000 | loss=2.6219 vocab=1.2043 token=1.7476 | 6.11 steps/s | VRAM=30.65 Go [2026-07-16 16:01:19] step 65600/1000000 | loss=2.6386 vocab=0.9611 token=1.7820 | 6.06 steps/s | VRAM=27.32 Go [2026-07-16 16:01:27] step 65650/1000000 | loss=2.6571 vocab=1.0035 token=1.9056 | 6.10 steps/s | VRAM=26.10 Go [2026-07-16 16:01:36] step 65700/1000000 | loss=2.6677 vocab=1.4835 token=1.6947 | 6.09 steps/s | VRAM=28.30 Go [2026-07-16 16:01:44] step 65750/1000000 | loss=2.5838 vocab=1.1118 token=1.4756 | 6.07 steps/s | VRAM=30.02 Go [2026-07-16 16:01:52] step 65800/1000000 | loss=2.6108 vocab=1.2350 token=1.9717 | 6.11 steps/s | VRAM=32.27 Go [2026-07-16 16:02:00] step 65850/1000000 | loss=2.5742 vocab=0.9200 token=1.8142 | 6.03 steps/s | VRAM=28.28 Go [2026-07-16 16:02:09] step 65900/1000000 | loss=2.6292 vocab=1.5175 token=1.8345 | 6.07 steps/s | VRAM=29.35 Go [2026-07-16 16:02:17] step 65950/1000000 | loss=2.5659 vocab=0.6558 token=1.5860 | 6.08 steps/s | VRAM=28.24 Go [2026-07-16 16:02:25] step 66000/1000000 | loss=2.5406 vocab=1.2201 token=2.0125 | 6.06 steps/s | VRAM=29.17 Go [2026-07-16 16:02:33] step 66050/1000000 | loss=2.5658 vocab=0.8831 token=1.5721 | 6.09 steps/s | VRAM=29.51 Go [2026-07-16 16:02:42] step 66100/1000000 | loss=2.6547 vocab=1.1038 token=1.9221 | 6.09 steps/s | VRAM=32.94 Go [2026-07-16 16:02:50] step 66150/1000000 | loss=2.7353 vocab=1.5096 token=1.3126 | 6.08 steps/s | VRAM=29.36 Go [2026-07-16 16:02:58] step 66200/1000000 | loss=2.6048 vocab=1.3778 token=1.5116 | 6.05 steps/s | VRAM=26.45 Go [2026-07-16 16:03:06] step 66250/1000000 | loss=2.6716 vocab=0.8424 token=1.3550 | 6.11 steps/s | VRAM=32.37 Go [2026-07-16 16:03:14] step 66300/1000000 | loss=2.5358 vocab=0.9984 token=1.4426 | 6.08 steps/s | VRAM=32.77 Go [2026-07-16 16:03:23] step 66350/1000000 | loss=2.6403 vocab=1.0997 token=1.3219 | 6.08 steps/s | VRAM=24.30 Go [2026-07-16 16:03:31] step 66400/1000000 | loss=2.5534 vocab=1.4620 token=1.5299 | 6.09 steps/s | VRAM=32.24 Go [2026-07-16 16:03:39] step 66450/1000000 | loss=2.5983 vocab=0.7946 token=1.6057 | 6.07 steps/s | VRAM=32.64 Go [2026-07-16 16:03:47] step 66500/1000000 | loss=2.6285 vocab=0.9364 token=1.4066 | 6.05 steps/s | VRAM=28.97 Go [2026-07-16 16:03:56] step 66550/1000000 | loss=2.5674 vocab=0.8718 token=1.5719 | 6.08 steps/s | VRAM=28.66 Go [2026-07-16 16:04:04] step 66600/1000000 | loss=2.6562 vocab=1.0082 token=1.7731 | 6.07 steps/s | VRAM=29.44 Go [2026-07-16 16:04:12] step 66650/1000000 | loss=2.5695 vocab=0.8761 token=1.1923 | 6.10 steps/s | VRAM=29.57 Go [2026-07-16 16:04:20] step 66700/1000000 | loss=2.5344 vocab=0.4571 token=0.8964 | 6.07 steps/s | VRAM=29.01 Go [2026-07-16 16:04:28] step 66750/1000000 | loss=2.4747 vocab=1.4020 token=1.6944 | 6.06 steps/s | VRAM=26.09 Go [2026-07-16 16:04:37] step 66800/1000000 | loss=2.5530 vocab=1.1980 token=1.2980 | 6.10 steps/s | VRAM=32.16 Go [2026-07-16 16:04:45] step 66850/1000000 | loss=2.6096 vocab=1.3861 token=1.6145 | 6.09 steps/s | VRAM=28.85 Go [2026-07-16 16:04:53] step 66900/1000000 | loss=2.6364 vocab=1.2496 token=1.7218 | 6.04 steps/s | VRAM=30.00 Go [2026-07-16 16:05:01] step 66950/1000000 | loss=2.6893 vocab=0.8633 token=1.4667 | 6.10 steps/s | VRAM=31.10 Go [2026-07-16 16:05:10] step 67000/1000000 | loss=2.6372 vocab=1.5492 token=1.6934 | 6.06 steps/s | VRAM=32.34 Go [2026-07-16 16:05:18] step 67050/1000000 | loss=2.5350 vocab=1.0818 token=1.6800 | 6.08 steps/s | VRAM=32.76 Go [2026-07-16 16:05:26] step 67100/1000000 | loss=2.5606 vocab=0.9140 token=2.0359 | 6.08 steps/s | VRAM=28.28 Go [2026-07-16 16:05:34] step 67150/1000000 | loss=2.6269 vocab=0.7632 token=1.2792 | 6.06 steps/s | VRAM=26.84 Go [2026-07-16 16:05:43] step 67200/1000000 | loss=2.6504 vocab=0.8114 token=1.2684 | 6.09 steps/s | VRAM=28.22 Go [2026-07-16 16:05:51] step 67250/1000000 | loss=2.7082 vocab=1.4193 token=1.6628 | 6.10 steps/s | VRAM=27.73 Go [2026-07-16 16:05:59] step 67300/1000000 | loss=2.5942 vocab=1.4515 token=1.6132 | 6.08 steps/s | VRAM=32.44 Go [2026-07-16 16:06:07] step 67350/1000000 | loss=2.6132 vocab=0.5601 token=1.4540 | 6.07 steps/s | VRAM=29.18 Go [2026-07-16 16:06:15] step 67400/1000000 | loss=2.6413 vocab=1.1347 token=1.4812 | 6.10 steps/s | VRAM=31.71 Go [2026-07-16 16:06:24] step 67450/1000000 | loss=2.6574 vocab=1.3240 token=1.8854 | 6.09 steps/s | VRAM=32.43 Go [2026-07-16 16:06:32] step 67500/1000000 | loss=2.5837 vocab=1.0389 token=1.7245 | 6.12 steps/s | VRAM=25.80 Go [2026-07-16 16:06:40] step 67550/1000000 | loss=2.6280 vocab=1.0609 token=1.5858 | 6.10 steps/s | VRAM=25.75 Go [2026-07-16 16:06:48] step 67600/1000000 | loss=2.6917 vocab=1.1500 token=1.8192 | 6.07 steps/s | VRAM=29.85 Go [2026-07-16 16:06:56] step 67650/1000000 | loss=2.5764 vocab=0.9715 token=1.5708 | 6.08 steps/s | VRAM=28.86 Go [2026-07-16 16:07:05] step 67700/1000000 | loss=2.5248 vocab=1.4969 token=1.5359 | 6.08 steps/s | VRAM=29.10 Go [2026-07-16 16:07:13] step 67750/1000000 | loss=2.7390 vocab=1.0156 token=1.6696 | 6.09 steps/s | VRAM=29.27 Go [2026-07-16 16:07:21] step 67800/1000000 | loss=2.5730 vocab=0.4851 token=1.4343 | 6.11 steps/s | VRAM=31.03 Go [2026-07-16 16:07:29] step 67850/1000000 | loss=2.6652 vocab=1.1126 token=1.5898 | 6.08 steps/s | VRAM=26.55 Go [2026-07-16 16:07:37] step 67900/1000000 | loss=2.5492 vocab=1.1560 token=1.5094 | 6.07 steps/s | VRAM=32.00 Go [2026-07-16 16:07:46] step 67950/1000000 | loss=2.6497 vocab=1.3814 token=1.6674 | 6.07 steps/s | VRAM=25.75 Go [2026-07-16 16:07:54] step 68000/1000000 | loss=2.4928 vocab=0.7482 token=1.3437 | 6.10 steps/s | VRAM=23.67 Go [2026-07-16 16:08:02] step 68050/1000000 | loss=2.6371 vocab=0.6547 token=1.6618 | 6.09 steps/s | VRAM=28.30 Go [2026-07-16 16:08:10] step 68100/1000000 | loss=2.5999 vocab=1.1007 token=1.3517 | 6.09 steps/s | VRAM=31.13 Go [2026-07-16 16:08:19] step 68150/1000000 | loss=2.6519 vocab=1.5485 token=1.5416 | 6.06 steps/s | VRAM=32.65 Go [2026-07-16 16:08:27] step 68200/1000000 | loss=2.6180 vocab=0.6455 token=1.5262 | 6.06 steps/s | VRAM=29.06 Go [2026-07-16 16:08:35] step 68250/1000000 | loss=2.6729 vocab=0.9923 token=1.8140 | 6.10 steps/s | VRAM=30.51 Go [2026-07-16 16:08:43] step 68300/1000000 | loss=2.6915 vocab=1.2738 token=1.4567 | 6.10 steps/s | VRAM=28.06 Go [2026-07-16 16:08:52] step 68350/1000000 | loss=2.5423 vocab=0.9091 token=1.6172 | 6.04 steps/s | VRAM=29.36 Go [2026-07-16 16:09:00] step 68400/1000000 | loss=2.5746 vocab=1.1593 token=1.6528 | 6.06 steps/s | VRAM=32.64 Go [2026-07-16 16:09:08] step 68450/1000000 | loss=2.7034 vocab=1.1409 token=1.5089 | 6.10 steps/s | VRAM=29.19 Go [2026-07-16 16:09:16] step 68500/1000000 | loss=2.5984 vocab=0.6831 token=1.4295 | 6.08 steps/s | VRAM=27.99 Go [2026-07-16 16:09:24] step 68550/1000000 | loss=2.6979 vocab=1.3957 token=1.7524 | 6.07 steps/s | VRAM=28.26 Go [2026-07-16 16:09:33] step 68600/1000000 | loss=2.6706 vocab=1.1596 token=1.4466 | 6.12 steps/s | VRAM=28.37 Go [2026-07-16 16:09:41] step 68650/1000000 | loss=2.6483 vocab=1.4081 token=1.5126 | 6.09 steps/s | VRAM=31.55 Go [2026-07-16 16:09:49] step 68700/1000000 | loss=2.4948 vocab=1.3931 token=1.7526 | 6.10 steps/s | VRAM=32.44 Go [2026-07-16 16:09:57] step 68750/1000000 | loss=2.5772 vocab=1.4919 token=1.7687 | 6.10 steps/s | VRAM=28.55 Go [2026-07-16 16:10:05] step 68800/1000000 | loss=2.5865 vocab=1.1122 token=1.6727 | 6.05 steps/s | VRAM=29.60 Go [2026-07-16 16:10:14] step 68850/1000000 | loss=2.6644 vocab=1.0856 token=1.5357 | 6.07 steps/s | VRAM=28.66 Go [2026-07-16 16:10:22] step 68900/1000000 | loss=2.4939 vocab=1.0278 token=1.1608 | 6.06 steps/s | VRAM=25.95 Go [2026-07-16 16:10:30] step 68950/1000000 | loss=2.6086 vocab=1.1264 token=1.6674 | 6.06 steps/s | VRAM=26.34 Go [2026-07-16 16:10:38] step 69000/1000000 | loss=2.6131 vocab=1.0142 token=1.7739 | 6.10 steps/s | VRAM=25.08 Go [2026-07-16 16:10:47] step 69050/1000000 | loss=2.6278 vocab=0.6127 token=1.6305 | 6.08 steps/s | VRAM=32.40 Go [2026-07-16 16:10:55] step 69100/1000000 | loss=2.5556 vocab=1.1464 token=1.6830 | 6.10 steps/s | VRAM=32.18 Go [2026-07-16 16:11:03] step 69150/1000000 | loss=2.7789 vocab=1.1685 token=1.5829 | 6.11 steps/s | VRAM=27.77 Go [2026-07-16 16:11:11] step 69200/1000000 | loss=2.6356 vocab=0.8097 token=1.6505 | 6.07 steps/s | VRAM=30.73 Go [2026-07-16 16:11:19] step 69250/1000000 | loss=2.6962 vocab=1.2701 token=1.7242 | 6.10 steps/s | VRAM=28.13 Go [2026-07-16 16:11:28] step 69300/1000000 | loss=2.5649 vocab=1.0295 token=1.2153 | 6.10 steps/s | VRAM=32.01 Go [2026-07-16 16:11:36] step 69350/1000000 | loss=2.5404 vocab=1.4059 token=1.6370 | 6.11 steps/s | VRAM=24.77 Go [2026-07-16 16:11:44] step 69400/1000000 | loss=2.5809 vocab=0.8997 token=1.3738 | 6.08 steps/s | VRAM=25.59 Go [2026-07-16 16:11:52] step 69450/1000000 | loss=2.6620 vocab=0.7960 token=1.6148 | 6.12 steps/s | VRAM=27.60 Go [2026-07-16 16:12:00] step 69500/1000000 | loss=2.5905 vocab=0.9218 token=1.6867 | 6.07 steps/s | VRAM=26.22 Go [2026-07-16 16:12:09] step 69550/1000000 | loss=2.4423 vocab=0.6174 token=1.4278 | 6.09 steps/s | VRAM=28.40 Go [2026-07-16 16:12:17] step 69600/1000000 | loss=2.6034 vocab=0.7257 token=1.6149 | 6.13 steps/s | VRAM=28.80 Go [2026-07-16 16:12:25] step 69650/1000000 | loss=2.5985 vocab=0.8449 token=1.4835 | 6.07 steps/s | VRAM=29.09 Go [2026-07-16 16:12:33] step 69700/1000000 | loss=2.6150 vocab=1.5389 token=1.6130 | 6.11 steps/s | VRAM=32.33 Go [2026-07-16 16:12:41] step 69750/1000000 | loss=2.5566 vocab=1.0556 token=1.9249 | 6.07 steps/s | VRAM=28.80 Go [2026-07-16 16:12:50] step 69800/1000000 | loss=2.6620 vocab=1.4929 token=2.1690 | 6.07 steps/s | VRAM=32.80 Go [2026-07-16 16:12:58] step 69850/1000000 | loss=2.5688 vocab=1.2192 token=1.8972 | 6.07 steps/s | VRAM=28.81 Go [2026-07-16 16:13:06] step 69900/1000000 | loss=2.5140 vocab=0.9547 token=1.8249 | 6.10 steps/s | VRAM=25.95 Go [2026-07-16 16:13:14] step 69950/1000000 | loss=2.6192 vocab=0.7250 token=1.3958 | 6.08 steps/s | VRAM=28.99 Go [2026-07-16 16:13:23] step 70000/1000000 | loss=2.5577 vocab=1.2034 token=1.4850 | 6.08 steps/s | VRAM=32.87 Go [2026-07-16 16:14:46] step 70050/1000000 | loss=2.5418 vocab=0.6791 token=1.4574 | 0.60 steps/s | VRAM=28.19 Go [2026-07-16 16:14:54] step 70100/1000000 | loss=2.6431 vocab=1.1711 token=1.5274 | 6.11 steps/s | VRAM=23.06 Go [2026-07-16 16:15:02] step 70150/1000000 | loss=2.6395 vocab=0.9765 token=1.5503 | 6.13 steps/s | VRAM=28.87 Go [2026-07-16 16:15:10] step 70200/1000000 | loss=2.6095 vocab=0.8388 token=1.5230 | 6.13 steps/s | VRAM=26.69 Go [2026-07-16 16:15:18] step 70250/1000000 | loss=2.5512 vocab=1.1297 token=1.7931 | 6.15 steps/s | VRAM=31.01 Go [2026-07-16 16:15:27] step 70300/1000000 | loss=2.7617 vocab=0.9084 token=1.7202 | 6.13 steps/s | VRAM=27.91 Go [2026-07-16 16:15:35] step 70350/1000000 | loss=2.6226 vocab=1.0171 token=1.3982 | 6.10 steps/s | VRAM=29.31 Go [2026-07-16 16:15:43] step 70400/1000000 | loss=2.5758 vocab=1.0984 token=1.2577 | 6.09 steps/s | VRAM=32.15 Go [2026-07-16 16:15:51] step 70450/1000000 | loss=2.5906 vocab=1.0184 token=1.8319 | 6.11 steps/s | VRAM=28.22 Go [2026-07-16 16:15:59] step 70500/1000000 | loss=2.6308 vocab=0.5982 token=1.4495 | 6.01 steps/s | VRAM=29.84 Go [2026-07-16 16:16:08] step 70550/1000000 | loss=2.6023 vocab=1.0425 token=1.6424 | 6.15 steps/s | VRAM=27.74 Go [2026-07-16 16:16:16] step 70600/1000000 | loss=2.6789 vocab=0.8843 token=1.3243 | 6.12 steps/s | VRAM=31.72 Go [2026-07-16 16:16:24] step 70650/1000000 | loss=2.5850 vocab=1.2250 token=1.7522 | 6.08 steps/s | VRAM=27.01 Go [2026-07-16 16:16:32] step 70700/1000000 | loss=2.6125 vocab=1.0454 token=1.6796 | 6.10 steps/s | VRAM=25.60 Go [2026-07-16 16:16:40] step 70750/1000000 | loss=2.4962 vocab=0.7972 token=1.5679 | 6.11 steps/s | VRAM=27.47 Go [2026-07-16 16:16:49] step 70800/1000000 | loss=2.7192 vocab=1.2995 token=1.5885 | 6.08 steps/s | VRAM=28.07 Go [2026-07-16 16:16:57] step 70850/1000000 | loss=2.7282 vocab=1.0499 token=1.4784 | 6.07 steps/s | VRAM=29.00 Go [2026-07-16 16:17:05] step 70900/1000000 | loss=2.5065 vocab=1.5266 token=1.8022 | 6.08 steps/s | VRAM=25.56 Go [2026-07-16 16:17:13] step 70950/1000000 | loss=2.5809 vocab=0.9483 token=1.9002 | 6.08 steps/s | VRAM=26.34 Go [2026-07-16 16:17:21] step 71000/1000000 | loss=2.5812 vocab=0.6546 token=1.4656 | 6.09 steps/s | VRAM=28.07 Go [2026-07-16 16:17:30] step 71050/1000000 | loss=2.6900 vocab=0.7688 token=1.6124 | 6.07 steps/s | VRAM=26.52 Go [2026-07-16 16:17:38] step 71100/1000000 | loss=2.7331 vocab=0.9067 token=2.0943 | 6.10 steps/s | VRAM=28.11 Go [2026-07-16 16:17:46] step 71150/1000000 | loss=2.6580 vocab=1.2616 token=1.4823 | 6.11 steps/s | VRAM=28.08 Go [2026-07-16 16:17:54] step 71200/1000000 | loss=2.5366 vocab=0.9151 token=1.4455 | 6.09 steps/s | VRAM=29.13 Go [2026-07-16 16:18:03] step 71250/1000000 | loss=2.6218 vocab=1.5524 token=1.7059 | 6.07 steps/s | VRAM=27.06 Go [2026-07-16 16:18:11] step 71300/1000000 | loss=2.6744 vocab=0.8539 token=1.4687 | 6.10 steps/s | VRAM=28.15 Go [2026-07-16 16:18:19] step 71350/1000000 | loss=2.5737 vocab=1.0917 token=1.4943 | 6.09 steps/s | VRAM=28.55 Go [2026-07-16 16:18:27] step 71400/1000000 | loss=2.5700 vocab=0.7477 token=0.9943 | 6.13 steps/s | VRAM=28.43 Go [2026-07-16 16:18:35] step 71450/1000000 | loss=2.5156 vocab=0.8900 token=1.4823 | 6.10 steps/s | VRAM=29.47 Go [2026-07-16 16:18:43] step 71500/1000000 | loss=2.5905 vocab=1.2646 token=1.3190 | 6.10 steps/s | VRAM=28.62 Go [2026-07-16 16:18:52] step 71550/1000000 | loss=2.4367 vocab=0.9462 token=1.5741 | 6.10 steps/s | VRAM=25.68 Go [2026-07-16 16:19:00] step 71600/1000000 | loss=2.7065 vocab=1.4831 token=1.4042 | 6.09 steps/s | VRAM=32.23 Go [2026-07-16 16:19:08] step 71650/1000000 | loss=2.5955 vocab=0.7409 token=1.1811 | 6.04 steps/s | VRAM=28.92 Go [2026-07-16 16:19:16] step 71700/1000000 | loss=2.5767 vocab=1.0697 token=1.3005 | 6.10 steps/s | VRAM=29.07 Go [2026-07-16 16:19:25] step 71750/1000000 | loss=2.4718 vocab=0.8688 token=1.3485 | 6.06 steps/s | VRAM=29.80 Go [2026-07-16 16:19:33] step 71800/1000000 | loss=2.5993 vocab=0.5874 token=1.4994 | 6.09 steps/s | VRAM=30.97 Go [2026-07-16 16:19:41] step 71850/1000000 | loss=2.6300 vocab=1.1368 token=1.7636 | 6.11 steps/s | VRAM=31.55 Go [2026-07-16 16:19:49] step 71900/1000000 | loss=2.6142 vocab=0.6219 token=1.6836 | 6.10 steps/s | VRAM=29.54 Go [2026-07-16 16:19:57] step 71950/1000000 | loss=2.6326 vocab=0.4640 token=1.3580 | 6.09 steps/s | VRAM=27.34 Go [2026-07-16 16:20:06] step 72000/1000000 | loss=2.5919 vocab=1.1731 token=1.9194 | 6.10 steps/s | VRAM=27.93 Go [2026-07-16 16:20:14] step 72050/1000000 | loss=2.6343 vocab=1.0195 token=1.0789 | 6.10 steps/s | VRAM=26.61 Go [2026-07-16 16:20:22] step 72100/1000000 | loss=2.5853 vocab=1.1766 token=1.6274 | 6.09 steps/s | VRAM=28.93 Go [2026-07-16 16:20:30] step 72150/1000000 | loss=2.5446 vocab=1.0474 token=1.7445 | 6.12 steps/s | VRAM=28.52 Go [2026-07-16 16:20:38] step 72200/1000000 | loss=2.7342 vocab=1.2962 token=1.4246 | 6.10 steps/s | VRAM=32.39 Go [2026-07-16 16:20:47] step 72250/1000000 | loss=2.6054 vocab=1.2239 token=1.6989 | 6.07 steps/s | VRAM=31.70 Go [2026-07-16 16:20:55] step 72300/1000000 | loss=2.6008 vocab=1.4041 token=1.7392 | 6.08 steps/s | VRAM=28.10 Go [2026-07-16 16:21:03] step 72350/1000000 | loss=2.6756 vocab=0.9124 token=1.6217 | 6.04 steps/s | VRAM=29.37 Go [2026-07-16 16:21:11] step 72400/1000000 | loss=2.6268 vocab=1.5481 token=1.4975 | 6.09 steps/s | VRAM=28.27 Go [2026-07-16 16:21:20] step 72450/1000000 | loss=2.6357 vocab=0.7515 token=1.3968 | 6.08 steps/s | VRAM=26.63 Go [2026-07-16 16:21:28] step 72500/1000000 | loss=2.5669 vocab=1.2058 token=1.5721 | 6.08 steps/s | VRAM=27.30 Go [2026-07-16 16:21:36] step 72550/1000000 | loss=2.6583 vocab=1.3246 token=1.6431 | 6.07 steps/s | VRAM=26.80 Go [2026-07-16 16:21:44] step 72600/1000000 | loss=2.5607 vocab=1.2947 token=1.5632 | 6.09 steps/s | VRAM=29.02 Go [2026-07-16 16:21:52] step 72650/1000000 | loss=2.6437 vocab=1.0974 token=1.5903 | 6.08 steps/s | VRAM=26.48 Go [2026-07-16 16:22:01] step 72700/1000000 | loss=2.5089 vocab=1.1747 token=1.5045 | 6.07 steps/s | VRAM=22.60 Go [2026-07-16 16:22:09] step 72750/1000000 | loss=2.7182 vocab=0.7396 token=1.4083 | 6.12 steps/s | VRAM=31.43 Go [2026-07-16 16:22:17] step 72800/1000000 | loss=2.6562 vocab=1.2378 token=1.7060 | 6.06 steps/s | VRAM=29.59 Go [2026-07-16 16:22:25] step 72850/1000000 | loss=2.5847 vocab=0.8703 token=1.1986 | 6.08 steps/s | VRAM=28.14 Go [2026-07-16 16:22:34] step 72900/1000000 | loss=2.5883 vocab=0.8223 token=1.0955 | 6.08 steps/s | VRAM=25.29 Go [2026-07-16 16:22:42] step 72950/1000000 | loss=2.5071 vocab=1.0218 token=1.6484 | 6.07 steps/s | VRAM=26.46 Go [2026-07-16 16:22:50] step 73000/1000000 | loss=2.5964 vocab=1.0690 token=1.6501 | 6.13 steps/s | VRAM=32.33 Go [2026-07-16 16:22:58] step 73050/1000000 | loss=2.5769 vocab=0.7622 token=1.2903 | 6.07 steps/s | VRAM=29.50 Go [2026-07-16 16:23:06] step 73100/1000000 | loss=2.5747 vocab=1.2310 token=1.7504 | 6.08 steps/s | VRAM=31.16 Go [2026-07-16 16:23:15] step 73150/1000000 | loss=2.5844 vocab=0.8851 token=1.8269 | 6.07 steps/s | VRAM=32.33 Go [2026-07-16 16:23:23] step 73200/1000000 | loss=2.7380 vocab=1.5579 token=1.6709 | 6.06 steps/s | VRAM=29.24 Go [2026-07-16 16:23:31] step 73250/1000000 | loss=2.6281 vocab=1.0271 token=1.6596 | 6.07 steps/s | VRAM=27.84 Go [2026-07-16 16:23:39] step 73300/1000000 | loss=2.5553 vocab=0.7626 token=1.5127 | 6.09 steps/s | VRAM=26.28 Go [2026-07-16 16:23:48] step 73350/1000000 | loss=2.5750 vocab=1.6325 token=1.6860 | 6.09 steps/s | VRAM=28.81 Go [2026-07-16 16:23:56] step 73400/1000000 | loss=2.5543 vocab=1.0951 token=2.0399 | 6.11 steps/s | VRAM=26.31 Go [2026-07-16 16:24:04] step 73450/1000000 | loss=2.6654 vocab=0.8931 token=1.6051 | 6.12 steps/s | VRAM=32.20 Go [2026-07-16 16:24:12] step 73500/1000000 | loss=2.5928 vocab=1.1602 token=1.9519 | 6.10 steps/s | VRAM=32.94 Go [2026-07-16 16:24:20] step 73550/1000000 | loss=2.6165 vocab=1.1781 token=1.4130 | 6.09 steps/s | VRAM=26.75 Go [2026-07-16 16:24:28] step 73600/1000000 | loss=2.4871 vocab=0.7512 token=1.2898 | 6.11 steps/s | VRAM=31.66 Go [2026-07-16 16:24:37] step 73650/1000000 | loss=2.6297 vocab=0.8622 token=1.4897 | 6.10 steps/s | VRAM=25.12 Go [2026-07-16 16:24:45] step 73700/1000000 | loss=2.5095 vocab=1.1667 token=1.6960 | 6.09 steps/s | VRAM=29.82 Go [2026-07-16 16:24:53] step 73750/1000000 | loss=2.5783 vocab=1.2374 token=1.3348 | 6.05 steps/s | VRAM=23.86 Go [2026-07-16 16:25:01] step 73800/1000000 | loss=2.6303 vocab=1.1266 token=1.3774 | 6.12 steps/s | VRAM=31.30 Go [2026-07-16 16:25:10] step 73850/1000000 | loss=2.6073 vocab=1.0292 token=1.2127 | 6.11 steps/s | VRAM=26.13 Go [2026-07-16 16:25:18] step 73900/1000000 | loss=2.5681 vocab=1.0138 token=1.3431 | 6.09 steps/s | VRAM=28.38 Go [2026-07-16 16:25:26] step 73950/1000000 | loss=2.5843 vocab=1.1246 token=1.6182 | 6.12 steps/s | VRAM=31.39 Go [2026-07-16 16:25:34] step 74000/1000000 | loss=2.5484 vocab=0.8008 token=1.4826 | 6.11 steps/s | VRAM=31.79 Go [2026-07-16 16:25:42] step 74050/1000000 | loss=2.6238 vocab=0.9779 token=1.6653 | 6.09 steps/s | VRAM=32.37 Go [2026-07-16 16:25:51] step 74100/1000000 | loss=2.5340 vocab=0.7566 token=1.5802 | 6.07 steps/s | VRAM=28.07 Go [2026-07-16 16:25:59] step 74150/1000000 | loss=2.4953 vocab=1.0075 token=1.2195 | 6.07 steps/s | VRAM=29.03 Go [2026-07-16 16:26:07] step 74200/1000000 | loss=2.6649 vocab=0.8352 token=1.2538 | 6.10 steps/s | VRAM=27.45 Go [2026-07-16 16:26:15] step 74250/1000000 | loss=2.6725 vocab=1.1174 token=1.7545 | 6.10 steps/s | VRAM=31.86 Go [2026-07-16 16:26:23] step 74300/1000000 | loss=2.6498 vocab=1.0939 token=1.5118 | 6.09 steps/s | VRAM=28.37 Go [2026-07-16 16:26:32] step 74350/1000000 | loss=2.5113 vocab=1.1299 token=1.4841 | 6.08 steps/s | VRAM=32.39 Go [2026-07-16 16:26:40] step 74400/1000000 | loss=2.5076 vocab=0.9415 token=1.9186 | 6.09 steps/s | VRAM=31.78 Go [2026-07-16 16:26:48] step 74450/1000000 | loss=2.5734 vocab=1.7403 token=1.6605 | 6.11 steps/s | VRAM=31.53 Go [2026-07-16 16:26:56] step 74500/1000000 | loss=2.5680 vocab=0.9857 token=1.6761 | 6.10 steps/s | VRAM=29.40 Go [2026-07-16 16:27:04] step 74550/1000000 | loss=2.5412 vocab=0.8477 token=1.4011 | 6.08 steps/s | VRAM=26.18 Go [2026-07-16 16:27:13] step 74600/1000000 | loss=2.6080 vocab=0.7576 token=1.6076 | 6.12 steps/s | VRAM=31.28 Go [2026-07-16 16:27:21] step 74650/1000000 | loss=2.6094 vocab=0.8492 token=1.6116 | 6.07 steps/s | VRAM=28.28 Go [2026-07-16 16:27:29] step 74700/1000000 | loss=2.5821 vocab=1.2349 token=1.4710 | 6.07 steps/s | VRAM=28.48 Go [2026-07-16 16:27:37] step 74750/1000000 | loss=2.6283 vocab=1.3610 token=1.8027 | 6.11 steps/s | VRAM=31.99 Go [2026-07-16 16:27:45] step 74800/1000000 | loss=2.6877 vocab=0.8048 token=1.1897 | 6.11 steps/s | VRAM=31.75 Go [2026-07-16 16:27:54] step 74850/1000000 | loss=2.7014 vocab=1.1469 token=1.7985 | 6.10 steps/s | VRAM=26.65 Go [2026-07-16 16:28:02] step 74900/1000000 | loss=2.5494 vocab=1.2340 token=1.8466 | 6.11 steps/s | VRAM=29.56 Go [2026-07-16 16:28:10] step 74950/1000000 | loss=2.5009 vocab=1.5092 token=1.6898 | 6.12 steps/s | VRAM=32.68 Go [2026-07-16 16:28:18] step 75000/1000000 | loss=2.5787 vocab=1.3890 token=1.5345 | 6.07 steps/s | VRAM=26.62 Go [2026-07-16 16:31:19] step 75050/1000000 | loss=2.5598 vocab=1.7985 token=1.3922 | 0.28 steps/s | VRAM=32.74 Go [2026-07-16 16:31:27] step 75100/1000000 | loss=2.5078 vocab=1.1000 token=1.3122 | 6.18 steps/s | VRAM=29.16 Go [2026-07-16 16:31:35] step 75150/1000000 | loss=2.5924 vocab=0.8729 token=1.7210 | 6.14 steps/s | VRAM=26.49 Go [2026-07-16 16:31:43] step 75200/1000000 | loss=2.6002 vocab=0.6521 token=1.1066 | 6.20 steps/s | VRAM=31.25 Go [2026-07-16 16:31:51] step 75250/1000000 | loss=2.5530 vocab=0.7181 token=1.2617 | 6.21 steps/s | VRAM=32.12 Go [2026-07-16 16:31:59] step 75300/1000000 | loss=2.6151 vocab=0.8489 token=1.3240 | 6.11 steps/s | VRAM=28.20 Go [2026-07-16 16:32:07] step 75350/1000000 | loss=2.6138 vocab=0.7733 token=1.5382 | 6.13 steps/s | VRAM=29.18 Go [2026-07-16 16:32:15] step 75400/1000000 | loss=2.5341 vocab=1.3409 token=1.8090 | 6.13 steps/s | VRAM=28.34 Go [2026-07-16 16:32:24] step 75450/1000000 | loss=2.5407 vocab=0.5833 token=1.6918 | 6.14 steps/s | VRAM=26.09 Go [2026-07-16 16:32:32] step 75500/1000000 | loss=2.5819 vocab=1.6600 token=1.5959 | 6.10 steps/s | VRAM=29.05 Go [2026-07-16 16:32:40] step 75550/1000000 | loss=2.4302 vocab=1.1393 token=1.4369 | 6.14 steps/s | VRAM=31.78 Go [2026-07-16 16:32:48] step 75600/1000000 | loss=2.5122 vocab=1.2984 token=1.5565 | 6.10 steps/s | VRAM=26.03 Go [2026-07-16 16:32:56] step 75650/1000000 | loss=2.5770 vocab=0.6830 token=1.5397 | 6.11 steps/s | VRAM=31.83 Go [2026-07-16 16:33:05] step 75700/1000000 | loss=2.6695 vocab=1.1811 token=1.8690 | 6.08 steps/s | VRAM=25.96 Go [2026-07-16 16:33:13] step 75750/1000000 | loss=2.5819 vocab=0.6784 token=1.0382 | 6.09 steps/s | VRAM=26.55 Go [2026-07-16 16:33:21] step 75800/1000000 | loss=2.6213 vocab=1.1572 token=1.6101 | 6.10 steps/s | VRAM=27.70 Go [2026-07-16 16:33:29] step 75850/1000000 | loss=2.5148 vocab=1.1240 token=1.5351 | 6.08 steps/s | VRAM=25.41 Go [2026-07-16 16:33:37] step 75900/1000000 | loss=2.5981 vocab=0.9351 token=1.8021 | 6.09 steps/s | VRAM=26.68 Go [2026-07-16 16:33:46] step 75950/1000000 | loss=2.6507 vocab=1.0253 token=1.7507 | 6.12 steps/s | VRAM=31.27 Go [2026-07-16 16:33:54] step 76000/1000000 | loss=2.5990 vocab=1.4335 token=1.6539 | 6.07 steps/s | VRAM=26.43 Go [2026-07-16 16:34:02] step 76050/1000000 | loss=2.6334 vocab=1.0641 token=1.7352 | 6.10 steps/s | VRAM=27.76 Go [2026-07-16 16:34:10] step 76100/1000000 | loss=2.6357 vocab=0.8330 token=1.8276 | 6.10 steps/s | VRAM=28.34 Go [2026-07-16 16:34:18] step 76150/1000000 | loss=2.6422 vocab=0.9745 token=0.9147 | 6.10 steps/s | VRAM=28.02 Go [2026-07-16 16:34:27] step 76200/1000000 | loss=2.4773 vocab=0.8364 token=1.4482 | 6.05 steps/s | VRAM=28.23 Go [2026-07-16 16:34:35] step 76250/1000000 | loss=2.6031 vocab=1.8517 token=1.8379 | 6.06 steps/s | VRAM=31.93 Go [2026-07-16 16:34:43] step 76300/1000000 | loss=2.5896 vocab=1.6832 token=1.3731 | 6.06 steps/s | VRAM=28.24 Go [2026-07-16 16:34:51] step 76350/1000000 | loss=2.6501 vocab=1.4732 token=1.7722 | 6.09 steps/s | VRAM=32.32 Go [2026-07-16 16:35:00] step 76400/1000000 | loss=2.6326 vocab=0.6776 token=2.0685 | 6.08 steps/s | VRAM=25.31 Go [2026-07-16 16:35:08] step 76450/1000000 | loss=2.5672 vocab=0.9594 token=1.0839 | 6.07 steps/s | VRAM=27.71 Go [2026-07-16 16:35:16] step 76500/1000000 | loss=2.6596 vocab=0.6361 token=2.0306 | 6.10 steps/s | VRAM=31.41 Go [2026-07-16 16:35:24] step 76550/1000000 | loss=2.5335 vocab=1.2167 token=1.6209 | 6.07 steps/s | VRAM=28.19 Go [2026-07-16 16:35:33] step 76600/1000000 | loss=2.4860 vocab=1.1185 token=1.5894 | 6.05 steps/s | VRAM=28.84 Go [2026-07-16 16:35:41] step 76650/1000000 | loss=2.4969 vocab=1.1300 token=1.6214 | 6.08 steps/s | VRAM=24.92 Go [2026-07-16 16:35:49] step 76700/1000000 | loss=2.6054 vocab=0.8406 token=1.8950 | 6.04 steps/s | VRAM=27.56 Go [2026-07-16 16:35:57] step 76750/1000000 | loss=2.4927 vocab=0.8784 token=1.7109 | 6.09 steps/s | VRAM=25.96 Go [2026-07-16 16:36:05] step 76800/1000000 | loss=2.4296 vocab=0.5855 token=1.2077 | 6.12 steps/s | VRAM=31.03 Go [2026-07-16 16:36:14] step 76850/1000000 | loss=2.5393 vocab=1.2919 token=1.4336 | 6.06 steps/s | VRAM=29.87 Go [2026-07-16 16:36:22] step 76900/1000000 | loss=2.6188 vocab=1.1348 token=1.6534 | 6.06 steps/s | VRAM=29.09 Go [2026-07-16 16:36:30] step 76950/1000000 | loss=2.5649 vocab=0.7748 token=1.0856 | 6.10 steps/s | VRAM=26.16 Go [2026-07-16 16:36:38] step 77000/1000000 | loss=2.6172 vocab=0.9103 token=1.4076 | 6.09 steps/s | VRAM=28.33 Go [2026-07-16 16:36:46] step 77050/1000000 | loss=2.6242 vocab=1.3050 token=1.6148 | 6.11 steps/s | VRAM=30.26 Go [2026-07-16 16:36:55] step 77100/1000000 | loss=2.5498 vocab=1.3823 token=1.5849 | 6.09 steps/s | VRAM=28.65 Go [2026-07-16 16:37:03] step 77150/1000000 | loss=2.6438 vocab=1.1367 token=1.7210 | 6.06 steps/s | VRAM=27.21 Go [2026-07-16 16:37:11] step 77200/1000000 | loss=2.6831 vocab=1.4057 token=1.7842 | 6.08 steps/s | VRAM=31.72 Go [2026-07-16 16:37:19] step 77250/1000000 | loss=2.5596 vocab=1.4309 token=1.5428 | 6.07 steps/s | VRAM=28.84 Go [2026-07-16 16:37:28] step 77300/1000000 | loss=2.5065 vocab=0.8828 token=1.0662 | 6.08 steps/s | VRAM=26.62 Go [2026-07-16 16:37:36] step 77350/1000000 | loss=2.6678 vocab=1.4291 token=1.4014 | 6.08 steps/s | VRAM=32.47 Go [2026-07-16 16:37:44] step 77400/1000000 | loss=2.6703 vocab=1.1571 token=1.6920 | 6.05 steps/s | VRAM=28.15 Go [2026-07-16 16:37:52] step 77450/1000000 | loss=2.6350 vocab=0.9638 token=1.2185 | 6.07 steps/s | VRAM=30.81 Go [2026-07-16 16:38:01] step 77500/1000000 | loss=2.6435 vocab=1.0952 token=1.4321 | 6.06 steps/s | VRAM=26.25 Go [2026-07-16 16:38:09] step 77550/1000000 | loss=2.4879 vocab=0.7943 token=1.1749 | 6.07 steps/s | VRAM=28.75 Go [2026-07-16 16:38:17] step 77600/1000000 | loss=2.6754 vocab=1.1325 token=1.6861 | 6.09 steps/s | VRAM=31.77 Go [2026-07-16 16:38:25] step 77650/1000000 | loss=2.5243 vocab=1.0668 token=1.4451 | 6.08 steps/s | VRAM=29.45 Go [2026-07-16 16:38:34] step 77700/1000000 | loss=2.4573 vocab=1.0680 token=1.6191 | 6.07 steps/s | VRAM=28.76 Go [2026-07-16 16:38:42] step 77750/1000000 | loss=2.6722 vocab=1.3048 token=1.6482 | 6.08 steps/s | VRAM=28.78 Go [2026-07-16 16:38:50] step 77800/1000000 | loss=2.6393 vocab=1.0398 token=1.4309 | 6.12 steps/s | VRAM=28.34 Go [2026-07-16 16:38:58] step 77850/1000000 | loss=2.5834 vocab=1.3333 token=1.4253 | 6.07 steps/s | VRAM=25.97 Go [2026-07-16 16:39:06] step 77900/1000000 | loss=2.4866 vocab=0.5878 token=1.3176 | 6.05 steps/s | VRAM=29.41 Go [2026-07-16 16:39:15] step 77950/1000000 | loss=2.5709 vocab=1.2594 token=1.4638 | 6.08 steps/s | VRAM=28.93 Go [2026-07-16 16:39:23] step 78000/1000000 | loss=2.5868 vocab=1.5465 token=1.6482 | 6.07 steps/s | VRAM=29.17 Go [2026-07-16 16:39:31] step 78050/1000000 | loss=2.5004 vocab=1.0955 token=1.6572 | 6.09 steps/s | VRAM=31.59 Go [2026-07-16 16:39:39] step 78100/1000000 | loss=2.6244 vocab=0.9818 token=1.3423 | 6.11 steps/s | VRAM=25.31 Go [2026-07-16 16:39:47] step 78150/1000000 | loss=2.5484 vocab=0.7476 token=1.6121 | 6.11 steps/s | VRAM=32.24 Go [2026-07-16 16:39:56] step 78200/1000000 | loss=2.5753 vocab=1.3005 token=1.5914 | 6.10 steps/s | VRAM=29.21 Go [2026-07-16 16:40:04] step 78250/1000000 | loss=2.4565 vocab=0.8689 token=1.5543 | 6.11 steps/s | VRAM=24.99 Go [2026-07-16 16:40:12] step 78300/1000000 | loss=2.4572 vocab=0.9320 token=1.6788 | 6.09 steps/s | VRAM=32.97 Go [2026-07-16 16:40:20] step 78350/1000000 | loss=2.6182 vocab=1.0865 token=1.7936 | 6.10 steps/s | VRAM=28.46 Go [2026-07-16 16:40:28] step 78400/1000000 | loss=2.6276 vocab=0.8360 token=1.4630 | 6.11 steps/s | VRAM=28.59 Go [2026-07-16 16:40:37] step 78450/1000000 | loss=2.6090 vocab=0.8740 token=1.0106 | 6.12 steps/s | VRAM=31.03 Go [2026-07-16 16:40:45] step 78500/1000000 | loss=2.6851 vocab=1.2069 token=1.6618 | 6.10 steps/s | VRAM=31.97 Go [2026-07-16 16:40:53] step 78550/1000000 | loss=2.5842 vocab=1.0239 token=1.5263 | 6.09 steps/s | VRAM=26.64 Go [2026-07-16 16:41:01] step 78600/1000000 | loss=2.6098 vocab=1.1251 token=1.3797 | 6.09 steps/s | VRAM=28.29 Go [2026-07-16 16:41:09] step 78650/1000000 | loss=2.6032 vocab=1.4539 token=1.6887 | 6.08 steps/s | VRAM=23.68 Go [2026-07-16 16:41:18] step 78700/1000000 | loss=2.6486 vocab=1.1579 token=1.5172 | 6.09 steps/s | VRAM=27.99 Go [2026-07-16 16:41:26] step 78750/1000000 | loss=2.6083 vocab=1.6830 token=1.7838 | 6.11 steps/s | VRAM=28.95 Go [2026-07-16 16:41:34] step 78800/1000000 | loss=2.5547 vocab=1.1715 token=1.6177 | 6.10 steps/s | VRAM=32.66 Go [2026-07-16 16:41:42] step 78850/1000000 | loss=2.5284 vocab=0.7455 token=1.5984 | 6.08 steps/s | VRAM=25.69 Go [2026-07-16 16:41:50] step 78900/1000000 | loss=2.5509 vocab=0.5283 token=1.2626 | 6.09 steps/s | VRAM=32.82 Go [2026-07-16 16:41:59] step 78950/1000000 | loss=2.7183 vocab=0.7268 token=2.0394 | 6.08 steps/s | VRAM=32.26 Go [2026-07-16 16:42:07] step 79000/1000000 | loss=2.6308 vocab=1.0996 token=1.5382 | 6.11 steps/s | VRAM=32.68 Go [2026-07-16 16:42:15] step 79050/1000000 | loss=2.5865 vocab=0.7642 token=1.6712 | 6.09 steps/s | VRAM=32.99 Go [2026-07-16 16:42:23] step 79100/1000000 | loss=2.4564 vocab=0.6994 token=1.4710 | 6.11 steps/s | VRAM=26.10 Go [2026-07-16 16:42:32] step 79150/1000000 | loss=2.5962 vocab=0.7378 token=1.5260 | 6.08 steps/s | VRAM=26.18 Go [2026-07-16 16:42:40] step 79200/1000000 | loss=2.6734 vocab=0.9899 token=1.4347 | 6.09 steps/s | VRAM=28.58 Go [2026-07-16 16:42:48] step 79250/1000000 | loss=2.5684 vocab=0.5371 token=1.3439 | 6.13 steps/s | VRAM=32.03 Go [2026-07-16 16:42:56] step 79300/1000000 | loss=2.5018 vocab=1.1602 token=0.8674 | 6.07 steps/s | VRAM=28.77 Go [2026-07-16 16:43:04] step 79350/1000000 | loss=2.6461 vocab=0.8353 token=1.6055 | 6.04 steps/s | VRAM=25.26 Go [2026-07-16 16:43:13] step 79400/1000000 | loss=2.6017 vocab=0.6767 token=1.0066 | 6.07 steps/s | VRAM=28.56 Go [2026-07-16 16:43:21] step 79450/1000000 | loss=2.5920 vocab=0.9740 token=1.7012 | 6.09 steps/s | VRAM=25.69 Go [2026-07-16 16:43:29] step 79500/1000000 | loss=2.6296 vocab=0.9814 token=1.8216 | 6.07 steps/s | VRAM=32.78 Go [2026-07-16 16:43:37] step 79550/1000000 | loss=2.5886 vocab=0.7052 token=1.4057 | 6.10 steps/s | VRAM=28.89 Go [2026-07-16 16:43:46] step 79600/1000000 | loss=2.5819 vocab=1.2205 token=1.4611 | 6.08 steps/s | VRAM=26.28 Go [2026-07-16 16:43:54] step 79650/1000000 | loss=2.6232 vocab=1.6632 token=1.6193 | 6.06 steps/s | VRAM=26.34 Go [2026-07-16 16:44:02] step 79700/1000000 | loss=2.5864 vocab=0.7602 token=1.2458 | 6.11 steps/s | VRAM=29.05 Go [2026-07-16 16:44:10] step 79750/1000000 | loss=2.6408 vocab=0.5496 token=1.3496 | 6.09 steps/s | VRAM=28.21 Go [2026-07-16 16:44:18] step 79800/1000000 | loss=2.5872 vocab=0.8612 token=1.2676 | 6.07 steps/s | VRAM=25.44 Go [2026-07-16 16:44:27] step 79850/1000000 | loss=2.6139 vocab=0.7479 token=1.0317 | 6.08 steps/s | VRAM=29.71 Go [2026-07-16 16:44:35] step 79900/1000000 | loss=2.5541 vocab=0.8863 token=1.2427 | 6.08 steps/s | VRAM=26.03 Go [2026-07-16 16:44:43] step 79950/1000000 | loss=2.4893 vocab=1.3653 token=1.1715 | 6.10 steps/s | VRAM=32.04 Go [2026-07-16 16:44:51] step 80000/1000000 | loss=2.5148 vocab=1.4201 token=1.4034 | 6.09 steps/s | VRAM=27.96 Go [2026-07-16 16:46:15] step 80050/1000000 | loss=2.4906 vocab=1.0571 token=1.5019 | 0.60 steps/s | VRAM=32.65 Go [2026-07-16 16:46:23] step 80100/1000000 | loss=2.5855 vocab=1.6202 token=1.6820 | 6.15 steps/s | VRAM=31.91 Go [2026-07-16 16:46:31] step 80150/1000000 | loss=2.4734 vocab=0.8815 token=1.3523 | 6.15 steps/s | VRAM=32.48 Go [2026-07-16 16:46:39] step 80200/1000000 | loss=2.5450 vocab=1.0146 token=1.3294 | 6.13 steps/s | VRAM=28.75 Go [2026-07-16 16:46:48] step 80250/1000000 | loss=2.6000 vocab=0.6813 token=1.3868 | 6.15 steps/s | VRAM=28.54 Go [2026-07-16 16:46:56] step 80300/1000000 | loss=2.5945 vocab=1.2119 token=1.4536 | 6.12 steps/s | VRAM=31.26 Go [2026-07-16 16:47:04] step 80350/1000000 | loss=2.5635 vocab=1.0600 token=1.7572 | 6.11 steps/s | VRAM=32.30 Go [2026-07-16 16:47:12] step 80400/1000000 | loss=2.5740 vocab=1.1114 token=1.5599 | 6.12 steps/s | VRAM=31.67 Go [2026-07-16 16:47:20] step 80450/1000000 | loss=2.6442 vocab=1.2685 token=1.7365 | 6.10 steps/s | VRAM=29.07 Go [2026-07-16 16:47:28] step 80500/1000000 | loss=2.5783 vocab=0.8657 token=1.3960 | 6.11 steps/s | VRAM=32.04 Go [2026-07-16 16:47:37] step 80550/1000000 | loss=2.5962 vocab=0.9787 token=1.7790 | 6.10 steps/s | VRAM=28.59 Go [2026-07-16 16:47:45] step 80600/1000000 | loss=2.5932 vocab=0.9803 token=1.9546 | 6.10 steps/s | VRAM=28.61 Go [2026-07-16 16:47:53] step 80650/1000000 | loss=2.5372 vocab=0.6953 token=1.2410 | 6.13 steps/s | VRAM=31.47 Go [2026-07-16 16:48:01] step 80700/1000000 | loss=2.4941 vocab=1.0834 token=1.5109 | 6.07 steps/s | VRAM=30.94 Go [2026-07-16 16:48:09] step 80750/1000000 | loss=2.5339 vocab=1.1086 token=1.2731 | 6.10 steps/s | VRAM=28.26 Go [2026-07-16 16:48:18] step 80800/1000000 | loss=2.6015 vocab=1.0100 token=1.6343 | 6.07 steps/s | VRAM=32.33 Go [2026-07-16 16:48:26] step 80850/1000000 | loss=2.5984 vocab=0.9272 token=1.0148 | 6.08 steps/s | VRAM=29.15 Go [2026-07-16 16:48:34] step 80900/1000000 | loss=2.5625 vocab=1.2404 token=1.6320 | 6.07 steps/s | VRAM=32.84 Go [2026-07-16 16:48:42] step 80950/1000000 | loss=2.5930 vocab=1.0390 token=1.4770 | 6.07 steps/s | VRAM=29.09 Go [2026-07-16 16:48:51] step 81000/1000000 | loss=2.5567 vocab=0.6197 token=1.3477 | 6.05 steps/s | VRAM=26.35 Go [2026-07-16 16:48:59] step 81050/1000000 | loss=2.5934 vocab=0.8368 token=1.6513 | 6.11 steps/s | VRAM=31.63 Go [2026-07-16 16:49:07] step 81100/1000000 | loss=2.7094 vocab=1.2761 token=1.6305 | 6.12 steps/s | VRAM=31.34 Go [2026-07-16 16:49:15] step 81150/1000000 | loss=2.6926 vocab=1.4741 token=1.9590 | 6.09 steps/s | VRAM=29.63 Go [2026-07-16 16:49:23] step 81200/1000000 | loss=2.6502 vocab=1.8246 token=1.8959 | 6.08 steps/s | VRAM=29.30 Go [2026-07-16 16:49:32] step 81250/1000000 | loss=2.6398 vocab=0.9084 token=1.5709 | 6.09 steps/s | VRAM=32.18 Go [2026-07-16 16:49:40] step 81300/1000000 | loss=2.4619 vocab=0.8567 token=1.5201 | 6.07 steps/s | VRAM=25.37 Go [2026-07-16 16:49:48] step 81350/1000000 | loss=2.6786 vocab=0.9613 token=1.5948 | 6.09 steps/s | VRAM=28.91 Go [2026-07-16 16:49:56] step 81400/1000000 | loss=2.5548 vocab=0.8426 token=1.2510 | 6.08 steps/s | VRAM=28.13 Go [2026-07-16 16:50:05] step 81450/1000000 | loss=2.7021 vocab=1.1767 token=1.7920 | 6.07 steps/s | VRAM=32.31 Go [2026-07-16 16:50:13] step 81500/1000000 | loss=2.6076 vocab=0.8695 token=1.0502 | 6.09 steps/s | VRAM=29.47 Go [2026-07-16 16:50:21] step 81550/1000000 | loss=2.6207 vocab=1.1933 token=1.6256 | 6.07 steps/s | VRAM=29.10 Go [2026-07-16 16:50:29] step 81600/1000000 | loss=2.5463 vocab=1.2893 token=1.5698 | 6.07 steps/s | VRAM=28.97 Go [2026-07-16 16:50:38] step 81650/1000000 | loss=2.5575 vocab=0.8963 token=1.5559 | 6.05 steps/s | VRAM=28.56 Go [2026-07-16 16:50:46] step 81700/1000000 | loss=2.5639 vocab=0.7630 token=1.3925 | 6.07 steps/s | VRAM=29.04 Go [2026-07-16 16:50:54] step 81750/1000000 | loss=2.5567 vocab=0.8625 token=1.1980 | 6.09 steps/s | VRAM=32.76 Go [2026-07-16 16:51:02] step 81800/1000000 | loss=2.5044 vocab=1.6968 token=1.6750 | 6.08 steps/s | VRAM=27.95 Go [2026-07-16 16:51:10] step 81850/1000000 | loss=2.5300 vocab=0.9873 token=1.3933 | 6.10 steps/s | VRAM=30.80 Go [2026-07-16 16:51:19] step 81900/1000000 | loss=2.5526 vocab=0.8145 token=1.5477 | 6.10 steps/s | VRAM=28.55 Go [2026-07-16 16:51:27] step 81950/1000000 | loss=2.5298 vocab=1.1541 token=1.6056 | 6.05 steps/s | VRAM=27.07 Go [2026-07-16 16:51:35] step 82000/1000000 | loss=2.5384 vocab=0.7088 token=1.4631 | 6.05 steps/s | VRAM=28.91 Go [2026-07-16 16:51:43] step 82050/1000000 | loss=2.5431 vocab=1.0586 token=1.2651 | 6.10 steps/s | VRAM=24.75 Go [2026-07-16 16:51:52] step 82100/1000000 | loss=2.5383 vocab=1.2728 token=1.5973 | 6.08 steps/s | VRAM=26.24 Go [2026-07-16 16:52:00] step 82150/1000000 | loss=2.5216 vocab=0.8732 token=1.7362 | 6.09 steps/s | VRAM=32.28 Go [2026-07-16 16:52:08] step 82200/1000000 | loss=2.6381 vocab=1.3338 token=1.4295 | 6.06 steps/s | VRAM=28.46 Go [2026-07-16 16:52:16] step 82250/1000000 | loss=2.4694 vocab=1.1897 token=1.6318 | 6.10 steps/s | VRAM=29.01 Go [2026-07-16 16:52:24] step 82300/1000000 | loss=2.6360 vocab=1.0345 token=1.4601 | 6.09 steps/s | VRAM=31.27 Go [2026-07-16 16:52:33] step 82350/1000000 | loss=2.7154 vocab=1.1688 token=1.6332 | 6.10 steps/s | VRAM=31.53 Go [2026-07-16 16:52:41] step 82400/1000000 | loss=2.5843 vocab=1.5214 token=1.3150 | 6.09 steps/s | VRAM=28.38 Go [2026-07-16 16:52:49] step 82450/1000000 | loss=2.4691 vocab=0.5471 token=1.5072 | 6.09 steps/s | VRAM=28.66 Go [2026-07-16 16:52:57] step 82500/1000000 | loss=2.6035 vocab=1.0478 token=1.5260 | 6.07 steps/s | VRAM=31.77 Go [2026-07-16 16:53:06] step 82550/1000000 | loss=2.5742 vocab=0.8635 token=1.3346 | 6.06 steps/s | VRAM=28.26 Go [2026-07-16 16:53:14] step 82600/1000000 | loss=2.6731 vocab=1.3300 token=1.3685 | 6.08 steps/s | VRAM=29.75 Go [2026-07-16 16:53:22] step 82650/1000000 | loss=2.6142 vocab=1.2327 token=1.2955 | 6.09 steps/s | VRAM=28.77 Go [2026-07-16 16:53:30] step 82700/1000000 | loss=2.5702 vocab=1.1032 token=1.7158 | 6.10 steps/s | VRAM=25.60 Go [2026-07-16 16:53:38] step 82750/1000000 | loss=2.6005 vocab=1.1426 token=1.4999 | 6.10 steps/s | VRAM=29.20 Go [2026-07-16 16:53:47] step 82800/1000000 | loss=2.4883 vocab=0.9654 token=1.5569 | 6.09 steps/s | VRAM=27.87 Go [2026-07-16 16:53:55] step 82850/1000000 | loss=2.4935 vocab=1.6544 token=1.8002 | 6.13 steps/s | VRAM=32.12 Go [2026-07-16 16:54:03] step 82900/1000000 | loss=2.5223 vocab=0.8463 token=1.5990 | 6.09 steps/s | VRAM=32.63 Go [2026-07-16 16:54:11] step 82950/1000000 | loss=2.6092 vocab=1.2949 token=1.7481 | 6.12 steps/s | VRAM=31.98 Go [2026-07-16 16:54:19] step 83000/1000000 | loss=2.5381 vocab=1.1264 token=1.7421 | 6.11 steps/s | VRAM=31.72 Go [2026-07-16 16:54:28] step 83050/1000000 | loss=2.5536 vocab=1.1713 token=1.5087 | 6.08 steps/s | VRAM=30.62 Go [2026-07-16 16:54:36] step 83100/1000000 | loss=2.3689 vocab=0.7201 token=1.3355 | 6.10 steps/s | VRAM=28.24 Go [2026-07-16 16:54:44] step 83150/1000000 | loss=2.5583 vocab=0.6868 token=1.2343 | 6.10 steps/s | VRAM=31.63 Go [2026-07-16 16:54:52] step 83200/1000000 | loss=2.5975 vocab=1.1446 token=1.4985 | 6.10 steps/s | VRAM=25.27 Go [2026-07-16 16:55:00] step 83250/1000000 | loss=2.5869 vocab=1.0420 token=1.3587 | 6.11 steps/s | VRAM=32.25 Go [2026-07-16 16:55:08] step 83300/1000000 | loss=2.5608 vocab=0.9693 token=1.6714 | 6.09 steps/s | VRAM=25.08 Go [2026-07-16 16:55:17] step 83350/1000000 | loss=2.5135 vocab=1.0021 token=1.4757 | 6.08 steps/s | VRAM=26.43 Go [2026-07-16 16:55:25] step 83400/1000000 | loss=2.5817 vocab=1.1072 token=1.5289 | 6.13 steps/s | VRAM=31.97 Go [2026-07-16 16:55:33] step 83450/1000000 | loss=2.5812 vocab=1.3511 token=1.9841 | 6.08 steps/s | VRAM=28.52 Go [2026-07-16 16:55:41] step 83500/1000000 | loss=2.5632 vocab=1.0278 token=1.8214 | 6.08 steps/s | VRAM=28.71 Go [2026-07-16 16:55:50] step 83550/1000000 | loss=2.5652 vocab=1.0392 token=1.3840 | 6.06 steps/s | VRAM=30.01 Go [2026-07-16 16:55:58] step 83600/1000000 | loss=2.5385 vocab=1.0819 token=1.5433 | 6.11 steps/s | VRAM=28.33 Go [2026-07-16 16:56:06] step 83650/1000000 | loss=2.5141 vocab=1.1521 token=1.5300 | 6.08 steps/s | VRAM=32.07 Go [2026-07-16 16:56:14] step 83700/1000000 | loss=2.5147 vocab=0.8800 token=1.5921 | 6.09 steps/s | VRAM=32.64 Go [2026-07-16 16:56:22] step 83750/1000000 | loss=2.6247 vocab=1.1045 token=1.4837 | 6.12 steps/s | VRAM=28.94 Go [2026-07-16 16:56:31] step 83800/1000000 | loss=2.5490 vocab=1.0707 token=1.5331 | 6.09 steps/s | VRAM=26.21 Go [2026-07-16 16:56:39] step 83850/1000000 | loss=2.6494 vocab=1.0796 token=1.4683 | 6.13 steps/s | VRAM=28.28 Go [2026-07-16 16:56:47] step 83900/1000000 | loss=2.5510 vocab=1.7118 token=1.3554 | 6.07 steps/s | VRAM=28.74 Go [2026-07-16 16:56:55] step 83950/1000000 | loss=2.6689 vocab=0.9290 token=1.2671 | 6.08 steps/s | VRAM=28.52 Go [2026-07-16 16:57:03] step 84000/1000000 | loss=2.6782 vocab=0.6829 token=1.4791 | 6.09 steps/s | VRAM=25.85 Go [2026-07-16 16:57:12] step 84050/1000000 | loss=2.6630 vocab=1.7196 token=1.6280 | 6.09 steps/s | VRAM=29.66 Go [2026-07-16 16:57:20] step 84100/1000000 | loss=2.6788 vocab=0.7138 token=1.5005 | 6.10 steps/s | VRAM=24.48 Go [2026-07-16 16:57:28] step 84150/1000000 | loss=2.5632 vocab=1.6775 token=1.8321 | 6.08 steps/s | VRAM=29.05 Go [2026-07-16 16:57:36] step 84200/1000000 | loss=2.4915 vocab=0.6904 token=1.6890 | 6.10 steps/s | VRAM=26.12 Go [2026-07-16 16:57:44] step 84250/1000000 | loss=2.5322 vocab=0.8338 token=1.2875 | 6.12 steps/s | VRAM=25.53 Go [2026-07-16 16:57:53] step 84300/1000000 | loss=2.6624 vocab=1.0961 token=1.7561 | 6.10 steps/s | VRAM=28.01 Go [2026-07-16 16:58:01] step 84350/1000000 | loss=2.5564 vocab=1.2456 token=1.6941 | 6.14 steps/s | VRAM=31.72 Go [2026-07-16 16:58:09] step 84400/1000000 | loss=2.6186 vocab=0.8583 token=1.6821 | 6.10 steps/s | VRAM=29.17 Go [2026-07-16 16:58:17] step 84450/1000000 | loss=2.4848 vocab=1.0616 token=1.6398 | 6.10 steps/s | VRAM=26.33 Go [2026-07-16 16:58:25] step 84500/1000000 | loss=2.6249 vocab=0.8983 token=1.5209 | 6.14 steps/s | VRAM=31.50 Go [2026-07-16 16:58:33] step 84550/1000000 | loss=2.5866 vocab=1.1432 token=1.6950 | 6.12 steps/s | VRAM=24.35 Go [2026-07-16 16:58:42] step 84600/1000000 | loss=2.5575 vocab=0.6152 token=1.3724 | 6.14 steps/s | VRAM=31.70 Go [2026-07-16 16:58:50] step 84650/1000000 | loss=2.5983 vocab=0.9852 token=1.6714 | 6.09 steps/s | VRAM=29.24 Go [2026-07-16 16:58:58] step 84700/1000000 | loss=2.4470 vocab=0.6251 token=1.4758 | 6.08 steps/s | VRAM=31.73 Go [2026-07-16 16:59:06] step 84750/1000000 | loss=2.5594 vocab=0.8811 token=1.3348 | 6.09 steps/s | VRAM=24.60 Go [2026-07-16 16:59:14] step 84800/1000000 | loss=2.4149 vocab=0.6032 token=1.3986 | 6.06 steps/s | VRAM=29.61 Go [2026-07-16 16:59:23] step 84850/1000000 | loss=2.5334 vocab=0.7829 token=1.2480 | 6.08 steps/s | VRAM=28.14 Go [2026-07-16 16:59:31] step 84900/1000000 | loss=2.6677 vocab=2.0237 token=1.4979 | 6.09 steps/s | VRAM=23.57 Go [2026-07-16 16:59:39] step 84950/1000000 | loss=2.4973 vocab=0.8482 token=1.3128 | 6.11 steps/s | VRAM=31.84 Go [2026-07-16 16:59:47] step 85000/1000000 | loss=2.5450 vocab=0.6000 token=1.0089 | 6.09 steps/s | VRAM=29.58 Go [2026-07-16 17:01:10] step 85050/1000000 | loss=2.6928 vocab=1.0404 token=1.1299 | 0.61 steps/s | VRAM=29.74 Go [2026-07-16 17:01:18] step 85100/1000000 | loss=2.5438 vocab=1.9055 token=1.9255 | 6.21 steps/s | VRAM=31.58 Go [2026-07-16 17:01:26] step 85150/1000000 | loss=2.4208 vocab=1.0898 token=1.4607 | 6.16 steps/s | VRAM=28.84 Go [2026-07-16 17:01:34] step 85200/1000000 | loss=2.6660 vocab=1.2073 token=1.7071 | 6.17 steps/s | VRAM=31.05 Go [2026-07-16 17:01:42] step 85250/1000000 | loss=2.5515 vocab=1.2226 token=1.5996 | 6.13 steps/s | VRAM=28.29 Go [2026-07-16 17:01:50] step 85300/1000000 | loss=2.5735 vocab=1.2585 token=1.4520 | 6.13 steps/s | VRAM=33.00 Go [2026-07-16 17:01:59] step 85350/1000000 | loss=2.5021 vocab=1.1213 token=1.8985 | 6.17 steps/s | VRAM=31.74 Go [2026-07-16 17:02:07] step 85400/1000000 | loss=2.6623 vocab=1.1316 token=1.3812 | 6.13 steps/s | VRAM=31.68 Go [2026-07-16 17:02:15] step 85450/1000000 | loss=2.5409 vocab=0.7269 token=1.6071 | 6.12 steps/s | VRAM=28.66 Go [2026-07-16 17:02:23] step 85500/1000000 | loss=2.6146 vocab=1.0648 token=1.6574 | 6.11 steps/s | VRAM=31.92 Go [2026-07-16 17:02:31] step 85550/1000000 | loss=2.5444 vocab=1.0535 token=1.4635 | 6.13 steps/s | VRAM=31.50 Go [2026-07-16 17:02:39] step 85600/1000000 | loss=2.5631 vocab=0.6483 token=1.2329 | 6.09 steps/s | VRAM=26.19 Go [2026-07-16 17:02:48] step 85650/1000000 | loss=2.4971 vocab=0.9723 token=1.4304 | 6.08 steps/s | VRAM=29.25 Go [2026-07-16 17:02:56] step 85700/1000000 | loss=2.4462 vocab=0.8940 token=1.4127 | 6.09 steps/s | VRAM=26.17 Go [2026-07-16 17:03:04] step 85750/1000000 | loss=2.4857 vocab=1.0376 token=1.7847 | 6.09 steps/s | VRAM=27.34 Go [2026-07-16 17:03:12] step 85800/1000000 | loss=2.6179 vocab=0.8076 token=1.3816 | 6.10 steps/s | VRAM=29.10 Go [2026-07-16 17:03:21] step 85850/1000000 | loss=2.5553 vocab=0.8990 token=1.5292 | 6.09 steps/s | VRAM=28.77 Go [2026-07-16 17:03:29] step 85900/1000000 | loss=2.6410 vocab=1.2553 token=1.4739 | 6.08 steps/s | VRAM=28.92 Go [2026-07-16 17:03:37] step 85950/1000000 | loss=2.6115 vocab=1.1248 token=1.4829 | 6.08 steps/s | VRAM=29.31 Go [2026-07-16 17:03:45] step 86000/1000000 | loss=2.4499 vocab=0.7078 token=1.6838 | 6.10 steps/s | VRAM=28.24 Go [2026-07-16 17:03:53] step 86050/1000000 | loss=2.5031 vocab=1.4014 token=1.1659 | 6.05 steps/s | VRAM=26.83 Go [2026-07-16 17:04:02] step 86100/1000000 | loss=2.5260 vocab=0.6149 token=1.3786 | 6.11 steps/s | VRAM=29.31 Go [2026-07-16 17:04:10] step 86150/1000000 | loss=2.6205 vocab=1.0489 token=1.8203 | 6.07 steps/s | VRAM=28.97 Go [2026-07-16 17:04:18] step 86200/1000000 | loss=2.5089 vocab=0.8303 token=1.4987 | 6.10 steps/s | VRAM=32.78 Go [2026-07-16 17:04:26] step 86250/1000000 | loss=2.5290 vocab=0.8220 token=1.4431 | 6.06 steps/s | VRAM=28.15 Go [2026-07-16 17:04:35] step 86300/1000000 | loss=2.5471 vocab=1.0779 token=1.3065 | 6.08 steps/s | VRAM=29.32 Go [2026-07-16 17:04:43] step 86350/1000000 | loss=2.5987 vocab=1.4085 token=1.6249 | 6.10 steps/s | VRAM=32.42 Go [2026-07-16 17:04:51] step 86400/1000000 | loss=2.5635 vocab=1.2321 token=1.4890 | 6.07 steps/s | VRAM=29.35 Go [2026-07-16 17:04:59] step 86450/1000000 | loss=2.5357 vocab=0.9959 token=1.6552 | 6.08 steps/s | VRAM=31.99 Go [2026-07-16 17:05:07] step 86500/1000000 | loss=2.5383 vocab=1.3505 token=1.5863 | 6.09 steps/s | VRAM=29.14 Go [2026-07-16 17:05:16] step 86550/1000000 | loss=2.5158 vocab=1.0074 token=1.6849 | 6.11 steps/s | VRAM=31.94 Go [2026-07-16 17:05:24] step 86600/1000000 | loss=2.5193 vocab=1.0882 token=1.3591 | 6.08 steps/s | VRAM=28.73 Go [2026-07-16 17:05:32] step 86650/1000000 | loss=2.4452 vocab=1.4636 token=1.7029 | 6.09 steps/s | VRAM=29.28 Go [2026-07-16 17:05:40] step 86700/1000000 | loss=2.5748 vocab=0.7970 token=1.4286 | 6.10 steps/s | VRAM=32.23 Go [2026-07-16 17:05:48] step 86750/1000000 | loss=2.5694 vocab=0.8176 token=1.6065 | 6.07 steps/s | VRAM=28.45 Go [2026-07-16 17:05:57] step 86800/1000000 | loss=2.6056 vocab=0.7388 token=1.3169 | 6.08 steps/s | VRAM=25.70 Go [2026-07-16 17:06:05] step 86850/1000000 | loss=2.5044 vocab=1.2693 token=1.3404 | 6.09 steps/s | VRAM=28.40 Go [2026-07-16 17:06:13] step 86900/1000000 | loss=2.5122 vocab=0.9513 token=1.3851 | 6.08 steps/s | VRAM=29.18 Go [2026-07-16 17:06:21] step 86950/1000000 | loss=2.4989 vocab=0.7566 token=1.2758 | 6.07 steps/s | VRAM=26.22 Go [2026-07-16 17:06:30] step 87000/1000000 | loss=2.4823 vocab=1.3089 token=1.5892 | 6.06 steps/s | VRAM=30.07 Go [2026-07-16 17:06:38] step 87050/1000000 | loss=2.6398 vocab=0.8953 token=1.8434 | 6.09 steps/s | VRAM=31.65 Go [2026-07-16 17:06:46] step 87100/1000000 | loss=2.6220 vocab=1.4580 token=1.5719 | 6.08 steps/s | VRAM=31.42 Go [2026-07-16 17:06:54] step 87150/1000000 | loss=2.5048 vocab=0.7757 token=1.3679 | 6.09 steps/s | VRAM=25.76 Go [2026-07-16 17:07:02] step 87200/1000000 | loss=2.4988 vocab=1.2304 token=1.2334 | 6.06 steps/s | VRAM=29.52 Go [2026-07-16 17:07:11] step 87250/1000000 | loss=2.4997 vocab=1.1689 token=1.4465 | 6.10 steps/s | VRAM=27.89 Go [2026-07-16 17:07:19] step 87300/1000000 | loss=2.5218 vocab=0.9313 token=1.7242 | 6.06 steps/s | VRAM=27.64 Go [2026-07-16 17:07:27] step 87350/1000000 | loss=2.5595 vocab=0.8779 token=1.5846 | 6.07 steps/s | VRAM=29.13 Go [2026-07-16 17:07:35] step 87400/1000000 | loss=2.6131 vocab=1.0825 token=1.9436 | 6.06 steps/s | VRAM=29.67 Go [2026-07-16 17:07:44] step 87450/1000000 | loss=2.5633 vocab=0.7758 token=1.3578 | 6.07 steps/s | VRAM=32.42 Go [2026-07-16 17:07:52] step 87500/1000000 | loss=2.5497 vocab=1.2059 token=1.5215 | 6.05 steps/s | VRAM=26.33 Go [2026-07-16 17:08:00] step 87550/1000000 | loss=2.5465 vocab=0.7346 token=1.3027 | 6.12 steps/s | VRAM=32.07 Go [2026-07-16 17:08:08] step 87600/1000000 | loss=2.6026 vocab=1.2697 token=1.2679 | 6.05 steps/s | VRAM=26.64 Go [2026-07-16 17:08:17] step 87650/1000000 | loss=2.5566 vocab=0.6450 token=1.7076 | 6.10 steps/s | VRAM=29.53 Go [2026-07-16 17:08:25] step 87700/1000000 | loss=2.5704 vocab=0.8431 token=1.4085 | 6.08 steps/s | VRAM=28.79 Go [2026-07-16 17:08:33] step 87750/1000000 | loss=2.6175 vocab=0.7328 token=1.6146 | 6.11 steps/s | VRAM=25.35 Go [2026-07-16 17:08:41] step 87800/1000000 | loss=2.6368 vocab=0.7951 token=1.7289 | 6.06 steps/s | VRAM=29.46 Go [2026-07-16 17:08:49] step 87850/1000000 | loss=2.5385 vocab=1.1120 token=1.5972 | 6.08 steps/s | VRAM=29.40 Go [2026-07-16 17:08:58] step 87900/1000000 | loss=2.5206 vocab=0.8251 token=1.3154 | 6.11 steps/s | VRAM=28.54 Go [2026-07-16 17:09:06] step 87950/1000000 | loss=2.5252 vocab=0.9706 token=1.3314 | 6.10 steps/s | VRAM=32.68 Go [2026-07-16 17:09:14] step 88000/1000000 | loss=2.6346 vocab=1.1197 token=1.9846 | 6.09 steps/s | VRAM=32.32 Go [2026-07-16 17:09:22] step 88050/1000000 | loss=2.5590 vocab=1.2350 token=1.7412 | 6.05 steps/s | VRAM=29.18 Go [2026-07-16 17:09:30] step 88100/1000000 | loss=2.4967 vocab=0.9584 token=1.6514 | 6.09 steps/s | VRAM=26.55 Go [2026-07-16 17:09:39] step 88150/1000000 | loss=2.4701 vocab=1.1236 token=1.3160 | 6.09 steps/s | VRAM=25.22 Go [2026-07-16 17:09:47] step 88200/1000000 | loss=2.4656 vocab=1.1738 token=1.6342 | 6.10 steps/s | VRAM=25.23 Go [2026-07-16 17:09:55] step 88250/1000000 | loss=2.5488 vocab=0.9072 token=1.6772 | 6.07 steps/s | VRAM=30.01 Go [2026-07-16 17:10:03] step 88300/1000000 | loss=2.6028 vocab=0.7514 token=1.0559 | 6.10 steps/s | VRAM=27.71 Go [2026-07-16 17:10:12] step 88350/1000000 | loss=2.5253 vocab=0.6826 token=1.4576 | 6.05 steps/s | VRAM=28.89 Go [2026-07-16 17:10:20] step 88400/1000000 | loss=2.4809 vocab=0.7415 token=1.8549 | 6.11 steps/s | VRAM=32.22 Go [2026-07-16 17:10:28] step 88450/1000000 | loss=2.4750 vocab=1.1614 token=1.9327 | 6.09 steps/s | VRAM=30.98 Go [2026-07-16 17:10:36] step 88500/1000000 | loss=2.6033 vocab=0.8349 token=1.3088 | 6.07 steps/s | VRAM=32.35 Go [2026-07-16 17:10:44] step 88550/1000000 | loss=2.5357 vocab=0.6811 token=1.4936 | 6.10 steps/s | VRAM=26.01 Go [2026-07-16 17:10:53] step 88600/1000000 | loss=2.5415 vocab=1.0815 token=1.4628 | 6.06 steps/s | VRAM=28.87 Go [2026-07-16 17:11:01] step 88650/1000000 | loss=2.6106 vocab=0.9951 token=1.5279 | 6.10 steps/s | VRAM=32.05 Go [2026-07-16 17:11:09] step 88700/1000000 | loss=2.4504 vocab=0.7043 token=1.4210 | 6.10 steps/s | VRAM=29.62 Go [2026-07-16 17:11:17] step 88750/1000000 | loss=2.4640 vocab=0.9468 token=1.7568 | 6.09 steps/s | VRAM=26.87 Go [2026-07-16 17:11:25] step 88800/1000000 | loss=2.5616 vocab=0.7712 token=1.3972 | 6.11 steps/s | VRAM=32.05 Go [2026-07-16 17:11:34] step 88850/1000000 | loss=2.5373 vocab=0.4230 token=1.3848 | 6.10 steps/s | VRAM=28.41 Go [2026-07-16 17:11:42] step 88900/1000000 | loss=2.5051 vocab=1.3041 token=1.6867 | 6.09 steps/s | VRAM=29.31 Go [2026-07-16 17:11:50] step 88950/1000000 | loss=2.4961 vocab=0.9536 token=1.6990 | 6.07 steps/s | VRAM=29.13 Go [2026-07-16 17:11:58] step 89000/1000000 | loss=2.5362 vocab=0.5765 token=1.2616 | 6.09 steps/s | VRAM=29.07 Go [2026-07-16 17:12:06] step 89050/1000000 | loss=2.6939 vocab=1.3202 token=1.5850 | 6.11 steps/s | VRAM=29.38 Go [2026-07-16 17:12:15] step 89100/1000000 | loss=2.6245 vocab=0.8904 token=1.5499 | 6.10 steps/s | VRAM=28.37 Go [2026-07-16 17:12:23] step 89150/1000000 | loss=2.6550 vocab=1.0706 token=1.7666 | 6.07 steps/s | VRAM=26.91 Go [2026-07-16 17:12:31] step 89200/1000000 | loss=2.6177 vocab=0.6758 token=1.6452 | 6.11 steps/s | VRAM=28.05 Go [2026-07-16 17:12:39] step 89250/1000000 | loss=2.4438 vocab=0.6191 token=1.3794 | 6.12 steps/s | VRAM=31.91 Go [2026-07-16 17:12:47] step 89300/1000000 | loss=2.5688 vocab=0.8368 token=1.3110 | 6.09 steps/s | VRAM=28.65 Go [2026-07-16 17:12:56] step 89350/1000000 | loss=2.5102 vocab=1.2277 token=1.4179 | 6.11 steps/s | VRAM=32.35 Go [2026-07-16 17:13:04] step 89400/1000000 | loss=2.5827 vocab=1.0002 token=1.4107 | 6.07 steps/s | VRAM=29.06 Go [2026-07-16 17:13:12] step 89450/1000000 | loss=2.5769 vocab=0.9308 token=1.6308 | 6.08 steps/s | VRAM=28.62 Go [2026-07-16 17:13:20] step 89500/1000000 | loss=2.5680 vocab=1.0929 token=1.5989 | 6.07 steps/s | VRAM=28.88 Go [2026-07-16 17:13:29] step 89550/1000000 | loss=2.6445 vocab=0.9647 token=1.4950 | 6.08 steps/s | VRAM=27.42 Go [2026-07-16 17:13:37] step 89600/1000000 | loss=2.5120 vocab=1.2773 token=1.2580 | 6.08 steps/s | VRAM=29.97 Go [2026-07-16 17:13:45] step 89650/1000000 | loss=2.5844 vocab=1.5202 token=1.3743 | 6.08 steps/s | VRAM=32.33 Go [2026-07-16 17:13:53] step 89700/1000000 | loss=2.5211 vocab=0.8180 token=1.1484 | 6.08 steps/s | VRAM=33.02 Go [2026-07-16 17:14:01] step 89750/1000000 | loss=2.6006 vocab=0.8851 token=1.3494 | 6.07 steps/s | VRAM=28.79 Go [2026-07-16 17:14:10] step 89800/1000000 | loss=2.5241 vocab=1.1295 token=1.4946 | 6.11 steps/s | VRAM=31.90 Go [2026-07-16 17:14:18] step 89850/1000000 | loss=2.4919 vocab=0.7421 token=1.3744 | 6.08 steps/s | VRAM=28.77 Go [2026-07-16 17:14:26] step 89900/1000000 | loss=2.5915 vocab=0.8974 token=1.8613 | 6.08 steps/s | VRAM=28.35 Go [2026-07-16 17:14:34] step 89950/1000000 | loss=2.4521 vocab=1.4005 token=1.9146 | 6.06 steps/s | VRAM=29.57 Go [2026-07-16 17:14:43] step 90000/1000000 | loss=2.5820 vocab=0.9180 token=1.7107 | 6.12 steps/s | VRAM=28.59 Go [2026-07-16 17:16:12] step 90050/1000000 | loss=2.5999 vocab=1.2600 token=2.0613 | 0.56 steps/s | VRAM=29.00 Go [2026-07-16 17:16:20] step 90100/1000000 | loss=2.5782 vocab=1.1614 token=1.3043 | 6.19 steps/s | VRAM=30.89 Go [2026-07-16 17:16:29] step 90150/1000000 | loss=2.5452 vocab=1.2573 token=1.9864 | 6.13 steps/s | VRAM=26.83 Go [2026-07-16 17:16:37] step 90200/1000000 | loss=2.5275 vocab=1.0053 token=1.6973 | 6.14 steps/s | VRAM=28.53 Go [2026-07-16 17:16:45] step 90250/1000000 | loss=2.4412 vocab=1.2558 token=1.5241 | 6.12 steps/s | VRAM=28.50 Go [2026-07-16 17:16:53] step 90300/1000000 | loss=2.5498 vocab=1.1034 token=1.5250 | 6.11 steps/s | VRAM=28.56 Go [2026-07-16 17:17:01] step 90350/1000000 | loss=2.5085 vocab=0.9330 token=1.4829 | 6.10 steps/s | VRAM=26.26 Go [2026-07-16 17:17:09] step 90400/1000000 | loss=2.4872 vocab=1.6529 token=1.3013 | 6.11 steps/s | VRAM=32.77 Go [2026-07-16 17:17:18] step 90450/1000000 | loss=2.5795 vocab=0.9838 token=1.8110 | 6.11 steps/s | VRAM=31.41 Go [2026-07-16 17:17:26] step 90500/1000000 | loss=2.5492 vocab=1.3248 token=1.7073 | 6.12 steps/s | VRAM=28.38 Go [2026-07-16 17:17:34] step 90550/1000000 | loss=2.5482 vocab=1.0333 token=1.3685 | 6.10 steps/s | VRAM=29.24 Go [2026-07-16 17:17:42] step 90600/1000000 | loss=2.5915 vocab=1.2404 token=1.0810 | 6.09 steps/s | VRAM=28.08 Go [2026-07-16 17:17:50] step 90650/1000000 | loss=2.4475 vocab=1.6941 token=1.4022 | 6.09 steps/s | VRAM=27.72 Go [2026-07-16 17:17:59] step 90700/1000000 | loss=2.4872 vocab=0.6854 token=1.2786 | 6.09 steps/s | VRAM=32.80 Go [2026-07-16 17:18:07] step 90750/1000000 | loss=2.5677 vocab=1.3455 token=1.8768 | 6.11 steps/s | VRAM=32.50 Go [2026-07-16 17:18:15] step 90800/1000000 | loss=2.4820 vocab=0.7940 token=1.8178 | 6.12 steps/s | VRAM=32.18 Go [2026-07-16 17:18:23] step 90850/1000000 | loss=2.5326 vocab=0.8160 token=1.3372 | 6.09 steps/s | VRAM=29.01 Go [2026-07-16 17:18:31] step 90900/1000000 | loss=2.5123 vocab=0.6810 token=1.7425 | 6.11 steps/s | VRAM=28.98 Go [2026-07-16 17:18:40] step 90950/1000000 | loss=2.5483 vocab=0.8518 token=1.2583 | 6.11 steps/s | VRAM=29.04 Go [2026-07-16 17:18:48] step 91000/1000000 | loss=2.5396 vocab=1.2018 token=1.6651 | 6.06 steps/s | VRAM=27.10 Go [2026-07-16 17:18:56] step 91050/1000000 | loss=2.5558 vocab=1.0936 token=1.9438 | 6.07 steps/s | VRAM=28.42 Go [2026-07-16 17:19:04] step 91100/1000000 | loss=2.6473 vocab=1.0609 token=1.7400 | 6.04 steps/s | VRAM=29.14 Go [2026-07-16 17:19:13] step 91150/1000000 | loss=2.4471 vocab=0.7588 token=1.4344 | 6.08 steps/s | VRAM=28.18 Go [2026-07-16 17:19:21] step 91200/1000000 | loss=2.5462 vocab=1.4060 token=1.5702 | 6.08 steps/s | VRAM=29.14 Go [2026-07-16 17:19:29] step 91250/1000000 | loss=2.5792 vocab=0.7724 token=1.9089 | 6.11 steps/s | VRAM=31.75 Go [2026-07-16 17:19:37] step 91300/1000000 | loss=2.5137 vocab=0.9744 token=1.5287 | 6.05 steps/s | VRAM=25.88 Go [2026-07-16 17:19:45] step 91350/1000000 | loss=2.6082 vocab=0.9541 token=1.7444 | 6.09 steps/s | VRAM=25.87 Go [2026-07-16 17:19:54] step 91400/1000000 | loss=2.5500 vocab=1.7362 token=1.7763 | 6.11 steps/s | VRAM=28.46 Go [2026-07-16 17:20:02] step 91450/1000000 | loss=2.5134 vocab=0.7925 token=1.6136 | 6.07 steps/s | VRAM=27.72 Go [2026-07-16 17:20:10] step 91500/1000000 | loss=2.5330 vocab=1.2406 token=1.7753 | 6.07 steps/s | VRAM=26.90 Go [2026-07-16 17:20:18] step 91550/1000000 | loss=2.4845 vocab=0.9012 token=1.4208 | 6.10 steps/s | VRAM=30.05 Go [2026-07-16 17:20:26] step 91600/1000000 | loss=2.4810 vocab=0.8113 token=1.1701 | 6.08 steps/s | VRAM=25.26 Go [2026-07-16 17:20:35] step 91650/1000000 | loss=2.5369 vocab=1.4608 token=1.7095 | 6.09 steps/s | VRAM=28.08 Go [2026-07-16 17:20:43] step 91700/1000000 | loss=2.5101 vocab=0.9031 token=1.3351 | 6.10 steps/s | VRAM=27.33 Go [2026-07-16 17:20:51] step 91750/1000000 | loss=2.5846 vocab=1.2546 token=1.7698 | 6.08 steps/s | VRAM=27.37 Go [2026-07-16 17:20:59] step 91800/1000000 | loss=2.4880 vocab=1.2935 token=1.3704 | 6.11 steps/s | VRAM=32.26 Go [2026-07-16 17:21:08] step 91850/1000000 | loss=2.6300 vocab=0.8747 token=1.3695 | 6.09 steps/s | VRAM=26.17 Go [2026-07-16 17:21:16] step 91900/1000000 | loss=2.4866 vocab=0.7462 token=1.3355 | 6.09 steps/s | VRAM=31.99 Go [2026-07-16 17:21:24] step 91950/1000000 | loss=2.6114 vocab=0.7775 token=1.2734 | 6.05 steps/s | VRAM=26.22 Go [2026-07-16 17:21:32] step 92000/1000000 | loss=2.4977 vocab=0.7182 token=1.3644 | 6.06 steps/s | VRAM=31.94 Go [2026-07-16 17:21:40] step 92050/1000000 | loss=2.5687 vocab=0.7852 token=1.4896 | 6.06 steps/s | VRAM=29.70 Go [2026-07-16 17:21:49] step 92100/1000000 | loss=2.5508 vocab=1.0588 token=1.4233 | 6.07 steps/s | VRAM=28.00 Go [2026-07-16 17:21:57] step 92150/1000000 | loss=2.4966 vocab=1.1046 token=1.2662 | 6.10 steps/s | VRAM=28.36 Go [2026-07-16 17:22:05] step 92200/1000000 | loss=2.4617 vocab=1.1065 token=1.6055 | 6.09 steps/s | VRAM=26.32 Go [2026-07-16 17:22:13] step 92250/1000000 | loss=2.6091 vocab=1.0922 token=1.6834 | 6.09 steps/s | VRAM=26.43 Go [2026-07-16 17:22:22] step 92300/1000000 | loss=2.4939 vocab=1.2140 token=1.8201 | 6.09 steps/s | VRAM=26.20 Go [2026-07-16 17:22:30] step 92350/1000000 | loss=2.5892 vocab=1.1900 token=1.7739 | 6.10 steps/s | VRAM=26.05 Go [2026-07-16 17:22:38] step 92400/1000000 | loss=2.5524 vocab=1.4860 token=1.8881 | 6.09 steps/s | VRAM=28.58 Go [2026-07-16 17:22:46] step 92450/1000000 | loss=2.5234 vocab=0.8068 token=1.2819 | 6.10 steps/s | VRAM=28.12 Go [2026-07-16 17:22:54] step 92500/1000000 | loss=2.5319 vocab=1.0022 token=1.5491 | 6.13 steps/s | VRAM=28.33 Go [2026-07-16 17:23:03] step 92550/1000000 | loss=2.5871 vocab=1.0350 token=1.5308 | 6.09 steps/s | VRAM=28.71 Go [2026-07-16 17:23:11] step 92600/1000000 | loss=2.5221 vocab=0.6559 token=1.6915 | 6.11 steps/s | VRAM=31.61 Go [2026-07-16 17:23:19] step 92650/1000000 | loss=2.4403 vocab=0.9491 token=1.3452 | 6.09 steps/s | VRAM=29.33 Go [2026-07-16 17:23:27] step 92700/1000000 | loss=2.5594 vocab=1.3037 token=1.6538 | 6.11 steps/s | VRAM=32.50 Go [2026-07-16 17:23:35] step 92750/1000000 | loss=2.5436 vocab=1.1287 token=1.4412 | 6.09 steps/s | VRAM=26.77 Go [2026-07-16 17:23:43] step 92800/1000000 | loss=2.5051 vocab=1.4121 token=1.7598 | 6.13 steps/s | VRAM=31.76 Go [2026-07-16 17:23:52] step 92850/1000000 | loss=2.5708 vocab=0.6687 token=1.6741 | 6.11 steps/s | VRAM=26.83 Go [2026-07-16 17:24:00] step 92900/1000000 | loss=2.6047 vocab=0.9647 token=1.6387 | 6.12 steps/s | VRAM=28.03 Go [2026-07-16 17:24:08] step 92950/1000000 | loss=2.5902 vocab=0.6126 token=1.4319 | 6.08 steps/s | VRAM=28.09 Go [2026-07-16 17:24:16] step 93000/1000000 | loss=2.5392 vocab=0.9825 token=1.5655 | 6.12 steps/s | VRAM=29.39 Go [2026-07-16 17:24:24] step 93050/1000000 | loss=2.6602 vocab=0.9368 token=1.4854 | 6.07 steps/s | VRAM=26.06 Go [2026-07-16 17:24:33] step 93100/1000000 | loss=2.6024 vocab=1.3807 token=1.8128 | 6.05 steps/s | VRAM=27.12 Go [2026-07-16 17:24:41] step 93150/1000000 | loss=2.4715 vocab=1.0412 token=1.4274 | 6.07 steps/s | VRAM=28.90 Go [2026-07-16 17:24:49] step 93200/1000000 | loss=2.4899 vocab=0.8186 token=1.5682 | 6.09 steps/s | VRAM=28.85 Go [2026-07-16 17:24:57] step 93250/1000000 | loss=2.5357 vocab=0.9818 token=1.2864 | 6.11 steps/s | VRAM=28.65 Go [2026-07-16 17:25:06] step 93300/1000000 | loss=2.5671 vocab=1.1279 token=1.5572 | 6.09 steps/s | VRAM=30.51 Go [2026-07-16 17:25:14] step 93350/1000000 | loss=2.5252 vocab=1.4196 token=1.6042 | 6.09 steps/s | VRAM=24.63 Go [2026-07-16 17:25:22] step 93400/1000000 | loss=2.4739 vocab=1.3487 token=1.5752 | 6.12 steps/s | VRAM=27.83 Go [2026-07-16 17:25:30] step 93450/1000000 | loss=2.6349 vocab=1.6896 token=1.6011 | 6.09 steps/s | VRAM=24.06 Go [2026-07-16 17:25:38] step 93500/1000000 | loss=2.7276 vocab=1.6835 token=1.8253 | 6.08 steps/s | VRAM=32.95 Go [2026-07-16 17:25:47] step 93550/1000000 | loss=2.5901 vocab=1.4817 token=1.6090 | 6.11 steps/s | VRAM=31.92 Go [2026-07-16 17:25:55] step 93600/1000000 | loss=2.5397 vocab=0.8694 token=1.2225 | 6.07 steps/s | VRAM=26.13 Go [2026-07-16 17:26:03] step 93650/1000000 | loss=2.6400 vocab=1.1581 token=1.5131 | 6.09 steps/s | VRAM=32.39 Go [2026-07-16 17:26:11] step 93700/1000000 | loss=2.5438 vocab=1.2756 token=1.6582 | 6.10 steps/s | VRAM=32.45 Go [2026-07-16 17:26:19] step 93750/1000000 | loss=2.5901 vocab=0.8754 token=1.3423 | 6.14 steps/s | VRAM=31.23 Go [2026-07-16 17:26:28] step 93800/1000000 | loss=2.5985 vocab=0.9709 token=1.4217 | 6.10 steps/s | VRAM=25.28 Go [2026-07-16 17:26:36] step 93850/1000000 | loss=2.4900 vocab=0.8881 token=1.5338 | 6.08 steps/s | VRAM=28.33 Go [2026-07-16 17:26:44] step 93900/1000000 | loss=2.5716 vocab=0.8588 token=1.5309 | 6.08 steps/s | VRAM=28.77 Go [2026-07-16 17:26:52] step 93950/1000000 | loss=2.6584 vocab=1.3196 token=1.6374 | 6.12 steps/s | VRAM=32.09 Go [2026-07-16 17:27:00] step 94000/1000000 | loss=2.5120 vocab=0.5870 token=1.7950 | 6.08 steps/s | VRAM=29.86 Go [2026-07-16 17:27:09] step 94050/1000000 | loss=2.6217 vocab=0.8349 token=1.6115 | 6.09 steps/s | VRAM=32.96 Go [2026-07-16 17:27:17] step 94100/1000000 | loss=2.6945 vocab=1.1649 token=1.5423 | 6.11 steps/s | VRAM=31.69 Go [2026-07-16 17:27:25] step 94150/1000000 | loss=2.5207 vocab=1.0358 token=1.5723 | 6.10 steps/s | VRAM=32.48 Go [2026-07-16 17:27:33] step 94200/1000000 | loss=2.5419 vocab=0.9725 token=1.5941 | 6.06 steps/s | VRAM=26.10 Go [2026-07-16 17:27:41] step 94250/1000000 | loss=2.5277 vocab=1.0672 token=1.8053 | 6.11 steps/s | VRAM=24.73 Go [2026-07-16 17:27:50] step 94300/1000000 | loss=2.5657 vocab=0.7453 token=1.2663 | 6.08 steps/s | VRAM=26.00 Go [2026-07-16 17:27:58] step 94350/1000000 | loss=2.5160 vocab=0.7545 token=1.3215 | 6.10 steps/s | VRAM=28.41 Go [2026-07-16 17:28:06] step 94400/1000000 | loss=2.5597 vocab=0.9774 token=1.7392 | 6.11 steps/s | VRAM=28.24 Go [2026-07-16 17:28:14] step 94450/1000000 | loss=2.5147 vocab=1.8003 token=2.0556 | 6.12 steps/s | VRAM=27.96 Go [2026-07-16 17:28:22] step 94500/1000000 | loss=2.4259 vocab=1.1830 token=1.3295 | 6.08 steps/s | VRAM=29.61 Go [2026-07-16 17:28:31] step 94550/1000000 | loss=2.5058 vocab=1.1579 token=1.3756 | 6.11 steps/s | VRAM=32.69 Go [2026-07-16 17:28:39] step 94600/1000000 | loss=2.5644 vocab=1.2773 token=1.6240 | 6.03 steps/s | VRAM=26.65 Go [2026-07-16 17:28:47] step 94650/1000000 | loss=2.5757 vocab=0.8371 token=1.4626 | 6.14 steps/s | VRAM=28.30 Go [2026-07-16 17:28:55] step 94700/1000000 | loss=2.5965 vocab=0.7460 token=1.1994 | 6.09 steps/s | VRAM=28.31 Go [2026-07-16 17:29:03] step 94750/1000000 | loss=2.3936 vocab=0.4771 token=1.2505 | 6.11 steps/s | VRAM=26.15 Go [2026-07-16 17:29:12] step 94800/1000000 | loss=2.4996 vocab=0.6204 token=1.4218 | 6.10 steps/s | VRAM=32.18 Go [2026-07-16 17:29:20] step 94850/1000000 | loss=2.6004 vocab=0.8313 token=1.5850 | 6.07 steps/s | VRAM=27.84 Go [2026-07-16 17:29:28] step 94900/1000000 | loss=2.4403 vocab=0.9910 token=1.2387 | 6.10 steps/s | VRAM=31.83 Go [2026-07-16 17:29:36] step 94950/1000000 | loss=2.6112 vocab=1.0611 token=1.7706 | 6.11 steps/s | VRAM=26.47 Go [2026-07-16 17:29:44] step 95000/1000000 | loss=2.5317 vocab=1.3431 token=1.5817 | 6.12 steps/s | VRAM=32.00 Go [2026-07-16 17:31:08] step 95050/1000000 | loss=2.5337 vocab=0.7158 token=1.2991 | 0.60 steps/s | VRAM=28.56 Go [2026-07-16 17:31:16] step 95100/1000000 | loss=2.4344 vocab=0.8379 token=1.4042 | 6.15 steps/s | VRAM=30.83 Go [2026-07-16 17:31:24] step 95150/1000000 | loss=2.4556 vocab=1.2986 token=1.5355 | 6.18 steps/s | VRAM=32.14 Go [2026-07-16 17:31:32] step 95200/1000000 | loss=2.5884 vocab=0.8418 token=1.5250 | 6.17 steps/s | VRAM=24.62 Go [2026-07-16 17:31:40] step 95250/1000000 | loss=2.5148 vocab=1.0322 token=1.4254 | 6.14 steps/s | VRAM=25.97 Go [2026-07-16 17:31:48] step 95300/1000000 | loss=2.6009 vocab=1.4711 token=2.1551 | 6.12 steps/s | VRAM=29.19 Go [2026-07-16 17:31:57] step 95350/1000000 | loss=2.6272 vocab=1.2833 token=1.7867 | 6.12 steps/s | VRAM=25.15 Go [2026-07-16 17:32:05] step 95400/1000000 | loss=2.6184 vocab=0.7181 token=1.5993 | 6.15 steps/s | VRAM=31.40 Go [2026-07-16 17:32:13] step 95450/1000000 | loss=2.5713 vocab=1.1719 token=1.4267 | 6.10 steps/s | VRAM=30.14 Go [2026-07-16 17:32:21] step 95500/1000000 | loss=2.5327 vocab=0.7053 token=1.7438 | 6.11 steps/s | VRAM=28.66 Go [2026-07-16 17:32:29] step 95550/1000000 | loss=2.5180 vocab=0.6373 token=1.3930 | 6.07 steps/s | VRAM=26.31 Go [2026-07-16 17:32:37] step 95600/1000000 | loss=2.5875 vocab=0.9974 token=1.4444 | 6.11 steps/s | VRAM=27.84 Go [2026-07-16 17:32:46] step 95650/1000000 | loss=2.5061 vocab=0.6157 token=1.4830 | 6.09 steps/s | VRAM=28.86 Go [2026-07-16 17:32:54] step 95700/1000000 | loss=2.5525 vocab=1.0567 token=1.3229 | 6.10 steps/s | VRAM=29.04 Go [2026-07-16 17:33:02] step 95750/1000000 | loss=2.5278 vocab=0.7559 token=1.3718 | 6.12 steps/s | VRAM=28.06 Go [2026-07-16 17:33:10] step 95800/1000000 | loss=2.5117 vocab=1.3601 token=1.6686 | 6.09 steps/s | VRAM=29.14 Go [2026-07-16 17:33:18] step 95850/1000000 | loss=2.6432 vocab=0.7949 token=1.2139 | 6.10 steps/s | VRAM=29.15 Go [2026-07-16 17:33:27] step 95900/1000000 | loss=2.4793 vocab=0.9446 token=1.3184 | 6.10 steps/s | VRAM=28.66 Go [2026-07-16 17:33:35] step 95950/1000000 | loss=2.6752 vocab=1.4968 token=1.5056 | 6.06 steps/s | VRAM=29.50 Go [2026-07-16 17:33:43] step 96000/1000000 | loss=2.4919 vocab=0.8813 token=1.5886 | 6.08 steps/s | VRAM=25.09 Go [2026-07-16 17:33:51] step 96050/1000000 | loss=2.5219 vocab=0.7648 token=1.3223 | 6.12 steps/s | VRAM=31.88 Go [2026-07-16 17:34:00] step 96100/1000000 | loss=2.5835 vocab=1.0365 token=1.4138 | 6.05 steps/s | VRAM=27.83 Go [2026-07-16 17:34:08] step 96150/1000000 | loss=2.4957 vocab=1.2823 token=2.0162 | 6.08 steps/s | VRAM=27.41 Go [2026-07-16 17:34:16] step 96200/1000000 | loss=2.5276 vocab=0.9314 token=1.9090 | 6.10 steps/s | VRAM=28.80 Go [2026-07-16 17:34:24] step 96250/1000000 | loss=2.5830 vocab=0.8508 token=1.2871 | 6.08 steps/s | VRAM=26.22 Go [2026-07-16 17:34:32] step 96300/1000000 | loss=2.4805 vocab=0.5578 token=1.5683 | 6.08 steps/s | VRAM=30.64 Go [2026-07-16 17:34:41] step 96350/1000000 | loss=2.5324 vocab=0.8398 token=1.4625 | 6.09 steps/s | VRAM=28.23 Go [2026-07-16 17:34:49] step 96400/1000000 | loss=2.5804 vocab=1.3273 token=1.4640 | 6.11 steps/s | VRAM=28.88 Go [2026-07-16 17:34:57] step 96450/1000000 | loss=2.5125 vocab=0.3885 token=1.4388 | 6.07 steps/s | VRAM=26.35 Go [2026-07-16 17:35:05] step 96500/1000000 | loss=2.5315 vocab=1.5238 token=1.9643 | 6.07 steps/s | VRAM=28.60 Go [2026-07-16 17:35:13] step 96550/1000000 | loss=2.4510 vocab=0.5944 token=1.4664 | 6.09 steps/s | VRAM=29.57 Go [2026-07-16 17:35:22] step 96600/1000000 | loss=2.5555 vocab=0.9961 token=1.7180 | 6.09 steps/s | VRAM=32.69 Go [2026-07-16 17:35:30] step 96650/1000000 | loss=2.6155 vocab=0.8523 token=1.4764 | 6.11 steps/s | VRAM=30.23 Go [2026-07-16 17:35:38] step 96700/1000000 | loss=2.5657 vocab=1.1860 token=1.5321 | 6.07 steps/s | VRAM=27.00 Go [2026-07-16 17:35:46] step 96750/1000000 | loss=2.5056 vocab=0.9120 token=1.5100 | 6.12 steps/s | VRAM=31.55 Go [2026-07-16 17:35:55] step 96800/1000000 | loss=2.5295 vocab=0.9121 token=1.9060 | 6.05 steps/s | VRAM=32.51 Go [2026-07-16 17:36:03] step 96850/1000000 | loss=2.4659 vocab=0.9316 token=1.5100 | 6.12 steps/s | VRAM=28.85 Go [2026-07-16 17:36:11] step 96900/1000000 | loss=2.5746 vocab=1.1015 token=1.5717 | 6.09 steps/s | VRAM=28.35 Go [2026-07-16 17:36:19] step 96950/1000000 | loss=2.4665 vocab=0.5278 token=1.3380 | 6.08 steps/s | VRAM=26.26 Go [2026-07-16 17:36:27] step 97000/1000000 | loss=2.5094 vocab=0.6086 token=1.4598 | 6.11 steps/s | VRAM=24.74 Go [2026-07-16 17:36:36] step 97050/1000000 | loss=2.5274 vocab=1.0570 token=1.6988 | 6.10 steps/s | VRAM=25.21 Go [2026-07-16 17:36:44] step 97100/1000000 | loss=2.5138 vocab=1.1550 token=1.4620 | 6.12 steps/s | VRAM=31.50 Go [2026-07-16 17:36:52] step 97150/1000000 | loss=2.5284 vocab=1.3267 token=1.5315 | 6.10 steps/s | VRAM=28.29 Go [2026-07-16 17:37:00] step 97200/1000000 | loss=2.4665 vocab=1.0318 token=1.7023 | 6.06 steps/s | VRAM=28.70 Go [2026-07-16 17:37:08] step 97250/1000000 | loss=2.6264 vocab=0.9431 token=1.2182 | 6.08 steps/s | VRAM=31.54 Go [2026-07-16 17:37:17] step 97300/1000000 | loss=2.5095 vocab=1.4181 token=1.9349 | 6.11 steps/s | VRAM=31.37 Go [2026-07-16 17:37:25] step 97350/1000000 | loss=2.4843 vocab=0.7416 token=1.5737 | 6.07 steps/s | VRAM=26.20 Go [2026-07-16 17:37:33] step 97400/1000000 | loss=2.5481 vocab=1.0693 token=1.4739 | 6.08 steps/s | VRAM=28.33 Go [2026-07-16 17:37:41] step 97450/1000000 | loss=2.4603 vocab=1.1013 token=1.4537 | 6.09 steps/s | VRAM=27.01 Go [2026-07-16 17:37:49] step 97500/1000000 | loss=2.5616 vocab=0.7820 token=1.2639 | 6.10 steps/s | VRAM=25.00 Go [2026-07-16 17:37:58] step 97550/1000000 | loss=2.5259 vocab=1.1318 token=1.6419 | 6.08 steps/s | VRAM=28.31 Go [2026-07-16 17:38:06] step 97600/1000000 | loss=2.6003 vocab=1.0343 token=1.5803 | 6.09 steps/s | VRAM=28.29 Go [2026-07-16 17:38:14] step 97650/1000000 | loss=2.5189 vocab=0.9340 token=1.4046 | 6.07 steps/s | VRAM=28.04 Go [2026-07-16 17:38:22] step 97700/1000000 | loss=2.5805 vocab=0.6046 token=1.3597 | 6.11 steps/s | VRAM=32.91 Go [2026-07-16 17:38:31] step 97750/1000000 | loss=2.5812 vocab=0.6685 token=1.1784 | 6.06 steps/s | VRAM=29.64 Go [2026-07-16 17:38:39] step 97800/1000000 | loss=2.5586 vocab=1.0092 token=1.3813 | 6.10 steps/s | VRAM=29.49 Go [2026-07-16 17:38:47] step 97850/1000000 | loss=2.5738 vocab=1.1295 token=1.8146 | 6.11 steps/s | VRAM=29.66 Go [2026-07-16 17:38:55] step 97900/1000000 | loss=2.5648 vocab=0.7078 token=1.4640 | 6.11 steps/s | VRAM=31.72 Go [2026-07-16 17:39:03] step 97950/1000000 | loss=2.5634 vocab=1.1779 token=1.5414 | 6.09 steps/s | VRAM=24.79 Go [2026-07-16 17:39:12] step 98000/1000000 | loss=2.5820 vocab=1.5494 token=1.6141 | 6.06 steps/s | VRAM=26.71 Go [2026-07-16 17:39:20] step 98050/1000000 | loss=2.5356 vocab=1.5354 token=1.4187 | 6.11 steps/s | VRAM=28.78 Go [2026-07-16 17:39:28] step 98100/1000000 | loss=2.5854 vocab=1.1423 token=1.5511 | 6.11 steps/s | VRAM=31.02 Go [2026-07-16 17:39:36] step 98150/1000000 | loss=2.4546 vocab=0.8207 token=1.1990 | 6.09 steps/s | VRAM=32.34 Go [2026-07-16 17:39:44] step 98200/1000000 | loss=2.4513 vocab=0.8359 token=1.3738 | 6.08 steps/s | VRAM=28.88 Go [2026-07-16 17:39:53] step 98250/1000000 | loss=2.5515 vocab=1.1816 token=1.5369 | 6.06 steps/s | VRAM=24.96 Go [2026-07-16 17:40:01] step 98300/1000000 | loss=2.6078 vocab=0.7050 token=1.3837 | 6.06 steps/s | VRAM=26.70 Go [2026-07-16 17:40:09] step 98350/1000000 | loss=2.5117 vocab=1.7907 token=1.6398 | 6.11 steps/s | VRAM=29.02 Go [2026-07-16 17:40:17] step 98400/1000000 | loss=2.4746 vocab=1.0727 token=1.6320 | 6.10 steps/s | VRAM=25.76 Go [2026-07-16 17:40:25] step 98450/1000000 | loss=2.4944 vocab=0.3646 token=1.2134 | 6.11 steps/s | VRAM=31.30 Go [2026-07-16 17:40:34] step 98500/1000000 | loss=2.4899 vocab=1.3464 token=1.5483 | 6.09 steps/s | VRAM=32.47 Go [2026-07-16 17:40:42] step 98550/1000000 | loss=2.4196 vocab=0.9760 token=1.5520 | 6.09 steps/s | VRAM=29.19 Go [2026-07-16 17:40:50] step 98600/1000000 | loss=2.4607 vocab=1.3725 token=2.0777 | 6.10 steps/s | VRAM=25.85 Go [2026-07-16 17:40:58] step 98650/1000000 | loss=2.5103 vocab=1.1970 token=1.7133 | 6.10 steps/s | VRAM=31.65 Go [2026-07-16 17:41:07] step 98700/1000000 | loss=2.5818 vocab=1.1789 token=1.7571 | 6.04 steps/s | VRAM=24.10 Go [2026-07-16 17:41:15] step 98750/1000000 | loss=2.5124 vocab=0.8386 token=1.5101 | 6.08 steps/s | VRAM=28.77 Go [2026-07-16 17:41:23] step 98800/1000000 | loss=2.6203 vocab=0.4940 token=1.1503 | 6.09 steps/s | VRAM=28.89 Go [2026-07-16 17:41:31] step 98850/1000000 | loss=2.4886 vocab=1.1268 token=1.2090 | 6.10 steps/s | VRAM=28.59 Go [2026-07-16 17:41:39] step 98900/1000000 | loss=2.5300 vocab=1.1114 token=1.4720 | 6.07 steps/s | VRAM=32.69 Go [2026-07-16 17:41:48] step 98950/1000000 | loss=2.6220 vocab=1.4693 token=1.6615 | 6.09 steps/s | VRAM=28.35 Go [2026-07-16 17:41:56] step 99000/1000000 | loss=2.4369 vocab=1.0392 token=1.4421 | 6.08 steps/s | VRAM=25.67 Go [2026-07-16 17:42:04] step 99050/1000000 | loss=2.5974 vocab=0.8093 token=1.4192 | 6.12 steps/s | VRAM=29.46 Go [2026-07-16 17:42:12] step 99100/1000000 | loss=2.4348 vocab=0.9544 token=1.3452 | 6.11 steps/s | VRAM=28.66 Go [2026-07-16 17:42:20] step 99150/1000000 | loss=2.5185 vocab=1.2199 token=1.8131 | 6.06 steps/s | VRAM=26.54 Go [2026-07-16 17:42:29] step 99200/1000000 | loss=2.4577 vocab=0.7800 token=1.5143 | 6.10 steps/s | VRAM=32.97 Go [2026-07-16 17:42:37] step 99250/1000000 | loss=2.4535 vocab=1.0745 token=1.4588 | 6.12 steps/s | VRAM=29.15 Go [2026-07-16 17:42:45] step 99300/1000000 | loss=2.5313 vocab=0.8953 token=1.3091 | 6.11 steps/s | VRAM=27.79 Go [2026-07-16 17:42:53] step 99350/1000000 | loss=2.5234 vocab=0.7117 token=1.3692 | 6.11 steps/s | VRAM=32.21 Go [2026-07-16 17:43:01] step 99400/1000000 | loss=2.5194 vocab=0.8664 token=1.4527 | 6.12 steps/s | VRAM=31.18 Go [2026-07-16 17:43:10] step 99450/1000000 | loss=2.5597 vocab=1.0976 token=1.3759 | 6.12 steps/s | VRAM=28.75 Go [2026-07-16 17:43:18] step 99500/1000000 | loss=2.5344 vocab=1.1509 token=1.5293 | 6.09 steps/s | VRAM=28.75 Go [2026-07-16 17:43:26] step 99550/1000000 | loss=2.5746 vocab=1.0049 token=1.4635 | 6.11 steps/s | VRAM=26.21 Go [2026-07-16 17:43:34] step 99600/1000000 | loss=2.3986 vocab=0.9965 token=1.6012 | 6.13 steps/s | VRAM=30.70 Go [2026-07-16 17:43:42] step 99650/1000000 | loss=2.5517 vocab=0.8654 token=1.8841 | 6.06 steps/s | VRAM=30.35 Go [2026-07-16 17:43:51] step 99700/1000000 | loss=2.5202 vocab=1.1768 token=1.4172 | 6.08 steps/s | VRAM=27.91 Go [2026-07-16 17:43:59] step 99750/1000000 | loss=2.4573 vocab=1.2584 token=1.5179 | 6.13 steps/s | VRAM=32.16 Go [2026-07-16 17:44:07] step 99800/1000000 | loss=2.4801 vocab=0.8159 token=1.1606 | 6.09 steps/s | VRAM=29.70 Go [2026-07-16 17:44:15] step 99850/1000000 | loss=2.4744 vocab=1.4269 token=1.5688 | 6.11 steps/s | VRAM=27.73 Go [2026-07-16 17:44:23] step 99900/1000000 | loss=2.4886 vocab=0.9404 token=1.2995 | 6.12 steps/s | VRAM=32.81 Go [2026-07-16 17:44:31] step 99950/1000000 | loss=2.4580 vocab=1.2184 token=1.7072 | 6.12 steps/s | VRAM=30.77 Go [2026-07-16 17:44:40] step 100000/1000000 | loss=2.4410 vocab=1.0698 token=1.4470 | 6.09 steps/s | VRAM=31.96 Go [2026-07-16 17:45:59] step 100050/1000000 | loss=2.4592 vocab=0.9154 token=1.0960 | 0.63 steps/s | VRAM=28.82 Go [2026-07-16 17:46:07] step 100100/1000000 | loss=2.5942 vocab=1.0969 token=1.6727 | 6.13 steps/s | VRAM=32.51 Go [2026-07-16 17:46:15] step 100150/1000000 | loss=2.5838 vocab=1.3749 token=1.5583 | 6.17 steps/s | VRAM=32.75 Go [2026-07-16 17:46:24] step 100200/1000000 | loss=2.4201 vocab=0.7555 token=1.5891 | 6.16 steps/s | VRAM=31.44 Go [2026-07-16 17:46:32] step 100250/1000000 | loss=2.4954 vocab=0.7797 token=1.6210 | 6.17 steps/s | VRAM=32.52 Go [2026-07-16 17:46:40] step 100300/1000000 | loss=2.5351 vocab=0.7245 token=1.4501 | 6.14 steps/s | VRAM=27.84 Go [2026-07-16 17:46:48] step 100350/1000000 | loss=2.4223 vocab=0.9875 token=1.4221 | 6.14 steps/s | VRAM=31.73 Go [2026-07-16 17:46:56] step 100400/1000000 | loss=2.4354 vocab=1.3269 token=1.3468 | 6.13 steps/s | VRAM=31.87 Go [2026-07-16 17:47:04] step 100450/1000000 | loss=2.4627 vocab=1.1111 token=1.4185 | 6.16 steps/s | VRAM=31.78 Go [2026-07-16 17:47:12] step 100500/1000000 | loss=2.4513 vocab=1.3810 token=1.8216 | 6.09 steps/s | VRAM=26.50 Go [2026-07-16 17:47:21] step 100550/1000000 | loss=2.4376 vocab=0.8592 token=1.4366 | 6.13 steps/s | VRAM=25.38 Go [2026-07-16 17:47:29] step 100600/1000000 | loss=2.5480 vocab=1.1978 token=1.8222 | 6.12 steps/s | VRAM=28.43 Go [2026-07-16 17:47:37] step 100650/1000000 | loss=2.4808 vocab=1.1264 token=1.2818 | 6.08 steps/s | VRAM=25.98 Go [2026-07-16 17:47:45] step 100700/1000000 | loss=2.4957 vocab=0.9197 token=1.1016 | 6.09 steps/s | VRAM=31.51 Go [2026-07-16 17:47:53] step 100750/1000000 | loss=2.5501 vocab=1.1896 token=1.6352 | 6.10 steps/s | VRAM=28.83 Go [2026-07-16 17:48:02] step 100800/1000000 | loss=2.5372 vocab=0.8644 token=1.2431 | 6.10 steps/s | VRAM=29.11 Go [2026-07-16 17:48:10] step 100850/1000000 | loss=2.5208 vocab=0.8020 token=1.3963 | 6.09 steps/s | VRAM=32.46 Go [2026-07-16 17:48:18] step 100900/1000000 | loss=2.5375 vocab=0.9364 token=1.4977 | 6.12 steps/s | VRAM=29.01 Go [2026-07-16 17:48:26] step 100950/1000000 | loss=2.4706 vocab=0.9106 token=1.4991 | 6.06 steps/s | VRAM=28.70 Go [2026-07-16 17:48:34] step 101000/1000000 | loss=2.5327 vocab=0.8558 token=1.6546 | 6.10 steps/s | VRAM=28.61 Go [2026-07-16 17:48:43] step 101050/1000000 | loss=2.5791 vocab=1.8276 token=1.7945 | 6.06 steps/s | VRAM=29.27 Go [2026-07-16 17:48:51] step 101100/1000000 | loss=2.6516 vocab=1.5941 token=1.7054 | 6.08 steps/s | VRAM=28.63 Go [2026-07-16 17:48:59] step 101150/1000000 | loss=2.5953 vocab=0.9595 token=1.2359 | 6.07 steps/s | VRAM=28.89 Go [2026-07-16 17:49:07] step 101200/1000000 | loss=2.5305 vocab=0.9214 token=1.3940 | 6.10 steps/s | VRAM=28.74 Go [2026-07-16 17:49:16] step 101250/1000000 | loss=2.4347 vocab=0.6202 token=1.4061 | 6.06 steps/s | VRAM=28.58 Go [2026-07-16 17:49:24] step 101300/1000000 | loss=2.6275 vocab=1.4759 token=1.8547 | 6.11 steps/s | VRAM=32.30 Go [2026-07-16 17:49:32] step 101350/1000000 | loss=2.5560 vocab=0.7558 token=1.4355 | 6.12 steps/s | VRAM=31.27 Go [2026-07-16 17:49:40] step 101400/1000000 | loss=2.5989 vocab=0.9943 token=1.5482 | 6.09 steps/s | VRAM=32.31 Go [2026-07-16 17:49:48] step 101450/1000000 | loss=2.5008 vocab=1.0798 token=1.9352 | 6.10 steps/s | VRAM=28.49 Go [2026-07-16 17:49:57] step 101500/1000000 | loss=2.6462 vocab=1.1635 token=1.5042 | 6.07 steps/s | VRAM=29.31 Go [2026-07-16 17:50:05] step 101550/1000000 | loss=2.6029 vocab=1.1195 token=1.4538 | 6.07 steps/s | VRAM=29.37 Go [2026-07-16 17:50:13] step 101600/1000000 | loss=2.5240 vocab=1.0910 token=1.3394 | 6.11 steps/s | VRAM=31.91 Go [2026-07-16 17:50:21] step 101650/1000000 | loss=2.5457 vocab=1.1654 token=1.8533 | 6.11 steps/s | VRAM=24.57 Go [2026-07-16 17:50:29] step 101700/1000000 | loss=2.5336 vocab=0.7719 token=0.9426 | 6.09 steps/s | VRAM=32.40 Go [2026-07-16 17:50:38] step 101750/1000000 | loss=2.4455 vocab=1.0882 token=1.6069 | 6.10 steps/s | VRAM=29.64 Go [2026-07-16 17:50:46] step 101800/1000000 | loss=2.5033 vocab=0.8013 token=1.3625 | 6.09 steps/s | VRAM=32.84 Go [2026-07-16 17:50:54] step 101850/1000000 | loss=2.5330 vocab=0.8140 token=1.6120 | 6.09 steps/s | VRAM=27.16 Go [2026-07-16 17:51:02] step 101900/1000000 | loss=2.3749 vocab=0.8439 token=1.2957 | 6.11 steps/s | VRAM=24.68 Go [2026-07-16 17:51:10] step 101950/1000000 | loss=2.5378 vocab=1.2057 token=1.6994 | 6.11 steps/s | VRAM=28.18 Go [2026-07-16 17:51:19] step 102000/1000000 | loss=2.5448 vocab=1.1638 token=1.3965 | 6.07 steps/s | VRAM=28.87 Go [2026-07-16 17:51:27] step 102050/1000000 | loss=2.4396 vocab=0.9138 token=1.4983 | 6.11 steps/s | VRAM=32.24 Go [2026-07-16 17:51:35] step 102100/1000000 | loss=2.6042 vocab=1.6948 token=1.7589 | 6.09 steps/s | VRAM=26.40 Go [2026-07-16 17:51:43] step 102150/1000000 | loss=2.5028 vocab=1.2103 token=1.6006 | 6.10 steps/s | VRAM=28.64 Go [2026-07-16 17:51:51] step 102200/1000000 | loss=2.5553 vocab=1.0779 token=1.3556 | 6.10 steps/s | VRAM=26.79 Go [2026-07-16 17:52:00] step 102250/1000000 | loss=2.5280 vocab=0.9735 token=1.7036 | 6.09 steps/s | VRAM=26.33 Go [2026-07-16 17:52:08] step 102300/1000000 | loss=2.4647 vocab=1.4284 token=1.5329 | 6.09 steps/s | VRAM=29.15 Go [2026-07-16 17:52:16] step 102350/1000000 | loss=2.4756 vocab=1.4086 token=0.9590 | 6.12 steps/s | VRAM=31.90 Go [2026-07-16 17:52:24] step 102400/1000000 | loss=2.5182 vocab=0.7885 token=1.2907 | 6.10 steps/s | VRAM=28.22 Go [2026-07-16 17:52:32] step 102450/1000000 | loss=2.5714 vocab=0.7457 token=1.3392 | 6.11 steps/s | VRAM=31.60 Go [2026-07-16 17:52:41] step 102500/1000000 | loss=2.4076 vocab=0.8437 token=1.5621 | 6.09 steps/s | VRAM=26.32 Go [2026-07-16 17:52:49] step 102550/1000000 | loss=2.7035 vocab=1.3454 token=1.3955 | 6.10 steps/s | VRAM=28.34 Go [2026-07-16 17:52:57] step 102600/1000000 | loss=2.6108 vocab=0.8976 token=1.5168 | 6.11 steps/s | VRAM=29.30 Go [2026-07-16 17:53:05] step 102650/1000000 | loss=2.5941 vocab=1.4471 token=1.5266 | 6.11 steps/s | VRAM=32.18 Go [2026-07-16 17:53:13] step 102700/1000000 | loss=2.5186 vocab=2.4555 token=1.2490 | 6.07 steps/s | VRAM=23.37 Go [2026-07-16 17:53:22] step 102750/1000000 | loss=2.5836 vocab=0.8010 token=1.7361 | 6.13 steps/s | VRAM=28.82 Go [2026-07-16 17:53:30] step 102800/1000000 | loss=2.5233 vocab=0.9976 token=1.7181 | 6.08 steps/s | VRAM=29.72 Go [2026-07-16 17:53:38] step 102850/1000000 | loss=2.5297 vocab=1.4036 token=1.4926 | 6.10 steps/s | VRAM=28.26 Go [2026-07-16 17:53:46] step 102900/1000000 | loss=2.5530 vocab=0.9051 token=1.2779 | 6.10 steps/s | VRAM=31.68 Go [2026-07-16 17:53:54] step 102950/1000000 | loss=2.5323 vocab=0.8419 token=1.5040 | 6.10 steps/s | VRAM=31.38 Go [2026-07-16 17:54:03] step 103000/1000000 | loss=2.5232 vocab=0.8198 token=1.4741 | 6.09 steps/s | VRAM=31.47 Go [2026-07-16 17:54:11] step 103050/1000000 | loss=2.5122 vocab=0.8963 token=1.1764 | 6.10 steps/s | VRAM=29.90 Go [2026-07-16 17:54:19] step 103100/1000000 | loss=2.5175 vocab=0.6067 token=1.4725 | 6.08 steps/s | VRAM=32.99 Go [2026-07-16 17:54:27] step 103150/1000000 | loss=2.5832 vocab=1.2777 token=1.3884 | 6.10 steps/s | VRAM=31.70 Go [2026-07-16 17:54:35] step 103200/1000000 | loss=2.5399 vocab=1.0390 token=1.5265 | 6.10 steps/s | VRAM=28.51 Go [2026-07-16 17:54:44] step 103250/1000000 | loss=2.4345 vocab=1.0844 token=1.1672 | 6.09 steps/s | VRAM=32.81 Go [2026-07-16 17:54:52] step 103300/1000000 | loss=2.5740 vocab=1.2429 token=1.4838 | 6.07 steps/s | VRAM=28.58 Go [2026-07-16 17:55:00] step 103350/1000000 | loss=2.5618 vocab=1.1902 token=1.8975 | 6.08 steps/s | VRAM=26.16 Go [2026-07-16 17:55:08] step 103400/1000000 | loss=2.5643 vocab=1.1273 token=1.5128 | 6.13 steps/s | VRAM=32.34 Go [2026-07-16 17:55:16] step 103450/1000000 | loss=2.4756 vocab=0.9303 token=1.4866 | 6.13 steps/s | VRAM=31.35 Go [2026-07-16 17:55:25] step 103500/1000000 | loss=2.5688 vocab=0.8849 token=1.4277 | 6.09 steps/s | VRAM=26.21 Go [2026-07-16 17:55:33] step 103550/1000000 | loss=2.5926 vocab=1.0144 token=1.3300 | 6.10 steps/s | VRAM=28.60 Go [2026-07-16 17:55:41] step 103600/1000000 | loss=2.5783 vocab=1.1350 token=1.7877 | 6.12 steps/s | VRAM=28.63 Go [2026-07-16 17:55:49] step 103650/1000000 | loss=2.5278 vocab=0.9183 token=1.6203 | 6.09 steps/s | VRAM=29.55 Go [2026-07-16 17:55:57] step 103700/1000000 | loss=2.5724 vocab=0.7223 token=1.4539 | 6.08 steps/s | VRAM=26.32 Go [2026-07-16 17:56:06] step 103750/1000000 | loss=2.5537 vocab=1.1343 token=1.3239 | 6.10 steps/s | VRAM=32.27 Go [2026-07-16 17:56:14] step 103800/1000000 | loss=2.4951 vocab=1.0775 token=1.5460 | 6.09 steps/s | VRAM=28.62 Go [2026-07-16 17:56:22] step 103850/1000000 | loss=2.4585 vocab=1.0049 token=1.3106 | 6.10 steps/s | VRAM=24.82 Go [2026-07-16 17:56:30] step 103900/1000000 | loss=2.5436 vocab=1.5810 token=1.4284 | 6.07 steps/s | VRAM=32.15 Go [2026-07-16 17:56:38] step 103950/1000000 | loss=2.4570 vocab=1.2986 token=1.1751 | 6.11 steps/s | VRAM=25.79 Go [2026-07-16 17:56:47] step 104000/1000000 | loss=2.5582 vocab=1.1431 token=1.3541 | 6.08 steps/s | VRAM=30.23 Go [2026-07-16 17:56:55] step 104050/1000000 | loss=2.4004 vocab=0.8029 token=1.2321 | 6.10 steps/s | VRAM=28.08 Go [2026-07-16 17:57:03] step 104100/1000000 | loss=2.5399 vocab=1.0825 token=1.3594 | 6.11 steps/s | VRAM=28.37 Go [2026-07-16 17:57:11] step 104150/1000000 | loss=2.5708 vocab=0.8672 token=1.7251 | 6.07 steps/s | VRAM=25.77 Go [2026-07-16 17:57:20] step 104200/1000000 | loss=2.4848 vocab=0.7908 token=1.3911 | 6.08 steps/s | VRAM=28.87 Go [2026-07-16 17:57:28] step 104250/1000000 | loss=2.5023 vocab=1.1458 token=1.4508 | 6.09 steps/s | VRAM=27.12 Go [2026-07-16 17:57:36] step 104300/1000000 | loss=2.5018 vocab=1.0802 token=1.3898 | 6.10 steps/s | VRAM=32.58 Go [2026-07-16 17:57:44] step 104350/1000000 | loss=2.5183 vocab=1.2480 token=1.6425 | 6.09 steps/s | VRAM=30.26 Go [2026-07-16 17:57:52] step 104400/1000000 | loss=2.4313 vocab=0.5429 token=1.3275 | 6.09 steps/s | VRAM=27.76 Go [2026-07-16 17:58:01] step 104450/1000000 | loss=2.5039 vocab=0.7224 token=1.4634 | 6.10 steps/s | VRAM=30.94 Go [2026-07-16 17:58:09] step 104500/1000000 | loss=2.5260 vocab=0.7134 token=1.2180 | 6.08 steps/s | VRAM=26.60 Go [2026-07-16 17:58:17] step 104550/1000000 | loss=2.3861 vocab=0.7987 token=1.6212 | 6.09 steps/s | VRAM=30.43 Go [2026-07-16 17:58:25] step 104600/1000000 | loss=2.5109 vocab=0.7105 token=1.2265 | 6.10 steps/s | VRAM=32.85 Go [2026-07-16 17:58:33] step 104650/1000000 | loss=2.5534 vocab=0.9086 token=1.2107 | 6.10 steps/s | VRAM=28.01 Go [2026-07-16 17:58:42] step 104700/1000000 | loss=2.5207 vocab=1.6208 token=1.5186 | 6.08 steps/s | VRAM=32.13 Go [2026-07-16 17:58:50] step 104750/1000000 | loss=2.6138 vocab=0.7895 token=1.2940 | 6.12 steps/s | VRAM=31.07 Go [2026-07-16 17:58:58] step 104800/1000000 | loss=2.5306 vocab=0.6294 token=1.3417 | 6.09 steps/s | VRAM=29.71 Go [2026-07-16 17:59:06] step 104850/1000000 | loss=2.5020 vocab=0.9018 token=1.3424 | 6.06 steps/s | VRAM=26.21 Go [2026-07-16 17:59:14] step 104900/1000000 | loss=2.4233 vocab=0.9716 token=1.4958 | 6.12 steps/s | VRAM=27.84 Go [2026-07-16 17:59:23] step 104950/1000000 | loss=2.4190 vocab=0.9915 token=1.4566 | 6.08 steps/s | VRAM=25.50 Go [2026-07-16 17:59:31] step 105000/1000000 | loss=2.5480 vocab=0.8372 token=1.4629 | 6.08 steps/s | VRAM=29.57 Go [2026-07-16 18:00:51] step 105050/1000000 | loss=2.6032 vocab=1.1450 token=1.7419 | 0.62 steps/s | VRAM=27.99 Go [2026-07-16 18:00:59] step 105100/1000000 | loss=2.5627 vocab=0.9654 token=1.3755 | 6.15 steps/s | VRAM=29.17 Go [2026-07-16 18:01:08] step 105150/1000000 | loss=2.4741 vocab=1.3003 token=1.6360 | 6.17 steps/s | VRAM=28.80 Go [2026-07-16 18:01:16] step 105200/1000000 | loss=2.4716 vocab=1.4634 token=1.4273 | 6.17 steps/s | VRAM=24.74 Go [2026-07-16 18:01:24] step 105250/1000000 | loss=2.5096 vocab=0.8749 token=1.5512 | 6.12 steps/s | VRAM=32.55 Go [2026-07-16 18:01:32] step 105300/1000000 | loss=2.5482 vocab=0.8227 token=1.5345 | 6.16 steps/s | VRAM=32.49 Go [2026-07-16 18:01:40] step 105350/1000000 | loss=2.5537 vocab=1.1653 token=1.7524 | 6.13 steps/s | VRAM=28.37 Go [2026-07-16 18:01:48] step 105400/1000000 | loss=2.5045 vocab=1.4817 token=1.9936 | 6.12 steps/s | VRAM=26.50 Go [2026-07-16 18:01:56] step 105450/1000000 | loss=2.5622 vocab=0.8581 token=1.7095 | 6.15 steps/s | VRAM=29.36 Go [2026-07-16 18:02:05] step 105500/1000000 | loss=2.5692 vocab=0.9948 token=1.3433 | 6.11 steps/s | VRAM=23.50 Go [2026-07-16 18:02:13] step 105550/1000000 | loss=2.6458 vocab=0.9495 token=1.6187 | 6.10 steps/s | VRAM=29.65 Go [2026-07-16 18:02:21] step 105600/1000000 | loss=2.4203 vocab=1.1128 token=1.4354 | 6.09 steps/s | VRAM=28.12 Go [2026-07-16 18:02:29] step 105650/1000000 | loss=2.6232 vocab=1.3366 token=1.3549 | 6.08 steps/s | VRAM=29.53 Go [2026-07-16 18:02:37] step 105700/1000000 | loss=2.4566 vocab=1.4154 token=1.5879 | 6.09 steps/s | VRAM=29.10 Go [2026-07-16 18:02:46] step 105750/1000000 | loss=2.5336 vocab=1.2246 token=1.7101 | 6.07 steps/s | VRAM=26.30 Go [2026-07-16 18:02:54] step 105800/1000000 | loss=2.5280 vocab=1.3244 token=1.6302 | 6.08 steps/s | VRAM=30.08 Go [2026-07-16 18:03:02] step 105850/1000000 | loss=2.5547 vocab=1.2134 token=1.4758 | 6.12 steps/s | VRAM=32.40 Go [2026-07-16 18:03:10] step 105900/1000000 | loss=2.5623 vocab=0.9835 token=1.4535 | 6.12 steps/s | VRAM=29.44 Go [2026-07-16 18:03:18] step 105950/1000000 | loss=2.5061 vocab=1.3161 token=1.7511 | 6.09 steps/s | VRAM=31.73 Go [2026-07-16 18:03:27] step 106000/1000000 | loss=2.4566 vocab=0.7460 token=1.4294 | 6.11 steps/s | VRAM=27.71 Go [2026-07-16 18:03:35] step 106050/1000000 | loss=2.6126 vocab=1.1722 token=1.4129 | 6.12 steps/s | VRAM=30.99 Go [2026-07-16 18:03:43] step 106100/1000000 | loss=2.5240 vocab=1.0640 token=1.7546 | 6.07 steps/s | VRAM=26.29 Go [2026-07-16 18:03:51] step 106150/1000000 | loss=2.3648 vocab=0.3927 token=0.8894 | 6.08 steps/s | VRAM=32.75 Go [2026-07-16 18:03:59] step 106200/1000000 | loss=2.4126 vocab=0.7832 token=1.7824 | 6.09 steps/s | VRAM=32.19 Go [2026-07-16 18:04:08] step 106250/1000000 | loss=2.5003 vocab=0.5818 token=1.1536 | 6.11 steps/s | VRAM=31.28 Go [2026-07-16 18:04:16] step 106300/1000000 | loss=2.4825 vocab=0.9208 token=1.3381 | 6.10 steps/s | VRAM=26.08 Go [2026-07-16 18:04:24] step 106350/1000000 | loss=2.4269 vocab=0.6666 token=1.3979 | 6.08 steps/s | VRAM=28.55 Go [2026-07-16 18:04:32] step 106400/1000000 | loss=2.4330 vocab=0.9725 token=1.4574 | 6.12 steps/s | VRAM=32.65 Go [2026-07-16 18:04:40] step 106450/1000000 | loss=2.5897 vocab=0.8634 token=1.4040 | 6.08 steps/s | VRAM=28.80 Go [2026-07-16 18:04:49] step 106500/1000000 | loss=2.5041 vocab=0.8576 token=1.3801 | 6.06 steps/s | VRAM=28.88 Go [2026-07-16 18:04:57] step 106550/1000000 | loss=2.4902 vocab=1.2021 token=1.7213 | 6.12 steps/s | VRAM=28.44 Go [2026-07-16 18:05:05] step 106600/1000000 | loss=2.4841 vocab=1.2809 token=1.2475 | 6.08 steps/s | VRAM=26.17 Go [2026-07-16 18:05:13] step 106650/1000000 | loss=2.4661 vocab=1.0339 token=1.6698 | 6.12 steps/s | VRAM=31.52 Go [2026-07-16 18:05:22] step 106700/1000000 | loss=2.5535 vocab=0.7032 token=1.3443 | 6.04 steps/s | VRAM=28.31 Go [2026-07-16 18:05:30] step 106750/1000000 | loss=2.4525 vocab=1.1683 token=1.8497 | 6.11 steps/s | VRAM=25.80 Go [2026-07-16 18:05:38] step 106800/1000000 | loss=2.4985 vocab=0.9850 token=1.4625 | 6.06 steps/s | VRAM=26.37 Go [2026-07-16 18:05:46] step 106850/1000000 | loss=2.5424 vocab=0.7428 token=1.4584 | 6.09 steps/s | VRAM=32.02 Go [2026-07-16 18:05:54] step 106900/1000000 | loss=2.5037 vocab=0.6254 token=1.3411 | 6.06 steps/s | VRAM=26.31 Go [2026-07-16 18:06:03] step 106950/1000000 | loss=2.4768 vocab=0.7122 token=1.3611 | 6.11 steps/s | VRAM=31.67 Go [2026-07-16 18:06:11] step 107000/1000000 | loss=2.5222 vocab=0.7162 token=1.6826 | 6.10 steps/s | VRAM=32.05 Go [2026-07-16 18:06:19] step 107050/1000000 | loss=2.4995 vocab=0.6204 token=1.4028 | 6.08 steps/s | VRAM=29.05 Go [2026-07-16 18:06:27] step 107100/1000000 | loss=2.5741 vocab=0.7639 token=1.7784 | 6.10 steps/s | VRAM=31.76 Go [2026-07-16 18:06:35] step 107150/1000000 | loss=2.5340 vocab=0.8710 token=1.2474 | 6.07 steps/s | VRAM=28.88 Go [2026-07-16 18:06:44] step 107200/1000000 | loss=2.5195 vocab=0.6028 token=1.0371 | 6.06 steps/s | VRAM=29.07 Go [2026-07-16 18:06:52] step 107250/1000000 | loss=2.4405 vocab=1.0584 token=1.5182 | 6.09 steps/s | VRAM=31.65 Go [2026-07-16 18:07:00] step 107300/1000000 | loss=2.4736 vocab=1.2696 token=1.4198 | 6.10 steps/s | VRAM=32.03 Go [2026-07-16 18:07:08] step 107350/1000000 | loss=2.5835 vocab=1.3245 token=1.9634 | 6.07 steps/s | VRAM=28.57 Go [2026-07-16 18:07:17] step 107400/1000000 | loss=2.3963 vocab=1.2182 token=1.2907 | 6.12 steps/s | VRAM=32.69 Go [2026-07-16 18:07:25] step 107450/1000000 | loss=2.4976 vocab=1.2241 token=1.6359 | 6.04 steps/s | VRAM=26.57 Go [2026-07-16 18:07:33] step 107500/1000000 | loss=2.5413 vocab=0.8362 token=1.3575 | 6.09 steps/s | VRAM=25.91 Go [2026-07-16 18:07:41] step 107550/1000000 | loss=2.6010 vocab=0.9252 token=1.3739 | 6.10 steps/s | VRAM=27.78 Go [2026-07-16 18:07:49] step 107600/1000000 | loss=2.6399 vocab=0.9495 token=1.2668 | 6.12 steps/s | VRAM=30.31 Go [2026-07-16 18:07:58] step 107650/1000000 | loss=2.4137 vocab=0.9856 token=1.2530 | 6.09 steps/s | VRAM=28.82 Go [2026-07-16 18:08:06] step 107700/1000000 | loss=2.4110 vocab=1.5096 token=1.7977 | 6.11 steps/s | VRAM=29.12 Go [2026-07-16 18:08:14] step 107750/1000000 | loss=2.5411 vocab=1.0756 token=1.5235 | 6.08 steps/s | VRAM=29.12 Go [2026-07-16 18:08:22] step 107800/1000000 | loss=2.5222 vocab=1.0868 token=1.3986 | 6.07 steps/s | VRAM=28.93 Go [2026-07-16 18:08:30] step 107850/1000000 | loss=2.4847 vocab=0.7557 token=1.3570 | 6.12 steps/s | VRAM=28.71 Go [2026-07-16 18:08:39] step 107900/1000000 | loss=2.5412 vocab=0.9594 token=1.4027 | 6.12 steps/s | VRAM=28.96 Go [2026-07-16 18:08:47] step 107950/1000000 | loss=2.5169 vocab=1.0455 token=1.3882 | 6.13 steps/s | VRAM=31.42 Go [2026-07-16 18:08:55] step 108000/1000000 | loss=2.4718 vocab=1.1359 token=1.4133 | 6.05 steps/s | VRAM=26.02 Go [2026-07-16 18:09:03] step 108050/1000000 | loss=2.4298 vocab=1.1225 token=1.7355 | 6.11 steps/s | VRAM=29.53 Go [2026-07-16 18:09:11] step 108100/1000000 | loss=2.5649 vocab=0.7332 token=1.6781 | 6.11 steps/s | VRAM=28.72 Go [2026-07-16 18:09:20] step 108150/1000000 | loss=2.4517 vocab=0.8553 token=1.0691 | 6.12 steps/s | VRAM=25.21 Go [2026-07-16 18:09:28] step 108200/1000000 | loss=2.4936 vocab=0.9810 token=1.3039 | 6.11 steps/s | VRAM=29.17 Go [2026-07-16 18:09:36] step 108250/1000000 | loss=2.5058 vocab=1.0281 token=1.5731 | 6.10 steps/s | VRAM=29.41 Go [2026-07-16 18:09:44] step 108300/1000000 | loss=2.4875 vocab=0.7879 token=1.0480 | 6.10 steps/s | VRAM=32.87 Go [2026-07-16 18:09:52] step 108350/1000000 | loss=2.4863 vocab=0.9488 token=1.3445 | 6.11 steps/s | VRAM=32.12 Go [2026-07-16 18:10:00] step 108400/1000000 | loss=2.5079 vocab=0.9739 token=1.5398 | 6.11 steps/s | VRAM=26.17 Go [2026-07-16 18:10:09] step 108450/1000000 | loss=2.4647 vocab=1.0376 token=1.3129 | 6.10 steps/s | VRAM=29.34 Go [2026-07-16 18:10:17] step 108500/1000000 | loss=2.4759 vocab=0.9904 token=1.5624 | 6.07 steps/s | VRAM=28.64 Go [2026-07-16 18:10:25] step 108550/1000000 | loss=2.4879 vocab=0.6614 token=1.7341 | 6.07 steps/s | VRAM=30.37 Go [2026-07-16 18:10:33] step 108600/1000000 | loss=2.4715 vocab=1.3862 token=1.5828 | 6.10 steps/s | VRAM=24.67 Go [2026-07-16 18:10:42] step 108650/1000000 | loss=2.5080 vocab=1.3612 token=1.9402 | 6.12 steps/s | VRAM=24.73 Go [2026-07-16 18:10:50] step 108700/1000000 | loss=2.4627 vocab=1.2279 token=1.8067 | 6.10 steps/s | VRAM=31.66 Go [2026-07-16 18:10:58] step 108750/1000000 | loss=2.5292 vocab=0.9309 token=1.3618 | 6.08 steps/s | VRAM=31.81 Go [2026-07-16 18:11:06] step 108800/1000000 | loss=2.4943 vocab=1.4095 token=1.4718 | 6.09 steps/s | VRAM=32.63 Go [2026-07-16 18:11:14] step 108850/1000000 | loss=2.5242 vocab=0.8454 token=1.4662 | 6.07 steps/s | VRAM=28.49 Go [2026-07-16 18:11:23] step 108900/1000000 | loss=2.4855 vocab=0.5543 token=1.1437 | 6.08 steps/s | VRAM=29.46 Go [2026-07-16 18:11:31] step 108950/1000000 | loss=2.4924 vocab=0.4381 token=1.1867 | 6.11 steps/s | VRAM=32.55 Go [2026-07-16 18:11:39] step 109000/1000000 | loss=2.3805 vocab=1.0588 token=1.2410 | 6.10 steps/s | VRAM=28.48 Go [2026-07-16 18:11:47] step 109050/1000000 | loss=2.4451 vocab=0.7531 token=1.6088 | 6.10 steps/s | VRAM=27.88 Go [2026-07-16 18:11:55] step 109100/1000000 | loss=2.3945 vocab=0.5440 token=1.3441 | 6.10 steps/s | VRAM=28.59 Go [2026-07-16 18:12:04] step 109150/1000000 | loss=2.4799 vocab=0.9068 token=1.3404 | 6.07 steps/s | VRAM=27.95 Go [2026-07-16 18:12:12] step 109200/1000000 | loss=2.5658 vocab=0.9223 token=1.4139 | 6.05 steps/s | VRAM=26.26 Go [2026-07-16 18:12:20] step 109250/1000000 | loss=2.5082 vocab=0.9485 token=1.4905 | 6.09 steps/s | VRAM=29.95 Go [2026-07-16 18:12:28] step 109300/1000000 | loss=2.4877 vocab=0.5780 token=0.6097 | 6.07 steps/s | VRAM=32.43 Go [2026-07-16 18:12:37] step 109350/1000000 | loss=2.4265 vocab=1.2200 token=1.4901 | 6.08 steps/s | VRAM=28.10 Go [2026-07-16 18:12:45] step 109400/1000000 | loss=2.4927 vocab=1.0025 token=1.1868 | 6.10 steps/s | VRAM=30.03 Go [2026-07-16 18:12:53] step 109450/1000000 | loss=2.4445 vocab=0.9078 token=1.6324 | 6.07 steps/s | VRAM=28.12 Go [2026-07-16 18:13:01] step 109500/1000000 | loss=2.5726 vocab=1.2269 token=1.2084 | 6.11 steps/s | VRAM=26.90 Go [2026-07-16 18:13:09] step 109550/1000000 | loss=2.5272 vocab=1.3912 token=1.6725 | 6.08 steps/s | VRAM=30.62 Go [2026-07-16 18:13:18] step 109600/1000000 | loss=2.5472 vocab=1.2486 token=1.5922 | 6.08 steps/s | VRAM=27.29 Go [2026-07-16 18:13:26] step 109650/1000000 | loss=2.5614 vocab=0.5904 token=1.5492 | 6.12 steps/s | VRAM=31.73 Go [2026-07-16 18:13:34] step 109700/1000000 | loss=2.4928 vocab=0.7635 token=1.1482 | 6.12 steps/s | VRAM=28.17 Go [2026-07-16 18:13:42] step 109750/1000000 | loss=2.4911 vocab=0.5464 token=1.4118 | 6.05 steps/s | VRAM=30.36 Go [2026-07-16 18:13:50] step 109800/1000000 | loss=2.3939 vocab=0.9882 token=1.4907 | 6.13 steps/s | VRAM=31.21 Go [2026-07-16 18:13:59] step 109850/1000000 | loss=2.4911 vocab=0.9222 token=1.8267 | 6.09 steps/s | VRAM=25.74 Go [2026-07-16 18:14:07] step 109900/1000000 | loss=2.4070 vocab=1.1895 token=1.7902 | 6.12 steps/s | VRAM=30.79 Go [2026-07-16 18:14:15] step 109950/1000000 | loss=2.4323 vocab=1.1906 token=1.6249 | 6.11 steps/s | VRAM=28.19 Go [2026-07-16 18:14:23] step 110000/1000000 | loss=2.5245 vocab=1.9185 token=1.3167 | 6.07 steps/s | VRAM=30.16 Go [2026-07-16 18:15:46] step 110050/1000000 | loss=2.4559 vocab=0.7400 token=1.5948 | 0.60 steps/s | VRAM=29.43 Go [2026-07-16 18:15:54] step 110100/1000000 | loss=2.4496 vocab=0.3521 token=1.2146 | 6.19 steps/s | VRAM=30.63 Go [2026-07-16 18:16:02] step 110150/1000000 | loss=2.4881 vocab=1.0896 token=1.4657 | 6.14 steps/s | VRAM=32.64 Go [2026-07-16 18:16:10] step 110200/1000000 | loss=2.4228 vocab=1.0403 token=1.4608 | 6.14 steps/s | VRAM=29.22 Go [2026-07-16 18:16:19] step 110250/1000000 | loss=2.5348 vocab=0.7427 token=1.6057 | 6.15 steps/s | VRAM=28.65 Go [2026-07-16 18:16:27] step 110300/1000000 | loss=2.4612 vocab=1.1821 token=1.5811 | 6.16 steps/s | VRAM=25.44 Go [2026-07-16 18:16:35] step 110350/1000000 | loss=2.5365 vocab=1.0541 token=1.3166 | 6.12 steps/s | VRAM=29.63 Go [2026-07-16 18:16:43] step 110400/1000000 | loss=2.5231 vocab=0.6105 token=1.2692 | 6.13 steps/s | VRAM=29.11 Go [2026-07-16 18:16:51] step 110450/1000000 | loss=2.4132 vocab=1.0677 token=1.2248 | 6.11 steps/s | VRAM=28.73 Go [2026-07-16 18:16:59] step 110500/1000000 | loss=2.5361 vocab=0.9776 token=1.5699 | 6.10 steps/s | VRAM=29.27 Go [2026-07-16 18:17:08] step 110550/1000000 | loss=2.5298 vocab=1.0280 token=1.3308 | 6.08 steps/s | VRAM=28.42 Go [2026-07-16 18:17:16] step 110600/1000000 | loss=2.6032 vocab=1.0447 token=1.6057 | 6.10 steps/s | VRAM=32.31 Go [2026-07-16 18:17:24] step 110650/1000000 | loss=2.3883 vocab=1.5212 token=1.6348 | 6.10 steps/s | VRAM=28.50 Go [2026-07-16 18:17:32] step 110700/1000000 | loss=2.5958 vocab=1.2114 token=1.3044 | 6.11 steps/s | VRAM=28.54 Go [2026-07-16 18:17:40] step 110750/1000000 | loss=2.4340 vocab=0.9524 token=1.3406 | 6.07 steps/s | VRAM=25.23 Go [2026-07-16 18:17:49] step 110800/1000000 | loss=2.4606 vocab=0.9480 token=1.3968 | 6.07 steps/s | VRAM=30.15 Go [2026-07-16 18:17:57] step 110850/1000000 | loss=2.4647 vocab=0.8697 token=1.7172 | 6.11 steps/s | VRAM=32.00 Go [2026-07-16 18:18:05] step 110900/1000000 | loss=2.4521 vocab=0.7071 token=1.3623 | 6.09 steps/s | VRAM=30.26 Go [2026-07-16 18:18:13] step 110950/1000000 | loss=2.5492 vocab=1.0018 token=1.3269 | 6.13 steps/s | VRAM=32.15 Go [2026-07-16 18:18:21] step 111000/1000000 | loss=2.4751 vocab=0.8634 token=1.5513 | 6.09 steps/s | VRAM=29.07 Go [2026-07-16 18:18:30] step 111050/1000000 | loss=2.5989 vocab=1.0372 token=1.5879 | 6.08 steps/s | VRAM=29.23 Go [2026-07-16 18:18:38] step 111100/1000000 | loss=2.5358 vocab=0.9991 token=1.6044 | 6.10 steps/s | VRAM=28.52 Go [2026-07-16 18:18:46] step 111150/1000000 | loss=2.5128 vocab=1.2378 token=1.2790 | 6.07 steps/s | VRAM=29.73 Go [2026-07-16 18:18:54] step 111200/1000000 | loss=2.5356 vocab=1.2319 token=1.4330 | 6.06 steps/s | VRAM=26.99 Go [2026-07-16 18:19:03] step 111250/1000000 | loss=2.6637 vocab=1.1033 token=1.0340 | 6.09 steps/s | VRAM=28.40 Go [2026-07-16 18:19:11] step 111300/1000000 | loss=2.5188 vocab=0.6360 token=1.4327 | 6.11 steps/s | VRAM=28.85 Go [2026-07-16 18:19:19] step 111350/1000000 | loss=2.4496 vocab=1.0087 token=1.4917 | 6.11 steps/s | VRAM=28.91 Go [2026-07-16 18:19:27] step 111400/1000000 | loss=2.3840 vocab=0.8608 token=1.6495 | 6.09 steps/s | VRAM=25.67 Go [2026-07-16 18:19:35] step 111450/1000000 | loss=2.5089 vocab=0.9783 token=1.6225 | 6.08 steps/s | VRAM=28.42 Go [2026-07-16 18:19:44] step 111500/1000000 | loss=2.4855 vocab=0.6674 token=1.4926 | 6.08 steps/s | VRAM=28.90 Go [2026-07-16 18:19:52] step 111550/1000000 | loss=2.4994 vocab=0.3293 token=0.9324 | 6.09 steps/s | VRAM=26.62 Go [2026-07-16 18:20:00] step 111600/1000000 | loss=2.5731 vocab=1.0612 token=1.4077 | 6.10 steps/s | VRAM=26.14 Go [2026-07-16 18:20:08] step 111650/1000000 | loss=2.4547 vocab=0.6246 token=1.1110 | 6.09 steps/s | VRAM=32.07 Go [2026-07-16 18:20:16] step 111700/1000000 | loss=2.4718 vocab=0.8363 token=1.1144 | 6.10 steps/s | VRAM=28.79 Go [2026-07-16 18:20:25] step 111750/1000000 | loss=2.5289 vocab=0.7738 token=1.7173 | 6.07 steps/s | VRAM=26.84 Go [2026-07-16 18:20:33] step 111800/1000000 | loss=2.6220 vocab=1.2370 token=1.7068 | 6.06 steps/s | VRAM=29.55 Go [2026-07-16 18:20:41] step 111850/1000000 | loss=2.5588 vocab=1.3913 token=1.8343 | 6.11 steps/s | VRAM=31.92 Go [2026-07-16 18:20:49] step 111900/1000000 | loss=2.5457 vocab=1.1245 token=1.5726 | 6.11 steps/s | VRAM=26.58 Go [2026-07-16 18:20:57] step 111950/1000000 | loss=2.4814 vocab=1.3284 token=1.5515 | 6.10 steps/s | VRAM=28.81 Go [2026-07-16 18:21:06] step 112000/1000000 | loss=2.5896 vocab=0.9987 token=1.7799 | 6.06 steps/s | VRAM=32.88 Go [2026-07-16 18:21:14] step 112050/1000000 | loss=2.4237 vocab=0.8354 token=1.5644 | 6.09 steps/s | VRAM=28.18 Go [2026-07-16 18:21:22] step 112100/1000000 | loss=2.4501 vocab=0.6738 token=1.2705 | 6.08 steps/s | VRAM=28.48 Go [2026-07-16 18:21:30] step 112150/1000000 | loss=2.4500 vocab=0.9137 token=1.5741 | 6.10 steps/s | VRAM=32.52 Go [2026-07-16 18:21:39] step 112200/1000000 | loss=2.4654 vocab=1.1988 token=1.5667 | 6.11 steps/s | VRAM=28.38 Go [2026-07-16 18:21:47] step 112250/1000000 | loss=2.4477 vocab=0.8446 token=1.3665 | 6.13 steps/s | VRAM=31.60 Go [2026-07-16 18:21:55] step 112300/1000000 | loss=2.4821 vocab=0.9055 token=1.4655 | 6.08 steps/s | VRAM=32.67 Go [2026-07-16 18:22:03] step 112350/1000000 | loss=2.5346 vocab=1.0488 token=1.5185 | 6.06 steps/s | VRAM=29.31 Go [2026-07-16 18:22:11] step 112400/1000000 | loss=2.4932 vocab=0.5411 token=1.3296 | 6.09 steps/s | VRAM=28.66 Go [2026-07-16 18:22:20] step 112450/1000000 | loss=2.5967 vocab=1.3354 token=1.5530 | 6.12 steps/s | VRAM=32.46 Go [2026-07-16 18:22:28] step 112500/1000000 | loss=2.5603 vocab=0.4560 token=0.9090 | 6.11 steps/s | VRAM=24.62 Go [2026-07-16 18:22:36] step 112550/1000000 | loss=2.4523 vocab=0.8254 token=1.4545 | 6.13 steps/s | VRAM=25.44 Go [2026-07-16 18:22:44] step 112600/1000000 | loss=2.5136 vocab=0.9421 token=1.6716 | 6.09 steps/s | VRAM=28.75 Go [2026-07-16 18:22:52] step 112650/1000000 | loss=2.3940 vocab=0.7504 token=1.1564 | 6.10 steps/s | VRAM=29.53 Go [2026-07-16 18:23:00] step 112700/1000000 | loss=2.4916 vocab=1.0302 token=1.3229 | 6.10 steps/s | VRAM=28.02 Go [2026-07-16 18:23:09] step 112750/1000000 | loss=2.5318 vocab=0.4778 token=1.3912 | 6.10 steps/s | VRAM=28.46 Go [2026-07-16 18:23:17] step 112800/1000000 | loss=2.5304 vocab=1.5253 token=1.6515 | 6.10 steps/s | VRAM=28.96 Go [2026-07-16 18:23:25] step 112850/1000000 | loss=2.4516 vocab=0.5805 token=1.4414 | 6.11 steps/s | VRAM=28.35 Go [2026-07-16 18:23:33] step 112900/1000000 | loss=2.5553 vocab=0.9009 token=1.6344 | 6.09 steps/s | VRAM=29.22 Go [2026-07-16 18:23:41] step 112950/1000000 | loss=2.5791 vocab=1.5553 token=1.7264 | 6.10 steps/s | VRAM=28.97 Go [2026-07-16 18:23:50] step 113000/1000000 | loss=2.4580 vocab=1.4745 token=1.4332 | 6.10 steps/s | VRAM=32.47 Go [2026-07-16 18:23:58] step 113050/1000000 | loss=2.5394 vocab=0.7005 token=1.5399 | 6.08 steps/s | VRAM=28.34 Go [2026-07-16 18:24:06] step 113100/1000000 | loss=2.5361 vocab=1.3910 token=2.0057 | 6.10 steps/s | VRAM=29.09 Go [2026-07-16 18:24:14] step 113150/1000000 | loss=2.5254 vocab=1.4910 token=1.5581 | 6.12 steps/s | VRAM=32.50 Go [2026-07-16 18:24:22] step 113200/1000000 | loss=2.4267 vocab=0.6326 token=1.6370 | 6.10 steps/s | VRAM=28.83 Go [2026-07-16 18:24:31] step 113250/1000000 | loss=2.5157 vocab=0.6937 token=1.1839 | 6.10 steps/s | VRAM=28.50 Go [2026-07-16 18:24:39] step 113300/1000000 | loss=2.5497 vocab=1.8967 token=1.6025 | 6.13 steps/s | VRAM=30.47 Go [2026-07-16 18:24:47] step 113350/1000000 | loss=2.4715 vocab=0.9315 token=1.3895 | 6.10 steps/s | VRAM=32.71 Go [2026-07-16 18:24:55] step 113400/1000000 | loss=2.4538 vocab=1.4866 token=1.3948 | 6.10 steps/s | VRAM=26.39 Go [2026-07-16 18:25:03] step 113450/1000000 | loss=2.4774 vocab=0.5293 token=1.2415 | 6.11 steps/s | VRAM=29.32 Go [2026-07-16 18:25:12] step 113500/1000000 | loss=2.5738 vocab=1.4901 token=1.7029 | 6.10 steps/s | VRAM=24.56 Go [2026-07-16 18:25:20] step 113550/1000000 | loss=2.5018 vocab=0.8299 token=1.3735 | 6.08 steps/s | VRAM=28.89 Go [2026-07-16 18:25:28] step 113600/1000000 | loss=2.4630 vocab=0.6343 token=1.4964 | 6.08 steps/s | VRAM=28.23 Go [2026-07-16 18:25:36] step 113650/1000000 | loss=2.4651 vocab=1.1353 token=0.9627 | 6.09 steps/s | VRAM=32.91 Go [2026-07-16 18:25:44] step 113700/1000000 | loss=2.4718 vocab=1.2120 token=1.1202 | 6.10 steps/s | VRAM=32.87 Go [2026-07-16 18:25:53] step 113750/1000000 | loss=2.4631 vocab=0.5512 token=1.3601 | 6.10 steps/s | VRAM=31.36 Go [2026-07-16 18:26:01] step 113800/1000000 | loss=2.5056 vocab=1.0512 token=1.2358 | 6.04 steps/s | VRAM=24.27 Go [2026-07-16 18:26:09] step 113850/1000000 | loss=2.5470 vocab=0.5872 token=1.6109 | 6.12 steps/s | VRAM=27.52 Go [2026-07-16 18:26:17] step 113900/1000000 | loss=2.5163 vocab=0.8738 token=1.5776 | 6.07 steps/s | VRAM=29.01 Go [2026-07-16 18:26:25] step 113950/1000000 | loss=2.4876 vocab=0.6097 token=1.1070 | 6.12 steps/s | VRAM=32.85 Go [2026-07-16 18:26:34] step 114000/1000000 | loss=2.5426 vocab=0.8679 token=1.2316 | 6.11 steps/s | VRAM=32.70 Go [2026-07-16 18:26:42] step 114050/1000000 | loss=2.4760 vocab=0.9495 token=1.4566 | 6.12 steps/s | VRAM=31.68 Go [2026-07-16 18:26:50] step 114100/1000000 | loss=2.4560 vocab=1.0861 token=1.4004 | 6.12 steps/s | VRAM=29.28 Go [2026-07-16 18:26:58] step 114150/1000000 | loss=2.5259 vocab=0.9446 token=1.4009 | 6.14 steps/s | VRAM=27.97 Go [2026-07-16 18:27:06] step 114200/1000000 | loss=2.5585 vocab=0.8110 token=1.4095 | 6.10 steps/s | VRAM=31.60 Go [2026-07-16 18:27:14] step 114250/1000000 | loss=2.4706 vocab=0.9746 token=1.4693 | 6.13 steps/s | VRAM=31.23 Go [2026-07-16 18:27:23] step 114300/1000000 | loss=2.4198 vocab=0.9195 token=1.5061 | 6.06 steps/s | VRAM=26.37 Go [2026-07-16 18:27:31] step 114350/1000000 | loss=2.4823 vocab=0.7143 token=1.5015 | 6.10 steps/s | VRAM=28.93 Go [2026-07-16 18:27:39] step 114400/1000000 | loss=2.5657 vocab=0.9113 token=1.5963 | 6.12 steps/s | VRAM=31.55 Go [2026-07-16 18:27:47] step 114450/1000000 | loss=2.4586 vocab=0.9950 token=1.4916 | 6.10 steps/s | VRAM=28.91 Go [2026-07-16 18:27:56] step 114500/1000000 | loss=2.4565 vocab=0.8666 token=1.6669 | 6.06 steps/s | VRAM=29.15 Go [2026-07-16 18:28:04] step 114550/1000000 | loss=2.4379 vocab=0.7055 token=1.4957 | 6.11 steps/s | VRAM=31.28 Go [2026-07-16 18:28:12] step 114600/1000000 | loss=2.4983 vocab=0.9081 token=1.6503 | 6.07 steps/s | VRAM=26.32 Go [2026-07-16 18:28:20] step 114650/1000000 | loss=2.6100 vocab=1.1181 token=2.0180 | 6.12 steps/s | VRAM=28.13 Go [2026-07-16 18:28:28] step 114700/1000000 | loss=2.4584 vocab=1.0310 token=1.3192 | 6.07 steps/s | VRAM=26.37 Go [2026-07-16 18:28:37] step 114750/1000000 | loss=2.4635 vocab=1.0178 token=1.4115 | 6.11 steps/s | VRAM=25.76 Go [2026-07-16 18:28:45] step 114800/1000000 | loss=2.4367 vocab=0.8782 token=1.6175 | 6.12 steps/s | VRAM=25.66 Go [2026-07-16 18:28:53] step 114850/1000000 | loss=2.4554 vocab=1.2468 token=1.1971 | 6.14 steps/s | VRAM=28.24 Go [2026-07-16 18:29:01] step 114900/1000000 | loss=2.5877 vocab=1.1188 token=1.2923 | 6.08 steps/s | VRAM=29.09 Go [2026-07-16 18:29:09] step 114950/1000000 | loss=2.6042 vocab=0.8212 token=1.7039 | 6.11 steps/s | VRAM=28.10 Go [2026-07-16 18:29:18] step 115000/1000000 | loss=2.5034 vocab=1.2482 token=1.5273 | 6.09 steps/s | VRAM=28.62 Go [2026-07-16 18:30:49] step 115050/1000000 | loss=2.4935 vocab=1.1505 token=1.2670 | 0.55 steps/s | VRAM=28.27 Go [2026-07-16 18:30:57] step 115100/1000000 | loss=2.5526 vocab=1.2911 token=1.3769 | 6.16 steps/s | VRAM=32.29 Go [2026-07-16 18:31:05] step 115150/1000000 | loss=2.4771 vocab=1.3517 token=1.3334 | 6.13 steps/s | VRAM=26.26 Go [2026-07-16 18:31:13] step 115200/1000000 | loss=2.5004 vocab=0.6804 token=1.1673 | 6.15 steps/s | VRAM=29.10 Go [2026-07-16 18:31:21] step 115250/1000000 | loss=2.5342 vocab=0.4974 token=0.8327 | 6.13 steps/s | VRAM=26.32 Go [2026-07-16 18:31:30] step 115300/1000000 | loss=2.5141 vocab=0.8849 token=1.4023 | 6.13 steps/s | VRAM=32.95 Go [2026-07-16 18:31:38] step 115350/1000000 | loss=2.3403 vocab=0.6156 token=1.2022 | 6.11 steps/s | VRAM=26.18 Go [2026-07-16 18:31:46] step 115400/1000000 | loss=2.5348 vocab=0.7762 token=1.4185 | 6.10 steps/s | VRAM=29.88 Go [2026-07-16 18:31:54] step 115450/1000000 | loss=2.5315 vocab=1.3560 token=1.6999 | 6.14 steps/s | VRAM=32.39 Go [2026-07-16 18:32:02] step 115500/1000000 | loss=2.4614 vocab=1.1039 token=1.5206 | 6.10 steps/s | VRAM=29.11 Go [2026-07-16 18:32:11] step 115550/1000000 | loss=2.6028 vocab=1.2802 token=1.6451 | 6.08 steps/s | VRAM=28.22 Go [2026-07-16 18:32:19] step 115600/1000000 | loss=2.5134 vocab=1.0723 token=1.0302 | 6.09 steps/s | VRAM=28.87 Go [2026-07-16 18:32:27] step 115650/1000000 | loss=2.5624 vocab=1.3667 token=1.7583 | 6.06 steps/s | VRAM=28.91 Go [2026-07-16 18:32:35] step 115700/1000000 | loss=2.4679 vocab=0.9510 token=1.7434 | 6.09 steps/s | VRAM=26.96 Go [2026-07-16 18:32:43] step 115750/1000000 | loss=2.4437 vocab=0.6645 token=1.3477 | 6.09 steps/s | VRAM=32.64 Go [2026-07-16 18:32:52] step 115800/1000000 | loss=2.4795 vocab=1.2253 token=1.4495 | 6.08 steps/s | VRAM=31.91 Go [2026-07-16 18:33:00] step 115850/1000000 | loss=2.5531 vocab=1.1515 token=1.4180 | 6.10 steps/s | VRAM=31.92 Go [2026-07-16 18:33:08] step 115900/1000000 | loss=2.5095 vocab=0.7719 token=1.3593 | 6.08 steps/s | VRAM=28.77 Go [2026-07-16 18:33:16] step 115950/1000000 | loss=2.5147 vocab=1.3736 token=1.4497 | 6.10 steps/s | VRAM=32.57 Go [2026-07-16 18:33:24] step 116000/1000000 | loss=2.5503 vocab=1.0987 token=1.7081 | 6.11 steps/s | VRAM=25.11 Go [2026-07-16 18:33:33] step 116050/1000000 | loss=2.5445 vocab=1.4358 token=1.7157 | 6.08 steps/s | VRAM=28.17 Go [2026-07-16 18:33:41] step 116100/1000000 | loss=2.5093 vocab=0.7289 token=1.6189 | 6.12 steps/s | VRAM=31.51 Go [2026-07-16 18:33:49] step 116150/1000000 | loss=2.4674 vocab=0.8901 token=1.5218 | 6.09 steps/s | VRAM=32.10 Go [2026-07-16 18:33:57] step 116200/1000000 | loss=2.4575 vocab=0.7525 token=1.1698 | 6.11 steps/s | VRAM=25.75 Go [2026-07-16 18:34:05] step 116250/1000000 | loss=2.5559 vocab=1.1637 token=1.7361 | 6.06 steps/s | VRAM=28.19 Go [2026-07-16 18:34:14] step 116300/1000000 | loss=2.5880 vocab=1.0855 token=1.3559 | 6.08 steps/s | VRAM=27.81 Go [2026-07-16 18:34:22] step 116350/1000000 | loss=2.4312 vocab=0.9709 token=1.4807 | 6.10 steps/s | VRAM=27.20 Go [2026-07-16 18:34:30] step 116400/1000000 | loss=2.5184 vocab=1.3932 token=1.5784 | 6.11 steps/s | VRAM=32.83 Go [2026-07-16 18:34:38] step 116450/1000000 | loss=2.4203 vocab=1.3620 token=1.8247 | 6.08 steps/s | VRAM=28.94 Go [2026-07-16 18:34:46] step 116500/1000000 | loss=2.5684 vocab=1.3973 token=1.2341 | 6.12 steps/s | VRAM=31.83 Go [2026-07-16 18:34:55] step 116550/1000000 | loss=2.4538 vocab=0.8099 token=1.0748 | 6.06 steps/s | VRAM=26.01 Go [2026-07-16 18:35:03] step 116600/1000000 | loss=2.4322 vocab=1.0028 token=1.1064 | 6.10 steps/s | VRAM=28.15 Go [2026-07-16 18:35:11] step 116650/1000000 | loss=2.5527 vocab=1.1640 token=1.2668 | 6.11 steps/s | VRAM=32.27 Go [2026-07-16 18:35:19] step 116700/1000000 | loss=2.4542 vocab=0.7262 token=1.2929 | 6.10 steps/s | VRAM=32.18 Go [2026-07-16 18:35:27] step 116750/1000000 | loss=2.5154 vocab=1.3637 token=1.6213 | 6.12 steps/s | VRAM=32.39 Go [2026-07-16 18:35:36] step 116800/1000000 | loss=2.4655 vocab=0.7104 token=1.1230 | 6.09 steps/s | VRAM=28.75 Go [2026-07-16 18:35:44] step 116850/1000000 | loss=2.5255 vocab=1.5086 token=1.6096 | 6.09 steps/s | VRAM=26.35 Go [2026-07-16 18:35:52] step 116900/1000000 | loss=2.5899 vocab=0.8291 token=1.5123 | 6.11 steps/s | VRAM=25.95 Go [2026-07-16 18:36:00] step 116950/1000000 | loss=2.4780 vocab=0.9272 token=1.6855 | 6.12 steps/s | VRAM=28.91 Go [2026-07-16 18:36:08] step 117000/1000000 | loss=2.6832 vocab=1.0256 token=1.2914 | 6.10 steps/s | VRAM=28.30 Go [2026-07-16 18:36:17] step 117050/1000000 | loss=2.5413 vocab=0.8210 token=1.3855 | 6.12 steps/s | VRAM=26.00 Go [2026-07-16 18:36:25] step 117100/1000000 | loss=2.5148 vocab=0.7881 token=1.5815 | 6.11 steps/s | VRAM=29.18 Go [2026-07-16 18:36:33] step 117150/1000000 | loss=2.5798 vocab=0.7595 token=1.1068 | 6.12 steps/s | VRAM=31.03 Go [2026-07-16 18:36:41] step 117200/1000000 | loss=2.4698 vocab=1.5160 token=1.6313 | 6.10 steps/s | VRAM=29.20 Go [2026-07-16 18:36:49] step 117250/1000000 | loss=2.4138 vocab=0.8200 token=1.4449 | 6.12 steps/s | VRAM=24.52 Go [2026-07-16 18:36:58] step 117300/1000000 | loss=2.5029 vocab=0.6428 token=1.2934 | 6.10 steps/s | VRAM=32.42 Go [2026-07-16 18:37:06] step 117350/1000000 | loss=2.5280 vocab=0.8352 token=1.6684 | 6.06 steps/s | VRAM=26.50 Go [2026-07-16 18:37:14] step 117400/1000000 | loss=2.4889 vocab=1.5043 token=1.7647 | 6.12 steps/s | VRAM=28.12 Go [2026-07-16 18:37:22] step 117450/1000000 | loss=2.4311 vocab=1.2483 token=1.5189 | 6.08 steps/s | VRAM=25.54 Go [2026-07-16 18:37:30] step 117500/1000000 | loss=2.5217 vocab=0.8366 token=1.4906 | 6.08 steps/s | VRAM=29.41 Go [2026-07-16 18:37:39] step 117550/1000000 | loss=2.6527 vocab=0.6620 token=1.4217 | 6.10 steps/s | VRAM=32.02 Go [2026-07-16 18:37:47] step 117600/1000000 | loss=2.6227 vocab=1.0953 token=1.7230 | 6.11 steps/s | VRAM=24.95 Go [2026-07-16 18:37:55] step 117650/1000000 | loss=2.5787 vocab=0.6882 token=1.2704 | 6.11 steps/s | VRAM=28.67 Go [2026-07-16 18:38:03] step 117700/1000000 | loss=2.3290 vocab=0.6655 token=1.2315 | 6.12 steps/s | VRAM=29.18 Go [2026-07-16 18:38:11] step 117750/1000000 | loss=2.4702 vocab=0.5950 token=1.3292 | 6.10 steps/s | VRAM=29.22 Go [2026-07-16 18:38:20] step 117800/1000000 | loss=2.4826 vocab=0.8400 token=1.1778 | 6.09 steps/s | VRAM=27.55 Go [2026-07-16 18:38:28] step 117850/1000000 | loss=2.5596 vocab=1.5261 token=1.4179 | 6.12 steps/s | VRAM=28.17 Go [2026-07-16 18:38:36] step 117900/1000000 | loss=2.5499 vocab=0.9111 token=1.2287 | 6.09 steps/s | VRAM=32.32 Go [2026-07-16 18:38:44] step 117950/1000000 | loss=2.5281 vocab=1.5552 token=1.5523 | 6.14 steps/s | VRAM=27.95 Go [2026-07-16 18:38:52] step 118000/1000000 | loss=2.4798 vocab=1.0201 token=1.3902 | 6.11 steps/s | VRAM=32.56 Go [2026-07-16 18:39:00] step 118050/1000000 | loss=2.4384 vocab=0.5721 token=1.3077 | 6.12 steps/s | VRAM=29.10 Go [2026-07-16 18:39:09] step 118100/1000000 | loss=2.5523 vocab=1.5208 token=1.7910 | 6.09 steps/s | VRAM=28.81 Go [2026-07-16 18:39:17] step 118150/1000000 | loss=2.4822 vocab=0.9210 token=1.8877 | 6.12 steps/s | VRAM=27.99 Go [2026-07-16 18:39:25] step 118200/1000000 | loss=2.3126 vocab=1.2059 token=1.5075 | 6.08 steps/s | VRAM=26.64 Go [2026-07-16 18:39:33] step 118250/1000000 | loss=2.4585 vocab=0.9319 token=1.2034 | 6.09 steps/s | VRAM=25.83 Go [2026-07-16 18:39:41] step 118300/1000000 | loss=2.6596 vocab=1.0974 token=1.8587 | 6.10 steps/s | VRAM=31.60 Go [2026-07-16 18:39:50] step 118350/1000000 | loss=2.4468 vocab=0.9576 token=1.4361 | 6.11 steps/s | VRAM=29.43 Go [2026-07-16 18:39:58] step 118400/1000000 | loss=2.5905 vocab=1.3169 token=1.6641 | 6.12 steps/s | VRAM=32.24 Go [2026-07-16 18:40:06] step 118450/1000000 | loss=2.5460 vocab=0.7098 token=1.5165 | 6.11 steps/s | VRAM=28.67 Go [2026-07-16 18:40:14] step 118500/1000000 | loss=2.4245 vocab=1.1015 token=1.7477 | 6.11 steps/s | VRAM=32.13 Go [2026-07-16 18:40:22] step 118550/1000000 | loss=2.5300 vocab=1.0600 token=1.5350 | 6.11 steps/s | VRAM=32.20 Go [2026-07-16 18:40:31] step 118600/1000000 | loss=2.6127 vocab=0.9811 token=1.6546 | 6.08 steps/s | VRAM=32.11 Go [2026-07-16 18:40:39] step 118650/1000000 | loss=2.4858 vocab=0.8854 token=1.5364 | 6.09 steps/s | VRAM=31.61 Go [2026-07-16 18:40:47] step 118700/1000000 | loss=2.4849 vocab=0.7202 token=1.1219 | 6.11 steps/s | VRAM=28.09 Go [2026-07-16 18:40:55] step 118750/1000000 | loss=2.5273 vocab=1.4256 token=1.6332 | 6.11 steps/s | VRAM=31.57 Go [2026-07-16 18:41:03] step 118800/1000000 | loss=2.3800 vocab=1.1791 token=1.9256 | 6.07 steps/s | VRAM=29.88 Go [2026-07-16 18:41:12] step 118850/1000000 | loss=2.4723 vocab=0.9395 token=1.4679 | 6.10 steps/s | VRAM=28.53 Go [2026-07-16 18:41:20] step 118900/1000000 | loss=2.4448 vocab=0.8411 token=1.6990 | 6.12 steps/s | VRAM=29.23 Go [2026-07-16 18:41:28] step 118950/1000000 | loss=2.4562 vocab=0.5713 token=1.6789 | 6.11 steps/s | VRAM=29.25 Go [2026-07-16 18:41:36] step 119000/1000000 | loss=2.4914 vocab=0.7065 token=1.4892 | 6.08 steps/s | VRAM=29.55 Go [2026-07-16 18:41:44] step 119050/1000000 | loss=2.4489 vocab=1.5271 token=1.7916 | 6.11 steps/s | VRAM=28.41 Go [2026-07-16 18:41:53] step 119100/1000000 | loss=2.3617 vocab=0.9120 token=1.2678 | 6.12 steps/s | VRAM=28.98 Go [2026-07-16 18:42:01] step 119150/1000000 | loss=2.5171 vocab=1.0424 token=1.3037 | 6.10 steps/s | VRAM=25.00 Go [2026-07-16 18:42:09] step 119200/1000000 | loss=2.4575 vocab=0.8524 token=1.4912 | 6.12 steps/s | VRAM=28.73 Go [2026-07-16 18:42:17] step 119250/1000000 | loss=2.4948 vocab=0.9756 token=1.1114 | 6.12 steps/s | VRAM=28.88 Go [2026-07-16 18:42:25] step 119300/1000000 | loss=2.4365 vocab=1.0542 token=1.5624 | 6.10 steps/s | VRAM=28.38 Go [2026-07-16 18:42:33] step 119350/1000000 | loss=2.4684 vocab=0.8739 token=1.3912 | 6.12 steps/s | VRAM=24.62 Go [2026-07-16 18:42:42] step 119400/1000000 | loss=2.4582 vocab=1.6737 token=1.4897 | 6.11 steps/s | VRAM=25.94 Go [2026-07-16 18:42:50] step 119450/1000000 | loss=2.5329 vocab=1.3663 token=1.6808 | 6.10 steps/s | VRAM=32.17 Go [2026-07-16 18:42:58] step 119500/1000000 | loss=2.5426 vocab=0.9468 token=1.4789 | 6.12 steps/s | VRAM=32.98 Go [2026-07-16 18:43:06] step 119550/1000000 | loss=2.6482 vocab=0.8852 token=1.6759 | 6.06 steps/s | VRAM=28.85 Go [2026-07-16 18:43:14] step 119600/1000000 | loss=2.5123 vocab=0.7538 token=1.4861 | 6.09 steps/s | VRAM=28.66 Go [2026-07-16 18:43:23] step 119650/1000000 | loss=2.5655 vocab=1.4438 token=1.4976 | 6.13 steps/s | VRAM=28.63 Go [2026-07-16 18:43:31] step 119700/1000000 | loss=2.4281 vocab=0.9543 token=1.6777 | 6.13 steps/s | VRAM=32.35 Go [2026-07-16 18:43:39] step 119750/1000000 | loss=2.4338 vocab=0.7863 token=1.0956 | 6.14 steps/s | VRAM=28.51 Go [2026-07-16 18:43:47] step 119800/1000000 | loss=2.6653 vocab=1.1296 token=1.5363 | 6.10 steps/s | VRAM=26.33 Go [2026-07-16 18:43:55] step 119850/1000000 | loss=2.4312 vocab=0.9241 token=1.5351 | 6.13 steps/s | VRAM=28.15 Go [2026-07-16 18:44:03] step 119900/1000000 | loss=2.4434 vocab=1.0679 token=1.3668 | 6.11 steps/s | VRAM=25.72 Go [2026-07-16 18:44:12] step 119950/1000000 | loss=2.4565 vocab=0.6474 token=1.3106 | 6.12 steps/s | VRAM=25.77 Go [2026-07-16 18:44:20] step 120000/1000000 | loss=2.5854 vocab=1.0969 token=1.7484 | 6.08 steps/s | VRAM=29.34 Go [2026-07-16 18:45:45] step 120050/1000000 | loss=2.3578 vocab=1.0619 token=1.5657 | 0.58 steps/s | VRAM=28.22 Go [2026-07-16 18:45:53] step 120100/1000000 | loss=2.5184 vocab=0.7989 token=1.3462 | 6.19 steps/s | VRAM=28.61 Go [2026-07-16 18:46:02] step 120150/1000000 | loss=2.6348 vocab=1.1571 token=1.3444 | 6.17 steps/s | VRAM=25.80 Go [2026-07-16 18:46:10] step 120200/1000000 | loss=2.4808 vocab=1.2701 token=1.2841 | 6.14 steps/s | VRAM=29.10 Go [2026-07-16 18:46:18] step 120250/1000000 | loss=2.5146 vocab=1.4183 token=1.6089 | 6.16 steps/s | VRAM=31.72 Go [2026-07-16 18:46:26] step 120300/1000000 | loss=2.5476 vocab=0.8848 token=1.7521 | 6.13 steps/s | VRAM=29.24 Go [2026-07-16 18:46:34] step 120350/1000000 | loss=2.5387 vocab=1.0406 token=1.7973 | 6.13 steps/s | VRAM=26.25 Go [2026-07-16 18:46:42] step 120400/1000000 | loss=2.5480 vocab=1.4389 token=1.7186 | 6.10 steps/s | VRAM=28.65 Go [2026-07-16 18:46:50] step 120450/1000000 | loss=2.5033 vocab=1.0783 token=1.5953 | 6.14 steps/s | VRAM=25.90 Go [2026-07-16 18:46:59] step 120500/1000000 | loss=2.4961 vocab=1.0470 token=1.7230 | 6.12 steps/s | VRAM=26.63 Go [2026-07-16 18:47:07] step 120550/1000000 | loss=2.3886 vocab=1.3193 token=1.7009 | 6.12 steps/s | VRAM=32.18 Go [2026-07-16 18:47:15] step 120600/1000000 | loss=2.4456 vocab=0.5124 token=1.0300 | 6.12 steps/s | VRAM=24.81 Go [2026-07-16 18:47:23] step 120650/1000000 | loss=2.4772 vocab=1.1238 token=1.3821 | 6.12 steps/s | VRAM=32.05 Go [2026-07-16 18:47:31] step 120700/1000000 | loss=2.4615 vocab=1.1987 token=1.3676 | 6.11 steps/s | VRAM=24.55 Go [2026-07-16 18:47:40] step 120750/1000000 | loss=2.4684 vocab=1.1009 token=1.5209 | 6.10 steps/s | VRAM=25.51 Go [2026-07-16 18:47:48] step 120800/1000000 | loss=2.4385 vocab=1.1319 token=1.2983 | 6.09 steps/s | VRAM=27.01 Go [2026-07-16 18:47:56] step 120850/1000000 | loss=2.4442 vocab=0.7012 token=1.2549 | 6.12 steps/s | VRAM=25.40 Go [2026-07-16 18:48:04] step 120900/1000000 | loss=2.5470 vocab=1.0529 token=1.4082 | 6.09 steps/s | VRAM=28.50 Go [2026-07-16 18:48:12] step 120950/1000000 | loss=2.4668 vocab=0.8531 token=1.4272 | 6.08 steps/s | VRAM=29.47 Go [2026-07-16 18:48:21] step 121000/1000000 | loss=2.5801 vocab=1.1710 token=1.6838 | 6.11 steps/s | VRAM=28.10 Go [2026-07-16 18:48:29] step 121050/1000000 | loss=2.5161 vocab=1.3046 token=1.9059 | 6.10 steps/s | VRAM=28.70 Go [2026-07-16 18:48:37] step 121100/1000000 | loss=2.4381 vocab=1.1245 token=1.6685 | 6.10 steps/s | VRAM=28.91 Go [2026-07-16 18:48:45] step 121150/1000000 | loss=2.5153 vocab=0.9140 token=1.3404 | 6.07 steps/s | VRAM=23.59 Go [2026-07-16 18:48:53] step 121200/1000000 | loss=2.5177 vocab=0.9070 token=1.4955 | 6.09 steps/s | VRAM=29.10 Go [2026-07-16 18:49:02] step 121250/1000000 | loss=2.4727 vocab=1.1623 token=1.5369 | 6.06 steps/s | VRAM=26.12 Go [2026-07-16 18:49:10] step 121300/1000000 | loss=2.4542 vocab=0.7945 token=1.5220 | 6.09 steps/s | VRAM=32.87 Go [2026-07-16 18:49:18] step 121350/1000000 | loss=2.5783 vocab=0.9557 token=1.0902 | 6.08 steps/s | VRAM=25.14 Go [2026-07-16 18:49:26] step 121400/1000000 | loss=2.4568 vocab=0.4041 token=1.1654 | 6.10 steps/s | VRAM=26.30 Go [2026-07-16 18:49:34] step 121450/1000000 | loss=2.4223 vocab=0.6511 token=1.7207 | 6.11 steps/s | VRAM=32.13 Go [2026-07-16 18:49:43] step 121500/1000000 | loss=2.5050 vocab=1.4544 token=1.3639 | 6.13 steps/s | VRAM=32.58 Go [2026-07-16 18:49:51] step 121550/1000000 | loss=2.5259 vocab=1.2930 token=1.6437 | 6.10 steps/s | VRAM=29.12 Go [2026-07-16 18:49:59] step 121600/1000000 | loss=2.4594 vocab=0.6645 token=1.1709 | 6.13 steps/s | VRAM=24.73 Go [2026-07-16 18:50:07] step 121650/1000000 | loss=2.5154 vocab=0.7371 token=1.8654 | 6.08 steps/s | VRAM=26.35 Go [2026-07-16 18:50:15] step 121700/1000000 | loss=2.5611 vocab=0.9221 token=1.4148 | 6.09 steps/s | VRAM=31.81 Go [2026-07-16 18:50:24] step 121750/1000000 | loss=2.5421 vocab=1.2060 token=1.7046 | 6.10 steps/s | VRAM=24.96 Go [2026-07-16 18:50:32] step 121800/1000000 | loss=2.4531 vocab=0.8711 token=1.5260 | 6.10 steps/s | VRAM=31.52 Go [2026-07-16 18:50:40] step 121850/1000000 | loss=2.4667 vocab=1.5050 token=1.7613 | 6.08 steps/s | VRAM=26.24 Go [2026-07-16 18:50:48] step 121900/1000000 | loss=2.5229 vocab=1.1844 token=1.3569 | 6.08 steps/s | VRAM=25.15 Go [2026-07-16 18:50:56] step 121950/1000000 | loss=2.5480 vocab=1.1841 token=1.3611 | 6.08 steps/s | VRAM=30.83 Go [2026-07-16 18:51:05] step 122000/1000000 | loss=2.4847 vocab=1.5366 token=1.8753 | 6.13 steps/s | VRAM=28.92 Go [2026-07-16 18:51:13] step 122050/1000000 | loss=2.4746 vocab=1.0917 token=1.2751 | 6.10 steps/s | VRAM=28.20 Go [2026-07-16 18:51:21] step 122100/1000000 | loss=2.5122 vocab=1.1335 token=1.6450 | 6.08 steps/s | VRAM=26.33 Go [2026-07-16 18:51:29] step 122150/1000000 | loss=2.6261 vocab=0.4724 token=1.3845 | 6.09 steps/s | VRAM=32.68 Go [2026-07-16 18:51:37] step 122200/1000000 | loss=2.4636 vocab=0.9825 token=1.3360 | 6.06 steps/s | VRAM=26.97 Go [2026-07-16 18:51:46] step 122250/1000000 | loss=2.4657 vocab=1.1145 token=1.3988 | 6.10 steps/s | VRAM=28.00 Go [2026-07-16 18:51:54] step 122300/1000000 | loss=2.4712 vocab=0.9844 token=1.5703 | 6.08 steps/s | VRAM=26.28 Go [2026-07-16 18:52:02] step 122350/1000000 | loss=2.5304 vocab=1.4537 token=1.2672 | 6.09 steps/s | VRAM=28.09 Go [2026-07-16 18:52:10] step 122400/1000000 | loss=2.5413 vocab=0.6032 token=1.3178 | 6.11 steps/s | VRAM=31.45 Go [2026-07-16 18:52:18] step 122450/1000000 | loss=2.5186 vocab=0.8176 token=1.2461 | 6.10 steps/s | VRAM=30.34 Go [2026-07-16 18:52:27] step 122500/1000000 | loss=2.4593 vocab=0.8622 token=1.3981 | 6.10 steps/s | VRAM=29.37 Go [2026-07-16 18:52:35] step 122550/1000000 | loss=2.4275 vocab=0.6829 token=1.1940 | 6.11 steps/s | VRAM=28.49 Go [2026-07-16 18:52:43] step 122600/1000000 | loss=2.5833 vocab=1.0553 token=1.1562 | 6.07 steps/s | VRAM=30.42 Go [2026-07-16 18:52:51] step 122650/1000000 | loss=2.5710 vocab=1.3891 token=1.5328 | 6.11 steps/s | VRAM=29.13 Go [2026-07-16 18:53:00] step 122700/1000000 | loss=2.5643 vocab=1.1011 token=1.4704 | 6.08 steps/s | VRAM=25.20 Go [2026-07-16 18:53:08] step 122750/1000000 | loss=2.4631 vocab=0.7817 token=1.6071 | 6.11 steps/s | VRAM=29.28 Go [2026-07-16 18:53:16] step 122800/1000000 | loss=2.5469 vocab=1.5503 token=1.7322 | 6.11 steps/s | VRAM=32.77 Go [2026-07-16 18:53:24] step 122850/1000000 | loss=2.5580 vocab=1.1803 token=1.5729 | 6.12 steps/s | VRAM=25.22 Go [2026-07-16 18:53:32] step 122900/1000000 | loss=2.5276 vocab=0.6931 token=1.4544 | 6.14 steps/s | VRAM=32.12 Go [2026-07-16 18:53:40] step 122950/1000000 | loss=2.5086 vocab=1.2098 token=1.9732 | 6.11 steps/s | VRAM=25.72 Go [2026-07-16 18:53:49] step 123000/1000000 | loss=2.5239 vocab=1.4301 token=1.7030 | 6.13 steps/s | VRAM=24.85 Go [2026-07-16 18:53:57] step 123050/1000000 | loss=2.4774 vocab=1.0396 token=1.6064 | 6.11 steps/s | VRAM=28.47 Go [2026-07-16 18:54:05] step 123100/1000000 | loss=2.4821 vocab=0.9850 token=1.3836 | 6.12 steps/s | VRAM=28.34 Go [2026-07-16 18:54:13] step 123150/1000000 | loss=2.4116 vocab=0.8819 token=1.2559 | 6.11 steps/s | VRAM=25.23 Go [2026-07-16 18:54:21] step 123200/1000000 | loss=2.4831 vocab=1.1051 token=1.4319 | 6.07 steps/s | VRAM=26.56 Go [2026-07-16 18:54:30] step 123250/1000000 | loss=2.5325 vocab=1.0164 token=1.3873 | 6.11 steps/s | VRAM=29.11 Go [2026-07-16 18:54:38] step 123300/1000000 | loss=2.4796 vocab=1.3816 token=1.4037 | 6.13 steps/s | VRAM=31.15 Go [2026-07-16 18:54:46] step 123350/1000000 | loss=2.4958 vocab=0.9988 token=1.5039 | 6.11 steps/s | VRAM=28.61 Go [2026-07-16 18:54:54] step 123400/1000000 | loss=2.5672 vocab=1.5137 token=1.5915 | 6.10 steps/s | VRAM=27.94 Go [2026-07-16 18:55:02] step 123450/1000000 | loss=2.5147 vocab=0.8370 token=1.3169 | 6.09 steps/s | VRAM=26.42 Go [2026-07-16 18:55:10] step 123500/1000000 | loss=2.4829 vocab=0.9482 token=1.4291 | 6.11 steps/s | VRAM=28.38 Go [2026-07-16 18:55:19] step 123550/1000000 | loss=2.5058 vocab=1.0643 token=1.7157 | 6.09 steps/s | VRAM=32.03 Go [2026-07-16 18:55:27] step 123600/1000000 | loss=2.4459 vocab=0.4677 token=1.0230 | 6.12 steps/s | VRAM=31.89 Go [2026-07-16 18:55:35] step 123650/1000000 | loss=2.2962 vocab=0.7772 token=1.1913 | 6.09 steps/s | VRAM=32.77 Go [2026-07-16 18:55:43] step 123700/1000000 | loss=2.4513 vocab=0.8330 token=1.2216 | 6.09 steps/s | VRAM=27.77 Go [2026-07-16 18:55:51] step 123750/1000000 | loss=2.4747 vocab=0.7239 token=0.9740 | 6.13 steps/s | VRAM=28.82 Go [2026-07-16 18:56:00] step 123800/1000000 | loss=2.4795 vocab=1.5962 token=1.6127 | 6.12 steps/s | VRAM=31.40 Go [2026-07-16 18:56:08] step 123850/1000000 | loss=2.4223 vocab=0.5047 token=1.3641 | 6.06 steps/s | VRAM=26.04 Go [2026-07-16 18:56:16] step 123900/1000000 | loss=2.5210 vocab=0.9613 token=1.8066 | 6.14 steps/s | VRAM=32.04 Go [2026-07-16 18:56:24] step 123950/1000000 | loss=2.4507 vocab=1.0680 token=1.4953 | 6.10 steps/s | VRAM=28.39 Go [2026-07-16 18:56:32] step 124000/1000000 | loss=2.4395 vocab=0.7128 token=1.1443 | 6.13 steps/s | VRAM=32.30 Go [2026-07-16 18:56:40] step 124050/1000000 | loss=2.3885 vocab=0.7644 token=1.4240 | 6.11 steps/s | VRAM=27.07 Go [2026-07-16 18:56:49] step 124100/1000000 | loss=2.5250 vocab=0.7604 token=1.4704 | 6.13 steps/s | VRAM=28.04 Go [2026-07-16 18:56:57] step 124150/1000000 | loss=2.5528 vocab=0.8382 token=1.1857 | 6.11 steps/s | VRAM=29.18 Go [2026-07-16 18:57:05] step 124200/1000000 | loss=2.5794 vocab=0.9582 token=1.6377 | 6.13 steps/s | VRAM=32.30 Go [2026-07-16 18:57:13] step 124250/1000000 | loss=2.4912 vocab=1.2780 token=1.5528 | 6.11 steps/s | VRAM=32.90 Go [2026-07-16 18:57:21] step 124300/1000000 | loss=2.4240 vocab=0.9029 token=1.3416 | 6.13 steps/s | VRAM=29.27 Go [2026-07-16 18:57:29] step 124350/1000000 | loss=2.4617 vocab=1.5846 token=1.5992 | 6.12 steps/s | VRAM=31.79 Go [2026-07-16 18:57:38] step 124400/1000000 | loss=2.5342 vocab=0.9871 token=1.5818 | 6.08 steps/s | VRAM=33.00 Go [2026-07-16 18:57:46] step 124450/1000000 | loss=2.3508 vocab=1.0925 token=1.5999 | 6.07 steps/s | VRAM=32.52 Go [2026-07-16 18:57:54] step 124500/1000000 | loss=2.6276 vocab=0.9643 token=1.7885 | 6.10 steps/s | VRAM=28.63 Go [2026-07-16 18:58:02] step 124550/1000000 | loss=2.5514 vocab=1.1349 token=1.3989 | 6.09 steps/s | VRAM=29.69 Go [2026-07-16 18:58:11] step 124600/1000000 | loss=2.3869 vocab=0.8197 token=1.2167 | 6.13 steps/s | VRAM=28.75 Go [2026-07-16 18:58:19] step 124650/1000000 | loss=2.4306 vocab=1.4851 token=1.3078 | 6.10 steps/s | VRAM=28.41 Go [2026-07-16 18:58:27] step 124700/1000000 | loss=2.4942 vocab=1.1835 token=1.7801 | 6.12 steps/s | VRAM=30.97 Go [2026-07-16 18:58:35] step 124750/1000000 | loss=2.5334 vocab=1.4025 token=1.3600 | 6.11 steps/s | VRAM=29.42 Go [2026-07-16 18:58:43] step 124800/1000000 | loss=2.4684 vocab=0.9271 token=1.7026 | 6.12 steps/s | VRAM=24.36 Go [2026-07-16 18:58:51] step 124850/1000000 | loss=2.5286 vocab=0.6955 token=1.2159 | 6.12 steps/s | VRAM=28.33 Go [2026-07-16 18:59:00] step 124900/1000000 | loss=2.4840 vocab=0.5880 token=1.5675 | 6.13 steps/s | VRAM=29.05 Go [2026-07-16 18:59:08] step 124950/1000000 | loss=2.5241 vocab=1.0792 token=1.2041 | 6.09 steps/s | VRAM=28.16 Go [2026-07-16 18:59:16] step 125000/1000000 | loss=2.5128 vocab=0.8601 token=1.2978 | 6.11 steps/s | VRAM=32.17 Go [2026-07-16 19:00:40] step 125050/1000000 | loss=2.5328 vocab=0.8076 token=1.5375 | 0.59 steps/s | VRAM=28.83 Go [2026-07-16 19:00:48] step 125100/1000000 | loss=2.4115 vocab=1.2039 token=1.4482 | 6.17 steps/s | VRAM=28.31 Go [2026-07-16 19:00:56] step 125150/1000000 | loss=2.3511 vocab=0.7832 token=1.2918 | 6.12 steps/s | VRAM=28.63 Go [2026-07-16 19:01:04] step 125200/1000000 | loss=2.5822 vocab=0.9631 token=1.5556 | 6.20 steps/s | VRAM=29.05 Go [2026-07-16 19:01:13] step 125250/1000000 | loss=2.5079 vocab=0.8014 token=1.4307 | 6.16 steps/s | VRAM=32.13 Go [2026-07-16 19:01:21] step 125300/1000000 | loss=2.4830 vocab=1.0506 token=1.6595 | 6.14 steps/s | VRAM=29.61 Go [2026-07-16 19:01:29] step 125350/1000000 | loss=2.5320 vocab=0.6599 token=1.5531 | 6.15 steps/s | VRAM=32.25 Go [2026-07-16 19:01:37] step 125400/1000000 | loss=2.5088 vocab=0.8245 token=1.6744 | 6.12 steps/s | VRAM=25.71 Go [2026-07-16 19:01:45] step 125450/1000000 | loss=2.3386 vocab=1.0613 token=1.4500 | 6.12 steps/s | VRAM=28.74 Go [2026-07-16 19:01:53] step 125500/1000000 | loss=2.5151 vocab=0.5103 token=0.9183 | 6.15 steps/s | VRAM=29.55 Go [2026-07-16 19:02:01] step 125550/1000000 | loss=2.4325 vocab=1.0134 token=1.1805 | 6.09 steps/s | VRAM=25.25 Go [2026-07-16 19:02:10] step 125600/1000000 | loss=2.5731 vocab=0.6157 token=1.3467 | 6.12 steps/s | VRAM=28.30 Go [2026-07-16 19:02:18] step 125650/1000000 | loss=2.6175 vocab=0.9737 token=1.4515 | 6.12 steps/s | VRAM=28.03 Go [2026-07-16 19:02:26] step 125700/1000000 | loss=2.6644 vocab=1.4148 token=1.5989 | 6.14 steps/s | VRAM=28.31 Go [2026-07-16 19:02:34] step 125750/1000000 | loss=2.4385 vocab=0.9040 token=1.6219 | 6.11 steps/s | VRAM=32.41 Go [2026-07-16 19:02:42] step 125800/1000000 | loss=2.4988 vocab=0.8875 token=1.3377 | 6.11 steps/s | VRAM=28.18 Go [2026-07-16 19:02:51] step 125850/1000000 | loss=2.4799 vocab=0.6812 token=1.1274 | 6.10 steps/s | VRAM=28.50 Go [2026-07-16 19:02:59] step 125900/1000000 | loss=2.5546 vocab=0.6702 token=1.1542 | 6.11 steps/s | VRAM=25.76 Go [2026-07-16 19:03:07] step 125950/1000000 | loss=2.4282 vocab=0.5353 token=1.4709 | 6.12 steps/s | VRAM=29.69 Go [2026-07-16 19:03:15] step 126000/1000000 | loss=2.3780 vocab=1.2257 token=1.6591 | 6.12 steps/s | VRAM=28.91 Go [2026-07-16 19:03:23] step 126050/1000000 | loss=2.3641 vocab=1.2367 token=1.7435 | 6.11 steps/s | VRAM=28.16 Go [2026-07-16 19:03:31] step 126100/1000000 | loss=2.4550 vocab=0.9385 token=1.3638 | 6.13 steps/s | VRAM=28.52 Go [2026-07-16 19:03:40] step 126150/1000000 | loss=2.4602 vocab=1.1578 token=1.5707 | 6.14 steps/s | VRAM=31.79 Go [2026-07-16 19:03:48] step 126200/1000000 | loss=2.5360 vocab=1.1089 token=1.1596 | 6.11 steps/s | VRAM=32.42 Go [2026-07-16 19:03:56] step 126250/1000000 | loss=2.4368 vocab=1.2017 token=1.4556 | 6.11 steps/s | VRAM=28.18 Go [2026-07-16 19:04:04] step 126300/1000000 | loss=2.5184 vocab=0.7118 token=1.6715 | 6.11 steps/s | VRAM=32.88 Go [2026-07-16 19:04:12] step 126350/1000000 | loss=2.5005 vocab=0.8697 token=1.6189 | 6.11 steps/s | VRAM=32.24 Go [2026-07-16 19:04:20] step 126400/1000000 | loss=2.4355 vocab=0.8460 token=1.6555 | 6.12 steps/s | VRAM=24.88 Go [2026-07-16 19:04:29] step 126450/1000000 | loss=2.4466 vocab=1.1530 token=1.1760 | 6.09 steps/s | VRAM=29.26 Go [2026-07-16 19:04:37] step 126500/1000000 | loss=2.4937 vocab=0.8622 token=1.5573 | 6.11 steps/s | VRAM=25.95 Go [2026-07-16 19:04:45] step 126550/1000000 | loss=2.4516 vocab=1.2051 token=1.6439 | 6.08 steps/s | VRAM=29.28 Go [2026-07-16 19:04:53] step 126600/1000000 | loss=2.4640 vocab=1.1761 token=1.7416 | 6.10 steps/s | VRAM=31.31 Go [2026-07-16 19:05:01] step 126650/1000000 | loss=2.4433 vocab=0.6866 token=1.4274 | 6.11 steps/s | VRAM=31.64 Go [2026-07-16 19:05:10] step 126700/1000000 | loss=2.5121 vocab=1.0413 token=1.6448 | 6.08 steps/s | VRAM=26.84 Go [2026-07-16 19:05:18] step 126750/1000000 | loss=2.4756 vocab=0.8876 token=1.1124 | 6.09 steps/s | VRAM=25.29 Go [2026-07-16 19:05:26] step 126800/1000000 | loss=2.4856 vocab=0.9162 token=1.5548 | 6.09 steps/s | VRAM=28.31 Go [2026-07-16 19:05:34] step 126850/1000000 | loss=2.3222 vocab=0.9286 token=1.2720 | 6.10 steps/s | VRAM=32.18 Go [2026-07-16 19:05:42] step 126900/1000000 | loss=2.4889 vocab=1.1124 token=1.3476 | 6.10 steps/s | VRAM=28.84 Go [2026-07-16 19:05:51] step 126950/1000000 | loss=2.4377 vocab=0.7369 token=1.3662 | 6.13 steps/s | VRAM=28.01 Go [2026-07-16 19:05:59] step 127000/1000000 | loss=2.5533 vocab=0.7384 token=1.5007 | 6.14 steps/s | VRAM=32.00 Go [2026-07-16 19:06:07] step 127050/1000000 | loss=2.5035 vocab=1.0982 token=1.3489 | 6.11 steps/s | VRAM=31.98 Go [2026-07-16 19:06:15] step 127100/1000000 | loss=2.3693 vocab=0.9014 token=1.5969 | 6.12 steps/s | VRAM=31.84 Go [2026-07-16 19:06:23] step 127150/1000000 | loss=2.4886 vocab=0.9773 token=1.1790 | 6.07 steps/s | VRAM=28.70 Go [2026-07-16 19:06:32] step 127200/1000000 | loss=2.4564 vocab=0.9578 token=1.3141 | 6.12 steps/s | VRAM=32.06 Go [2026-07-16 19:06:40] step 127250/1000000 | loss=2.4489 vocab=1.2265 token=1.8533 | 6.12 steps/s | VRAM=31.44 Go [2026-07-16 19:06:48] step 127300/1000000 | loss=2.4970 vocab=1.3117 token=1.2348 | 6.09 steps/s | VRAM=25.85 Go [2026-07-16 19:06:56] step 127350/1000000 | loss=2.4824 vocab=0.8822 token=1.5076 | 6.13 steps/s | VRAM=28.75 Go [2026-07-16 19:07:04] step 127400/1000000 | loss=2.4444 vocab=0.9122 token=1.6860 | 6.11 steps/s | VRAM=29.14 Go [2026-07-16 19:07:12] step 127450/1000000 | loss=2.4476 vocab=1.1584 token=1.7380 | 6.11 steps/s | VRAM=29.42 Go [2026-07-16 19:07:21] step 127500/1000000 | loss=2.5194 vocab=1.9515 token=1.7175 | 6.12 steps/s | VRAM=32.60 Go [2026-07-16 19:07:29] step 127550/1000000 | loss=2.4190 vocab=1.1251 token=1.7135 | 6.11 steps/s | VRAM=25.34 Go [2026-07-16 19:07:37] step 127600/1000000 | loss=2.5105 vocab=1.4052 token=1.1971 | 6.08 steps/s | VRAM=27.03 Go [2026-07-16 19:07:45] step 127650/1000000 | loss=2.5147 vocab=0.8615 token=1.6882 | 6.14 steps/s | VRAM=31.67 Go [2026-07-16 19:07:53] step 127700/1000000 | loss=2.4736 vocab=1.5071 token=1.4711 | 6.11 steps/s | VRAM=30.09 Go [2026-07-16 19:08:02] step 127750/1000000 | loss=2.4655 vocab=0.8596 token=1.4134 | 6.11 steps/s | VRAM=28.49 Go [2026-07-16 19:08:10] step 127800/1000000 | loss=2.5218 vocab=1.0995 token=1.4366 | 6.13 steps/s | VRAM=31.91 Go [2026-07-16 19:08:18] step 127850/1000000 | loss=2.4231 vocab=1.0935 token=1.6988 | 6.11 steps/s | VRAM=28.22 Go [2026-07-16 19:08:26] step 127900/1000000 | loss=2.3911 vocab=1.1014 token=1.3865 | 6.09 steps/s | VRAM=26.65 Go [2026-07-16 19:08:34] step 127950/1000000 | loss=2.4729 vocab=0.9875 token=1.5290 | 6.11 steps/s | VRAM=28.68 Go [2026-07-16 19:08:42] step 128000/1000000 | loss=2.6635 vocab=0.8146 token=1.0264 | 6.14 steps/s | VRAM=32.07 Go [2026-07-16 19:08:51] step 128050/1000000 | loss=2.5173 vocab=0.8239 token=1.2326 | 6.07 steps/s | VRAM=29.42 Go [2026-07-16 19:08:59] step 128100/1000000 | loss=2.3333 vocab=0.9661 token=1.2936 | 6.14 steps/s | VRAM=28.12 Go [2026-07-16 19:09:07] step 128150/1000000 | loss=2.5365 vocab=0.6322 token=1.4623 | 6.10 steps/s | VRAM=32.14 Go [2026-07-16 19:09:15] step 128200/1000000 | loss=2.3887 vocab=0.9515 token=1.6036 | 6.12 steps/s | VRAM=28.82 Go [2026-07-16 19:09:23] step 128250/1000000 | loss=2.5281 vocab=1.0384 token=1.3076 | 6.09 steps/s | VRAM=27.06 Go [2026-07-16 19:09:32] step 128300/1000000 | loss=2.4082 vocab=0.8095 token=1.6096 | 6.13 steps/s | VRAM=27.93 Go [2026-07-16 19:09:40] step 128350/1000000 | loss=2.4704 vocab=0.9000 token=1.2240 | 6.09 steps/s | VRAM=26.96 Go [2026-07-16 19:09:48] step 128400/1000000 | loss=2.4226 vocab=1.0395 token=1.5599 | 6.12 steps/s | VRAM=24.64 Go [2026-07-16 19:09:56] step 128450/1000000 | loss=2.5476 vocab=1.2076 token=1.4111 | 6.14 steps/s | VRAM=24.98 Go [2026-07-16 19:10:04] step 128500/1000000 | loss=2.5136 vocab=1.1804 token=1.5492 | 6.12 steps/s | VRAM=28.52 Go [2026-07-16 19:10:12] step 128550/1000000 | loss=2.4284 vocab=1.1580 token=1.2642 | 6.12 steps/s | VRAM=29.20 Go [2026-07-16 19:10:21] step 128600/1000000 | loss=2.3954 vocab=0.6342 token=1.0067 | 6.13 steps/s | VRAM=32.54 Go [2026-07-16 19:10:29] step 128650/1000000 | loss=2.5226 vocab=1.0227 token=1.7403 | 6.13 steps/s | VRAM=28.73 Go [2026-07-16 19:10:37] step 128700/1000000 | loss=2.4287 vocab=1.1258 token=1.6721 | 6.13 steps/s | VRAM=25.34 Go [2026-07-16 19:10:45] step 128750/1000000 | loss=2.5525 vocab=1.2034 token=2.2392 | 6.14 steps/s | VRAM=26.02 Go [2026-07-16 19:10:53] step 128800/1000000 | loss=2.4638 vocab=1.0696 token=1.4126 | 6.12 steps/s | VRAM=27.93 Go [2026-07-16 19:11:01] step 128850/1000000 | loss=2.4953 vocab=0.7744 token=1.4442 | 6.11 steps/s | VRAM=28.88 Go [2026-07-16 19:11:10] step 128900/1000000 | loss=2.5159 vocab=0.8275 token=1.5718 | 6.14 steps/s | VRAM=28.49 Go [2026-07-16 19:11:18] step 128950/1000000 | loss=2.4924 vocab=1.1704 token=1.8542 | 6.12 steps/s | VRAM=32.60 Go [2026-07-16 19:11:26] step 129000/1000000 | loss=2.4470 vocab=1.1374 token=1.4980 | 6.13 steps/s | VRAM=28.28 Go [2026-07-16 19:11:34] step 129050/1000000 | loss=2.5719 vocab=1.6576 token=1.5833 | 6.16 steps/s | VRAM=27.45 Go [2026-07-16 19:11:42] step 129100/1000000 | loss=2.3242 vocab=0.9910 token=1.6154 | 6.15 steps/s | VRAM=26.32 Go [2026-07-16 19:11:50] step 129150/1000000 | loss=2.5106 vocab=1.3132 token=1.5840 | 6.15 steps/s | VRAM=32.46 Go [2026-07-16 19:11:58] step 129200/1000000 | loss=2.4623 vocab=1.1313 token=1.3818 | 6.12 steps/s | VRAM=25.81 Go [2026-07-16 19:12:07] step 129250/1000000 | loss=2.5792 vocab=0.6821 token=1.3847 | 6.13 steps/s | VRAM=30.17 Go [2026-07-16 19:12:15] step 129300/1000000 | loss=2.4607 vocab=0.8798 token=1.4239 | 6.13 steps/s | VRAM=26.62 Go [2026-07-16 19:12:23] step 129350/1000000 | loss=2.3496 vocab=1.2945 token=1.8430 | 6.15 steps/s | VRAM=31.94 Go [2026-07-16 19:12:31] step 129400/1000000 | loss=2.4483 vocab=1.0829 token=1.5880 | 6.10 steps/s | VRAM=29.19 Go [2026-07-16 19:12:39] step 129450/1000000 | loss=2.3788 vocab=0.7824 token=1.5554 | 6.14 steps/s | VRAM=31.61 Go [2026-07-16 19:12:47] step 129500/1000000 | loss=2.3706 vocab=1.4248 token=1.6005 | 6.14 steps/s | VRAM=28.12 Go [2026-07-16 19:12:56] step 129550/1000000 | loss=2.5760 vocab=1.1422 token=1.7261 | 6.12 steps/s | VRAM=31.20 Go [2026-07-16 19:13:04] step 129600/1000000 | loss=2.5389 vocab=1.2386 token=1.3887 | 6.13 steps/s | VRAM=32.28 Go [2026-07-16 19:13:12] step 129650/1000000 | loss=2.4228 vocab=1.4321 token=1.6219 | 6.11 steps/s | VRAM=32.48 Go [2026-07-16 19:13:20] step 129700/1000000 | loss=2.4559 vocab=0.7592 token=0.7833 | 6.12 steps/s | VRAM=28.80 Go [2026-07-16 19:13:28] step 129750/1000000 | loss=2.4837 vocab=0.9088 token=1.6234 | 6.10 steps/s | VRAM=28.89 Go [2026-07-16 19:13:36] step 129800/1000000 | loss=2.4134 vocab=0.8503 token=1.3583 | 6.13 steps/s | VRAM=31.69 Go [2026-07-16 19:13:45] step 129850/1000000 | loss=2.5100 vocab=1.0043 token=1.1245 | 6.10 steps/s | VRAM=26.42 Go [2026-07-16 19:13:53] step 129900/1000000 | loss=2.4417 vocab=0.5867 token=1.4777 | 6.09 steps/s | VRAM=25.85 Go [2026-07-16 19:14:01] step 129950/1000000 | loss=2.5451 vocab=0.8890 token=1.1875 | 6.14 steps/s | VRAM=32.34 Go [2026-07-16 19:14:09] step 130000/1000000 | loss=2.5312 vocab=0.9960 token=1.3371 | 6.12 steps/s | VRAM=32.89 Go [2026-07-16 19:15:28] step 130050/1000000 | loss=2.4899 vocab=0.8428 token=1.5515 | 0.63 steps/s | VRAM=32.64 Go [2026-07-16 19:15:36] step 130100/1000000 | loss=2.4260 vocab=1.1068 token=1.6021 | 6.17 steps/s | VRAM=32.56 Go [2026-07-16 19:15:44] step 130150/1000000 | loss=2.4092 vocab=1.2527 token=1.6385 | 6.17 steps/s | VRAM=28.42 Go [2026-07-16 19:15:52] step 130200/1000000 | loss=2.4068 vocab=1.1931 token=1.4845 | 6.16 steps/s | VRAM=29.40 Go [2026-07-16 19:16:01] step 130250/1000000 | loss=2.5248 vocab=0.9392 token=1.4845 | 6.15 steps/s | VRAM=28.41 Go [2026-07-16 19:16:09] step 130300/1000000 | loss=2.4672 vocab=0.7497 token=1.5439 | 6.16 steps/s | VRAM=28.97 Go [2026-07-16 19:16:17] step 130350/1000000 | loss=2.4732 vocab=0.7376 token=1.0967 | 6.15 steps/s | VRAM=31.46 Go [2026-07-16 19:16:25] step 130400/1000000 | loss=2.5090 vocab=0.9749 token=1.2610 | 6.13 steps/s | VRAM=28.74 Go [2026-07-16 19:16:33] step 130450/1000000 | loss=2.4467 vocab=1.1346 token=1.3464 | 6.10 steps/s | VRAM=29.39 Go [2026-07-16 19:16:41] step 130500/1000000 | loss=2.4504 vocab=1.0951 token=1.4319 | 6.12 steps/s | VRAM=29.58 Go [2026-07-16 19:16:50] step 130550/1000000 | loss=2.3950 vocab=0.9197 token=1.1381 | 6.12 steps/s | VRAM=28.46 Go [2026-07-16 19:16:58] step 130600/1000000 | loss=2.4684 vocab=1.3550 token=1.4991 | 6.11 steps/s | VRAM=28.48 Go [2026-07-16 19:17:06] step 130650/1000000 | loss=2.5446 vocab=1.0612 token=1.4685 | 6.10 steps/s | VRAM=27.92 Go [2026-07-16 19:17:14] step 130700/1000000 | loss=2.5163 vocab=0.7826 token=1.5256 | 6.11 steps/s | VRAM=28.99 Go [2026-07-16 19:17:22] step 130750/1000000 | loss=2.4112 vocab=1.0194 token=1.4326 | 6.10 steps/s | VRAM=26.64 Go [2026-07-16 19:17:30] step 130800/1000000 | loss=2.4386 vocab=1.1554 token=1.5800 | 6.10 steps/s | VRAM=29.14 Go [2026-07-16 19:17:39] step 130850/1000000 | loss=2.3987 vocab=0.8555 token=1.4466 | 6.08 steps/s | VRAM=32.93 Go [2026-07-16 19:17:47] step 130900/1000000 | loss=2.4161 vocab=1.0549 token=1.6472 | 6.08 steps/s | VRAM=28.34 Go [2026-07-16 19:17:55] step 130950/1000000 | loss=2.5752 vocab=1.2855 token=1.6830 | 6.11 steps/s | VRAM=32.39 Go [2026-07-16 19:18:03] step 131000/1000000 | loss=2.4375 vocab=1.1350 token=1.4251 | 6.09 steps/s | VRAM=29.48 Go [2026-07-16 19:18:12] step 131050/1000000 | loss=2.5832 vocab=1.3917 token=1.7884 | 6.10 steps/s | VRAM=29.07 Go [2026-07-16 19:18:20] step 131100/1000000 | loss=2.5721 vocab=1.2315 token=1.7051 | 6.12 steps/s | VRAM=31.32 Go [2026-07-16 19:18:28] step 131150/1000000 | loss=2.5267 vocab=1.1623 token=1.7009 | 6.06 steps/s | VRAM=30.09 Go [2026-07-16 19:18:36] step 131200/1000000 | loss=2.5237 vocab=1.0965 token=1.3877 | 6.11 steps/s | VRAM=31.73 Go [2026-07-16 19:18:44] step 131250/1000000 | loss=2.5281 vocab=0.6945 token=0.9592 | 6.09 steps/s | VRAM=26.44 Go [2026-07-16 19:18:52] step 131300/1000000 | loss=2.4661 vocab=0.8022 token=1.6188 | 6.13 steps/s | VRAM=32.31 Go [2026-07-16 19:19:01] step 131350/1000000 | loss=2.5053 vocab=0.8600 token=1.4659 | 6.12 steps/s | VRAM=32.64 Go [2026-07-16 19:19:09] step 131400/1000000 | loss=2.5186 vocab=0.8025 token=1.3253 | 6.13 steps/s | VRAM=28.60 Go [2026-07-16 19:19:17] step 131450/1000000 | loss=2.3685 vocab=0.9651 token=2.0580 | 6.10 steps/s | VRAM=25.47 Go [2026-07-16 19:19:25] step 131500/1000000 | loss=2.4657 vocab=0.7532 token=0.8185 | 6.10 steps/s | VRAM=24.79 Go [2026-07-16 19:19:33] step 131550/1000000 | loss=2.3782 vocab=0.6903 token=1.6462 | 6.12 steps/s | VRAM=25.88 Go [2026-07-16 19:19:42] step 131600/1000000 | loss=2.5124 vocab=1.2828 token=1.6650 | 6.09 steps/s | VRAM=28.51 Go [2026-07-16 19:19:50] step 131650/1000000 | loss=2.4091 vocab=0.8451 token=1.1585 | 6.08 steps/s | VRAM=31.25 Go [2026-07-16 19:19:58] step 131700/1000000 | loss=2.5153 vocab=1.0101 token=1.5116 | 6.10 steps/s | VRAM=26.43 Go [2026-07-16 19:20:06] step 131750/1000000 | loss=2.4489 vocab=1.2482 token=1.3746 | 6.11 steps/s | VRAM=28.93 Go [2026-07-16 19:20:14] step 131800/1000000 | loss=2.4160 vocab=0.9376 token=1.2805 | 6.12 steps/s | VRAM=32.23 Go [2026-07-16 19:20:23] step 131850/1000000 | loss=2.5166 vocab=1.0478 token=1.7082 | 6.09 steps/s | VRAM=27.60 Go [2026-07-16 19:20:31] step 131900/1000000 | loss=2.4818 vocab=1.1167 token=1.4288 | 6.13 steps/s | VRAM=28.57 Go [2026-07-16 19:20:39] step 131950/1000000 | loss=2.4449 vocab=0.8523 token=1.2756 | 6.14 steps/s | VRAM=30.52 Go [2026-07-16 19:20:47] step 132000/1000000 | loss=2.6124 vocab=0.8795 token=1.9496 | 6.09 steps/s | VRAM=28.98 Go [2026-07-16 19:20:55] step 132050/1000000 | loss=2.4958 vocab=1.0202 token=1.1014 | 6.07 steps/s | VRAM=32.95 Go [2026-07-16 19:21:04] step 132100/1000000 | loss=2.3866 vocab=0.9944 token=1.7680 | 6.11 steps/s | VRAM=28.59 Go [2026-07-16 19:21:12] step 132150/1000000 | loss=2.5552 vocab=1.0818 token=1.4087 | 6.10 steps/s | VRAM=32.15 Go [2026-07-16 19:21:20] step 132200/1000000 | loss=2.3991 vocab=0.6823 token=1.1324 | 6.11 steps/s | VRAM=28.69 Go [2026-07-16 19:21:28] step 132250/1000000 | loss=2.4967 vocab=0.8744 token=1.0890 | 6.11 steps/s | VRAM=26.34 Go [2026-07-16 19:21:36] step 132300/1000000 | loss=2.5261 vocab=0.5957 token=1.5774 | 6.08 steps/s | VRAM=26.08 Go [2026-07-16 19:21:45] step 132350/1000000 | loss=2.4447 vocab=0.9694 token=1.3001 | 6.08 steps/s | VRAM=26.16 Go [2026-07-16 19:21:53] step 132400/1000000 | loss=2.4506 vocab=0.8777 token=1.2515 | 6.11 steps/s | VRAM=32.09 Go [2026-07-16 19:22:01] step 132450/1000000 | loss=2.5377 vocab=1.0371 token=1.0069 | 6.15 steps/s | VRAM=31.75 Go [2026-07-16 19:22:09] step 132500/1000000 | loss=2.4503 vocab=0.9446 token=1.3701 | 6.11 steps/s | VRAM=26.86 Go [2026-07-16 19:22:17] step 132550/1000000 | loss=2.5337 vocab=1.1159 token=1.6257 | 6.12 steps/s | VRAM=29.31 Go [2026-07-16 19:22:25] step 132600/1000000 | loss=2.3369 vocab=1.0119 token=1.2976 | 6.13 steps/s | VRAM=32.61 Go [2026-07-16 19:22:34] step 132650/1000000 | loss=2.4980 vocab=0.7382 token=1.7527 | 6.09 steps/s | VRAM=26.44 Go [2026-07-16 19:22:42] step 132700/1000000 | loss=2.4181 vocab=1.1680 token=1.7237 | 6.12 steps/s | VRAM=30.00 Go [2026-07-16 19:22:50] step 132750/1000000 | loss=2.5085 vocab=1.0865 token=1.0484 | 6.10 steps/s | VRAM=25.30 Go [2026-07-16 19:22:58] step 132800/1000000 | loss=2.4922 vocab=1.7029 token=2.1198 | 6.11 steps/s | VRAM=28.10 Go [2026-07-16 19:23:06] step 132850/1000000 | loss=2.6187 vocab=0.7700 token=1.4801 | 6.10 steps/s | VRAM=29.23 Go [2026-07-16 19:23:14] step 132900/1000000 | loss=2.3455 vocab=1.5979 token=1.6249 | 6.11 steps/s | VRAM=28.06 Go [2026-07-16 19:23:23] step 132950/1000000 | loss=2.4744 vocab=0.8368 token=1.3866 | 6.10 steps/s | VRAM=32.17 Go [2026-07-16 19:23:31] step 133000/1000000 | loss=2.4455 vocab=0.6337 token=1.3597 | 6.09 steps/s | VRAM=29.66 Go [2026-07-16 19:23:39] step 133050/1000000 | loss=2.3925 vocab=1.3213 token=1.5275 | 6.07 steps/s | VRAM=26.37 Go [2026-07-16 19:23:47] step 133100/1000000 | loss=2.3874 vocab=0.9173 token=1.7348 | 6.14 steps/s | VRAM=28.84 Go [2026-07-16 19:23:55] step 133150/1000000 | loss=2.4778 vocab=1.1269 token=1.4089 | 6.13 steps/s | VRAM=28.32 Go [2026-07-16 19:24:04] step 133200/1000000 | loss=2.4519 vocab=0.8947 token=1.7001 | 6.13 steps/s | VRAM=31.41 Go [2026-07-16 19:24:12] step 133250/1000000 | loss=2.4799 vocab=1.4125 token=1.6432 | 6.11 steps/s | VRAM=28.00 Go [2026-07-16 19:24:20] step 133300/1000000 | loss=2.5525 vocab=0.8293 token=1.5776 | 6.10 steps/s | VRAM=29.07 Go [2026-07-16 19:24:28] step 133350/1000000 | loss=2.4723 vocab=0.8920 token=0.8084 | 6.12 steps/s | VRAM=28.54 Go [2026-07-16 19:24:36] step 133400/1000000 | loss=2.3454 vocab=1.1902 token=1.4736 | 6.12 steps/s | VRAM=26.06 Go [2026-07-16 19:24:45] step 133450/1000000 | loss=2.4974 vocab=1.1348 token=1.7481 | 6.10 steps/s | VRAM=29.14 Go [2026-07-16 19:24:53] step 133500/1000000 | loss=2.4820 vocab=1.1602 token=1.4996 | 6.08 steps/s | VRAM=30.14 Go [2026-07-16 19:25:01] step 133550/1000000 | loss=2.4468 vocab=1.4768 token=1.7433 | 6.09 steps/s | VRAM=28.98 Go [2026-07-16 19:25:09] step 133600/1000000 | loss=2.5928 vocab=1.0413 token=1.3056 | 6.11 steps/s | VRAM=28.39 Go [2026-07-16 19:25:17] step 133650/1000000 | loss=2.4899 vocab=0.8434 token=1.4348 | 6.10 steps/s | VRAM=29.67 Go [2026-07-16 19:25:26] step 133700/1000000 | loss=2.5552 vocab=1.0970 token=1.4749 | 6.10 steps/s | VRAM=26.18 Go [2026-07-16 19:25:34] step 133750/1000000 | loss=2.4842 vocab=1.1154 token=1.4973 | 6.14 steps/s | VRAM=27.63 Go [2026-07-16 19:25:42] step 133800/1000000 | loss=2.5589 vocab=0.8926 token=1.7266 | 6.09 steps/s | VRAM=26.05 Go [2026-07-16 19:25:50] step 133850/1000000 | loss=2.4722 vocab=0.6301 token=1.6664 | 6.13 steps/s | VRAM=25.38 Go [2026-07-16 19:25:58] step 133900/1000000 | loss=2.4752 vocab=0.8127 token=1.1875 | 6.14 steps/s | VRAM=28.51 Go [2026-07-16 19:26:06] step 133950/1000000 | loss=2.4193 vocab=0.9550 token=1.7122 | 6.13 steps/s | VRAM=28.70 Go [2026-07-16 19:26:15] step 134000/1000000 | loss=2.5808 vocab=0.7236 token=1.5930 | 6.13 steps/s | VRAM=28.49 Go [2026-07-16 19:26:23] step 134050/1000000 | loss=2.4222 vocab=0.8648 token=1.3857 | 6.11 steps/s | VRAM=28.80 Go [2026-07-16 19:26:31] step 134100/1000000 | loss=2.4795 vocab=1.1451 token=1.6013 | 6.09 steps/s | VRAM=30.31 Go [2026-07-16 19:26:39] step 134150/1000000 | loss=2.4729 vocab=1.2155 token=1.0909 | 6.10 steps/s | VRAM=29.09 Go [2026-07-16 19:26:47] step 134200/1000000 | loss=2.3844 vocab=1.3141 token=1.6402 | 6.11 steps/s | VRAM=29.87 Go [2026-07-16 19:26:55] step 134250/1000000 | loss=2.5361 vocab=0.8640 token=1.0154 | 6.12 steps/s | VRAM=32.32 Go [2026-07-16 19:27:04] step 134300/1000000 | loss=2.4989 vocab=0.9042 token=1.0583 | 6.11 steps/s | VRAM=27.84 Go [2026-07-16 19:27:12] step 134350/1000000 | loss=2.4701 vocab=1.2062 token=1.5402 | 6.10 steps/s | VRAM=28.28 Go [2026-07-16 19:27:20] step 134400/1000000 | loss=2.4078 vocab=0.7232 token=1.3381 | 6.09 steps/s | VRAM=30.63 Go [2026-07-16 19:27:28] step 134450/1000000 | loss=2.5282 vocab=1.5591 token=1.9862 | 6.12 steps/s | VRAM=26.27 Go [2026-07-16 19:27:36] step 134500/1000000 | loss=2.5624 vocab=0.9705 token=1.3458 | 6.13 steps/s | VRAM=31.68 Go [2026-07-16 19:27:45] step 134550/1000000 | loss=2.4210 vocab=0.8626 token=1.1299 | 6.11 steps/s | VRAM=29.33 Go [2026-07-16 19:27:53] step 134600/1000000 | loss=2.3653 vocab=1.0671 token=1.7098 | 6.14 steps/s | VRAM=31.72 Go [2026-07-16 19:28:01] step 134650/1000000 | loss=2.4476 vocab=0.7589 token=1.2922 | 6.11 steps/s | VRAM=28.59 Go [2026-07-16 19:28:09] step 134700/1000000 | loss=2.4712 vocab=1.0693 token=1.4996 | 6.11 steps/s | VRAM=25.65 Go [2026-07-16 19:28:17] step 134750/1000000 | loss=2.5319 vocab=0.9313 token=1.1205 | 6.13 steps/s | VRAM=26.28 Go [2026-07-16 19:28:25] step 134800/1000000 | loss=2.4634 vocab=1.1532 token=1.4685 | 6.11 steps/s | VRAM=29.17 Go [2026-07-16 19:28:34] step 134850/1000000 | loss=2.4083 vocab=1.4582 token=1.5673 | 6.12 steps/s | VRAM=32.64 Go [2026-07-16 19:28:42] step 134900/1000000 | loss=2.5643 vocab=0.8790 token=1.4627 | 6.09 steps/s | VRAM=32.02 Go [2026-07-16 19:28:50] step 134950/1000000 | loss=2.5315 vocab=0.9065 token=1.7139 | 6.11 steps/s | VRAM=28.84 Go [2026-07-16 19:28:58] step 135000/1000000 | loss=2.4278 vocab=1.2438 token=1.1484 | 6.05 steps/s | VRAM=26.89 Go [2026-07-16 19:30:19] step 135050/1000000 | loss=2.4463 vocab=1.0403 token=1.2245 | 0.62 steps/s | VRAM=28.04 Go [2026-07-16 19:30:27] step 135100/1000000 | loss=2.5279 vocab=0.4371 token=1.0027 | 6.14 steps/s | VRAM=23.49 Go [2026-07-16 19:30:35] step 135150/1000000 | loss=2.3882 vocab=1.4634 token=1.7676 | 6.19 steps/s | VRAM=25.48 Go [2026-07-16 19:30:43] step 135200/1000000 | loss=2.5215 vocab=1.0980 token=1.3703 | 6.19 steps/s | VRAM=27.76 Go [2026-07-16 19:30:51] step 135250/1000000 | loss=2.4102 vocab=0.7035 token=1.4398 | 6.17 steps/s | VRAM=29.08 Go [2026-07-16 19:30:59] step 135300/1000000 | loss=2.6338 vocab=0.6729 token=1.4216 | 6.13 steps/s | VRAM=32.35 Go [2026-07-16 19:31:08] step 135350/1000000 | loss=2.4903 vocab=0.9329 token=1.0336 | 6.15 steps/s | VRAM=31.59 Go [2026-07-16 19:31:16] step 135400/1000000 | loss=2.4807 vocab=0.9048 token=1.1788 | 6.16 steps/s | VRAM=28.91 Go [2026-07-16 19:31:24] step 135450/1000000 | loss=2.4169 vocab=0.6640 token=1.4296 | 6.13 steps/s | VRAM=28.42 Go [2026-07-16 19:31:32] step 135500/1000000 | loss=2.3737 vocab=0.9111 token=1.2920 | 6.10 steps/s | VRAM=26.32 Go [2026-07-16 19:31:40] step 135550/1000000 | loss=2.3477 vocab=0.6990 token=1.1275 | 6.11 steps/s | VRAM=29.10 Go [2026-07-16 19:31:48] step 135600/1000000 | loss=2.6255 vocab=1.0005 token=1.4694 | 6.12 steps/s | VRAM=26.34 Go [2026-07-16 19:31:57] step 135650/1000000 | loss=2.4082 vocab=0.8315 token=1.4061 | 6.11 steps/s | VRAM=28.54 Go [2026-07-16 19:32:05] step 135700/1000000 | loss=2.4954 vocab=0.6574 token=1.1119 | 6.13 steps/s | VRAM=27.89 Go [2026-07-16 19:32:13] step 135750/1000000 | loss=2.4857 vocab=1.3747 token=1.5522 | 6.11 steps/s | VRAM=28.02 Go [2026-07-16 19:32:21] step 135800/1000000 | loss=2.4384 vocab=0.8988 token=1.2979 | 6.13 steps/s | VRAM=29.33 Go [2026-07-16 19:32:29] step 135850/1000000 | loss=2.4403 vocab=0.8380 token=1.3661 | 6.15 steps/s | VRAM=31.63 Go [2026-07-16 19:32:37] step 135900/1000000 | loss=2.5265 vocab=1.1524 token=1.1673 | 6.13 steps/s | VRAM=31.82 Go [2026-07-16 19:32:46] step 135950/1000000 | loss=2.4749 vocab=1.0325 token=1.8389 | 6.12 steps/s | VRAM=29.62 Go [2026-07-16 19:32:54] step 136000/1000000 | loss=2.4130 vocab=1.2856 token=1.2183 | 6.12 steps/s | VRAM=31.85 Go [2026-07-16 19:33:02] step 136050/1000000 | loss=2.4357 vocab=1.3389 token=1.4884 | 6.12 steps/s | VRAM=28.85 Go [2026-07-16 19:33:10] step 136100/1000000 | loss=2.4198 vocab=0.8676 token=1.2797 | 6.10 steps/s | VRAM=24.31 Go [2026-07-16 19:33:18] step 136150/1000000 | loss=2.4084 vocab=0.8703 token=1.6223 | 6.12 steps/s | VRAM=32.52 Go [2026-07-16 19:33:26] step 136200/1000000 | loss=2.4338 vocab=1.2596 token=1.7161 | 6.12 steps/s | VRAM=26.90 Go [2026-07-16 19:33:35] step 136250/1000000 | loss=2.5064 vocab=0.9422 token=1.6400 | 6.12 steps/s | VRAM=25.20 Go [2026-07-16 19:33:43] step 136300/1000000 | loss=2.4553 vocab=0.6268 token=1.2310 | 6.14 steps/s | VRAM=31.22 Go [2026-07-16 19:33:51] step 136350/1000000 | loss=2.5173 vocab=1.1091 token=1.5220 | 6.13 steps/s | VRAM=31.27 Go [2026-07-16 19:33:59] step 136400/1000000 | loss=2.4858 vocab=1.0013 token=1.5308 | 6.12 steps/s | VRAM=31.78 Go [2026-07-16 19:34:07] step 136450/1000000 | loss=2.5055 vocab=1.0797 token=1.4180 | 6.11 steps/s | VRAM=28.67 Go [2026-07-16 19:34:15] step 136500/1000000 | loss=2.4812 vocab=0.9344 token=1.5584 | 6.12 steps/s | VRAM=26.35 Go [2026-07-16 19:34:24] step 136550/1000000 | loss=2.5864 vocab=0.7647 token=0.7134 | 6.09 steps/s | VRAM=29.03 Go [2026-07-16 19:34:32] step 136600/1000000 | loss=2.4818 vocab=1.1160 token=1.6895 | 6.11 steps/s | VRAM=28.94 Go [2026-07-16 19:34:40] step 136650/1000000 | loss=2.3828 vocab=1.3661 token=1.5919 | 6.10 steps/s | VRAM=28.23 Go [2026-07-16 19:34:48] step 136700/1000000 | loss=2.5080 vocab=0.9759 token=1.2357 | 6.15 steps/s | VRAM=24.58 Go [2026-07-16 19:34:56] step 136750/1000000 | loss=2.4963 vocab=1.2741 token=1.5837 | 6.12 steps/s | VRAM=24.88 Go [2026-07-16 19:35:04] step 136800/1000000 | loss=2.4605 vocab=0.7988 token=1.7278 | 6.17 steps/s | VRAM=31.80 Go [2026-07-16 19:35:13] step 136850/1000000 | loss=2.3333 vocab=1.0117 token=1.4888 | 6.08 steps/s | VRAM=29.61 Go [2026-07-16 19:35:21] step 136900/1000000 | loss=2.4712 vocab=1.2305 token=1.1101 | 6.12 steps/s | VRAM=32.12 Go [2026-07-16 19:35:29] step 136950/1000000 | loss=2.4173 vocab=0.5809 token=1.0398 | 6.10 steps/s | VRAM=29.78 Go [2026-07-16 19:35:37] step 137000/1000000 | loss=2.4178 vocab=1.1478 token=1.3954 | 6.12 steps/s | VRAM=28.55 Go [2026-07-16 19:35:45] step 137050/1000000 | loss=2.4862 vocab=1.2969 token=1.3944 | 6.10 steps/s | VRAM=32.18 Go [2026-07-16 19:35:54] step 137100/1000000 | loss=2.5417 vocab=1.1845 token=1.2249 | 6.08 steps/s | VRAM=29.51 Go [2026-07-16 19:36:02] step 137150/1000000 | loss=2.5069 vocab=1.0578 token=1.6640 | 6.16 steps/s | VRAM=31.27 Go [2026-07-16 19:36:10] step 137200/1000000 | loss=2.5089 vocab=1.2430 token=1.7401 | 6.13 steps/s | VRAM=30.71 Go [2026-07-16 19:36:18] step 137250/1000000 | loss=2.5032 vocab=1.2233 token=1.6736 | 6.11 steps/s | VRAM=32.29 Go [2026-07-16 19:36:26] step 137300/1000000 | loss=2.4900 vocab=0.8813 token=1.4450 | 6.13 steps/s | VRAM=32.24 Go [2026-07-16 19:36:34] step 137350/1000000 | loss=2.4646 vocab=0.9234 token=1.1891 | 6.11 steps/s | VRAM=26.60 Go [2026-07-16 19:36:43] step 137400/1000000 | loss=2.4016 vocab=1.1925 token=1.2706 | 6.10 steps/s | VRAM=31.46 Go [2026-07-16 19:36:51] step 137450/1000000 | loss=2.4769 vocab=0.6553 token=1.3391 | 6.13 steps/s | VRAM=29.74 Go [2026-07-16 19:36:59] step 137500/1000000 | loss=2.4238 vocab=0.9334 token=1.5699 | 6.11 steps/s | VRAM=25.87 Go [2026-07-16 19:37:07] step 137550/1000000 | loss=2.4170 vocab=1.0376 token=1.8971 | 6.13 steps/s | VRAM=31.13 Go [2026-07-16 19:37:15] step 137600/1000000 | loss=2.4398 vocab=0.8297 token=1.2223 | 6.12 steps/s | VRAM=28.45 Go [2026-07-16 19:37:23] step 137650/1000000 | loss=2.5443 vocab=0.9581 token=1.4662 | 6.11 steps/s | VRAM=26.28 Go [2026-07-16 19:37:32] step 137700/1000000 | loss=2.4236 vocab=1.4480 token=1.5146 | 6.14 steps/s | VRAM=32.04 Go [2026-07-16 19:37:40] step 137750/1000000 | loss=2.5478 vocab=1.0983 token=1.4183 | 6.10 steps/s | VRAM=32.80 Go [2026-07-16 19:37:48] step 137800/1000000 | loss=2.4636 vocab=1.0913 token=1.9513 | 6.10 steps/s | VRAM=32.14 Go [2026-07-16 19:37:56] step 137850/1000000 | loss=2.4507 vocab=1.3718 token=1.2004 | 6.13 steps/s | VRAM=31.59 Go [2026-07-16 19:38:04] step 137900/1000000 | loss=2.3598 vocab=0.7297 token=1.0931 | 6.14 steps/s | VRAM=32.77 Go [2026-07-16 19:38:12] step 137950/1000000 | loss=2.4472 vocab=0.8598 token=1.3312 | 6.14 steps/s | VRAM=28.27 Go [2026-07-16 19:38:21] step 138000/1000000 | loss=2.3825 vocab=0.8206 token=1.3500 | 6.12 steps/s | VRAM=32.18 Go [2026-07-16 19:38:29] step 138050/1000000 | loss=2.4199 vocab=1.0054 token=1.2570 | 6.15 steps/s | VRAM=24.71 Go [2026-07-16 19:38:37] step 138100/1000000 | loss=2.4143 vocab=1.0246 token=1.3174 | 6.13 steps/s | VRAM=32.98 Go [2026-07-16 19:38:45] step 138150/1000000 | loss=2.5422 vocab=1.1020 token=1.5788 | 6.08 steps/s | VRAM=29.89 Go [2026-07-16 19:38:53] step 138200/1000000 | loss=2.4610 vocab=1.2060 token=1.3668 | 6.16 steps/s | VRAM=32.42 Go [2026-07-16 19:39:01] step 138250/1000000 | loss=2.5157 vocab=0.9537 token=1.6791 | 6.09 steps/s | VRAM=28.80 Go [2026-07-16 19:39:10] step 138300/1000000 | loss=2.4770 vocab=1.1478 token=1.5590 | 6.12 steps/s | VRAM=28.89 Go [2026-07-16 19:39:18] step 138350/1000000 | loss=2.3831 vocab=1.1308 token=1.2593 | 6.07 steps/s | VRAM=27.27 Go [2026-07-16 19:39:26] step 138400/1000000 | loss=2.5335 vocab=0.7050 token=1.6605 | 6.15 steps/s | VRAM=27.06 Go [2026-07-16 19:39:34] step 138450/1000000 | loss=2.3354 vocab=0.5323 token=1.4701 | 6.13 steps/s | VRAM=28.08 Go [2026-07-16 19:39:42] step 138500/1000000 | loss=2.5115 vocab=1.3123 token=1.3921 | 6.09 steps/s | VRAM=26.24 Go [2026-07-16 19:39:50] step 138550/1000000 | loss=2.4560 vocab=1.6374 token=1.7171 | 6.15 steps/s | VRAM=28.20 Go [2026-07-16 19:39:59] step 138600/1000000 | loss=2.4579 vocab=0.8150 token=1.2608 | 6.14 steps/s | VRAM=32.12 Go [2026-07-16 19:40:07] step 138650/1000000 | loss=2.5446 vocab=0.9254 token=1.4697 | 6.12 steps/s | VRAM=28.80 Go [2026-07-16 19:40:15] step 138700/1000000 | loss=2.5163 vocab=1.1838 token=1.5755 | 6.12 steps/s | VRAM=31.97 Go [2026-07-16 19:40:23] step 138750/1000000 | loss=2.4914 vocab=0.9704 token=1.7772 | 6.13 steps/s | VRAM=25.45 Go [2026-07-16 19:40:31] step 138800/1000000 | loss=2.4899 vocab=1.4685 token=2.0992 | 6.12 steps/s | VRAM=28.55 Go [2026-07-16 19:40:39] step 138850/1000000 | loss=2.4540 vocab=0.7517 token=1.7242 | 6.11 steps/s | VRAM=32.41 Go [2026-07-16 19:40:48] step 138900/1000000 | loss=2.4971 vocab=0.8163 token=1.6320 | 6.09 steps/s | VRAM=29.22 Go [2026-07-16 19:40:56] step 138950/1000000 | loss=2.4566 vocab=1.3029 token=1.5305 | 6.12 steps/s | VRAM=30.76 Go [2026-07-16 19:41:04] step 139000/1000000 | loss=2.4546 vocab=1.1781 token=1.6155 | 6.09 steps/s | VRAM=26.86 Go [2026-07-16 19:41:12] step 139050/1000000 | loss=2.5169 vocab=1.4019 token=1.0560 | 6.15 steps/s | VRAM=31.56 Go [2026-07-16 19:41:20] step 139100/1000000 | loss=2.4517 vocab=0.8639 token=1.0396 | 6.10 steps/s | VRAM=30.34 Go [2026-07-16 19:41:29] step 139150/1000000 | loss=2.5130 vocab=1.0320 token=1.5648 | 6.14 steps/s | VRAM=31.88 Go [2026-07-16 19:41:37] step 139200/1000000 | loss=2.4954 vocab=0.9861 token=1.6101 | 6.12 steps/s | VRAM=28.68 Go [2026-07-16 19:41:45] step 139250/1000000 | loss=2.4173 vocab=0.6040 token=1.3311 | 6.12 steps/s | VRAM=26.20 Go [2026-07-16 19:41:53] step 139300/1000000 | loss=2.3190 vocab=1.2605 token=1.6826 | 6.14 steps/s | VRAM=31.12 Go [2026-07-16 19:42:01] step 139350/1000000 | loss=2.3681 vocab=1.1029 token=1.6002 | 6.13 steps/s | VRAM=28.40 Go [2026-07-16 19:42:09] step 139400/1000000 | loss=2.5671 vocab=1.2141 token=1.2857 | 6.12 steps/s | VRAM=26.35 Go [2026-07-16 19:42:17] step 139450/1000000 | loss=2.5383 vocab=0.7261 token=1.4985 | 6.12 steps/s | VRAM=28.68 Go [2026-07-16 19:42:26] step 139500/1000000 | loss=2.4763 vocab=1.1767 token=1.1666 | 6.14 steps/s | VRAM=26.49 Go [2026-07-16 19:42:34] step 139550/1000000 | loss=2.4640 vocab=0.5801 token=1.4331 | 6.15 steps/s | VRAM=25.07 Go [2026-07-16 19:42:42] step 139600/1000000 | loss=2.4535 vocab=0.7942 token=1.3688 | 6.10 steps/s | VRAM=28.68 Go [2026-07-16 19:42:50] step 139650/1000000 | loss=2.4555 vocab=1.1901 token=1.8469 | 6.14 steps/s | VRAM=29.11 Go [2026-07-16 19:42:58] step 139700/1000000 | loss=2.4678 vocab=1.0807 token=1.3594 | 6.12 steps/s | VRAM=31.49 Go [2026-07-16 19:43:06] step 139750/1000000 | loss=2.4763 vocab=1.1289 token=1.4864 | 6.13 steps/s | VRAM=32.25 Go [2026-07-16 19:43:15] step 139800/1000000 | loss=2.5143 vocab=1.1217 token=1.4643 | 6.12 steps/s | VRAM=25.58 Go [2026-07-16 19:43:23] step 139850/1000000 | loss=2.5232 vocab=1.9292 token=1.6193 | 6.13 steps/s | VRAM=29.16 Go [2026-07-16 19:43:31] step 139900/1000000 | loss=2.4714 vocab=0.9332 token=1.4993 | 6.13 steps/s | VRAM=31.37 Go [2026-07-16 19:43:39] step 139950/1000000 | loss=2.5129 vocab=1.4023 token=1.7683 | 6.13 steps/s | VRAM=28.30 Go [2026-07-16 19:43:47] step 140000/1000000 | loss=2.4151 vocab=0.7604 token=1.3385 | 6.09 steps/s | VRAM=32.61 Go [2026-07-16 19:45:12] step 140050/1000000 | loss=2.5190 vocab=0.7234 token=1.7933 | 0.59 steps/s | VRAM=25.46 Go [2026-07-16 19:51:34] step 140100/1000000 | loss=0.4772 vocab=0.9720 token=1.2354 | 30.55 steps/s | VRAM=31.32 Go [2026-07-16 19:51:42] step 140150/1000000 | loss=2.4705 vocab=0.7781 token=1.4710 | 6.26 steps/s | VRAM=30.05 Go [2026-07-16 19:51:50] step 140200/1000000 | loss=2.5507 vocab=0.9592 token=1.6996 | 6.24 steps/s | VRAM=30.14 Go [2026-07-16 19:51:58] step 140250/1000000 | loss=2.3795 vocab=0.9347 token=0.8278 | 6.20 steps/s | VRAM=30.14 Go [2026-07-16 19:52:06] step 140300/1000000 | loss=2.4815 vocab=1.1919 token=1.5572 | 6.21 steps/s | VRAM=27.18 Go [2026-07-16 19:52:14] step 140350/1000000 | loss=2.3692 vocab=1.4636 token=1.6188 | 6.17 steps/s | VRAM=30.63 Go [2026-07-16 19:52:22] step 140400/1000000 | loss=2.4059 vocab=0.6037 token=1.5841 | 6.16 steps/s | VRAM=30.86 Go [2026-07-16 19:52:30] step 140450/1000000 | loss=2.3947 vocab=0.7161 token=1.3553 | 6.18 steps/s | VRAM=29.93 Go [2026-07-16 19:52:38] step 140500/1000000 | loss=2.4713 vocab=1.0840 token=1.4357 | 6.17 steps/s | VRAM=30.03 Go [2026-07-16 19:52:46] step 140550/1000000 | loss=2.4484 vocab=0.9259 token=1.6703 | 6.16 steps/s | VRAM=30.55 Go [2026-07-16 19:52:54] step 140600/1000000 | loss=2.3771 vocab=0.7871 token=1.1751 | 6.17 steps/s | VRAM=30.26 Go [2026-07-16 19:53:02] step 140650/1000000 | loss=2.5439 vocab=0.9507 token=1.6980 | 6.13 steps/s | VRAM=27.22 Go [2026-07-16 19:53:11] step 140700/1000000 | loss=2.3498 vocab=1.1199 token=1.2569 | 6.16 steps/s | VRAM=29.94 Go [2026-07-16 19:53:19] step 140750/1000000 | loss=2.3526 vocab=0.5315 token=1.5304 | 6.13 steps/s | VRAM=30.75 Go [2026-07-16 19:53:27] step 140800/1000000 | loss=2.3063 vocab=0.4281 token=0.9392 | 6.13 steps/s | VRAM=27.80 Go [2026-07-16 19:53:35] step 140850/1000000 | loss=2.4147 vocab=0.6306 token=1.3498 | 6.08 steps/s | VRAM=30.47 Go [2026-07-16 19:53:43] step 140900/1000000 | loss=2.4766 vocab=0.9888 token=1.3036 | 6.14 steps/s | VRAM=29.83 Go [2026-07-16 19:53:51] step 140950/1000000 | loss=2.5077 vocab=0.9023 token=1.3644 | 6.13 steps/s | VRAM=30.23 Go [2026-07-16 19:54:00] step 141000/1000000 | loss=2.3858 vocab=0.9066 token=1.4549 | 6.09 steps/s | VRAM=27.47 Go [2026-07-16 19:54:08] step 141050/1000000 | loss=2.4781 vocab=1.2512 token=1.9372 | 6.12 steps/s | VRAM=29.94 Go [2026-07-16 19:54:16] step 141100/1000000 | loss=2.4397 vocab=0.5490 token=1.3964 | 6.12 steps/s | VRAM=30.44 Go [2026-07-16 19:54:24] step 141150/1000000 | loss=2.5156 vocab=1.0495 token=1.4361 | 6.12 steps/s | VRAM=30.84 Go [2026-07-16 19:54:32] step 141200/1000000 | loss=2.4110 vocab=1.0687 token=1.3758 | 6.11 steps/s | VRAM=29.71 Go [2026-07-16 19:54:40] step 141250/1000000 | loss=2.4051 vocab=0.9518 token=1.2376 | 6.12 steps/s | VRAM=30.16 Go [2026-07-16 19:54:49] step 141300/1000000 | loss=2.4061 vocab=1.1436 token=1.7071 | 6.06 steps/s | VRAM=30.51 Go [2026-07-16 19:54:57] step 141350/1000000 | loss=2.4998 vocab=1.0313 token=1.3970 | 6.12 steps/s | VRAM=30.10 Go [2026-07-16 19:55:05] step 141400/1000000 | loss=2.4719 vocab=0.6988 token=1.2416 | 6.13 steps/s | VRAM=30.46 Go [2026-07-16 19:55:13] step 141450/1000000 | loss=2.4510 vocab=0.9003 token=1.4963 | 6.12 steps/s | VRAM=30.14 Go [2026-07-16 19:55:21] step 141500/1000000 | loss=2.3788 vocab=0.8824 token=1.5225 | 6.13 steps/s | VRAM=30.57 Go [2026-07-16 19:55:30] step 141550/1000000 | loss=2.2860 vocab=0.6150 token=1.2632 | 6.11 steps/s | VRAM=29.96 Go [2026-07-16 19:55:38] step 141600/1000000 | loss=2.4419 vocab=1.2154 token=1.2750 | 6.12 steps/s | VRAM=27.56 Go [2026-07-16 19:55:46] step 141650/1000000 | loss=2.3353 vocab=0.6583 token=1.0955 | 6.11 steps/s | VRAM=30.40 Go [2026-07-16 19:55:54] step 141700/1000000 | loss=2.4860 vocab=0.7242 token=1.4112 | 6.10 steps/s | VRAM=29.65 Go [2026-07-16 19:56:02] step 141750/1000000 | loss=2.3841 vocab=0.6855 token=1.4073 | 6.11 steps/s | VRAM=32.56 Go [2026-07-16 19:56:10] step 141800/1000000 | loss=2.4782 vocab=0.9655 token=1.3461 | 6.11 steps/s | VRAM=30.23 Go [2026-07-16 19:56:19] step 141850/1000000 | loss=2.4422 vocab=1.1847 token=1.6203 | 6.10 steps/s | VRAM=30.28 Go [2026-07-16 19:56:27] step 141900/1000000 | loss=2.4466 vocab=0.6662 token=1.6336 | 6.13 steps/s | VRAM=30.45 Go [2026-07-16 19:56:35] step 141950/1000000 | loss=2.4740 vocab=0.7299 token=1.6634 | 6.12 steps/s | VRAM=30.21 Go [2026-07-16 19:56:43] step 142000/1000000 | loss=2.5267 vocab=1.1732 token=1.6158 | 6.14 steps/s | VRAM=30.34 Go [2026-07-16 19:56:51] step 142050/1000000 | loss=2.3937 vocab=1.1138 token=1.6402 | 6.13 steps/s | VRAM=29.94 Go [2026-07-16 19:57:00] step 142100/1000000 | loss=2.4347 vocab=0.6660 token=1.3910 | 6.09 steps/s | VRAM=30.53 Go [2026-07-16 19:57:08] step 142150/1000000 | loss=2.4695 vocab=1.0980 token=1.6239 | 6.14 steps/s | VRAM=30.39 Go [2026-07-16 19:57:16] step 142200/1000000 | loss=2.3833 vocab=1.3443 token=1.8231 | 6.11 steps/s | VRAM=30.33 Go [2026-07-16 19:57:24] step 142250/1000000 | loss=2.3375 vocab=1.2194 token=1.7326 | 6.09 steps/s | VRAM=30.62 Go [2026-07-16 19:57:32] step 142300/1000000 | loss=2.3933 vocab=1.0655 token=1.2607 | 6.12 steps/s | VRAM=30.54 Go [2026-07-16 19:57:40] step 142350/1000000 | loss=2.3286 vocab=1.2230 token=1.3708 | 6.13 steps/s | VRAM=30.43 Go [2026-07-16 19:57:49] step 142400/1000000 | loss=2.5107 vocab=0.8776 token=1.5430 | 6.10 steps/s | VRAM=30.06 Go [2026-07-17 03:43:38] step 140100/1000000 | loss=0.4840 vocab=0.8951 token=1.3507 | 23.66 steps/s | VRAM=29.90 Go [2026-07-17 03:43:46] step 140150/1000000 | loss=2.4470 vocab=0.9932 token=1.4463 | 6.21 steps/s | VRAM=32.07 Go [2026-07-17 03:43:54] step 140200/1000000 | loss=2.4048 vocab=0.9327 token=1.4116 | 6.22 steps/s | VRAM=26.01 Go [2026-07-17 03:44:02] step 140250/1000000 | loss=2.4021 vocab=1.1747 token=1.1796 | 6.18 steps/s | VRAM=28.77 Go [2026-07-17 03:44:10] step 140300/1000000 | loss=2.4806 vocab=0.8838 token=1.4198 | 6.16 steps/s | VRAM=28.92 Go [2026-07-17 03:44:18] step 140350/1000000 | loss=2.4687 vocab=1.4559 token=1.3700 | 6.14 steps/s | VRAM=26.51 Go [2026-07-17 03:44:26] step 140400/1000000 | loss=2.3692 vocab=0.8370 token=1.8319 | 6.17 steps/s | VRAM=26.44 Go [2026-07-17 03:44:35] step 140450/1000000 | loss=2.5601 vocab=0.6726 token=1.3343 | 6.18 steps/s | VRAM=24.88 Go [2026-07-17 03:44:43] step 140500/1000000 | loss=2.4749 vocab=0.7875 token=1.3655 | 6.15 steps/s | VRAM=28.89 Go [2026-07-17 03:44:51] step 140550/1000000 | loss=2.4227 vocab=0.9950 token=1.6336 | 6.15 steps/s | VRAM=32.38 Go [2026-07-17 03:44:59] step 140600/1000000 | loss=2.5270 vocab=1.0366 token=1.3626 | 6.13 steps/s | VRAM=31.78 Go [2026-07-17 03:45:07] step 140650/1000000 | loss=2.4848 vocab=0.6744 token=1.4560 | 6.12 steps/s | VRAM=28.07 Go [2026-07-17 03:45:15] step 140700/1000000 | loss=2.4970 vocab=1.0690 token=1.5249 | 6.12 steps/s | VRAM=28.99 Go [2026-07-17 03:45:23] step 140750/1000000 | loss=2.4223 vocab=1.2871 token=1.8597 | 6.11 steps/s | VRAM=28.83 Go [2026-07-17 03:45:32] step 140800/1000000 | loss=2.4602 vocab=1.5849 token=1.5611 | 6.10 steps/s | VRAM=28.29 Go [2026-07-17 03:45:40] step 140850/1000000 | loss=2.4203 vocab=0.9707 token=1.4081 | 6.11 steps/s | VRAM=31.90 Go [2026-07-17 03:45:48] step 140900/1000000 | loss=2.4461 vocab=1.0386 token=1.2026 | 6.09 steps/s | VRAM=32.00 Go [2026-07-17 03:45:56] step 140950/1000000 | loss=2.5536 vocab=0.9095 token=1.3283 | 6.07 steps/s | VRAM=28.59 Go [2026-07-17 03:46:04] step 141000/1000000 | loss=2.4796 vocab=0.7850 token=1.4914 | 6.10 steps/s | VRAM=29.33 Go [2026-07-17 03:46:13] step 141050/1000000 | loss=2.4654 vocab=1.4142 token=1.9795 | 6.07 steps/s | VRAM=32.40 Go [2026-07-17 03:46:21] step 141100/1000000 | loss=2.4289 vocab=1.0450 token=1.6039 | 6.07 steps/s | VRAM=28.72 Go [2026-07-17 03:46:29] step 141150/1000000 | loss=2.3586 vocab=0.6528 token=1.5221 | 6.11 steps/s | VRAM=31.71 Go [2026-07-17 03:46:37] step 141200/1000000 | loss=2.5453 vocab=0.3427 token=0.9177 | 6.11 steps/s | VRAM=25.10 Go [2026-07-17 03:46:46] step 141250/1000000 | loss=2.4205 vocab=0.7776 token=1.2808 | 6.08 steps/s | VRAM=30.73 Go [2026-07-17 03:46:54] step 141300/1000000 | loss=2.5234 vocab=1.0474 token=1.4173 | 6.08 steps/s | VRAM=27.97 Go [2026-07-17 03:47:02] step 141350/1000000 | loss=2.4062 vocab=0.7498 token=1.4527 | 6.05 steps/s | VRAM=30.01 Go [2026-07-17 03:47:10] step 141400/1000000 | loss=2.4500 vocab=0.9975 token=1.2782 | 6.06 steps/s | VRAM=24.48 Go [2026-07-17 03:47:19] step 141450/1000000 | loss=2.4869 vocab=0.8702 token=1.3958 | 6.10 steps/s | VRAM=31.89 Go [2026-07-17 03:47:27] step 141500/1000000 | loss=2.4458 vocab=0.7505 token=1.3028 | 6.05 steps/s | VRAM=29.86 Go [2026-07-17 03:47:35] step 141550/1000000 | loss=2.5457 vocab=0.8775 token=1.6822 | 6.09 steps/s | VRAM=32.11 Go [2026-07-17 03:47:43] step 141600/1000000 | loss=2.4097 vocab=1.1297 token=1.5553 | 6.06 steps/s | VRAM=28.11 Go [2026-07-17 03:47:51] step 141650/1000000 | loss=2.3565 vocab=0.4733 token=1.2769 | 6.07 steps/s | VRAM=26.38 Go [2026-07-17 03:48:00] step 141700/1000000 | loss=2.3426 vocab=0.5213 token=1.4205 | 6.10 steps/s | VRAM=32.33 Go [2026-07-17 03:48:08] step 141750/1000000 | loss=2.4386 vocab=0.8698 token=1.4702 | 6.07 steps/s | VRAM=25.99 Go [2026-07-17 03:48:16] step 141800/1000000 | loss=2.4334 vocab=0.8998 token=1.4574 | 6.07 steps/s | VRAM=28.29 Go [2026-07-17 03:48:24] step 141850/1000000 | loss=2.4882 vocab=1.5972 token=2.1147 | 6.13 steps/s | VRAM=28.48 Go [2026-07-17 03:48:33] step 141900/1000000 | loss=2.4810 vocab=0.7024 token=1.1543 | 6.08 steps/s | VRAM=32.58 Go [2026-07-17 03:48:41] step 141950/1000000 | loss=2.3678 vocab=0.7614 token=1.3078 | 6.08 steps/s | VRAM=28.70 Go [2026-07-17 03:48:49] step 142000/1000000 | loss=2.3352 vocab=1.1336 token=1.6436 | 6.06 steps/s | VRAM=32.37 Go [2026-07-17 03:48:57] step 142050/1000000 | loss=2.3995 vocab=0.7616 token=1.3982 | 6.06 steps/s | VRAM=26.55 Go [2026-07-17 03:49:05] step 142100/1000000 | loss=2.3883 vocab=0.5526 token=1.3625 | 6.08 steps/s | VRAM=28.90 Go [2026-07-17 03:49:14] step 142150/1000000 | loss=2.4436 vocab=0.9317 token=1.5580 | 6.09 steps/s | VRAM=28.00 Go [2026-07-17 03:49:22] step 142200/1000000 | loss=2.4357 vocab=0.9811 token=1.4212 | 6.09 steps/s | VRAM=28.58 Go [2026-07-17 03:49:30] step 142250/1000000 | loss=2.4414 vocab=0.7064 token=1.6174 | 6.10 steps/s | VRAM=28.62 Go [2026-07-17 03:49:38] step 142300/1000000 | loss=2.4377 vocab=1.1537 token=1.5925 | 6.09 steps/s | VRAM=28.62 Go [2026-07-17 03:49:47] step 142350/1000000 | loss=2.4911 vocab=1.2815 token=2.1805 | 6.09 steps/s | VRAM=32.18 Go [2026-07-17 03:49:55] step 142400/1000000 | loss=2.5703 vocab=1.1074 token=1.6013 | 6.11 steps/s | VRAM=31.98 Go [2026-07-17 03:50:03] step 142450/1000000 | loss=2.5444 vocab=1.0756 token=1.3848 | 6.09 steps/s | VRAM=32.66 Go [2026-07-17 03:50:11] step 142500/1000000 | loss=2.5301 vocab=1.3054 token=1.6423 | 6.08 steps/s | VRAM=25.97 Go [2026-07-17 03:50:19] step 142550/1000000 | loss=2.4238 vocab=0.8271 token=1.3209 | 6.06 steps/s | VRAM=27.30 Go [2026-07-17 03:50:28] step 142600/1000000 | loss=2.4540 vocab=0.9489 token=1.1559 | 6.10 steps/s | VRAM=31.50 Go [2026-07-17 03:50:36] step 142650/1000000 | loss=2.5277 vocab=0.9397 token=1.7837 | 6.07 steps/s | VRAM=32.82 Go [2026-07-17 03:50:44] step 142700/1000000 | loss=2.4621 vocab=0.8359 token=1.3401 | 6.09 steps/s | VRAM=29.12 Go [2026-07-17 03:50:52] step 142750/1000000 | loss=2.4110 vocab=1.5589 token=1.4913 | 6.07 steps/s | VRAM=28.90 Go [2026-07-17 03:51:00] step 142800/1000000 | loss=2.3574 vocab=1.0988 token=1.2655 | 6.09 steps/s | VRAM=28.67 Go [2026-07-17 03:51:09] step 142850/1000000 | loss=2.4156 vocab=1.0758 token=1.5393 | 6.06 steps/s | VRAM=26.68 Go [2026-07-17 03:51:17] step 142900/1000000 | loss=2.5106 vocab=0.7243 token=1.3534 | 6.10 steps/s | VRAM=26.56 Go [2026-07-17 03:51:25] step 142950/1000000 | loss=2.4545 vocab=0.9889 token=1.8721 | 6.11 steps/s | VRAM=28.17 Go [2026-07-17 03:51:33] step 143000/1000000 | loss=2.4516 vocab=0.8614 token=1.5974 | 6.06 steps/s | VRAM=29.15 Go [2026-07-17 03:51:42] step 143050/1000000 | loss=2.4165 vocab=1.1344 token=1.3531 | 6.09 steps/s | VRAM=32.12 Go [2026-07-17 03:51:50] step 143100/1000000 | loss=2.4551 vocab=1.2846 token=1.3089 | 6.05 steps/s | VRAM=26.30 Go [2026-07-17 03:51:58] step 143150/1000000 | loss=2.3883 vocab=1.1384 token=1.4428 | 6.07 steps/s | VRAM=26.66 Go [2026-07-17 03:52:06] step 143200/1000000 | loss=2.3751 vocab=0.8535 token=1.1845 | 6.11 steps/s | VRAM=32.11 Go [2026-07-17 03:52:14] step 143250/1000000 | loss=2.5144 vocab=1.5963 token=1.7736 | 6.09 steps/s | VRAM=28.11 Go [2026-07-17 03:52:23] step 143300/1000000 | loss=2.3909 vocab=0.9903 token=1.4233 | 6.07 steps/s | VRAM=25.79 Go [2026-07-17 03:52:31] step 143350/1000000 | loss=2.4541 vocab=1.3776 token=1.5737 | 6.07 steps/s | VRAM=26.52 Go [2026-07-17 03:52:39] step 143400/1000000 | loss=2.4809 vocab=1.3711 token=1.6664 | 6.10 steps/s | VRAM=28.54 Go [2026-07-17 03:52:47] step 143450/1000000 | loss=2.3488 vocab=1.1756 token=1.7983 | 6.11 steps/s | VRAM=28.30 Go [2026-07-17 03:52:56] step 143500/1000000 | loss=2.4728 vocab=1.2138 token=1.2982 | 6.08 steps/s | VRAM=27.11 Go [2026-07-17 03:53:04] step 143550/1000000 | loss=2.5556 vocab=0.8531 token=1.3852 | 6.13 steps/s | VRAM=32.54 Go [2026-07-17 03:53:12] step 143600/1000000 | loss=2.4521 vocab=0.8459 token=1.3109 | 6.11 steps/s | VRAM=32.34 Go [2026-07-17 03:53:20] step 143650/1000000 | loss=2.4229 vocab=1.3883 token=1.6600 | 6.09 steps/s | VRAM=27.65 Go [2026-07-17 03:53:28] step 143700/1000000 | loss=2.5396 vocab=1.1907 token=1.5639 | 6.08 steps/s | VRAM=28.45 Go [2026-07-17 03:53:37] step 143750/1000000 | loss=2.4922 vocab=0.7450 token=1.4687 | 6.09 steps/s | VRAM=26.00 Go [2026-07-17 03:53:45] step 143800/1000000 | loss=2.3929 vocab=1.1223 token=1.4630 | 6.09 steps/s | VRAM=28.57 Go [2026-07-17 03:53:53] step 143850/1000000 | loss=2.4544 vocab=1.2450 token=1.8125 | 6.09 steps/s | VRAM=26.23 Go [2026-07-17 03:54:01] step 143900/1000000 | loss=2.3997 vocab=1.3909 token=1.5623 | 6.08 steps/s | VRAM=31.34 Go [2026-07-17 03:54:09] step 143950/1000000 | loss=2.4503 vocab=1.2961 token=1.8941 | 6.07 steps/s | VRAM=29.09 Go [2026-07-17 03:54:18] step 144000/1000000 | loss=2.4505 vocab=0.8680 token=1.3681 | 6.10 steps/s | VRAM=28.37 Go [2026-07-17 03:54:26] step 144050/1000000 | loss=2.3835 vocab=1.1733 token=1.8591 | 6.08 steps/s | VRAM=29.41 Go [2026-07-17 03:54:34] step 144100/1000000 | loss=2.4174 vocab=0.7520 token=1.4964 | 6.04 steps/s | VRAM=28.18 Go [2026-07-17 03:54:42] step 144150/1000000 | loss=2.3723 vocab=0.8258 token=1.4909 | 6.10 steps/s | VRAM=28.48 Go [2026-07-17 03:54:51] step 144200/1000000 | loss=2.4713 vocab=1.1413 token=1.6130 | 6.08 steps/s | VRAM=29.26 Go [2026-07-17 03:54:59] step 144250/1000000 | loss=2.4160 vocab=0.7702 token=1.2064 | 6.07 steps/s | VRAM=28.60 Go [2026-07-17 03:55:07] step 144300/1000000 | loss=2.5060 vocab=0.9611 token=1.4475 | 6.05 steps/s | VRAM=28.39 Go [2026-07-17 03:55:15] step 144350/1000000 | loss=2.3987 vocab=1.4618 token=1.6792 | 6.08 steps/s | VRAM=24.88 Go [2026-07-17 03:55:24] step 144400/1000000 | loss=2.5317 vocab=1.0673 token=1.3585 | 6.07 steps/s | VRAM=28.19 Go [2026-07-17 03:55:32] step 144450/1000000 | loss=2.2923 vocab=1.4065 token=1.7878 | 6.09 steps/s | VRAM=32.79 Go [2026-07-17 03:55:40] step 144500/1000000 | loss=2.4085 vocab=0.8824 token=1.4319 | 6.06 steps/s | VRAM=28.24 Go [2026-07-17 03:55:48] step 144550/1000000 | loss=2.3332 vocab=0.8978 token=1.0952 | 6.10 steps/s | VRAM=28.99 Go [2026-07-17 03:55:56] step 144600/1000000 | loss=2.3279 vocab=0.5448 token=1.0505 | 6.11 steps/s | VRAM=28.38 Go [2026-07-17 03:56:05] step 144650/1000000 | loss=2.5075 vocab=1.5784 token=1.5555 | 6.09 steps/s | VRAM=28.50 Go [2026-07-17 03:56:13] step 144700/1000000 | loss=2.5273 vocab=0.5327 token=1.3862 | 6.07 steps/s | VRAM=26.39 Go [2026-07-17 03:56:21] step 144750/1000000 | loss=2.4703 vocab=1.8810 token=1.6263 | 6.08 steps/s | VRAM=29.73 Go [2026-07-17 03:56:29] step 144800/1000000 | loss=2.4159 vocab=1.0263 token=1.0938 | 6.04 steps/s | VRAM=26.10 Go [2026-07-17 03:56:38] step 144850/1000000 | loss=2.5874 vocab=1.5908 token=1.7119 | 6.06 steps/s | VRAM=28.81 Go [2026-07-17 03:56:46] step 144900/1000000 | loss=2.3514 vocab=1.0170 token=1.3924 | 6.07 steps/s | VRAM=25.61 Go [2026-07-17 03:56:54] step 144950/1000000 | loss=2.5265 vocab=0.7714 token=1.4158 | 6.10 steps/s | VRAM=26.25 Go [2026-07-17 03:57:02] step 145000/1000000 | loss=2.4006 vocab=0.9110 token=1.4206 | 6.09 steps/s | VRAM=25.85 Go [2026-07-17 03:58:21] step 145050/1000000 | loss=2.5084 vocab=1.5001 token=1.6738 | 0.64 steps/s | VRAM=31.83 Go [2026-07-17 03:58:29] step 145100/1000000 | loss=2.4224 vocab=0.7806 token=1.4827 | 6.17 steps/s | VRAM=28.11 Go [2026-07-17 03:58:37] step 145150/1000000 | loss=2.4687 vocab=0.7977 token=1.3484 | 6.13 steps/s | VRAM=29.25 Go [2026-07-17 03:58:45] step 145200/1000000 | loss=2.3135 vocab=0.6922 token=0.9199 | 6.14 steps/s | VRAM=28.16 Go [2026-07-17 03:58:53] step 145250/1000000 | loss=2.4652 vocab=1.1642 token=1.6272 | 6.12 steps/s | VRAM=29.24 Go [2026-07-17 03:59:01] step 145300/1000000 | loss=2.5121 vocab=1.1688 token=1.5882 | 6.10 steps/s | VRAM=25.71 Go [2026-07-17 03:59:10] step 145350/1000000 | loss=2.4212 vocab=1.1293 token=1.7385 | 6.12 steps/s | VRAM=29.52 Go [2026-07-17 03:59:18] step 145400/1000000 | loss=2.4198 vocab=0.5925 token=1.3270 | 6.09 steps/s | VRAM=29.36 Go [2026-07-17 03:59:26] step 145450/1000000 | loss=2.4973 vocab=0.9957 token=1.3494 | 6.10 steps/s | VRAM=25.98 Go [2026-07-17 03:59:34] step 145500/1000000 | loss=2.4455 vocab=1.5533 token=1.6405 | 6.12 steps/s | VRAM=32.59 Go [2026-07-17 03:59:42] step 145550/1000000 | loss=2.4287 vocab=0.7899 token=0.9576 | 6.10 steps/s | VRAM=32.93 Go [2026-07-17 03:59:51] step 145600/1000000 | loss=2.4478 vocab=0.9600 token=1.1761 | 6.09 steps/s | VRAM=32.50 Go [2026-07-17 03:59:59] step 145650/1000000 | loss=2.4723 vocab=1.1416 token=2.0597 | 6.07 steps/s | VRAM=29.67 Go [2026-07-17 04:00:07] step 145700/1000000 | loss=2.4463 vocab=1.4996 token=1.8845 | 6.08 steps/s | VRAM=28.21 Go [2026-07-17 04:00:15] step 145750/1000000 | loss=2.3902 vocab=0.6298 token=1.2999 | 6.08 steps/s | VRAM=32.38 Go [2026-07-17 04:00:24] step 145800/1000000 | loss=2.4147 vocab=1.2298 token=1.7274 | 6.08 steps/s | VRAM=28.74 Go [2026-07-17 04:00:32] step 145850/1000000 | loss=2.4316 vocab=1.1346 token=1.4621 | 6.05 steps/s | VRAM=28.06 Go [2026-07-17 04:00:40] step 145900/1000000 | loss=2.4022 vocab=0.9540 token=1.2983 | 6.08 steps/s | VRAM=25.70 Go [2026-07-17 04:00:48] step 145950/1000000 | loss=2.3386 vocab=0.9895 token=1.3490 | 6.10 steps/s | VRAM=32.85 Go [2026-07-17 04:00:56] step 146000/1000000 | loss=2.4715 vocab=1.4005 token=1.8325 | 6.06 steps/s | VRAM=24.87 Go [2026-07-17 04:01:05] step 146050/1000000 | loss=2.5015 vocab=0.7302 token=1.1588 | 6.09 steps/s | VRAM=29.35 Go [2026-07-17 04:01:13] step 146100/1000000 | loss=2.5540 vocab=1.4920 token=1.2909 | 6.09 steps/s | VRAM=31.95 Go [2026-07-17 04:01:21] step 146150/1000000 | loss=2.4325 vocab=1.1046 token=1.3224 | 6.08 steps/s | VRAM=24.90 Go [2026-07-17 04:01:29] step 146200/1000000 | loss=2.4701 vocab=1.0035 token=1.8286 | 6.10 steps/s | VRAM=32.43 Go [2026-07-17 04:01:38] step 146250/1000000 | loss=2.4269 vocab=0.4300 token=0.5986 | 6.09 steps/s | VRAM=29.15 Go [2026-07-17 04:01:46] step 146300/1000000 | loss=2.4445 vocab=0.7249 token=1.1186 | 6.09 steps/s | VRAM=29.13 Go [2026-07-17 04:01:54] step 146350/1000000 | loss=2.4868 vocab=1.3221 token=1.5581 | 6.10 steps/s | VRAM=28.35 Go [2026-07-17 04:02:02] step 146400/1000000 | loss=2.3481 vocab=0.5744 token=1.2111 | 6.12 steps/s | VRAM=31.39 Go [2026-07-17 04:02:10] step 146450/1000000 | loss=2.3700 vocab=0.8711 token=1.6020 | 6.10 steps/s | VRAM=32.07 Go [2026-07-17 04:02:19] step 146500/1000000 | loss=2.4792 vocab=0.8669 token=1.4230 | 6.08 steps/s | VRAM=25.69 Go [2026-07-17 04:02:27] step 146550/1000000 | loss=2.4787 vocab=1.1084 token=1.4574 | 6.11 steps/s | VRAM=31.26 Go [2026-07-17 04:02:35] step 146600/1000000 | loss=2.4589 vocab=1.3198 token=1.7350 | 6.08 steps/s | VRAM=32.41 Go [2026-07-17 04:02:43] step 146650/1000000 | loss=2.4412 vocab=1.6326 token=1.3911 | 6.05 steps/s | VRAM=28.99 Go [2026-07-17 04:02:51] step 146700/1000000 | loss=2.5555 vocab=0.9156 token=0.9721 | 6.07 steps/s | VRAM=26.54 Go [2026-07-17 04:03:00] step 146750/1000000 | loss=2.3845 vocab=1.0182 token=1.5810 | 6.10 steps/s | VRAM=32.87 Go [2026-07-17 04:03:08] step 146800/1000000 | loss=2.4120 vocab=1.0582 token=1.6903 | 6.09 steps/s | VRAM=32.22 Go [2026-07-17 04:03:16] step 146850/1000000 | loss=2.3683 vocab=0.9758 token=1.6194 | 6.09 steps/s | VRAM=28.40 Go [2026-07-17 04:03:24] step 146900/1000000 | loss=2.4427 vocab=0.7592 token=1.5990 | 6.12 steps/s | VRAM=32.00 Go [2026-07-17 04:03:32] step 146950/1000000 | loss=2.3940 vocab=1.1014 token=1.4033 | 6.08 steps/s | VRAM=28.77 Go [2026-07-17 04:03:41] step 147000/1000000 | loss=2.5001 vocab=0.8509 token=1.2259 | 6.07 steps/s | VRAM=26.61 Go [2026-07-17 04:03:49] step 147050/1000000 | loss=2.5020 vocab=0.8630 token=1.3785 | 6.11 steps/s | VRAM=32.56 Go [2026-07-17 04:03:57] step 147100/1000000 | loss=2.4738 vocab=0.7164 token=1.4032 | 6.09 steps/s | VRAM=28.29 Go [2026-07-17 04:04:05] step 147150/1000000 | loss=2.4390 vocab=0.7688 token=1.0761 | 6.11 steps/s | VRAM=25.24 Go [2026-07-17 04:04:13] step 147200/1000000 | loss=2.4325 vocab=1.3901 token=1.6796 | 6.11 steps/s | VRAM=31.86 Go [2026-07-17 04:04:22] step 147250/1000000 | loss=2.3155 vocab=0.9658 token=1.4234 | 6.07 steps/s | VRAM=28.23 Go [2026-07-17 04:04:30] step 147300/1000000 | loss=2.4734 vocab=0.7917 token=1.0617 | 6.06 steps/s | VRAM=29.83 Go [2026-07-17 04:04:38] step 147350/1000000 | loss=2.5353 vocab=0.8715 token=1.1971 | 6.09 steps/s | VRAM=28.03 Go [2026-07-17 04:04:46] step 147400/1000000 | loss=2.4551 vocab=0.6248 token=1.3405 | 6.08 steps/s | VRAM=29.62 Go [2026-07-17 04:04:55] step 147450/1000000 | loss=2.3638 vocab=0.6764 token=1.4571 | 6.10 steps/s | VRAM=29.42 Go [2026-07-17 04:05:03] step 147500/1000000 | loss=2.3267 vocab=0.9683 token=1.6813 | 6.12 steps/s | VRAM=30.62 Go [2026-07-17 04:05:11] step 147550/1000000 | loss=2.4350 vocab=1.2888 token=1.7577 | 6.08 steps/s | VRAM=26.98 Go [2026-07-17 04:05:19] step 147600/1000000 | loss=2.3718 vocab=0.9759 token=1.5025 | 6.06 steps/s | VRAM=28.57 Go [2026-07-17 04:05:27] step 147650/1000000 | loss=2.3682 vocab=0.9595 token=1.7365 | 6.09 steps/s | VRAM=26.34 Go [2026-07-17 04:05:36] step 147700/1000000 | loss=2.4335 vocab=1.0978 token=1.7257 | 6.06 steps/s | VRAM=26.94 Go [2026-07-17 04:05:44] step 147750/1000000 | loss=2.4267 vocab=0.9835 token=1.4869 | 6.08 steps/s | VRAM=27.95 Go [2026-07-17 04:05:52] step 147800/1000000 | loss=2.3333 vocab=0.7313 token=1.2101 | 6.08 steps/s | VRAM=26.11 Go [2026-07-17 04:06:00] step 147850/1000000 | loss=2.4526 vocab=0.5980 token=1.3778 | 6.06 steps/s | VRAM=28.98 Go [2026-07-17 04:06:09] step 147900/1000000 | loss=2.5053 vocab=0.9929 token=1.1077 | 6.07 steps/s | VRAM=25.26 Go [2026-07-17 04:06:17] step 147950/1000000 | loss=2.4718 vocab=1.0189 token=1.8083 | 6.11 steps/s | VRAM=31.79 Go [2026-07-17 04:06:25] step 148000/1000000 | loss=2.3866 vocab=0.8729 token=1.5549 | 6.09 steps/s | VRAM=24.69 Go [2026-07-17 04:06:33] step 148050/1000000 | loss=2.4200 vocab=0.7374 token=1.1702 | 6.04 steps/s | VRAM=26.62 Go [2026-07-17 04:06:41] step 148100/1000000 | loss=2.4402 vocab=0.4583 token=1.6290 | 6.09 steps/s | VRAM=31.53 Go [2026-07-17 04:06:50] step 148150/1000000 | loss=2.3145 vocab=1.4695 token=1.6110 | 6.02 steps/s | VRAM=26.53 Go [2026-07-17 04:06:58] step 148200/1000000 | loss=2.4296 vocab=1.0217 token=1.4805 | 6.08 steps/s | VRAM=28.39 Go [2026-07-17 04:07:06] step 148250/1000000 | loss=2.5000 vocab=0.6370 token=1.4336 | 6.09 steps/s | VRAM=31.68 Go [2026-07-17 04:07:14] step 148300/1000000 | loss=2.4072 vocab=1.5007 token=1.5189 | 6.09 steps/s | VRAM=29.38 Go [2026-07-17 04:07:23] step 148350/1000000 | loss=2.3833 vocab=1.3436 token=1.3514 | 6.10 steps/s | VRAM=29.31 Go [2026-07-17 04:07:31] step 148400/1000000 | loss=2.4813 vocab=1.4719 token=1.5515 | 6.07 steps/s | VRAM=29.96 Go [2026-07-17 04:07:39] step 148450/1000000 | loss=2.5176 vocab=0.9497 token=1.4522 | 6.11 steps/s | VRAM=31.33 Go [2026-07-17 04:07:47] step 148500/1000000 | loss=2.3809 vocab=0.8876 token=1.3812 | 6.09 steps/s | VRAM=26.59 Go [2026-07-17 04:07:55] step 148550/1000000 | loss=2.5161 vocab=1.0256 token=1.6323 | 6.08 steps/s | VRAM=28.42 Go [2026-07-17 04:08:04] step 148600/1000000 | loss=2.4404 vocab=1.2282 token=1.6818 | 6.06 steps/s | VRAM=26.63 Go [2026-07-17 04:08:12] step 148650/1000000 | loss=2.3897 vocab=0.9058 token=1.4134 | 6.11 steps/s | VRAM=27.76 Go [2026-07-17 04:08:20] step 148700/1000000 | loss=2.3715 vocab=1.0151 token=1.7324 | 6.09 steps/s | VRAM=26.45 Go [2026-07-17 04:08:28] step 148750/1000000 | loss=2.4832 vocab=0.8192 token=1.5514 | 6.07 steps/s | VRAM=26.32 Go [2026-07-17 04:08:36] step 148800/1000000 | loss=2.4155 vocab=0.7232 token=1.2912 | 6.14 steps/s | VRAM=25.13 Go [2026-07-17 04:08:45] step 148850/1000000 | loss=2.4058 vocab=0.9927 token=1.7286 | 6.07 steps/s | VRAM=29.30 Go [2026-07-17 04:08:53] step 148900/1000000 | loss=2.3590 vocab=0.7458 token=1.2254 | 6.09 steps/s | VRAM=29.20 Go [2026-07-17 04:09:01] step 148950/1000000 | loss=2.4271 vocab=1.4314 token=1.9223 | 6.08 steps/s | VRAM=26.12 Go [2026-07-17 04:09:09] step 149000/1000000 | loss=2.4975 vocab=1.0948 token=1.6145 | 6.07 steps/s | VRAM=32.75 Go [2026-07-17 04:09:18] step 149050/1000000 | loss=2.4376 vocab=1.3814 token=1.6871 | 6.10 steps/s | VRAM=31.50 Go [2026-07-17 04:09:26] step 149100/1000000 | loss=2.2715 vocab=1.2248 token=1.4000 | 6.08 steps/s | VRAM=32.62 Go [2026-07-17 04:09:34] step 149150/1000000 | loss=2.3965 vocab=0.9777 token=1.8148 | 6.10 steps/s | VRAM=28.02 Go [2026-07-17 04:09:42] step 149200/1000000 | loss=2.4248 vocab=0.9748 token=1.9671 | 6.12 steps/s | VRAM=27.87 Go [2026-07-17 04:09:50] step 149250/1000000 | loss=2.5359 vocab=1.2242 token=1.4587 | 6.07 steps/s | VRAM=29.19 Go [2026-07-17 04:09:59] step 149300/1000000 | loss=2.4883 vocab=0.9582 token=1.5424 | 6.03 steps/s | VRAM=25.03 Go [2026-07-17 04:10:07] step 149350/1000000 | loss=2.4313 vocab=0.8233 token=1.3316 | 6.11 steps/s | VRAM=28.20 Go [2026-07-17 04:10:15] step 149400/1000000 | loss=2.4969 vocab=1.5002 token=1.6097 | 6.09 steps/s | VRAM=31.65 Go [2026-07-17 04:10:23] step 149450/1000000 | loss=2.4410 vocab=1.5169 token=1.3798 | 6.11 steps/s | VRAM=30.67 Go [2026-07-17 04:10:31] step 149500/1000000 | loss=2.4095 vocab=0.8668 token=1.2440 | 6.08 steps/s | VRAM=29.69 Go [2026-07-17 04:10:40] step 149550/1000000 | loss=2.4259 vocab=1.3880 token=1.6749 | 6.08 steps/s | VRAM=25.60 Go [2026-07-17 04:10:48] step 149600/1000000 | loss=2.4399 vocab=0.7870 token=1.1470 | 6.09 steps/s | VRAM=32.03 Go [2026-07-17 04:10:56] step 149650/1000000 | loss=2.4610 vocab=1.0665 token=1.4736 | 6.06 steps/s | VRAM=28.36 Go [2026-07-17 04:11:04] step 149700/1000000 | loss=2.4563 vocab=0.8017 token=1.5315 | 6.09 steps/s | VRAM=32.03 Go [2026-07-17 04:11:13] step 149750/1000000 | loss=2.3939 vocab=1.0913 token=1.2033 | 6.05 steps/s | VRAM=26.20 Go [2026-07-17 04:11:21] step 149800/1000000 | loss=2.5213 vocab=0.5830 token=1.1404 | 6.09 steps/s | VRAM=31.43 Go [2026-07-17 04:11:29] step 149850/1000000 | loss=2.3709 vocab=1.3569 token=1.5402 | 6.11 steps/s | VRAM=24.92 Go [2026-07-17 04:11:37] step 149900/1000000 | loss=2.4941 vocab=1.0665 token=1.6159 | 6.11 steps/s | VRAM=28.75 Go [2026-07-17 04:11:45] step 149950/1000000 | loss=2.5849 vocab=0.9192 token=1.4906 | 6.09 steps/s | VRAM=26.30 Go [2026-07-17 04:11:54] step 150000/1000000 | loss=2.3879 vocab=1.0124 token=1.3181 | 6.09 steps/s | VRAM=31.75 Go [2026-07-17 04:13:24] step 150050/1000000 | loss=2.4605 vocab=0.5983 token=1.2781 | 0.55 steps/s | VRAM=28.27 Go [2026-07-17 04:13:32] step 150100/1000000 | loss=2.4899 vocab=0.9258 token=1.4943 | 6.20 steps/s | VRAM=27.78 Go [2026-07-17 04:13:41] step 150150/1000000 | loss=2.4543 vocab=0.7538 token=1.2279 | 6.16 steps/s | VRAM=25.26 Go [2026-07-17 04:13:49] step 150200/1000000 | loss=2.4951 vocab=0.5729 token=1.1124 | 6.10 steps/s | VRAM=27.10 Go [2026-07-17 04:13:57] step 150250/1000000 | loss=2.4024 vocab=1.1415 token=1.7999 | 6.13 steps/s | VRAM=25.18 Go [2026-07-17 04:14:05] step 150300/1000000 | loss=2.3989 vocab=0.9823 token=0.7601 | 6.12 steps/s | VRAM=25.49 Go [2026-07-17 04:14:13] step 150350/1000000 | loss=2.4726 vocab=1.3250 token=1.6962 | 6.10 steps/s | VRAM=32.14 Go [2026-07-17 04:14:22] step 150400/1000000 | loss=2.5120 vocab=1.0078 token=1.2442 | 6.05 steps/s | VRAM=28.23 Go [2026-07-17 04:14:30] step 150450/1000000 | loss=2.3759 vocab=0.8065 token=1.6005 | 6.08 steps/s | VRAM=30.03 Go [2026-07-17 04:14:38] step 150500/1000000 | loss=2.4296 vocab=1.2674 token=1.1877 | 6.09 steps/s | VRAM=28.00 Go [2026-07-17 04:14:46] step 150550/1000000 | loss=2.4335 vocab=1.2417 token=1.4814 | 6.10 steps/s | VRAM=28.08 Go [2026-07-17 04:14:54] step 150600/1000000 | loss=2.5056 vocab=1.2255 token=1.4607 | 6.10 steps/s | VRAM=27.80 Go [2026-07-17 04:15:03] step 150650/1000000 | loss=2.3960 vocab=0.6549 token=1.1681 | 6.08 steps/s | VRAM=32.69 Go [2026-07-17 04:15:11] step 150700/1000000 | loss=2.4379 vocab=0.4779 token=1.2570 | 6.07 steps/s | VRAM=28.29 Go [2026-07-17 04:15:19] step 150750/1000000 | loss=2.2601 vocab=0.7861 token=1.2129 | 6.11 steps/s | VRAM=31.55 Go [2026-07-17 04:15:27] step 150800/1000000 | loss=2.4166 vocab=0.4880 token=1.0328 | 6.04 steps/s | VRAM=28.19 Go [2026-07-17 04:15:35] step 150850/1000000 | loss=2.5100 vocab=0.9130 token=1.4653 | 6.09 steps/s | VRAM=28.36 Go [2026-07-17 04:15:44] step 150900/1000000 | loss=2.5074 vocab=1.4435 token=1.5145 | 6.05 steps/s | VRAM=32.16 Go [2026-07-17 04:15:52] step 150950/1000000 | loss=2.5095 vocab=0.8925 token=1.2031 | 6.10 steps/s | VRAM=31.52 Go [2026-07-17 04:16:00] step 151000/1000000 | loss=2.4182 vocab=0.7067 token=1.2547 | 6.07 steps/s | VRAM=28.70 Go [2026-07-17 04:16:08] step 151050/1000000 | loss=2.3246 vocab=0.7793 token=1.5781 | 6.09 steps/s | VRAM=25.93 Go [2026-07-17 04:16:17] step 151100/1000000 | loss=2.4993 vocab=1.0953 token=1.8440 | 6.07 steps/s | VRAM=30.19 Go [2026-07-17 04:16:25] step 151150/1000000 | loss=2.4297 vocab=0.9710 token=1.5656 | 6.07 steps/s | VRAM=28.68 Go [2026-07-17 04:16:33] step 151200/1000000 | loss=2.3606 vocab=0.9154 token=1.9774 | 6.09 steps/s | VRAM=32.21 Go [2026-07-17 04:16:41] step 151250/1000000 | loss=2.4852 vocab=0.6155 token=1.4249 | 6.08 steps/s | VRAM=28.02 Go [2026-07-17 04:16:50] step 151300/1000000 | loss=2.4886 vocab=0.7622 token=1.4616 | 6.07 steps/s | VRAM=29.10 Go [2026-07-17 04:16:58] step 151350/1000000 | loss=2.4953 vocab=1.0711 token=1.7742 | 6.08 steps/s | VRAM=32.04 Go [2026-07-17 04:17:06] step 151400/1000000 | loss=2.3933 vocab=0.2922 token=1.1657 | 6.07 steps/s | VRAM=28.65 Go [2026-07-17 04:17:14] step 151450/1000000 | loss=2.4362 vocab=1.0464 token=1.5994 | 6.06 steps/s | VRAM=28.28 Go [2026-07-17 04:17:22] step 151500/1000000 | loss=2.5257 vocab=1.1123 token=1.3876 | 6.09 steps/s | VRAM=32.44 Go [2026-07-17 04:17:31] step 151550/1000000 | loss=2.4705 vocab=0.9769 token=1.6755 | 6.06 steps/s | VRAM=29.88 Go [2026-07-17 04:17:39] step 151600/1000000 | loss=2.5594 vocab=1.2420 token=1.7290 | 6.09 steps/s | VRAM=28.07 Go [2026-07-17 04:17:47] step 151650/1000000 | loss=2.3933 vocab=1.0552 token=1.5061 | 6.07 steps/s | VRAM=29.27 Go [2026-07-17 04:17:55] step 151700/1000000 | loss=2.3780 vocab=0.8824 token=1.3520 | 6.08 steps/s | VRAM=28.90 Go [2026-07-17 04:18:04] step 151750/1000000 | loss=2.4771 vocab=0.9660 token=1.7719 | 6.06 steps/s | VRAM=25.98 Go [2026-07-17 04:18:12] step 151800/1000000 | loss=2.5216 vocab=1.3064 token=1.6826 | 6.09 steps/s | VRAM=28.11 Go [2026-07-17 04:18:20] step 151850/1000000 | loss=2.3450 vocab=0.9080 token=1.3019 | 6.05 steps/s | VRAM=26.26 Go [2026-07-17 04:18:28] step 151900/1000000 | loss=2.4506 vocab=1.0262 token=2.1227 | 6.07 steps/s | VRAM=28.09 Go [2026-07-17 04:18:37] step 151950/1000000 | loss=2.3408 vocab=1.2048 token=1.6773 | 6.10 steps/s | VRAM=28.26 Go [2026-07-17 04:18:45] step 152000/1000000 | loss=2.3776 vocab=0.9762 token=1.4659 | 6.08 steps/s | VRAM=23.73 Go [2026-07-17 04:18:53] step 152050/1000000 | loss=2.3963 vocab=1.1563 token=1.6441 | 6.13 steps/s | VRAM=31.28 Go [2026-07-17 04:19:01] step 152100/1000000 | loss=2.4314 vocab=0.7527 token=1.2278 | 6.09 steps/s | VRAM=32.38 Go [2026-07-17 04:19:09] step 152150/1000000 | loss=2.4811 vocab=0.6025 token=1.2872 | 6.04 steps/s | VRAM=25.95 Go [2026-07-17 04:19:18] step 152200/1000000 | loss=2.5362 vocab=0.8564 token=1.5593 | 6.10 steps/s | VRAM=32.74 Go [2026-07-17 04:19:26] step 152250/1000000 | loss=2.3124 vocab=1.4220 token=1.2869 | 6.07 steps/s | VRAM=28.28 Go [2026-07-17 04:19:34] step 152300/1000000 | loss=2.4529 vocab=0.9856 token=1.7759 | 6.05 steps/s | VRAM=32.25 Go [2026-07-17 04:19:42] step 152350/1000000 | loss=2.3361 vocab=1.6609 token=1.2668 | 6.06 steps/s | VRAM=32.92 Go [2026-07-17 04:19:51] step 152400/1000000 | loss=2.4554 vocab=1.0874 token=1.4006 | 6.09 steps/s | VRAM=28.34 Go [2026-07-17 04:19:59] step 152450/1000000 | loss=2.4280 vocab=1.1604 token=1.3059 | 6.06 steps/s | VRAM=25.66 Go [2026-07-17 04:20:07] step 152500/1000000 | loss=2.4585 vocab=0.6415 token=1.3229 | 6.10 steps/s | VRAM=26.00 Go [2026-07-17 04:20:15] step 152550/1000000 | loss=2.4536 vocab=1.0644 token=1.6583 | 6.09 steps/s | VRAM=32.29 Go [2026-07-17 04:20:23] step 152600/1000000 | loss=2.4640 vocab=0.5249 token=1.1763 | 6.12 steps/s | VRAM=32.51 Go [2026-07-17 04:20:32] step 152650/1000000 | loss=2.4381 vocab=0.5793 token=1.6774 | 6.13 steps/s | VRAM=31.60 Go [2026-07-17 04:20:40] step 152700/1000000 | loss=2.4980 vocab=1.2014 token=1.5573 | 6.10 steps/s | VRAM=28.83 Go [2026-07-17 04:20:48] step 152750/1000000 | loss=2.4830 vocab=1.2820 token=1.4318 | 6.05 steps/s | VRAM=29.46 Go [2026-07-17 04:20:56] step 152800/1000000 | loss=2.4749 vocab=1.3533 token=1.4200 | 6.07 steps/s | VRAM=33.00 Go [2026-07-17 04:21:04] step 152850/1000000 | loss=2.4053 vocab=1.0763 token=1.4856 | 6.09 steps/s | VRAM=32.69 Go [2026-07-17 04:21:13] step 152900/1000000 | loss=2.4647 vocab=0.7096 token=1.5170 | 6.08 steps/s | VRAM=26.63 Go [2026-07-17 04:21:21] step 152950/1000000 | loss=2.5737 vocab=0.6753 token=1.2266 | 6.08 steps/s | VRAM=28.56 Go [2026-07-17 04:21:29] step 153000/1000000 | loss=2.3858 vocab=1.1076 token=1.3712 | 6.06 steps/s | VRAM=29.26 Go [2026-07-17 04:21:37] step 153050/1000000 | loss=2.2383 vocab=1.1729 token=1.3991 | 6.09 steps/s | VRAM=26.54 Go [2026-07-17 04:21:46] step 153100/1000000 | loss=2.4986 vocab=1.5689 token=1.9418 | 6.09 steps/s | VRAM=32.73 Go [2026-07-17 04:21:54] step 153150/1000000 | loss=2.4307 vocab=1.2929 token=1.4728 | 6.10 steps/s | VRAM=24.96 Go [2026-07-17 04:22:02] step 153200/1000000 | loss=2.5095 vocab=1.1409 token=1.6566 | 6.06 steps/s | VRAM=28.71 Go [2026-07-17 04:22:10] step 153250/1000000 | loss=2.4707 vocab=0.8547 token=1.4653 | 6.07 steps/s | VRAM=26.60 Go [2026-07-17 04:22:19] step 153300/1000000 | loss=2.3481 vocab=1.2333 token=1.4786 | 6.05 steps/s | VRAM=28.81 Go [2026-07-17 04:22:27] step 153350/1000000 | loss=2.5151 vocab=2.6293 token=1.6941 | 6.08 steps/s | VRAM=26.04 Go [2026-07-17 04:22:35] step 153400/1000000 | loss=2.4595 vocab=1.0364 token=1.2859 | 6.09 steps/s | VRAM=31.48 Go [2026-07-17 04:22:43] step 153450/1000000 | loss=2.4812 vocab=0.5575 token=1.5484 | 6.07 steps/s | VRAM=28.01 Go [2026-07-17 04:22:51] step 153500/1000000 | loss=2.5361 vocab=0.9625 token=1.4872 | 6.10 steps/s | VRAM=30.82 Go [2026-07-17 04:23:00] step 153550/1000000 | loss=2.4812 vocab=1.0755 token=1.6669 | 6.11 steps/s | VRAM=29.91 Go [2026-07-17 04:23:08] step 153600/1000000 | loss=2.5247 vocab=0.5152 token=1.3668 | 6.10 steps/s | VRAM=28.85 Go [2026-07-17 04:23:16] step 153650/1000000 | loss=2.3529 vocab=0.7991 token=1.1196 | 6.12 steps/s | VRAM=29.35 Go [2026-07-17 04:23:24] step 153700/1000000 | loss=2.5319 vocab=1.1485 token=1.5628 | 6.10 steps/s | VRAM=28.80 Go [2026-07-17 04:23:32] step 153750/1000000 | loss=2.4870 vocab=0.8610 token=1.5139 | 6.10 steps/s | VRAM=28.97 Go [2026-07-17 04:23:41] step 153800/1000000 | loss=2.4462 vocab=1.6318 token=1.6512 | 6.10 steps/s | VRAM=28.08 Go [2026-07-17 04:23:49] step 153850/1000000 | loss=2.5129 vocab=1.2041 token=1.1701 | 6.09 steps/s | VRAM=26.46 Go [2026-07-17 04:23:57] step 153900/1000000 | loss=2.3216 vocab=0.9450 token=1.2807 | 6.11 steps/s | VRAM=31.91 Go [2026-07-17 04:24:05] step 153950/1000000 | loss=2.5072 vocab=1.0548 token=1.8700 | 6.09 steps/s | VRAM=24.26 Go [2026-07-17 04:24:13] step 154000/1000000 | loss=2.5333 vocab=1.5454 token=1.6324 | 6.12 steps/s | VRAM=28.65 Go [2026-07-17 04:24:22] step 154050/1000000 | loss=2.3689 vocab=1.0936 token=1.2840 | 6.08 steps/s | VRAM=28.46 Go [2026-07-17 04:24:30] step 154100/1000000 | loss=2.2784 vocab=0.9704 token=1.5768 | 6.07 steps/s | VRAM=25.70 Go [2026-07-17 04:24:38] step 154150/1000000 | loss=2.4250 vocab=0.7132 token=1.0769 | 6.12 steps/s | VRAM=32.21 Go [2026-07-17 04:24:46] step 154200/1000000 | loss=2.5609 vocab=0.8360 token=1.9437 | 6.08 steps/s | VRAM=26.81 Go [2026-07-17 04:24:54] step 154250/1000000 | loss=2.4722 vocab=1.0219 token=1.4916 | 6.11 steps/s | VRAM=25.82 Go [2026-07-17 04:25:03] step 154300/1000000 | loss=2.3556 vocab=0.4701 token=1.4944 | 6.12 steps/s | VRAM=26.48 Go [2026-07-17 04:25:11] step 154350/1000000 | loss=2.4383 vocab=1.2467 token=1.5070 | 6.09 steps/s | VRAM=28.68 Go [2026-07-17 04:25:19] step 154400/1000000 | loss=2.4566 vocab=0.7701 token=1.5198 | 6.12 steps/s | VRAM=29.48 Go [2026-07-17 04:25:27] step 154450/1000000 | loss=2.3497 vocab=0.7433 token=1.2996 | 6.10 steps/s | VRAM=25.44 Go [2026-07-17 04:25:35] step 154500/1000000 | loss=2.5620 vocab=0.8558 token=1.4539 | 6.13 steps/s | VRAM=26.81 Go [2026-07-17 04:25:43] step 154550/1000000 | loss=2.5189 vocab=0.8690 token=1.5385 | 6.12 steps/s | VRAM=30.22 Go [2026-07-17 04:25:52] step 154600/1000000 | loss=2.4652 vocab=0.9458 token=1.3200 | 6.10 steps/s | VRAM=28.22 Go [2026-07-17 04:26:00] step 154650/1000000 | loss=2.3773 vocab=0.9803 token=1.1682 | 6.08 steps/s | VRAM=28.63 Go [2026-07-17 04:26:08] step 154700/1000000 | loss=2.3789 vocab=0.9647 token=1.7371 | 6.13 steps/s | VRAM=25.80 Go [2026-07-17 04:26:16] step 154750/1000000 | loss=2.4383 vocab=0.6628 token=1.2953 | 6.09 steps/s | VRAM=25.58 Go [2026-07-17 04:26:24] step 154800/1000000 | loss=2.4550 vocab=1.2512 token=1.6160 | 6.13 steps/s | VRAM=28.60 Go [2026-07-17 04:26:33] step 154850/1000000 | loss=2.5172 vocab=0.6296 token=1.3814 | 6.07 steps/s | VRAM=27.85 Go [2026-07-17 04:26:41] step 154900/1000000 | loss=2.5420 vocab=0.7853 token=1.5497 | 6.11 steps/s | VRAM=32.61 Go [2026-07-17 04:26:49] step 154950/1000000 | loss=2.4211 vocab=1.0366 token=1.4747 | 6.11 steps/s | VRAM=29.15 Go [2026-07-17 04:26:57] step 155000/1000000 | loss=2.4100 vocab=1.3169 token=1.7607 | 6.12 steps/s | VRAM=31.55 Go [2026-07-17 04:28:16] step 155050/1000000 | loss=2.4265 vocab=0.8403 token=1.2243 | 0.63 steps/s | VRAM=29.25 Go [2026-07-17 04:28:24] step 155100/1000000 | loss=2.3999 vocab=0.5687 token=1.1694 | 6.18 steps/s | VRAM=28.72 Go [2026-07-17 04:28:32] step 155150/1000000 | loss=2.4373 vocab=0.7087 token=1.3167 | 6.12 steps/s | VRAM=29.39 Go [2026-07-17 04:28:40] step 155200/1000000 | loss=2.3123 vocab=0.3625 token=1.0587 | 6.16 steps/s | VRAM=31.56 Go [2026-07-17 04:28:49] step 155250/1000000 | loss=2.4722 vocab=1.3317 token=1.4737 | 6.16 steps/s | VRAM=28.91 Go [2026-07-17 04:28:57] step 155300/1000000 | loss=2.4301 vocab=0.7537 token=1.1531 | 6.12 steps/s | VRAM=26.92 Go [2026-07-17 04:29:05] step 155350/1000000 | loss=2.4042 vocab=0.8875 token=1.9070 | 6.14 steps/s | VRAM=32.75 Go [2026-07-17 04:29:13] step 155400/1000000 | loss=2.3519 vocab=0.4004 token=1.0716 | 6.14 steps/s | VRAM=32.39 Go [2026-07-17 04:29:21] step 155450/1000000 | loss=2.3569 vocab=1.1017 token=2.3004 | 6.11 steps/s | VRAM=26.02 Go [2026-07-17 04:29:29] step 155500/1000000 | loss=2.5555 vocab=1.0619 token=1.4005 | 6.10 steps/s | VRAM=28.61 Go [2026-07-17 04:29:38] step 155550/1000000 | loss=2.4602 vocab=0.8949 token=1.2701 | 6.12 steps/s | VRAM=31.65 Go [2026-07-17 04:29:46] step 155600/1000000 | loss=2.3911 vocab=0.7999 token=1.4730 | 6.09 steps/s | VRAM=29.61 Go [2026-07-17 04:29:54] step 155650/1000000 | loss=2.4838 vocab=0.9584 token=1.2957 | 6.07 steps/s | VRAM=32.74 Go [2026-07-17 04:30:02] step 155700/1000000 | loss=2.2544 vocab=0.8483 token=1.1651 | 6.09 steps/s | VRAM=28.57 Go [2026-07-17 04:30:10] step 155750/1000000 | loss=2.4366 vocab=1.0067 token=1.6793 | 6.09 steps/s | VRAM=29.60 Go [2026-07-17 04:30:19] step 155800/1000000 | loss=2.5045 vocab=1.4852 token=2.0458 | 6.08 steps/s | VRAM=28.81 Go [2026-07-17 04:30:27] step 155850/1000000 | loss=2.3248 vocab=0.7941 token=1.6030 | 6.07 steps/s | VRAM=25.82 Go [2026-07-17 04:30:35] step 155900/1000000 | loss=2.5219 vocab=0.8379 token=1.7946 | 6.10 steps/s | VRAM=28.84 Go [2026-07-17 04:30:43] step 155950/1000000 | loss=2.4485 vocab=0.9863 token=1.5311 | 6.10 steps/s | VRAM=31.04 Go [2026-07-17 04:30:51] step 156000/1000000 | loss=2.4888 vocab=0.9183 token=1.5157 | 6.08 steps/s | VRAM=28.89 Go [2026-07-17 04:31:00] step 156050/1000000 | loss=2.4020 vocab=1.1251 token=1.5833 | 6.10 steps/s | VRAM=29.46 Go [2026-07-17 04:31:08] step 156100/1000000 | loss=2.5323 vocab=0.8893 token=1.3912 | 6.10 steps/s | VRAM=25.60 Go [2026-07-17 04:31:16] step 156150/1000000 | loss=2.4350 vocab=0.6688 token=1.3750 | 6.11 steps/s | VRAM=28.59 Go [2026-07-17 04:31:24] step 156200/1000000 | loss=2.4053 vocab=1.0384 token=1.5260 | 6.09 steps/s | VRAM=29.08 Go [2026-07-17 04:31:32] step 156250/1000000 | loss=2.4075 vocab=0.6093 token=1.2885 | 6.08 steps/s | VRAM=25.78 Go [2026-07-17 04:31:41] step 156300/1000000 | loss=2.4184 vocab=0.7410 token=1.4033 | 6.08 steps/s | VRAM=25.29 Go [2026-07-17 04:31:49] step 156350/1000000 | loss=2.4144 vocab=0.9814 token=1.3238 | 6.12 steps/s | VRAM=31.75 Go [2026-07-17 04:31:57] step 156400/1000000 | loss=2.4248 vocab=0.7147 token=1.6100 | 6.08 steps/s | VRAM=26.30 Go [2026-07-17 04:32:05] step 156450/1000000 | loss=2.5162 vocab=0.9682 token=1.3843 | 6.09 steps/s | VRAM=28.26 Go [2026-07-17 04:32:14] step 156500/1000000 | loss=2.5134 vocab=1.0019 token=1.4900 | 6.06 steps/s | VRAM=26.26 Go [2026-07-17 04:32:22] step 156550/1000000 | loss=2.4512 vocab=0.9938 token=1.6462 | 6.06 steps/s | VRAM=28.53 Go [2026-07-17 04:32:30] step 156600/1000000 | loss=2.4683 vocab=1.1362 token=1.5337 | 6.09 steps/s | VRAM=28.27 Go [2026-07-17 04:32:38] step 156650/1000000 | loss=2.4346 vocab=0.9543 token=1.2274 | 6.09 steps/s | VRAM=29.17 Go [2026-07-17 04:32:46] step 156700/1000000 | loss=2.3307 vocab=0.8770 token=1.1376 | 6.12 steps/s | VRAM=31.92 Go [2026-07-17 04:32:55] step 156750/1000000 | loss=2.3590 vocab=0.7617 token=1.4225 | 6.10 steps/s | VRAM=32.20 Go [2026-07-17 04:33:03] step 156800/1000000 | loss=2.4891 vocab=1.0427 token=1.3996 | 6.08 steps/s | VRAM=26.27 Go [2026-07-17 04:33:11] step 156850/1000000 | loss=2.3285 vocab=0.9624 token=0.9732 | 6.09 steps/s | VRAM=28.27 Go [2026-07-17 04:33:19] step 156900/1000000 | loss=2.4566 vocab=0.7856 token=1.8108 | 6.09 steps/s | VRAM=28.12 Go [2026-07-17 04:33:27] step 156950/1000000 | loss=2.3002 vocab=0.7396 token=0.9865 | 6.10 steps/s | VRAM=24.43 Go [2026-07-17 04:33:36] step 157000/1000000 | loss=2.4553 vocab=1.4752 token=1.4202 | 6.13 steps/s | VRAM=27.86 Go [2026-07-17 04:33:44] step 157050/1000000 | loss=2.3034 vocab=1.1592 token=1.6013 | 6.10 steps/s | VRAM=29.48 Go [2026-07-17 04:33:52] step 157100/1000000 | loss=2.4346 vocab=1.4536 token=1.3133 | 6.11 steps/s | VRAM=31.84 Go [2026-07-17 04:34:00] step 157150/1000000 | loss=2.4613 vocab=1.4333 token=1.6381 | 6.11 steps/s | VRAM=24.47 Go [2026-07-17 04:34:08] step 157200/1000000 | loss=2.5978 vocab=1.3403 token=1.5579 | 6.09 steps/s | VRAM=29.46 Go [2026-07-17 04:34:17] step 157250/1000000 | loss=2.3761 vocab=0.7151 token=1.3720 | 6.06 steps/s | VRAM=26.34 Go [2026-07-17 04:34:25] step 157300/1000000 | loss=2.3269 vocab=0.8007 token=1.3619 | 6.12 steps/s | VRAM=28.55 Go [2026-07-17 04:34:33] step 157350/1000000 | loss=2.4060 vocab=0.7965 token=1.4559 | 6.08 steps/s | VRAM=28.86 Go [2026-07-17 04:34:41] step 157400/1000000 | loss=2.3819 vocab=0.8991 token=1.1651 | 6.07 steps/s | VRAM=30.49 Go [2026-07-17 04:34:49] step 157450/1000000 | loss=2.5279 vocab=1.5426 token=2.2318 | 6.11 steps/s | VRAM=31.61 Go [2026-07-17 04:34:58] step 157500/1000000 | loss=2.4173 vocab=0.7504 token=1.4852 | 6.11 steps/s | VRAM=25.57 Go [2026-07-17 04:35:06] step 157550/1000000 | loss=2.4881 vocab=0.6889 token=1.5493 | 6.09 steps/s | VRAM=28.67 Go [2026-07-17 04:35:14] step 157600/1000000 | loss=2.4111 vocab=1.2272 token=1.7287 | 6.11 steps/s | VRAM=28.97 Go [2026-07-17 04:35:22] step 157650/1000000 | loss=2.4803 vocab=0.8581 token=1.7716 | 6.11 steps/s | VRAM=27.95 Go [2026-07-17 04:35:30] step 157700/1000000 | loss=2.4665 vocab=0.9081 token=1.4114 | 6.10 steps/s | VRAM=32.23 Go [2026-07-17 04:35:39] step 157750/1000000 | loss=2.3707 vocab=1.3326 token=1.6615 | 6.08 steps/s | VRAM=28.94 Go [2026-07-17 04:35:47] step 157800/1000000 | loss=2.4849 vocab=0.6008 token=1.0890 | 6.08 steps/s | VRAM=28.93 Go [2026-07-17 04:35:55] step 157850/1000000 | loss=2.4051 vocab=1.6128 token=1.9834 | 6.10 steps/s | VRAM=28.14 Go [2026-07-17 04:36:03] step 157900/1000000 | loss=2.5001 vocab=0.8618 token=1.6846 | 6.11 steps/s | VRAM=27.88 Go [2026-07-17 04:36:11] step 157950/1000000 | loss=2.4280 vocab=1.2164 token=1.4728 | 6.11 steps/s | VRAM=31.92 Go [2026-07-17 04:36:20] step 158000/1000000 | loss=2.4572 vocab=0.7936 token=1.5014 | 6.06 steps/s | VRAM=28.78 Go [2026-07-17 04:36:28] step 158050/1000000 | loss=2.5251 vocab=0.6495 token=1.4936 | 6.08 steps/s | VRAM=26.53 Go [2026-07-17 04:36:36] step 158100/1000000 | loss=2.2877 vocab=0.9183 token=1.3828 | 6.11 steps/s | VRAM=32.25 Go [2026-07-17 04:36:44] step 158150/1000000 | loss=2.3995 vocab=1.1556 token=1.5422 | 6.10 steps/s | VRAM=32.58 Go [2026-07-17 04:36:52] step 158200/1000000 | loss=2.4555 vocab=1.2257 token=1.3164 | 6.10 steps/s | VRAM=28.09 Go [2026-07-17 04:37:01] step 158250/1000000 | loss=2.3738 vocab=0.9713 token=1.3501 | 6.10 steps/s | VRAM=29.18 Go [2026-07-17 04:37:09] step 158300/1000000 | loss=2.3288 vocab=0.7306 token=1.2713 | 6.10 steps/s | VRAM=25.98 Go [2026-07-17 04:37:17] step 158350/1000000 | loss=2.3876 vocab=0.9967 token=1.4636 | 6.10 steps/s | VRAM=28.26 Go [2026-07-17 04:37:25] step 158400/1000000 | loss=2.5035 vocab=0.8702 token=1.4033 | 6.12 steps/s | VRAM=32.58 Go [2026-07-17 04:37:33] step 158450/1000000 | loss=2.3450 vocab=0.4302 token=1.0985 | 6.12 steps/s | VRAM=28.72 Go [2026-07-17 04:37:41] step 158500/1000000 | loss=2.4507 vocab=1.1504 token=1.4665 | 6.16 steps/s | VRAM=31.59 Go [2026-07-17 04:37:50] step 158550/1000000 | loss=2.3430 vocab=0.7777 token=1.1345 | 6.14 steps/s | VRAM=25.63 Go [2026-07-17 04:37:58] step 158600/1000000 | loss=2.5146 vocab=1.2523 token=1.4832 | 6.12 steps/s | VRAM=28.58 Go [2026-07-17 04:38:06] step 158650/1000000 | loss=2.3582 vocab=0.6824 token=1.6761 | 6.14 steps/s | VRAM=32.19 Go [2026-07-17 04:38:14] step 158700/1000000 | loss=2.4148 vocab=1.2818 token=1.4558 | 6.11 steps/s | VRAM=28.79 Go [2026-07-17 04:38:22] step 158750/1000000 | loss=2.4096 vocab=0.5985 token=1.3541 | 6.09 steps/s | VRAM=31.88 Go [2026-07-17 04:38:31] step 158800/1000000 | loss=2.4635 vocab=1.5038 token=1.8362 | 6.08 steps/s | VRAM=26.80 Go [2026-07-17 04:38:39] step 158850/1000000 | loss=2.3030 vocab=0.9950 token=1.3402 | 6.13 steps/s | VRAM=31.49 Go [2026-07-17 04:38:47] step 158900/1000000 | loss=2.4226 vocab=0.7131 token=1.1197 | 6.09 steps/s | VRAM=28.98 Go [2026-07-17 04:38:55] step 158950/1000000 | loss=2.3048 vocab=1.1910 token=1.6101 | 6.11 steps/s | VRAM=28.55 Go [2026-07-17 04:39:03] step 159000/1000000 | loss=2.4036 vocab=0.8590 token=1.3595 | 6.09 steps/s | VRAM=28.72 Go [2026-07-17 04:39:12] step 159050/1000000 | loss=2.4806 vocab=1.2471 token=1.2373 | 6.10 steps/s | VRAM=31.94 Go [2026-07-17 04:39:20] step 159100/1000000 | loss=2.4669 vocab=1.2990 token=1.8383 | 6.11 steps/s | VRAM=27.96 Go [2026-07-17 04:39:28] step 159150/1000000 | loss=2.5015 vocab=0.9067 token=1.5388 | 6.05 steps/s | VRAM=27.06 Go [2026-07-17 04:39:36] step 159200/1000000 | loss=2.3043 vocab=0.9258 token=1.4120 | 6.09 steps/s | VRAM=28.89 Go [2026-07-17 04:39:44] step 159250/1000000 | loss=2.4656 vocab=0.8130 token=1.3421 | 6.09 steps/s | VRAM=32.46 Go [2026-07-17 04:39:53] step 159300/1000000 | loss=2.3919 vocab=1.3104 token=1.6935 | 6.09 steps/s | VRAM=29.31 Go [2026-07-17 04:40:01] step 159350/1000000 | loss=2.2898 vocab=1.0559 token=1.7268 | 6.07 steps/s | VRAM=26.48 Go [2026-07-17 04:40:09] step 159400/1000000 | loss=2.3681 vocab=0.8173 token=1.7264 | 6.13 steps/s | VRAM=31.33 Go [2026-07-17 04:40:17] step 159450/1000000 | loss=2.4764 vocab=0.9090 token=1.4955 | 6.09 steps/s | VRAM=29.52 Go [2026-07-17 04:40:25] step 159500/1000000 | loss=2.4579 vocab=0.9153 token=1.8131 | 6.08 steps/s | VRAM=28.95 Go [2026-07-17 04:40:34] step 159550/1000000 | loss=2.4134 vocab=1.0322 token=1.4484 | 6.10 steps/s | VRAM=28.40 Go [2026-07-17 04:40:42] step 159600/1000000 | loss=2.4239 vocab=1.1947 token=1.6224 | 6.09 steps/s | VRAM=31.88 Go [2026-07-17 04:40:50] step 159650/1000000 | loss=2.4083 vocab=0.4000 token=1.0584 | 6.09 steps/s | VRAM=26.26 Go [2026-07-17 04:40:58] step 159700/1000000 | loss=2.4625 vocab=0.5811 token=1.2363 | 6.11 steps/s | VRAM=28.23 Go [2026-07-17 04:41:06] step 159750/1000000 | loss=2.4256 vocab=0.9197 token=1.4859 | 6.10 steps/s | VRAM=32.76 Go [2026-07-17 04:41:15] step 159800/1000000 | loss=2.4431 vocab=0.6851 token=1.0875 | 6.09 steps/s | VRAM=26.17 Go [2026-07-17 04:41:23] step 159850/1000000 | loss=2.5336 vocab=0.7833 token=1.1291 | 6.09 steps/s | VRAM=28.99 Go [2026-07-17 04:41:31] step 159900/1000000 | loss=2.4889 vocab=1.2448 token=1.6939 | 6.13 steps/s | VRAM=24.32 Go [2026-07-17 04:41:39] step 159950/1000000 | loss=2.3194 vocab=0.9038 token=0.7217 | 6.09 steps/s | VRAM=28.68 Go [2026-07-17 04:41:47] step 160000/1000000 | loss=2.4509 vocab=0.7495 token=1.0946 | 6.11 steps/s | VRAM=32.28 Go [2026-07-17 04:43:07] step 160050/1000000 | loss=2.5380 vocab=1.0757 token=1.7312 | 0.63 steps/s | VRAM=31.59 Go [2026-07-17 04:43:15] step 160100/1000000 | loss=2.4058 vocab=1.4672 token=2.0044 | 6.18 steps/s | VRAM=28.24 Go [2026-07-17 04:43:24] step 160150/1000000 | loss=2.4113 vocab=0.6945 token=1.5899 | 6.14 steps/s | VRAM=32.64 Go [2026-07-17 04:43:32] step 160200/1000000 | loss=2.3260 vocab=1.1641 token=1.5795 | 6.16 steps/s | VRAM=28.14 Go [2026-07-17 04:43:40] step 160250/1000000 | loss=2.4209 vocab=0.6966 token=1.3506 | 6.15 steps/s | VRAM=28.95 Go [2026-07-17 04:43:48] step 160300/1000000 | loss=2.4748 vocab=0.7464 token=1.5997 | 6.10 steps/s | VRAM=29.86 Go [2026-07-17 04:43:56] step 160350/1000000 | loss=2.4206 vocab=0.8147 token=1.4899 | 6.17 steps/s | VRAM=31.41 Go [2026-07-17 04:44:04] step 160400/1000000 | loss=2.4059 vocab=0.6569 token=1.5477 | 6.08 steps/s | VRAM=32.27 Go [2026-07-17 04:44:13] step 160450/1000000 | loss=2.2938 vocab=0.9928 token=1.2707 | 6.11 steps/s | VRAM=28.73 Go [2026-07-17 04:44:21] step 160500/1000000 | loss=2.3748 vocab=0.7357 token=1.3454 | 6.12 steps/s | VRAM=29.03 Go [2026-07-17 04:44:29] step 160550/1000000 | loss=2.4034 vocab=1.0746 token=1.7164 | 6.11 steps/s | VRAM=29.08 Go [2026-07-17 04:44:37] step 160600/1000000 | loss=2.4707 vocab=1.1245 token=1.7842 | 6.09 steps/s | VRAM=26.38 Go [2026-07-17 04:44:45] step 160650/1000000 | loss=2.4092 vocab=0.9551 token=1.6269 | 6.11 steps/s | VRAM=31.80 Go [2026-07-17 04:44:54] step 160700/1000000 | loss=2.4152 vocab=0.8435 token=1.4434 | 6.08 steps/s | VRAM=28.96 Go [2026-07-17 04:45:02] step 160750/1000000 | loss=2.4204 vocab=1.1332 token=1.7345 | 6.10 steps/s | VRAM=31.33 Go [2026-07-17 04:45:10] step 160800/1000000 | loss=2.4700 vocab=1.0258 token=1.6642 | 6.09 steps/s | VRAM=28.85 Go [2026-07-17 04:45:18] step 160850/1000000 | loss=2.3960 vocab=1.5893 token=1.5872 | 6.09 steps/s | VRAM=29.36 Go [2026-07-17 04:45:26] step 160900/1000000 | loss=2.3502 vocab=0.6858 token=1.0885 | 6.12 steps/s | VRAM=29.33 Go [2026-07-17 04:45:35] step 160950/1000000 | loss=2.3688 vocab=0.7193 token=1.0260 | 6.10 steps/s | VRAM=32.41 Go [2026-07-17 04:45:43] step 161000/1000000 | loss=2.4008 vocab=0.9065 token=1.3528 | 6.11 steps/s | VRAM=29.00 Go [2026-07-17 04:45:51] step 161050/1000000 | loss=2.4131 vocab=0.8610 token=1.0970 | 6.08 steps/s | VRAM=24.84 Go [2026-07-17 04:45:59] step 161100/1000000 | loss=2.3686 vocab=0.8223 token=1.3196 | 6.07 steps/s | VRAM=30.10 Go [2026-07-17 04:46:07] step 161150/1000000 | loss=2.5000 vocab=0.7771 token=1.3627 | 6.08 steps/s | VRAM=32.46 Go [2026-07-17 04:46:16] step 161200/1000000 | loss=2.4153 vocab=0.5751 token=1.0487 | 6.08 steps/s | VRAM=28.30 Go [2026-07-17 04:46:24] step 161250/1000000 | loss=2.4355 vocab=1.1798 token=1.5222 | 6.08 steps/s | VRAM=25.00 Go [2026-07-17 04:46:32] step 161300/1000000 | loss=2.5512 vocab=1.3636 token=1.6668 | 6.10 steps/s | VRAM=25.52 Go [2026-07-17 04:46:40] step 161350/1000000 | loss=2.3159 vocab=1.0861 token=1.4624 | 6.10 steps/s | VRAM=25.05 Go [2026-07-17 04:46:48] step 161400/1000000 | loss=2.4955 vocab=1.1443 token=1.4144 | 6.09 steps/s | VRAM=29.14 Go [2026-07-17 04:46:57] step 161450/1000000 | loss=2.4010 vocab=0.7817 token=1.4260 | 6.09 steps/s | VRAM=31.29 Go [2026-07-17 04:47:05] step 161500/1000000 | loss=2.4082 vocab=0.8409 token=1.1454 | 6.07 steps/s | VRAM=28.84 Go [2026-07-17 04:47:13] step 161550/1000000 | loss=2.3904 vocab=1.2185 token=1.2789 | 6.09 steps/s | VRAM=29.98 Go [2026-07-17 04:47:21] step 161600/1000000 | loss=2.3694 vocab=1.1608 token=1.5224 | 6.10 steps/s | VRAM=28.45 Go [2026-07-17 04:47:29] step 161650/1000000 | loss=2.4758 vocab=1.0745 token=1.4535 | 6.09 steps/s | VRAM=29.26 Go [2026-07-17 04:47:38] step 161700/1000000 | loss=2.4380 vocab=1.2161 token=1.3636 | 6.07 steps/s | VRAM=29.84 Go [2026-07-17 04:47:46] step 161750/1000000 | loss=2.4118 vocab=1.3423 token=1.7311 | 6.09 steps/s | VRAM=28.79 Go [2026-07-17 04:47:54] step 161800/1000000 | loss=2.4390 vocab=0.9004 token=1.2790 | 6.07 steps/s | VRAM=26.52 Go [2026-07-17 04:48:02] step 161850/1000000 | loss=2.5525 vocab=1.1792 token=1.4953 | 6.09 steps/s | VRAM=29.79 Go [2026-07-17 04:48:11] step 161900/1000000 | loss=2.3524 vocab=0.6266 token=1.3183 | 6.09 steps/s | VRAM=31.49 Go [2026-07-17 04:48:19] step 161950/1000000 | loss=2.4254 vocab=1.0681 token=0.8756 | 6.07 steps/s | VRAM=28.66 Go [2026-07-17 04:48:27] step 162000/1000000 | loss=2.4359 vocab=1.3237 token=1.1365 | 6.09 steps/s | VRAM=28.41 Go [2026-07-17 04:48:35] step 162050/1000000 | loss=2.3487 vocab=0.9557 token=1.6897 | 6.10 steps/s | VRAM=32.35 Go [2026-07-17 04:48:43] step 162100/1000000 | loss=2.3915 vocab=0.7003 token=1.4205 | 6.10 steps/s | VRAM=28.85 Go [2026-07-17 04:48:52] step 162150/1000000 | loss=2.1700 vocab=0.6340 token=1.4086 | 6.07 steps/s | VRAM=26.57 Go [2026-07-17 04:49:00] step 162200/1000000 | loss=2.4699 vocab=0.8721 token=1.3583 | 6.10 steps/s | VRAM=32.48 Go [2026-07-17 04:49:08] step 162250/1000000 | loss=2.3986 vocab=0.9499 token=1.8261 | 6.12 steps/s | VRAM=31.93 Go [2026-07-17 04:49:16] step 162300/1000000 | loss=2.4489 vocab=0.5496 token=1.2129 | 6.12 steps/s | VRAM=31.93 Go [2026-07-17 04:49:24] step 162350/1000000 | loss=2.4803 vocab=0.5857 token=0.7928 | 6.12 steps/s | VRAM=31.59 Go [2026-07-17 04:49:33] step 162400/1000000 | loss=2.5302 vocab=1.2100 token=1.3807 | 6.13 steps/s | VRAM=32.73 Go [2026-07-17 04:49:41] step 162450/1000000 | loss=2.3988 vocab=1.1899 token=1.4906 | 6.08 steps/s | VRAM=29.29 Go [2026-07-17 04:49:49] step 162500/1000000 | loss=2.4991 vocab=1.1053 token=1.4588 | 6.09 steps/s | VRAM=27.27 Go [2026-07-17 04:49:57] step 162550/1000000 | loss=2.4042 vocab=0.6041 token=1.3348 | 6.11 steps/s | VRAM=27.54 Go [2026-07-17 04:50:05] step 162600/1000000 | loss=2.5228 vocab=1.0375 token=1.3644 | 6.10 steps/s | VRAM=29.18 Go [2026-07-17 04:50:14] step 162650/1000000 | loss=2.4180 vocab=0.7883 token=1.2136 | 6.10 steps/s | VRAM=29.75 Go [2026-07-17 04:50:22] step 162700/1000000 | loss=2.4443 vocab=0.9976 token=1.1800 | 6.10 steps/s | VRAM=26.46 Go [2026-07-17 04:50:30] step 162750/1000000 | loss=2.2698 vocab=1.0164 token=2.0077 | 6.13 steps/s | VRAM=23.87 Go [2026-07-17 04:50:38] step 162800/1000000 | loss=2.4163 vocab=0.6906 token=1.1096 | 6.11 steps/s | VRAM=31.91 Go [2026-07-17 04:50:46] step 162850/1000000 | loss=2.4324 vocab=0.9953 token=1.7032 | 6.10 steps/s | VRAM=29.32 Go [2026-07-17 04:50:54] step 162900/1000000 | loss=2.4071 vocab=0.7992 token=1.4460 | 6.09 steps/s | VRAM=31.68 Go [2026-07-17 04:51:03] step 162950/1000000 | loss=2.4506 vocab=0.7861 token=1.4067 | 6.12 steps/s | VRAM=31.31 Go [2026-07-17 04:51:11] step 163000/1000000 | loss=2.4051 vocab=0.9913 token=1.5027 | 6.08 steps/s | VRAM=32.89 Go [2026-07-17 04:51:19] step 163050/1000000 | loss=2.5315 vocab=1.0374 token=1.3456 | 6.11 steps/s | VRAM=28.71 Go [2026-07-17 04:51:27] step 163100/1000000 | loss=2.5737 vocab=0.7470 token=1.4653 | 6.11 steps/s | VRAM=28.57 Go [2026-07-17 04:51:35] step 163150/1000000 | loss=2.4057 vocab=0.7806 token=1.4506 | 6.07 steps/s | VRAM=32.94 Go [2026-07-17 04:51:44] step 163200/1000000 | loss=2.5016 vocab=0.8571 token=1.6750 | 6.06 steps/s | VRAM=25.98 Go [2026-07-17 04:51:52] step 163250/1000000 | loss=2.4081 vocab=1.1742 token=1.6637 | 6.09 steps/s | VRAM=30.77 Go [2026-07-17 04:52:00] step 163300/1000000 | loss=2.3397 vocab=0.8818 token=1.4148 | 6.08 steps/s | VRAM=30.34 Go [2026-07-17 04:52:08] step 163350/1000000 | loss=2.4715 vocab=0.7178 token=1.2878 | 6.09 steps/s | VRAM=26.36 Go [2026-07-17 04:52:17] step 163400/1000000 | loss=2.4600 vocab=0.6591 token=1.4624 | 6.11 steps/s | VRAM=32.39 Go [2026-07-17 04:52:25] step 163450/1000000 | loss=2.5039 vocab=1.1972 token=1.6990 | 6.10 steps/s | VRAM=29.52 Go [2026-07-17 04:52:33] step 163500/1000000 | loss=2.4326 vocab=1.3918 token=1.8456 | 6.09 steps/s | VRAM=25.53 Go [2026-07-17 04:52:41] step 163550/1000000 | loss=2.4127 vocab=0.6982 token=1.4225 | 6.11 steps/s | VRAM=29.98 Go [2026-07-17 04:52:49] step 163600/1000000 | loss=2.4283 vocab=0.9443 token=1.2605 | 6.10 steps/s | VRAM=28.70 Go [2026-07-17 04:52:58] step 163650/1000000 | loss=2.4808 vocab=0.9045 token=1.7758 | 6.09 steps/s | VRAM=25.24 Go [2026-07-17 04:53:06] step 163700/1000000 | loss=2.4227 vocab=1.1791 token=1.6199 | 6.08 steps/s | VRAM=28.96 Go [2026-07-17 04:53:14] step 163750/1000000 | loss=2.4560 vocab=1.2360 token=1.5642 | 6.09 steps/s | VRAM=29.12 Go [2026-07-17 04:53:22] step 163800/1000000 | loss=2.3698 vocab=0.6809 token=1.4374 | 6.10 steps/s | VRAM=27.81 Go [2026-07-17 04:53:30] step 163850/1000000 | loss=2.3448 vocab=0.7926 token=1.2898 | 6.13 steps/s | VRAM=29.09 Go [2026-07-17 04:53:38] step 163900/1000000 | loss=2.4835 vocab=1.4691 token=1.9300 | 6.14 steps/s | VRAM=32.03 Go [2026-07-17 04:53:47] step 163950/1000000 | loss=2.4010 vocab=1.3949 token=1.0434 | 6.09 steps/s | VRAM=28.65 Go [2026-07-17 04:53:55] step 164000/1000000 | loss=2.4217 vocab=0.5462 token=1.1713 | 6.09 steps/s | VRAM=32.47 Go [2026-07-17 04:54:03] step 164050/1000000 | loss=2.3963 vocab=0.8482 token=1.0030 | 6.11 steps/s | VRAM=28.61 Go [2026-07-17 04:54:11] step 164100/1000000 | loss=2.2676 vocab=1.0322 token=1.7652 | 6.09 steps/s | VRAM=26.13 Go [2026-07-17 04:54:20] step 164150/1000000 | loss=2.3409 vocab=1.0975 token=1.3309 | 6.10 steps/s | VRAM=30.38 Go [2026-07-17 04:54:28] step 164200/1000000 | loss=2.4510 vocab=0.9288 token=1.4541 | 6.10 steps/s | VRAM=27.92 Go [2026-07-17 04:54:36] step 164250/1000000 | loss=2.4853 vocab=1.1902 token=1.5032 | 6.06 steps/s | VRAM=26.30 Go [2026-07-17 04:54:44] step 164300/1000000 | loss=2.5785 vocab=1.0880 token=1.6522 | 6.09 steps/s | VRAM=28.28 Go [2026-07-17 04:54:52] step 164350/1000000 | loss=2.2441 vocab=0.7882 token=1.6110 | 6.08 steps/s | VRAM=26.56 Go [2026-07-17 04:55:01] step 164400/1000000 | loss=2.5095 vocab=1.9658 token=1.2807 | 6.09 steps/s | VRAM=29.37 Go [2026-07-17 04:55:09] step 164450/1000000 | loss=2.4341 vocab=1.3012 token=1.6773 | 6.09 steps/s | VRAM=24.77 Go [2026-07-17 04:55:17] step 164500/1000000 | loss=2.3670 vocab=0.9862 token=1.7206 | 6.13 steps/s | VRAM=32.30 Go [2026-07-17 04:55:25] step 164550/1000000 | loss=2.4446 vocab=0.7032 token=1.5390 | 6.11 steps/s | VRAM=32.35 Go [2026-07-17 04:55:33] step 164600/1000000 | loss=2.3135 vocab=1.2264 token=1.5679 | 6.11 steps/s | VRAM=25.17 Go [2026-07-17 04:55:42] step 164650/1000000 | loss=2.3449 vocab=0.6220 token=1.6054 | 6.11 steps/s | VRAM=28.85 Go [2026-07-17 04:55:50] step 164700/1000000 | loss=2.4178 vocab=1.3480 token=1.5641 | 6.11 steps/s | VRAM=32.75 Go [2026-07-17 04:55:58] step 164750/1000000 | loss=2.5524 vocab=0.9077 token=1.2607 | 6.11 steps/s | VRAM=26.89 Go [2026-07-17 04:56:06] step 164800/1000000 | loss=2.4439 vocab=1.1352 token=1.6926 | 6.10 steps/s | VRAM=26.16 Go [2026-07-17 04:56:14] step 164850/1000000 | loss=2.3738 vocab=0.9782 token=1.3543 | 6.10 steps/s | VRAM=29.89 Go [2026-07-17 04:56:23] step 164900/1000000 | loss=2.4666 vocab=1.0150 token=1.2530 | 6.08 steps/s | VRAM=25.98 Go [2026-07-17 04:56:31] step 164950/1000000 | loss=2.4708 vocab=1.0249 token=1.4519 | 6.08 steps/s | VRAM=26.52 Go [2026-07-17 04:56:39] step 165000/1000000 | loss=2.4595 vocab=1.1009 token=1.6966 | 6.09 steps/s | VRAM=25.38 Go [2026-07-17 04:58:15] step 165050/1000000 | loss=2.4579 vocab=1.7477 token=2.0568 | 0.52 steps/s | VRAM=29.27 Go [2026-07-17 04:58:23] step 165100/1000000 | loss=2.3954 vocab=0.8141 token=1.7062 | 6.18 steps/s | VRAM=25.61 Go [2026-07-17 04:58:32] step 165150/1000000 | loss=2.4843 vocab=1.0920 token=1.1962 | 6.14 steps/s | VRAM=26.82 Go [2026-07-17 04:58:40] step 165200/1000000 | loss=2.4833 vocab=1.2022 token=1.4282 | 6.17 steps/s | VRAM=28.20 Go [2026-07-17 04:58:48] step 165250/1000000 | loss=2.5372 vocab=0.9598 token=1.6119 | 6.18 steps/s | VRAM=32.30 Go [2026-07-17 04:58:56] step 165300/1000000 | loss=2.4275 vocab=1.2579 token=1.5933 | 6.15 steps/s | VRAM=32.29 Go [2026-07-17 04:59:04] step 165350/1000000 | loss=2.3505 vocab=0.7898 token=1.0551 | 6.13 steps/s | VRAM=25.07 Go [2026-07-17 04:59:12] step 165400/1000000 | loss=2.4749 vocab=1.0986 token=1.2689 | 6.14 steps/s | VRAM=30.51 Go [2026-07-17 04:59:20] step 165450/1000000 | loss=2.3968 vocab=1.0624 token=1.3856 | 6.12 steps/s | VRAM=27.29 Go [2026-07-17 04:59:29] step 165500/1000000 | loss=2.3682 vocab=0.5692 token=1.4068 | 6.13 steps/s | VRAM=28.01 Go [2026-07-17 04:59:37] step 165550/1000000 | loss=2.3792 vocab=1.2887 token=1.5323 | 6.12 steps/s | VRAM=30.63 Go [2026-07-17 04:59:45] step 165600/1000000 | loss=2.4042 vocab=1.3492 token=1.5652 | 6.10 steps/s | VRAM=29.73 Go [2026-07-17 04:59:53] step 165650/1000000 | loss=2.4050 vocab=1.0970 token=1.9822 | 6.11 steps/s | VRAM=29.11 Go [2026-07-17 05:00:01] step 165700/1000000 | loss=2.4557 vocab=0.5238 token=1.0981 | 6.12 steps/s | VRAM=31.01 Go [2026-07-17 05:00:09] step 165750/1000000 | loss=2.4321 vocab=0.9929 token=1.6949 | 6.11 steps/s | VRAM=30.39 Go [2026-07-17 05:00:18] step 165800/1000000 | loss=2.4018 vocab=1.0562 token=1.4496 | 6.10 steps/s | VRAM=28.59 Go [2026-07-17 05:00:26] step 165850/1000000 | loss=2.3677 vocab=0.9842 token=1.2945 | 6.07 steps/s | VRAM=29.76 Go [2026-07-17 05:00:34] step 165900/1000000 | loss=2.3476 vocab=0.7360 token=1.3700 | 6.10 steps/s | VRAM=30.04 Go [2026-07-17 05:00:42] step 165950/1000000 | loss=2.5231 vocab=0.9157 token=1.2168 | 6.09 steps/s | VRAM=27.79 Go [2026-07-17 05:00:50] step 166000/1000000 | loss=2.3449 vocab=0.8080 token=1.4025 | 6.11 steps/s | VRAM=29.71 Go [2026-07-17 05:00:59] step 166050/1000000 | loss=2.4293 vocab=1.1359 token=1.2263 | 6.12 steps/s | VRAM=31.95 Go [2026-07-17 05:01:07] step 166100/1000000 | loss=2.4424 vocab=1.1387 token=1.7014 | 6.09 steps/s | VRAM=28.32 Go [2026-07-17 05:01:15] step 166150/1000000 | loss=2.4036 vocab=0.9722 token=1.5772 | 6.10 steps/s | VRAM=32.25 Go [2026-07-17 05:01:23] step 166200/1000000 | loss=2.2980 vocab=0.7072 token=1.6298 | 6.12 steps/s | VRAM=27.47 Go [2026-07-17 05:01:31] step 166250/1000000 | loss=2.4456 vocab=1.8522 token=1.4014 | 6.10 steps/s | VRAM=25.61 Go [2026-07-17 05:01:40] step 166300/1000000 | loss=2.4009 vocab=1.0200 token=1.4091 | 6.10 steps/s | VRAM=25.03 Go [2026-07-17 05:01:48] step 166350/1000000 | loss=2.4524 vocab=0.6725 token=1.1187 | 6.08 steps/s | VRAM=26.54 Go [2026-07-17 05:01:56] step 166400/1000000 | loss=2.4885 vocab=1.0921 token=1.6306 | 6.11 steps/s | VRAM=32.26 Go [2026-07-17 05:02:04] step 166450/1000000 | loss=2.4572 vocab=1.0492 token=1.5651 | 6.08 steps/s | VRAM=26.91 Go [2026-07-17 05:02:12] step 166500/1000000 | loss=2.4634 vocab=1.3687 token=1.6077 | 6.11 steps/s | VRAM=28.97 Go [2026-07-17 05:02:21] step 166550/1000000 | loss=2.4810 vocab=1.0866 token=1.4417 | 6.11 steps/s | VRAM=32.44 Go [2026-07-17 05:02:29] step 166600/1000000 | loss=2.4391 vocab=1.4088 token=1.6502 | 6.11 steps/s | VRAM=32.11 Go [2026-07-17 05:02:37] step 166650/1000000 | loss=2.3893 vocab=0.8871 token=1.8188 | 6.11 steps/s | VRAM=29.38 Go [2026-07-17 05:02:45] step 166700/1000000 | loss=2.4093 vocab=0.9938 token=1.3253 | 6.10 steps/s | VRAM=28.48 Go [2026-07-17 05:02:53] step 166750/1000000 | loss=2.4733 vocab=1.3664 token=1.4888 | 6.10 steps/s | VRAM=26.00 Go [2026-07-17 05:03:02] step 166800/1000000 | loss=2.4010 vocab=1.0241 token=1.7841 | 6.10 steps/s | VRAM=30.94 Go [2026-07-17 05:03:10] step 166850/1000000 | loss=2.4375 vocab=0.6592 token=1.6398 | 6.09 steps/s | VRAM=29.34 Go [2026-07-17 05:03:18] step 166900/1000000 | loss=2.5509 vocab=1.4029 token=1.6938 | 6.10 steps/s | VRAM=31.94 Go [2026-07-17 05:03:26] step 166950/1000000 | loss=2.4270 vocab=0.8021 token=1.2896 | 6.14 steps/s | VRAM=28.92 Go [2026-07-17 05:03:34] step 167000/1000000 | loss=2.4002 vocab=1.1895 token=1.6981 | 6.08 steps/s | VRAM=29.75 Go [2026-07-17 05:03:42] step 167050/1000000 | loss=2.3842 vocab=0.7291 token=1.2815 | 6.13 steps/s | VRAM=27.29 Go [2026-07-17 05:03:51] step 167100/1000000 | loss=2.4199 vocab=0.6761 token=1.5669 | 6.10 steps/s | VRAM=27.16 Go [2026-07-17 05:03:59] step 167150/1000000 | loss=2.4747 vocab=0.7533 token=1.5453 | 6.08 steps/s | VRAM=28.72 Go [2026-07-17 05:04:07] step 167200/1000000 | loss=2.3908 vocab=1.4483 token=1.6925 | 6.12 steps/s | VRAM=25.83 Go [2026-07-17 05:04:15] step 167250/1000000 | loss=2.4753 vocab=0.9350 token=1.2489 | 6.13 steps/s | VRAM=23.71 Go [2026-07-17 05:04:23] step 167300/1000000 | loss=2.3126 vocab=0.7505 token=1.1553 | 6.14 steps/s | VRAM=28.72 Go [2026-07-17 05:04:32] step 167350/1000000 | loss=2.3622 vocab=0.5181 token=1.1808 | 6.10 steps/s | VRAM=29.03 Go [2026-07-17 05:04:40] step 167400/1000000 | loss=2.4584 vocab=1.1478 token=1.7833 | 6.12 steps/s | VRAM=32.65 Go [2026-07-17 05:04:48] step 167450/1000000 | loss=2.4614 vocab=0.5571 token=1.7813 | 6.13 steps/s | VRAM=32.38 Go [2026-07-17 05:04:56] step 167500/1000000 | loss=2.4679 vocab=0.8376 token=1.2699 | 6.06 steps/s | VRAM=32.69 Go [2026-07-17 05:05:04] step 167550/1000000 | loss=2.4406 vocab=0.7666 token=1.8189 | 6.11 steps/s | VRAM=31.44 Go [2026-07-17 05:05:12] step 167600/1000000 | loss=2.4622 vocab=1.1090 token=1.2920 | 6.09 steps/s | VRAM=28.77 Go [2026-07-17 05:05:21] step 167650/1000000 | loss=2.4364 vocab=1.1606 token=1.3766 | 6.12 steps/s | VRAM=27.98 Go [2026-07-17 05:05:29] step 167700/1000000 | loss=2.4802 vocab=1.0276 token=1.5084 | 6.10 steps/s | VRAM=27.30 Go [2026-07-17 05:05:37] step 167750/1000000 | loss=2.3992 vocab=0.5139 token=1.0920 | 6.14 steps/s | VRAM=31.76 Go [2026-07-17 05:05:45] step 167800/1000000 | loss=2.4483 vocab=0.9108 token=1.4479 | 6.13 steps/s | VRAM=31.53 Go [2026-07-17 05:05:53] step 167850/1000000 | loss=2.3554 vocab=0.9940 token=1.3760 | 6.14 steps/s | VRAM=31.32 Go [2026-07-17 05:06:01] step 167900/1000000 | loss=2.2972 vocab=0.6635 token=1.2924 | 6.16 steps/s | VRAM=31.54 Go [2026-07-17 05:06:10] step 167950/1000000 | loss=2.3037 vocab=0.5551 token=1.1714 | 6.12 steps/s | VRAM=29.34 Go [2026-07-17 05:06:18] step 168000/1000000 | loss=2.4018 vocab=1.0940 token=1.6867 | 6.09 steps/s | VRAM=29.30 Go [2026-07-17 05:06:26] step 168050/1000000 | loss=2.4864 vocab=0.9930 token=1.4074 | 6.09 steps/s | VRAM=23.36 Go [2026-07-17 05:06:34] step 168100/1000000 | loss=2.4331 vocab=1.3273 token=1.6125 | 6.13 steps/s | VRAM=32.38 Go [2026-07-17 05:06:42] step 168150/1000000 | loss=2.3945 vocab=0.9164 token=1.5792 | 6.11 steps/s | VRAM=24.23 Go [2026-07-17 05:06:51] step 168200/1000000 | loss=2.3241 vocab=1.0084 token=1.4851 | 6.09 steps/s | VRAM=32.76 Go [2026-07-17 05:06:59] step 168250/1000000 | loss=2.4068 vocab=0.9156 token=1.5071 | 6.13 steps/s | VRAM=32.54 Go [2026-07-17 05:07:07] step 168300/1000000 | loss=2.4179 vocab=1.1690 token=1.2905 | 6.13 steps/s | VRAM=28.27 Go [2026-07-17 05:07:15] step 168350/1000000 | loss=2.4557 vocab=0.9906 token=1.5685 | 6.14 steps/s | VRAM=31.82 Go [2026-07-17 05:07:23] step 168400/1000000 | loss=2.3943 vocab=0.9272 token=1.3005 | 6.12 steps/s | VRAM=25.91 Go [2026-07-17 05:07:31] step 168450/1000000 | loss=2.3543 vocab=0.9803 token=1.2293 | 6.11 steps/s | VRAM=32.18 Go [2026-07-17 05:07:40] step 168500/1000000 | loss=2.5292 vocab=1.4759 token=1.5343 | 6.13 steps/s | VRAM=31.73 Go [2026-07-17 05:07:48] step 168550/1000000 | loss=2.3652 vocab=1.1735 token=1.1298 | 6.12 steps/s | VRAM=31.85 Go [2026-07-17 05:07:56] step 168600/1000000 | loss=2.4399 vocab=1.4646 token=1.3718 | 6.10 steps/s | VRAM=26.90 Go [2026-07-17 05:08:04] step 168650/1000000 | loss=2.3301 vocab=0.6207 token=1.3664 | 6.10 steps/s | VRAM=28.16 Go [2026-07-17 05:08:12] step 168700/1000000 | loss=2.3977 vocab=0.7979 token=1.7171 | 6.08 steps/s | VRAM=32.05 Go [2026-07-17 05:08:21] step 168750/1000000 | loss=2.3971 vocab=0.9826 token=1.9720 | 6.09 steps/s | VRAM=29.04 Go [2026-07-17 05:08:29] step 168800/1000000 | loss=2.4900 vocab=1.1334 token=1.4984 | 6.10 steps/s | VRAM=32.51 Go [2026-07-17 05:08:37] step 168850/1000000 | loss=2.3424 vocab=1.0682 token=1.3322 | 6.12 steps/s | VRAM=23.01 Go [2026-07-17 05:08:45] step 168900/1000000 | loss=2.4769 vocab=1.2989 token=1.4255 | 6.10 steps/s | VRAM=32.81 Go [2026-07-17 05:08:53] step 168950/1000000 | loss=2.3931 vocab=0.9983 token=1.1331 | 6.09 steps/s | VRAM=28.82 Go [2026-07-17 05:09:01] step 169000/1000000 | loss=2.4574 vocab=0.9526 token=1.5072 | 6.09 steps/s | VRAM=28.09 Go [2026-07-17 05:09:10] step 169050/1000000 | loss=2.4289 vocab=0.6985 token=0.9890 | 6.05 steps/s | VRAM=24.15 Go [2026-07-17 05:09:18] step 169100/1000000 | loss=2.3807 vocab=0.6067 token=1.3248 | 6.13 steps/s | VRAM=32.70 Go [2026-07-17 05:09:26] step 169150/1000000 | loss=2.4657 vocab=1.0143 token=1.5889 | 6.09 steps/s | VRAM=29.20 Go [2026-07-17 05:09:34] step 169200/1000000 | loss=2.4634 vocab=0.8737 token=1.5650 | 6.09 steps/s | VRAM=25.22 Go [2026-07-17 05:09:42] step 169250/1000000 | loss=2.4744 vocab=0.7176 token=1.2864 | 6.14 steps/s | VRAM=28.45 Go [2026-07-17 05:09:51] step 169300/1000000 | loss=2.3048 vocab=0.7194 token=1.2551 | 6.13 steps/s | VRAM=32.99 Go [2026-07-17 05:09:59] step 169350/1000000 | loss=2.4953 vocab=1.2056 token=1.3114 | 6.10 steps/s | VRAM=28.21 Go [2026-07-17 05:10:07] step 169400/1000000 | loss=2.4061 vocab=0.9642 token=1.3449 | 6.12 steps/s | VRAM=28.04 Go [2026-07-17 05:10:15] step 169450/1000000 | loss=2.5441 vocab=1.1640 token=1.7599 | 6.10 steps/s | VRAM=26.86 Go [2026-07-17 05:10:23] step 169500/1000000 | loss=2.4708 vocab=0.8339 token=1.3599 | 6.09 steps/s | VRAM=29.77 Go [2026-07-17 05:10:32] step 169550/1000000 | loss=2.5044 vocab=1.1538 token=1.7309 | 6.11 steps/s | VRAM=28.38 Go [2026-07-17 05:10:40] step 169600/1000000 | loss=2.4566 vocab=1.2186 token=1.4393 | 6.08 steps/s | VRAM=29.09 Go [2026-07-17 05:10:48] step 169650/1000000 | loss=2.3935 vocab=2.0092 token=1.6304 | 6.09 steps/s | VRAM=28.17 Go [2026-07-17 05:10:56] step 169700/1000000 | loss=2.4587 vocab=0.5272 token=1.0830 | 6.14 steps/s | VRAM=28.51 Go [2026-07-17 05:11:04] step 169750/1000000 | loss=2.5212 vocab=1.0888 token=1.2248 | 6.09 steps/s | VRAM=27.79 Go [2026-07-17 05:11:13] step 169800/1000000 | loss=2.4012 vocab=0.9687 token=1.4018 | 6.09 steps/s | VRAM=26.30 Go [2026-07-17 05:11:21] step 169850/1000000 | loss=2.3527 vocab=0.7769 token=1.5564 | 6.15 steps/s | VRAM=24.99 Go [2026-07-17 05:11:29] step 169900/1000000 | loss=2.3384 vocab=0.7032 token=1.2672 | 6.09 steps/s | VRAM=26.37 Go [2026-07-17 05:11:37] step 169950/1000000 | loss=2.3618 vocab=1.1382 token=1.5536 | 6.10 steps/s | VRAM=32.29 Go [2026-07-17 05:11:45] step 170000/1000000 | loss=2.4078 vocab=1.1726 token=1.5066 | 6.11 steps/s | VRAM=30.55 Go [2026-07-17 05:13:07] step 170050/1000000 | loss=2.5074 vocab=0.9358 token=1.2918 | 0.61 steps/s | VRAM=27.04 Go [2026-07-17 05:13:15] step 170100/1000000 | loss=2.4064 vocab=0.8216 token=1.2468 | 6.18 steps/s | VRAM=28.28 Go [2026-07-17 05:13:23] step 170150/1000000 | loss=2.4332 vocab=0.8332 token=1.5314 | 6.17 steps/s | VRAM=28.81 Go [2026-07-17 05:13:31] step 170200/1000000 | loss=2.3386 vocab=0.9401 token=1.6062 | 6.17 steps/s | VRAM=31.84 Go [2026-07-17 05:13:39] step 170250/1000000 | loss=2.4619 vocab=0.8992 token=1.2246 | 6.17 steps/s | VRAM=28.44 Go [2026-07-17 05:13:48] step 170300/1000000 | loss=2.4517 vocab=1.3985 token=1.7736 | 6.15 steps/s | VRAM=26.30 Go [2026-07-17 05:13:56] step 170350/1000000 | loss=2.4890 vocab=1.5837 token=1.6982 | 6.18 steps/s | VRAM=32.04 Go [2026-07-17 05:14:04] step 170400/1000000 | loss=2.4578 vocab=0.8977 token=1.6205 | 6.16 steps/s | VRAM=31.91 Go [2026-07-17 05:14:12] step 170450/1000000 | loss=2.3668 vocab=1.0302 token=1.9273 | 6.11 steps/s | VRAM=29.65 Go [2026-07-17 05:14:20] step 170500/1000000 | loss=2.3239 vocab=0.9454 token=1.4590 | 6.12 steps/s | VRAM=32.39 Go [2026-07-17 05:14:28] step 170550/1000000 | loss=2.2833 vocab=0.4174 token=1.0365 | 6.14 steps/s | VRAM=31.08 Go [2026-07-17 05:14:36] step 170600/1000000 | loss=2.4659 vocab=0.9340 token=1.4305 | 6.15 steps/s | VRAM=28.65 Go [2026-07-17 05:14:45] step 170650/1000000 | loss=2.4812 vocab=0.8531 token=1.8671 | 6.11 steps/s | VRAM=26.18 Go [2026-07-17 05:14:53] step 170700/1000000 | loss=2.4784 vocab=0.9187 token=1.6495 | 6.10 steps/s | VRAM=26.20 Go [2026-07-17 05:15:01] step 170750/1000000 | loss=2.4442 vocab=0.4626 token=0.9137 | 6.09 steps/s | VRAM=28.22 Go [2026-07-17 05:15:09] step 170800/1000000 | loss=2.3934 vocab=0.9273 token=1.7308 | 6.13 steps/s | VRAM=26.61 Go [2026-07-17 05:15:17] step 170850/1000000 | loss=2.3938 vocab=1.0298 token=1.3963 | 6.11 steps/s | VRAM=27.77 Go [2026-07-17 05:15:26] step 170900/1000000 | loss=2.4645 vocab=0.7973 token=1.4688 | 6.12 steps/s | VRAM=31.70 Go [2026-07-17 05:15:34] step 170950/1000000 | loss=2.4771 vocab=0.8980 token=1.5998 | 6.09 steps/s | VRAM=32.14 Go [2026-07-17 05:15:42] step 171000/1000000 | loss=2.4044 vocab=0.6516 token=1.4810 | 6.11 steps/s | VRAM=29.00 Go [2026-07-17 05:15:50] step 171050/1000000 | loss=2.4120 vocab=1.1701 token=1.2671 | 6.14 steps/s | VRAM=29.09 Go [2026-07-17 05:15:58] step 171100/1000000 | loss=2.4247 vocab=1.4335 token=1.5195 | 6.11 steps/s | VRAM=28.43 Go [2026-07-17 05:16:06] step 171150/1000000 | loss=2.4132 vocab=0.5946 token=1.4605 | 6.12 steps/s | VRAM=28.43 Go [2026-07-17 05:16:15] step 171200/1000000 | loss=2.3872 vocab=0.8587 token=1.2617 | 6.09 steps/s | VRAM=28.66 Go [2026-07-17 05:16:23] step 171250/1000000 | loss=2.3230 vocab=1.2471 token=1.4096 | 6.13 steps/s | VRAM=29.05 Go [2026-07-17 05:16:31] step 171300/1000000 | loss=2.4793 vocab=0.9907 token=1.7538 | 6.12 steps/s | VRAM=25.88 Go [2026-07-17 05:16:39] step 171350/1000000 | loss=2.3559 vocab=0.7840 token=1.3463 | 6.10 steps/s | VRAM=26.47 Go [2026-07-17 05:16:47] step 171400/1000000 | loss=2.3740 vocab=0.5686 token=0.7382 | 6.12 steps/s | VRAM=28.77 Go [2026-07-17 05:16:56] step 171450/1000000 | loss=2.3606 vocab=1.2225 token=1.1268 | 6.11 steps/s | VRAM=32.94 Go [2026-07-17 05:17:04] step 171500/1000000 | loss=2.3331 vocab=0.8923 token=1.8123 | 6.12 steps/s | VRAM=27.93 Go [2026-07-17 05:17:12] step 171550/1000000 | loss=2.3896 vocab=0.9788 token=1.5472 | 6.05 steps/s | VRAM=26.24 Go [2026-07-17 05:17:20] step 171600/1000000 | loss=2.4650 vocab=1.2717 token=1.8492 | 6.12 steps/s | VRAM=25.77 Go [2026-07-17 05:17:28] step 171650/1000000 | loss=2.4650 vocab=0.4834 token=0.9743 | 6.10 steps/s | VRAM=30.08 Go [2026-07-17 05:17:37] step 171700/1000000 | loss=2.4274 vocab=0.8873 token=0.9478 | 6.09 steps/s | VRAM=26.56 Go [2026-07-17 05:17:45] step 171750/1000000 | loss=2.4607 vocab=1.2676 token=1.4501 | 6.08 steps/s | VRAM=26.55 Go [2026-07-17 05:17:53] step 171800/1000000 | loss=2.5340 vocab=1.1672 token=1.6124 | 6.11 steps/s | VRAM=27.96 Go [2026-07-17 05:18:01] step 171850/1000000 | loss=2.3589 vocab=0.9859 token=1.9352 | 6.13 steps/s | VRAM=28.64 Go [2026-07-17 05:18:09] step 171900/1000000 | loss=2.4551 vocab=1.1039 token=1.3883 | 6.10 steps/s | VRAM=29.41 Go [2026-07-17 05:18:17] step 171950/1000000 | loss=2.5049 vocab=1.1973 token=1.8215 | 6.12 steps/s | VRAM=29.04 Go [2026-07-17 05:18:26] step 172000/1000000 | loss=2.3516 vocab=0.6936 token=1.6823 | 6.09 steps/s | VRAM=26.50 Go [2026-07-17 05:18:34] step 172050/1000000 | loss=2.4438 vocab=0.7852 token=1.0806 | 6.10 steps/s | VRAM=26.08 Go [2026-07-17 05:18:42] step 172100/1000000 | loss=2.3695 vocab=0.9955 token=1.2043 | 6.13 steps/s | VRAM=28.52 Go [2026-07-17 05:18:50] step 172150/1000000 | loss=2.5018 vocab=0.7890 token=1.4190 | 6.10 steps/s | VRAM=29.51 Go [2026-07-17 05:18:58] step 172200/1000000 | loss=2.3678 vocab=1.1786 token=1.2936 | 6.13 steps/s | VRAM=31.58 Go [2026-07-17 05:19:07] step 172250/1000000 | loss=2.2518 vocab=1.1807 token=1.8040 | 6.12 steps/s | VRAM=25.95 Go [2026-07-17 05:19:15] step 172300/1000000 | loss=2.4758 vocab=0.9711 token=1.5503 | 6.14 steps/s | VRAM=31.61 Go [2026-07-17 05:19:23] step 172350/1000000 | loss=2.4179 vocab=1.2163 token=1.7808 | 6.11 steps/s | VRAM=32.07 Go [2026-07-17 05:19:31] step 172400/1000000 | loss=2.3075 vocab=1.0380 token=1.4638 | 6.08 steps/s | VRAM=30.19 Go [2026-07-17 05:19:39] step 172450/1000000 | loss=2.4174 vocab=0.8310 token=1.1731 | 6.12 steps/s | VRAM=29.58 Go [2026-07-17 05:19:47] step 172500/1000000 | loss=2.3994 vocab=1.0587 token=1.4605 | 6.14 steps/s | VRAM=28.62 Go [2026-07-17 05:19:56] step 172550/1000000 | loss=2.4183 vocab=0.8842 token=1.4708 | 6.11 steps/s | VRAM=27.93 Go [2026-07-17 05:20:04] step 172600/1000000 | loss=2.4294 vocab=1.0217 token=1.3994 | 6.11 steps/s | VRAM=26.11 Go [2026-07-17 05:20:12] step 172650/1000000 | loss=2.3943 vocab=0.5085 token=0.8447 | 6.10 steps/s | VRAM=26.13 Go [2026-07-17 05:20:20] step 172700/1000000 | loss=2.3643 vocab=0.8825 token=1.3336 | 6.10 steps/s | VRAM=25.53 Go [2026-07-17 05:20:28] step 172750/1000000 | loss=2.4891 vocab=1.1264 token=1.7573 | 6.16 steps/s | VRAM=25.60 Go [2026-07-17 05:20:36] step 172800/1000000 | loss=2.4198 vocab=1.0008 token=1.4239 | 6.11 steps/s | VRAM=29.72 Go [2026-07-17 05:20:45] step 172850/1000000 | loss=2.4674 vocab=0.7210 token=1.5120 | 6.12 steps/s | VRAM=25.55 Go [2026-07-17 05:20:53] step 172900/1000000 | loss=2.4012 vocab=1.4498 token=1.3830 | 6.12 steps/s | VRAM=30.14 Go [2026-07-17 05:21:01] step 172950/1000000 | loss=2.4946 vocab=1.0466 token=1.3997 | 6.13 steps/s | VRAM=32.58 Go [2026-07-17 05:21:09] step 173000/1000000 | loss=2.3834 vocab=0.8736 token=1.4410 | 6.10 steps/s | VRAM=26.33 Go [2026-07-17 05:21:17] step 173050/1000000 | loss=2.4062 vocab=1.1178 token=1.5211 | 6.09 steps/s | VRAM=26.00 Go [2026-07-17 05:21:26] step 173100/1000000 | loss=2.2818 vocab=0.8283 token=1.1366 | 6.11 steps/s | VRAM=26.87 Go [2026-07-17 05:21:34] step 173150/1000000 | loss=2.4216 vocab=0.7416 token=1.1136 | 6.07 steps/s | VRAM=26.77 Go [2026-07-17 05:21:42] step 173200/1000000 | loss=2.4265 vocab=1.2811 token=1.5243 | 6.11 steps/s | VRAM=29.31 Go [2026-07-17 05:21:50] step 173250/1000000 | loss=2.4312 vocab=1.3692 token=1.4813 | 6.08 steps/s | VRAM=28.63 Go [2026-07-17 05:21:58] step 173300/1000000 | loss=2.3314 vocab=1.1079 token=1.6053 | 6.09 steps/s | VRAM=29.14 Go [2026-07-17 05:22:07] step 173350/1000000 | loss=2.4514 vocab=1.0554 token=1.3970 | 6.08 steps/s | VRAM=29.54 Go [2026-07-17 05:22:15] step 173400/1000000 | loss=2.4301 vocab=0.8247 token=1.4726 | 6.06 steps/s | VRAM=29.05 Go [2026-07-17 05:22:23] step 173450/1000000 | loss=2.3158 vocab=0.9187 token=1.1633 | 6.08 steps/s | VRAM=26.16 Go [2026-07-17 05:22:31] step 173500/1000000 | loss=2.2918 vocab=0.8232 token=1.0874 | 6.09 steps/s | VRAM=31.03 Go [2026-07-17 05:22:39] step 173550/1000000 | loss=2.3221 vocab=0.3291 token=1.0834 | 6.11 steps/s | VRAM=28.38 Go [2026-07-17 05:22:48] step 173600/1000000 | loss=2.3737 vocab=1.0130 token=1.6560 | 6.12 steps/s | VRAM=31.22 Go [2026-07-17 05:22:56] step 173650/1000000 | loss=2.4430 vocab=1.8134 token=1.7084 | 6.11 steps/s | VRAM=28.83 Go [2026-07-17 05:23:04] step 173700/1000000 | loss=2.4280 vocab=0.7617 token=1.2369 | 6.10 steps/s | VRAM=27.96 Go [2026-07-17 05:23:12] step 173750/1000000 | loss=2.4538 vocab=0.9035 token=1.7052 | 6.11 steps/s | VRAM=32.05 Go [2026-07-17 05:23:20] step 173800/1000000 | loss=2.4582 vocab=1.5138 token=1.7675 | 6.08 steps/s | VRAM=25.15 Go [2026-07-17 05:23:29] step 173850/1000000 | loss=2.3224 vocab=0.9648 token=1.5104 | 6.08 steps/s | VRAM=29.25 Go [2026-07-17 05:23:37] step 173900/1000000 | loss=2.4077 vocab=0.6485 token=1.3068 | 6.07 steps/s | VRAM=27.86 Go [2026-07-17 05:23:45] step 173950/1000000 | loss=2.4992 vocab=0.9540 token=1.2575 | 6.15 steps/s | VRAM=30.59 Go [2026-07-17 05:23:53] step 174000/1000000 | loss=2.4433 vocab=0.6455 token=1.0021 | 6.12 steps/s | VRAM=31.69 Go [2026-07-17 05:24:01] step 174050/1000000 | loss=2.2827 vocab=1.8350 token=1.3836 | 6.11 steps/s | VRAM=28.95 Go [2026-07-17 05:24:10] step 174100/1000000 | loss=2.3917 vocab=0.5476 token=1.4610 | 6.12 steps/s | VRAM=24.69 Go [2026-07-17 05:24:18] step 174150/1000000 | loss=2.4530 vocab=1.0880 token=1.7194 | 6.12 steps/s | VRAM=30.84 Go [2026-07-17 05:24:26] step 174200/1000000 | loss=2.3294 vocab=1.1781 token=1.3510 | 6.12 steps/s | VRAM=25.90 Go [2026-07-17 05:24:34] step 174250/1000000 | loss=2.4005 vocab=1.0229 token=1.6393 | 6.11 steps/s | VRAM=28.40 Go [2026-07-17 05:24:42] step 174300/1000000 | loss=2.3281 vocab=0.6532 token=1.3730 | 6.09 steps/s | VRAM=26.41 Go [2026-07-17 05:24:50] step 174350/1000000 | loss=2.3368 vocab=0.7388 token=1.0629 | 6.10 steps/s | VRAM=26.21 Go [2026-07-17 05:24:59] step 174400/1000000 | loss=2.4890 vocab=0.6583 token=1.1816 | 6.12 steps/s | VRAM=32.07 Go [2026-07-17 05:25:07] step 174450/1000000 | loss=2.3845 vocab=0.7316 token=1.0872 | 6.06 steps/s | VRAM=23.68 Go [2026-07-17 05:25:15] step 174500/1000000 | loss=2.4305 vocab=0.5619 token=1.5733 | 6.12 steps/s | VRAM=28.11 Go [2026-07-17 05:25:23] step 174550/1000000 | loss=2.3671 vocab=1.0853 token=1.4965 | 6.10 steps/s | VRAM=29.86 Go [2026-07-17 05:25:31] step 174600/1000000 | loss=2.4245 vocab=0.7626 token=1.1444 | 6.10 steps/s | VRAM=28.78 Go [2026-07-17 05:25:40] step 174650/1000000 | loss=2.4880 vocab=1.1166 token=1.5769 | 6.11 steps/s | VRAM=30.85 Go [2026-07-17 05:25:48] step 174700/1000000 | loss=2.4760 vocab=0.9223 token=1.3310 | 6.11 steps/s | VRAM=31.97 Go [2026-07-17 05:25:56] step 174750/1000000 | loss=2.4069 vocab=0.9033 token=1.0697 | 6.12 steps/s | VRAM=32.31 Go [2026-07-17 05:26:04] step 174800/1000000 | loss=2.3465 vocab=0.6762 token=1.0725 | 6.10 steps/s | VRAM=28.05 Go [2026-07-17 05:26:12] step 174850/1000000 | loss=2.3002 vocab=0.5961 token=1.2443 | 6.07 steps/s | VRAM=28.97 Go [2026-07-17 05:26:21] step 174900/1000000 | loss=2.3287 vocab=0.7259 token=1.4119 | 6.11 steps/s | VRAM=25.74 Go [2026-07-17 05:26:29] step 174950/1000000 | loss=2.4241 vocab=0.9489 token=1.6209 | 6.11 steps/s | VRAM=29.17 Go [2026-07-17 05:26:37] step 175000/1000000 | loss=2.2718 vocab=0.8610 token=1.6978 | 6.12 steps/s | VRAM=31.96 Go [2026-07-17 05:27:57] step 175050/1000000 | loss=2.3673 vocab=1.0024 token=1.4128 | 0.63 steps/s | VRAM=32.98 Go [2026-07-17 05:28:05] step 175100/1000000 | loss=2.4003 vocab=0.9726 token=1.6561 | 6.15 steps/s | VRAM=28.70 Go [2026-07-17 05:28:13] step 175150/1000000 | loss=2.3649 vocab=0.5663 token=1.2174 | 6.15 steps/s | VRAM=26.84 Go [2026-07-17 05:28:21] step 175200/1000000 | loss=2.4856 vocab=1.2681 token=1.5559 | 6.15 steps/s | VRAM=28.96 Go [2026-07-17 05:28:29] step 175250/1000000 | loss=2.3893 vocab=0.9502 token=1.3678 | 6.18 steps/s | VRAM=32.32 Go [2026-07-17 05:28:37] step 175300/1000000 | loss=2.3592 vocab=0.6504 token=0.8671 | 6.14 steps/s | VRAM=25.34 Go [2026-07-17 05:28:45] step 175350/1000000 | loss=2.3582 vocab=0.7319 token=1.1576 | 6.16 steps/s | VRAM=24.42 Go [2026-07-17 05:28:54] step 175400/1000000 | loss=2.4085 vocab=0.7555 token=1.0066 | 6.11 steps/s | VRAM=30.35 Go [2026-07-17 05:29:02] step 175450/1000000 | loss=2.4509 vocab=1.2662 token=1.6222 | 6.11 steps/s | VRAM=32.20 Go [2026-07-17 05:29:10] step 175500/1000000 | loss=2.4410 vocab=1.0521 token=1.3415 | 6.11 steps/s | VRAM=32.52 Go [2026-07-17 05:29:18] step 175550/1000000 | loss=2.2695 vocab=0.6156 token=0.8634 | 6.10 steps/s | VRAM=32.17 Go [2026-07-17 05:29:26] step 175600/1000000 | loss=2.4626 vocab=0.8175 token=1.4993 | 6.08 steps/s | VRAM=28.19 Go [2026-07-17 05:29:35] step 175650/1000000 | loss=2.4501 vocab=1.1661 token=1.5394 | 6.09 steps/s | VRAM=25.89 Go [2026-07-17 05:29:43] step 175700/1000000 | loss=2.3731 vocab=0.5230 token=1.0629 | 6.08 steps/s | VRAM=29.87 Go [2026-07-17 05:29:51] step 175750/1000000 | loss=2.4033 vocab=1.2757 token=1.5481 | 6.09 steps/s | VRAM=29.11 Go [2026-07-17 05:29:59] step 175800/1000000 | loss=2.3869 vocab=1.3003 token=1.4589 | 6.09 steps/s | VRAM=29.26 Go [2026-07-17 05:30:07] step 175850/1000000 | loss=2.4171 vocab=1.1462 token=1.4703 | 6.09 steps/s | VRAM=28.69 Go [2026-07-17 05:30:16] step 175900/1000000 | loss=2.4479 vocab=1.1143 token=1.3101 | 6.11 steps/s | VRAM=32.30 Go [2026-07-17 05:30:24] step 175950/1000000 | loss=2.2498 vocab=1.1499 token=1.7015 | 6.08 steps/s | VRAM=32.63 Go [2026-07-17 05:30:32] step 176000/1000000 | loss=2.3423 vocab=0.7137 token=1.4954 | 6.10 steps/s | VRAM=26.47 Go [2026-07-17 05:30:40] step 176050/1000000 | loss=2.4751 vocab=0.8264 token=1.7271 | 6.09 steps/s | VRAM=29.55 Go [2026-07-17 05:30:48] step 176100/1000000 | loss=2.4323 vocab=0.8407 token=1.3979 | 6.12 steps/s | VRAM=31.63 Go [2026-07-17 05:30:57] step 176150/1000000 | loss=2.3548 vocab=1.1002 token=1.4204 | 6.10 steps/s | VRAM=28.71 Go [2026-07-17 05:31:05] step 176200/1000000 | loss=2.3788 vocab=0.7111 token=1.5090 | 6.09 steps/s | VRAM=32.98 Go [2026-07-17 05:31:13] step 176250/1000000 | loss=2.4100 vocab=1.0382 token=1.0717 | 6.10 steps/s | VRAM=28.00 Go [2026-07-17 05:31:21] step 176300/1000000 | loss=2.4426 vocab=0.8899 token=1.1628 | 6.08 steps/s | VRAM=28.07 Go [2026-07-17 05:31:29] step 176350/1000000 | loss=2.4872 vocab=0.8382 token=1.1658 | 6.11 steps/s | VRAM=31.49 Go [2026-07-17 05:31:38] step 176400/1000000 | loss=2.4518 vocab=1.2274 token=1.5799 | 6.10 steps/s | VRAM=28.64 Go [2026-07-17 05:31:46] step 176450/1000000 | loss=2.4140 vocab=1.2023 token=1.4589 | 6.10 steps/s | VRAM=31.41 Go [2026-07-17 05:31:54] step 176500/1000000 | loss=2.3616 vocab=0.7097 token=1.4168 | 6.08 steps/s | VRAM=26.09 Go [2026-07-17 05:32:02] step 176550/1000000 | loss=2.4264 vocab=0.9116 token=1.4278 | 6.11 steps/s | VRAM=31.86 Go [2026-07-17 05:32:10] step 176600/1000000 | loss=2.4303 vocab=1.4087 token=1.7048 | 6.11 steps/s | VRAM=31.38 Go [2026-07-17 05:32:19] step 176650/1000000 | loss=2.4442 vocab=0.8977 token=1.1865 | 6.09 steps/s | VRAM=29.10 Go [2026-07-17 05:32:27] step 176700/1000000 | loss=2.3793 vocab=1.1254 token=1.7225 | 6.09 steps/s | VRAM=29.12 Go [2026-07-17 05:32:35] step 176750/1000000 | loss=2.4620 vocab=0.5001 token=1.3115 | 6.09 steps/s | VRAM=29.10 Go [2026-07-17 05:32:43] step 176800/1000000 | loss=2.3496 vocab=1.1731 token=1.4063 | 6.09 steps/s | VRAM=28.98 Go [2026-07-17 05:32:51] step 176850/1000000 | loss=2.2785 vocab=0.6529 token=1.3265 | 6.09 steps/s | VRAM=28.35 Go [2026-07-17 05:33:00] step 176900/1000000 | loss=2.3047 vocab=1.0344 token=1.4149 | 6.09 steps/s | VRAM=28.93 Go [2026-07-17 05:33:08] step 176950/1000000 | loss=2.3836 vocab=0.7833 token=1.0053 | 6.09 steps/s | VRAM=32.11 Go [2026-07-17 05:33:16] step 177000/1000000 | loss=2.3994 vocab=1.4333 token=1.6074 | 6.07 steps/s | VRAM=28.32 Go [2026-07-17 05:33:24] step 177050/1000000 | loss=2.2681 vocab=0.6871 token=1.3030 | 6.11 steps/s | VRAM=31.01 Go [2026-07-17 05:33:32] step 177100/1000000 | loss=2.4332 vocab=0.9843 token=1.3543 | 6.12 steps/s | VRAM=31.59 Go [2026-07-17 05:33:41] step 177150/1000000 | loss=2.3681 vocab=0.8955 token=1.1528 | 6.12 steps/s | VRAM=28.57 Go [2026-07-17 05:33:49] step 177200/1000000 | loss=2.4622 vocab=0.7399 token=1.2013 | 6.05 steps/s | VRAM=26.98 Go [2026-07-17 05:33:57] step 177250/1000000 | loss=2.4269 vocab=0.7787 token=1.4462 | 6.11 steps/s | VRAM=28.51 Go [2026-07-17 05:34:05] step 177300/1000000 | loss=2.4648 vocab=1.4302 token=1.8843 | 6.11 steps/s | VRAM=29.04 Go [2026-07-17 05:34:13] step 177350/1000000 | loss=2.3872 vocab=1.8086 token=1.4532 | 6.10 steps/s | VRAM=31.81 Go [2026-07-17 05:34:22] step 177400/1000000 | loss=2.3890 vocab=1.4208 token=1.6552 | 6.07 steps/s | VRAM=29.56 Go [2026-07-17 05:34:30] step 177450/1000000 | loss=2.4094 vocab=1.0590 token=1.3076 | 6.10 steps/s | VRAM=25.73 Go [2026-07-17 05:34:38] step 177500/1000000 | loss=2.4145 vocab=1.0083 token=1.2630 | 6.10 steps/s | VRAM=32.80 Go [2026-07-17 05:34:46] step 177550/1000000 | loss=2.4375 vocab=0.6938 token=1.0592 | 6.10 steps/s | VRAM=26.88 Go [2026-07-17 05:34:55] step 177600/1000000 | loss=2.4131 vocab=1.1233 token=1.4484 | 6.09 steps/s | VRAM=31.39 Go [2026-07-17 05:35:03] step 177650/1000000 | loss=2.5081 vocab=1.1680 token=1.6347 | 6.12 steps/s | VRAM=31.58 Go [2026-07-17 05:35:11] step 177700/1000000 | loss=2.4251 vocab=1.1130 token=1.3245 | 6.13 steps/s | VRAM=32.25 Go [2026-07-17 05:35:19] step 177750/1000000 | loss=2.3938 vocab=1.7303 token=2.0572 | 6.06 steps/s | VRAM=26.29 Go [2026-07-17 05:35:27] step 177800/1000000 | loss=2.4066 vocab=1.2415 token=1.3812 | 6.10 steps/s | VRAM=29.70 Go [2026-07-17 05:35:35] step 177850/1000000 | loss=2.3239 vocab=1.2441 token=1.2380 | 6.13 steps/s | VRAM=27.91 Go [2026-07-17 05:35:44] step 177900/1000000 | loss=2.5034 vocab=1.4028 token=1.6016 | 6.10 steps/s | VRAM=26.59 Go [2026-07-17 05:35:52] step 177950/1000000 | loss=2.4068 vocab=1.0009 token=1.5432 | 6.09 steps/s | VRAM=26.48 Go [2026-07-17 05:36:00] step 178000/1000000 | loss=2.3648 vocab=1.2253 token=1.4745 | 6.14 steps/s | VRAM=27.99 Go [2026-07-17 05:36:08] step 178050/1000000 | loss=2.3870 vocab=0.8111 token=1.3168 | 6.09 steps/s | VRAM=26.38 Go [2026-07-17 05:36:16] step 178100/1000000 | loss=2.3406 vocab=1.0798 token=1.5099 | 6.12 steps/s | VRAM=28.70 Go [2026-07-17 05:36:25] step 178150/1000000 | loss=2.3666 vocab=1.1790 token=1.7913 | 6.10 steps/s | VRAM=26.23 Go [2026-07-17 05:36:33] step 178200/1000000 | loss=2.4472 vocab=0.9859 token=1.3520 | 6.13 steps/s | VRAM=28.62 Go [2026-07-17 05:36:41] step 178250/1000000 | loss=2.3569 vocab=1.3104 token=1.0150 | 6.14 steps/s | VRAM=28.66 Go [2026-07-17 05:36:49] step 178300/1000000 | loss=2.3819 vocab=0.8598 token=1.3073 | 6.11 steps/s | VRAM=25.96 Go [2026-07-17 05:36:57] step 178350/1000000 | loss=2.4226 vocab=1.1136 token=1.2824 | 6.13 steps/s | VRAM=27.84 Go [2026-07-17 05:37:05] step 178400/1000000 | loss=2.4147 vocab=0.8212 token=1.2529 | 6.11 steps/s | VRAM=27.99 Go [2026-07-17 05:37:14] step 178450/1000000 | loss=2.3400 vocab=0.6873 token=1.3329 | 6.11 steps/s | VRAM=29.85 Go [2026-07-17 05:37:22] step 178500/1000000 | loss=2.3618 vocab=0.7740 token=1.3410 | 6.12 steps/s | VRAM=27.77 Go [2026-07-17 05:37:30] step 178550/1000000 | loss=2.3645 vocab=1.1407 token=2.0675 | 6.10 steps/s | VRAM=29.08 Go [2026-07-17 05:37:38] step 178600/1000000 | loss=2.4009 vocab=0.8229 token=1.1932 | 6.09 steps/s | VRAM=26.20 Go [2026-07-17 05:37:46] step 178650/1000000 | loss=2.4898 vocab=0.8916 token=1.4396 | 6.11 steps/s | VRAM=25.91 Go [2026-07-17 05:37:54] step 178700/1000000 | loss=2.4553 vocab=1.0652 token=1.5484 | 6.14 steps/s | VRAM=31.68 Go [2026-07-17 05:38:03] step 178750/1000000 | loss=2.3437 vocab=0.7672 token=1.2803 | 6.12 steps/s | VRAM=31.89 Go [2026-07-17 05:38:11] step 178800/1000000 | loss=2.3364 vocab=1.2425 token=1.3730 | 6.13 steps/s | VRAM=28.98 Go [2026-07-17 05:38:19] step 178850/1000000 | loss=2.4309 vocab=1.0987 token=1.7062 | 6.09 steps/s | VRAM=28.50 Go [2026-07-17 05:38:27] step 178900/1000000 | loss=2.3460 vocab=1.0576 token=1.5381 | 6.09 steps/s | VRAM=32.94 Go [2026-07-17 05:38:35] step 178950/1000000 | loss=2.4418 vocab=1.2694 token=1.6702 | 6.10 steps/s | VRAM=29.29 Go [2026-07-17 05:38:44] step 179000/1000000 | loss=2.3136 vocab=0.9208 token=1.7222 | 6.08 steps/s | VRAM=26.34 Go [2026-07-17 05:38:52] step 179050/1000000 | loss=2.4260 vocab=0.8144 token=1.3301 | 6.10 steps/s | VRAM=32.77 Go [2026-07-17 05:39:00] step 179100/1000000 | loss=2.4002 vocab=0.7169 token=1.1457 | 6.13 steps/s | VRAM=30.67 Go [2026-07-17 05:39:08] step 179150/1000000 | loss=2.4070 vocab=1.8642 token=1.7572 | 6.09 steps/s | VRAM=25.62 Go [2026-07-17 05:39:16] step 179200/1000000 | loss=2.3253 vocab=0.6636 token=1.3631 | 6.09 steps/s | VRAM=28.02 Go [2026-07-17 05:39:25] step 179250/1000000 | loss=2.3671 vocab=0.9298 token=1.5811 | 6.11 steps/s | VRAM=28.25 Go [2026-07-17 05:39:33] step 179300/1000000 | loss=2.5126 vocab=0.9995 token=1.4799 | 6.09 steps/s | VRAM=28.49 Go [2026-07-17 05:39:41] step 179350/1000000 | loss=2.4201 vocab=1.0703 token=1.5436 | 6.08 steps/s | VRAM=32.93 Go [2026-07-17 05:39:49] step 179400/1000000 | loss=2.4278 vocab=0.9973 token=1.4505 | 6.12 steps/s | VRAM=32.85 Go [2026-07-17 05:39:57] step 179450/1000000 | loss=2.3918 vocab=0.6350 token=1.3760 | 6.07 steps/s | VRAM=24.66 Go [2026-07-17 05:40:06] step 179500/1000000 | loss=2.2612 vocab=0.7284 token=0.9232 | 6.10 steps/s | VRAM=28.18 Go [2026-07-17 05:40:14] step 179550/1000000 | loss=2.2905 vocab=0.8625 token=1.0411 | 6.11 steps/s | VRAM=25.60 Go [2026-07-17 05:40:22] step 179600/1000000 | loss=2.3841 vocab=0.7247 token=1.3921 | 6.08 steps/s | VRAM=28.98 Go [2026-07-17 05:40:30] step 179650/1000000 | loss=2.3801 vocab=0.7043 token=1.2032 | 6.11 steps/s | VRAM=29.21 Go [2026-07-17 05:40:38] step 179700/1000000 | loss=2.4545 vocab=0.9357 token=1.2291 | 6.12 steps/s | VRAM=28.40 Go [2026-07-17 05:40:47] step 179750/1000000 | loss=2.3409 vocab=1.0489 token=1.2258 | 6.11 steps/s | VRAM=28.37 Go [2026-07-17 05:40:55] step 179800/1000000 | loss=2.4670 vocab=1.0322 token=1.3557 | 6.11 steps/s | VRAM=32.09 Go [2026-07-17 05:41:03] step 179850/1000000 | loss=2.4689 vocab=0.8898 token=1.6590 | 6.10 steps/s | VRAM=28.07 Go [2026-07-17 05:41:11] step 179900/1000000 | loss=2.3859 vocab=1.4067 token=1.3229 | 6.12 steps/s | VRAM=32.40 Go [2026-07-17 05:41:19] step 179950/1000000 | loss=2.3802 vocab=0.6687 token=1.2406 | 6.13 steps/s | VRAM=31.85 Go [2026-07-17 05:41:27] step 180000/1000000 | loss=2.3736 vocab=1.4900 token=1.5594 | 6.12 steps/s | VRAM=31.84 Go [2026-07-17 05:43:01] step 180050/1000000 | loss=2.3973 vocab=1.6545 token=1.2395 | 0.53 steps/s | VRAM=31.72 Go [2026-07-17 05:43:09] step 180100/1000000 | loss=2.4413 vocab=0.6779 token=1.0680 | 6.16 steps/s | VRAM=30.18 Go [2026-07-17 05:43:17] step 180150/1000000 | loss=2.3076 vocab=1.1643 token=1.5053 | 6.14 steps/s | VRAM=29.30 Go [2026-07-17 05:43:26] step 180200/1000000 | loss=2.4178 vocab=1.0065 token=1.4326 | 6.15 steps/s | VRAM=32.18 Go [2026-07-17 05:43:34] step 180250/1000000 | loss=2.4119 vocab=0.7157 token=1.0178 | 6.14 steps/s | VRAM=32.82 Go [2026-07-17 05:43:42] step 180300/1000000 | loss=2.3997 vocab=0.8385 token=1.6058 | 6.12 steps/s | VRAM=25.86 Go [2026-07-17 05:43:50] step 180350/1000000 | loss=2.6646 vocab=0.9949 token=1.3074 | 6.12 steps/s | VRAM=27.96 Go [2026-07-17 05:43:58] step 180400/1000000 | loss=2.4984 vocab=1.0173 token=1.3969 | 6.12 steps/s | VRAM=32.59 Go [2026-07-17 05:44:06] step 180450/1000000 | loss=2.4469 vocab=0.7637 token=1.2385 | 6.10 steps/s | VRAM=28.67 Go [2026-07-17 05:44:15] step 180500/1000000 | loss=2.4612 vocab=1.2356 token=1.6694 | 6.13 steps/s | VRAM=27.46 Go [2026-07-17 05:44:23] step 180550/1000000 | loss=2.4521 vocab=1.3190 token=1.4090 | 6.10 steps/s | VRAM=28.47 Go [2026-07-17 05:44:31] step 180600/1000000 | loss=2.4252 vocab=0.8494 token=1.4102 | 6.06 steps/s | VRAM=26.07 Go [2026-07-17 05:44:39] step 180650/1000000 | loss=2.5399 vocab=0.8056 token=1.6688 | 6.11 steps/s | VRAM=26.01 Go [2026-07-17 05:44:47] step 180700/1000000 | loss=2.2565 vocab=0.6512 token=1.0170 | 6.07 steps/s | VRAM=28.25 Go [2026-07-17 05:44:56] step 180750/1000000 | loss=2.3555 vocab=0.7446 token=1.5375 | 6.09 steps/s | VRAM=28.63 Go [2026-07-17 05:45:04] step 180800/1000000 | loss=2.4342 vocab=0.6912 token=1.3478 | 6.06 steps/s | VRAM=26.72 Go [2026-07-17 05:45:12] step 180850/1000000 | loss=2.4051 vocab=1.2262 token=1.3538 | 6.11 steps/s | VRAM=24.70 Go [2026-07-17 05:45:20] step 180900/1000000 | loss=2.4441 vocab=1.4637 token=1.6437 | 6.11 steps/s | VRAM=29.14 Go [2026-07-17 05:45:28] step 180950/1000000 | loss=2.4351 vocab=0.7741 token=1.3283 | 6.10 steps/s | VRAM=31.61 Go [2026-07-17 05:45:37] step 181000/1000000 | loss=2.2895 vocab=0.5996 token=1.2631 | 6.10 steps/s | VRAM=31.29 Go [2026-07-17 05:45:45] step 181050/1000000 | loss=2.3987 vocab=1.1796 token=1.4262 | 6.07 steps/s | VRAM=29.73 Go [2026-07-17 05:45:53] step 181100/1000000 | loss=2.4499 vocab=0.9827 token=1.6709 | 6.11 steps/s | VRAM=31.98 Go [2026-07-17 05:46:01] step 181150/1000000 | loss=2.4578 vocab=0.9719 token=1.7050 | 6.09 steps/s | VRAM=28.76 Go [2026-07-17 05:46:10] step 181200/1000000 | loss=2.3472 vocab=1.0444 token=1.5840 | 6.06 steps/s | VRAM=28.73 Go [2026-07-17 05:46:18] step 181250/1000000 | loss=2.4347 vocab=0.7165 token=1.4488 | 6.10 steps/s | VRAM=29.31 Go [2026-07-17 05:46:26] step 181300/1000000 | loss=2.4397 vocab=1.1013 token=1.5898 | 6.06 steps/s | VRAM=26.74 Go [2026-07-17 05:46:34] step 181350/1000000 | loss=2.4957 vocab=0.9171 token=1.1935 | 6.11 steps/s | VRAM=30.85 Go [2026-07-17 05:46:42] step 181400/1000000 | loss=2.4348 vocab=0.8591 token=1.2334 | 6.08 steps/s | VRAM=32.46 Go [2026-07-17 05:46:51] step 181450/1000000 | loss=2.4309 vocab=0.5599 token=1.1196 | 6.10 steps/s | VRAM=28.10 Go [2026-07-17 05:46:59] step 181500/1000000 | loss=2.3886 vocab=1.1943 token=1.5750 | 6.12 steps/s | VRAM=29.08 Go [2026-07-17 05:47:07] step 181550/1000000 | loss=2.4718 vocab=0.9374 token=1.3147 | 6.09 steps/s | VRAM=32.09 Go [2026-07-17 05:47:15] step 181600/1000000 | loss=2.4202 vocab=1.2899 token=1.2308 | 6.07 steps/s | VRAM=29.06 Go [2026-07-17 05:47:23] step 181650/1000000 | loss=2.4133 vocab=0.6358 token=1.5389 | 6.09 steps/s | VRAM=25.89 Go [2026-07-17 05:47:32] step 181700/1000000 | loss=2.3539 vocab=1.5993 token=1.5397 | 6.08 steps/s | VRAM=29.43 Go [2026-07-17 05:47:40] step 181750/1000000 | loss=2.4830 vocab=1.1179 token=1.1132 | 6.10 steps/s | VRAM=28.74 Go [2026-07-17 05:47:48] step 181800/1000000 | loss=2.3326 vocab=0.8910 token=1.3927 | 6.09 steps/s | VRAM=26.13 Go [2026-07-17 05:47:56] step 181850/1000000 | loss=2.3557 vocab=0.9544 token=1.2765 | 6.10 steps/s | VRAM=28.22 Go [2026-07-17 05:48:04] step 181900/1000000 | loss=2.4713 vocab=1.1929 token=1.5246 | 6.07 steps/s | VRAM=32.69 Go [2026-07-17 05:48:13] step 181950/1000000 | loss=2.3402 vocab=1.2603 token=1.3457 | 6.08 steps/s | VRAM=25.96 Go [2026-07-17 05:48:21] step 182000/1000000 | loss=2.4440 vocab=0.8487 token=1.6160 | 6.09 steps/s | VRAM=28.32 Go [2026-07-17 05:48:29] step 182050/1000000 | loss=2.3131 vocab=0.9206 token=1.2013 | 6.05 steps/s | VRAM=29.20 Go [2026-07-17 05:48:37] step 182100/1000000 | loss=2.5906 vocab=0.8941 token=1.4532 | 6.07 steps/s | VRAM=28.62 Go [2026-07-17 05:48:46] step 182150/1000000 | loss=2.4736 vocab=0.5575 token=1.3097 | 6.13 steps/s | VRAM=31.71 Go [2026-07-17 05:48:54] step 182200/1000000 | loss=2.4193 vocab=0.8367 token=1.3012 | 6.13 steps/s | VRAM=28.46 Go [2026-07-17 05:49:02] step 182250/1000000 | loss=2.4087 vocab=1.4221 token=1.7584 | 6.10 steps/s | VRAM=32.22 Go [2026-07-17 05:49:10] step 182300/1000000 | loss=2.3826 vocab=0.5332 token=1.3382 | 6.10 steps/s | VRAM=28.24 Go [2026-07-17 05:49:18] step 182350/1000000 | loss=2.4826 vocab=0.7856 token=1.7874 | 6.12 steps/s | VRAM=24.69 Go [2026-07-17 05:49:27] step 182400/1000000 | loss=2.4181 vocab=0.9486 token=1.6614 | 6.09 steps/s | VRAM=28.96 Go [2026-07-17 05:49:35] step 182450/1000000 | loss=2.3874 vocab=0.9963 token=1.5509 | 6.12 steps/s | VRAM=32.14 Go [2026-07-17 05:49:43] step 182500/1000000 | loss=2.4233 vocab=1.2174 token=1.1504 | 6.08 steps/s | VRAM=26.38 Go [2026-07-17 05:49:51] step 182550/1000000 | loss=2.3779 vocab=0.4789 token=1.4553 | 6.09 steps/s | VRAM=28.47 Go [2026-07-17 05:49:59] step 182600/1000000 | loss=2.3608 vocab=1.0677 token=1.2045 | 6.11 steps/s | VRAM=28.40 Go [2026-07-17 05:50:07] step 182650/1000000 | loss=2.3932 vocab=1.5494 token=1.9115 | 6.11 steps/s | VRAM=33.01 Go [2026-07-17 05:50:16] step 182700/1000000 | loss=2.4698 vocab=0.6739 token=1.5309 | 6.07 steps/s | VRAM=32.86 Go [2026-07-17 05:50:24] step 182750/1000000 | loss=2.4313 vocab=0.9508 token=1.3933 | 6.11 steps/s | VRAM=28.14 Go [2026-07-17 05:50:32] step 182800/1000000 | loss=2.4237 vocab=1.2686 token=1.6984 | 6.08 steps/s | VRAM=25.14 Go [2026-07-17 05:50:40] step 182850/1000000 | loss=2.4406 vocab=0.7933 token=1.1890 | 6.12 steps/s | VRAM=28.79 Go [2026-07-17 05:50:48] step 182900/1000000 | loss=2.3797 vocab=0.6866 token=1.4002 | 6.09 steps/s | VRAM=28.36 Go [2026-07-17 05:50:57] step 182950/1000000 | loss=2.3389 vocab=0.6655 token=1.2800 | 6.13 steps/s | VRAM=27.36 Go [2026-07-17 05:51:05] step 183000/1000000 | loss=2.5162 vocab=0.7417 token=1.7561 | 6.07 steps/s | VRAM=30.39 Go [2026-07-17 05:51:13] step 183050/1000000 | loss=2.4057 vocab=0.9079 token=1.3768 | 6.11 steps/s | VRAM=28.97 Go [2026-07-17 05:51:21] step 183100/1000000 | loss=2.4448 vocab=0.8716 token=1.2481 | 6.12 steps/s | VRAM=29.05 Go [2026-07-17 05:51:29] step 183150/1000000 | loss=2.4502 vocab=0.7799 token=1.4172 | 6.11 steps/s | VRAM=32.47 Go [2026-07-17 05:51:38] step 183200/1000000 | loss=2.4608 vocab=0.4501 token=1.1847 | 6.09 steps/s | VRAM=26.40 Go [2026-07-17 05:51:46] step 183250/1000000 | loss=2.4081 vocab=0.9683 token=1.1680 | 6.16 steps/s | VRAM=31.63 Go [2026-07-17 05:51:54] step 183300/1000000 | loss=2.4980 vocab=1.3784 token=1.8158 | 6.09 steps/s | VRAM=29.90 Go [2026-07-17 05:52:02] step 183350/1000000 | loss=2.4182 vocab=1.0192 token=1.2029 | 6.09 steps/s | VRAM=29.73 Go [2026-07-17 05:52:10] step 183400/1000000 | loss=2.4102 vocab=1.0977 token=1.3383 | 6.09 steps/s | VRAM=29.15 Go [2026-07-17 05:52:19] step 183450/1000000 | loss=2.4574 vocab=1.7610 token=1.1164 | 6.12 steps/s | VRAM=29.11 Go [2026-07-17 05:52:27] step 183500/1000000 | loss=2.3201 vocab=0.6367 token=1.1263 | 6.09 steps/s | VRAM=28.88 Go [2026-07-17 05:52:35] step 183550/1000000 | loss=2.3072 vocab=1.1095 token=1.6300 | 6.12 steps/s | VRAM=28.68 Go [2026-07-17 05:52:43] step 183600/1000000 | loss=2.3897 vocab=1.4662 token=1.5493 | 6.12 steps/s | VRAM=31.22 Go [2026-07-17 05:52:51] step 183650/1000000 | loss=2.5704 vocab=1.1539 token=1.7380 | 6.07 steps/s | VRAM=25.84 Go [2026-07-17 05:53:00] step 183700/1000000 | loss=2.4393 vocab=1.0253 token=1.1187 | 6.08 steps/s | VRAM=29.13 Go [2026-07-17 05:53:08] step 183750/1000000 | loss=2.3641 vocab=0.9608 token=1.1746 | 6.08 steps/s | VRAM=26.57 Go [2026-07-17 05:53:16] step 183800/1000000 | loss=2.5218 vocab=1.8984 token=1.1112 | 6.11 steps/s | VRAM=27.73 Go [2026-07-17 05:53:24] step 183850/1000000 | loss=2.3469 vocab=0.7736 token=1.5030 | 6.12 steps/s | VRAM=29.78 Go [2026-07-17 05:53:32] step 183900/1000000 | loss=2.5367 vocab=1.0366 token=1.6607 | 6.10 steps/s | VRAM=32.11 Go [2026-07-17 05:53:41] step 183950/1000000 | loss=2.3716 vocab=0.5623 token=1.1825 | 6.11 steps/s | VRAM=28.93 Go [2026-07-17 05:53:49] step 184000/1000000 | loss=2.4377 vocab=0.8279 token=1.7244 | 6.10 steps/s | VRAM=25.93 Go [2026-07-17 05:53:57] step 184050/1000000 | loss=2.3757 vocab=0.9907 token=1.2194 | 6.13 steps/s | VRAM=32.55 Go [2026-07-17 05:54:05] step 184100/1000000 | loss=2.3775 vocab=0.8062 token=1.0221 | 6.12 steps/s | VRAM=28.56 Go [2026-07-17 05:54:13] step 184150/1000000 | loss=2.4311 vocab=0.9310 token=1.6175 | 6.13 steps/s | VRAM=25.31 Go [2026-07-17 05:54:21] step 184200/1000000 | loss=2.4341 vocab=0.6969 token=1.2824 | 6.09 steps/s | VRAM=29.04 Go [2026-07-17 05:54:30] step 184250/1000000 | loss=2.4812 vocab=1.3176 token=1.3584 | 6.07 steps/s | VRAM=26.40 Go [2026-07-17 05:54:38] step 184300/1000000 | loss=2.4393 vocab=1.2230 token=1.2116 | 6.09 steps/s | VRAM=31.49 Go [2026-07-17 05:54:46] step 184350/1000000 | loss=2.4595 vocab=1.0638 token=1.3415 | 6.12 steps/s | VRAM=29.35 Go [2026-07-17 05:54:54] step 184400/1000000 | loss=2.4127 vocab=0.9459 token=1.3513 | 6.12 steps/s | VRAM=32.13 Go [2026-07-17 05:55:02] step 184450/1000000 | loss=2.4911 vocab=1.3619 token=1.7220 | 6.11 steps/s | VRAM=28.78 Go [2026-07-17 05:55:11] step 184500/1000000 | loss=2.2832 vocab=1.0176 token=1.4096 | 6.11 steps/s | VRAM=32.13 Go [2026-07-17 05:55:19] step 184550/1000000 | loss=2.3792 vocab=0.9486 token=1.3281 | 6.11 steps/s | VRAM=29.18 Go [2026-07-17 05:55:27] step 184600/1000000 | loss=2.4309 vocab=1.4482 token=1.6793 | 6.07 steps/s | VRAM=32.95 Go [2026-07-17 05:55:35] step 184650/1000000 | loss=2.4125 vocab=1.0194 token=1.3576 | 6.08 steps/s | VRAM=32.39 Go [2026-07-17 05:55:43] step 184700/1000000 | loss=2.4516 vocab=1.1861 token=1.5009 | 6.10 steps/s | VRAM=28.38 Go [2026-07-17 05:55:52] step 184750/1000000 | loss=2.3766 vocab=0.6193 token=1.5479 | 6.10 steps/s | VRAM=25.29 Go [2026-07-17 05:56:00] step 184800/1000000 | loss=2.3619 vocab=0.8820 token=1.3943 | 6.10 steps/s | VRAM=32.99 Go [2026-07-17 05:56:08] step 184850/1000000 | loss=2.3863 vocab=1.5480 token=1.6919 | 6.08 steps/s | VRAM=28.50 Go [2026-07-17 05:56:16] step 184900/1000000 | loss=2.3856 vocab=0.9256 token=1.1959 | 6.08 steps/s | VRAM=29.31 Go [2026-07-17 05:56:24] step 184950/1000000 | loss=2.3630 vocab=0.9297 token=1.6485 | 6.09 steps/s | VRAM=28.81 Go [2026-07-17 05:56:33] step 185000/1000000 | loss=2.4956 vocab=0.9944 token=1.6768 | 6.11 steps/s | VRAM=26.25 Go [2026-07-17 05:57:53] step 185050/1000000 | loss=2.4102 vocab=1.3034 token=1.3624 | 0.62 steps/s | VRAM=31.54 Go [2026-07-17 05:58:01] step 185100/1000000 | loss=2.4435 vocab=0.8433 token=1.3588 | 6.15 steps/s | VRAM=25.88 Go [2026-07-17 05:58:10] step 185150/1000000 | loss=2.4092 vocab=1.3111 token=1.7284 | 6.17 steps/s | VRAM=26.63 Go [2026-07-17 05:58:18] step 185200/1000000 | loss=2.5320 vocab=2.0408 token=1.8695 | 6.19 steps/s | VRAM=27.66 Go [2026-07-17 05:58:26] step 185250/1000000 | loss=2.3608 vocab=0.4159 token=1.0282 | 6.13 steps/s | VRAM=25.90 Go [2026-07-17 05:58:34] step 185300/1000000 | loss=2.3935 vocab=0.8901 token=1.7829 | 6.13 steps/s | VRAM=25.11 Go [2026-07-17 05:58:42] step 185350/1000000 | loss=2.3752 vocab=0.9459 token=1.5100 | 6.11 steps/s | VRAM=28.20 Go [2026-07-17 05:58:50] step 185400/1000000 | loss=2.3692 vocab=0.9222 token=1.3946 | 6.14 steps/s | VRAM=25.63 Go [2026-07-17 05:58:58] step 185450/1000000 | loss=2.4173 vocab=0.6469 token=1.3847 | 6.13 steps/s | VRAM=25.01 Go [2026-07-17 05:59:07] step 185500/1000000 | loss=2.5024 vocab=1.1135 token=1.3117 | 6.11 steps/s | VRAM=28.96 Go [2026-07-17 05:59:15] step 185550/1000000 | loss=2.4283 vocab=1.1793 token=1.3276 | 6.12 steps/s | VRAM=29.00 Go [2026-07-17 05:59:23] step 185600/1000000 | loss=2.5513 vocab=1.3627 token=1.4494 | 6.08 steps/s | VRAM=24.39 Go [2026-07-17 05:59:31] step 185650/1000000 | loss=2.4314 vocab=0.8516 token=1.0139 | 6.10 steps/s | VRAM=32.38 Go [2026-07-17 05:59:39] step 185700/1000000 | loss=2.3629 vocab=1.1251 token=1.3554 | 6.08 steps/s | VRAM=26.16 Go [2026-07-17 05:59:48] step 185750/1000000 | loss=2.4214 vocab=0.6115 token=1.6190 | 6.09 steps/s | VRAM=28.29 Go [2026-07-17 05:59:56] step 185800/1000000 | loss=2.4310 vocab=0.9859 token=1.1597 | 6.10 steps/s | VRAM=27.57 Go [2026-07-17 06:00:04] step 185850/1000000 | loss=2.4029 vocab=1.4928 token=1.3024 | 6.11 steps/s | VRAM=28.03 Go [2026-07-17 06:00:12] step 185900/1000000 | loss=2.4381 vocab=0.7929 token=1.2977 | 6.06 steps/s | VRAM=26.22 Go [2026-07-17 06:00:20] step 185950/1000000 | loss=2.4257 vocab=0.6799 token=1.4101 | 6.11 steps/s | VRAM=28.18 Go [2026-07-17 06:00:29] step 186000/1000000 | loss=2.3602 vocab=0.8866 token=1.4194 | 6.07 steps/s | VRAM=32.88 Go [2026-07-17 06:00:37] step 186050/1000000 | loss=2.4850 vocab=1.2204 token=1.6371 | 6.08 steps/s | VRAM=29.24 Go [2026-07-17 06:00:45] step 186100/1000000 | loss=2.4124 vocab=0.6944 token=1.0382 | 6.09 steps/s | VRAM=26.53 Go [2026-07-17 06:00:53] step 186150/1000000 | loss=2.4880 vocab=1.3415 token=2.0623 | 6.13 steps/s | VRAM=24.47 Go [2026-07-17 06:01:01] step 186200/1000000 | loss=2.4661 vocab=0.7198 token=0.9904 | 6.13 steps/s | VRAM=33.00 Go [2026-07-17 06:01:10] step 186250/1000000 | loss=2.4273 vocab=0.6677 token=1.3410 | 6.09 steps/s | VRAM=28.11 Go [2026-07-17 06:01:18] step 186300/1000000 | loss=2.4284 vocab=1.2106 token=1.6070 | 6.08 steps/s | VRAM=31.73 Go [2026-07-17 06:01:26] step 186350/1000000 | loss=2.3431 vocab=1.5197 token=1.5786 | 6.11 steps/s | VRAM=25.74 Go [2026-07-17 06:01:34] step 186400/1000000 | loss=2.3201 vocab=0.8162 token=1.3469 | 6.12 steps/s | VRAM=31.32 Go [2026-07-17 06:01:42] step 186450/1000000 | loss=2.3272 vocab=1.1039 token=1.6550 | 6.09 steps/s | VRAM=29.38 Go [2026-07-17 06:01:51] step 186500/1000000 | loss=2.3577 vocab=0.8161 token=1.2003 | 6.10 steps/s | VRAM=26.41 Go [2026-07-17 06:01:59] step 186550/1000000 | loss=2.3569 vocab=1.3444 token=1.3712 | 6.10 steps/s | VRAM=28.10 Go [2026-07-17 06:02:07] step 186600/1000000 | loss=2.4223 vocab=0.7304 token=1.4990 | 6.12 steps/s | VRAM=32.47 Go [2026-07-17 06:02:15] step 186650/1000000 | loss=2.2257 vocab=0.8183 token=1.1934 | 6.11 steps/s | VRAM=29.09 Go [2026-07-17 06:02:23] step 186700/1000000 | loss=2.3318 vocab=1.1407 token=1.2684 | 6.09 steps/s | VRAM=25.93 Go [2026-07-17 06:02:32] step 186750/1000000 | loss=2.4264 vocab=1.2112 token=1.3846 | 6.14 steps/s | VRAM=32.61 Go [2026-07-17 06:02:40] step 186800/1000000 | loss=2.3621 vocab=1.1027 token=1.4746 | 6.11 steps/s | VRAM=29.63 Go [2026-07-17 06:02:48] step 186850/1000000 | loss=2.3841 vocab=1.6012 token=1.6620 | 6.12 steps/s | VRAM=24.72 Go [2026-07-17 06:02:56] step 186900/1000000 | loss=2.4884 vocab=1.2100 token=1.5825 | 6.11 steps/s | VRAM=32.25 Go [2026-07-17 06:03:04] step 186950/1000000 | loss=2.4267 vocab=0.7920 token=1.5169 | 6.11 steps/s | VRAM=31.35 Go [2026-07-17 06:03:12] step 187000/1000000 | loss=2.4603 vocab=0.8288 token=1.3234 | 6.08 steps/s | VRAM=26.05 Go [2026-07-17 06:03:21] step 187050/1000000 | loss=2.4176 vocab=1.3908 token=1.2446 | 6.07 steps/s | VRAM=25.96 Go [2026-07-17 06:03:29] step 187100/1000000 | loss=2.5259 vocab=1.8372 token=1.5784 | 6.11 steps/s | VRAM=25.97 Go [2026-07-17 06:03:37] step 187150/1000000 | loss=2.4810 vocab=1.5461 token=1.4264 | 6.09 steps/s | VRAM=29.95 Go [2026-07-17 06:03:45] step 187200/1000000 | loss=2.4499 vocab=1.0220 token=1.5045 | 6.12 steps/s | VRAM=31.69 Go [2026-07-17 06:03:53] step 187250/1000000 | loss=2.4572 vocab=0.7404 token=1.0098 | 6.12 steps/s | VRAM=28.70 Go [2026-07-17 06:04:02] step 187300/1000000 | loss=2.4266 vocab=0.7897 token=1.6241 | 6.10 steps/s | VRAM=32.28 Go [2026-07-17 06:04:10] step 187350/1000000 | loss=2.4870 vocab=1.0040 token=1.4444 | 6.11 steps/s | VRAM=29.29 Go [2026-07-17 06:04:18] step 187400/1000000 | loss=2.5187 vocab=0.7669 token=1.2630 | 6.13 steps/s | VRAM=31.65 Go [2026-07-17 06:04:26] step 187450/1000000 | loss=2.4528 vocab=0.9537 token=1.3178 | 6.11 steps/s | VRAM=32.84 Go [2026-07-17 06:04:34] step 187500/1000000 | loss=2.4882 vocab=1.5068 token=1.3848 | 6.10 steps/s | VRAM=31.61 Go [2026-07-17 06:04:43] step 187550/1000000 | loss=2.3408 vocab=0.5421 token=1.2418 | 6.11 steps/s | VRAM=26.22 Go [2026-07-17 06:04:51] step 187600/1000000 | loss=2.4128 vocab=0.4462 token=1.3492 | 6.12 steps/s | VRAM=30.72 Go [2026-07-17 06:04:59] step 187650/1000000 | loss=2.3838 vocab=0.6821 token=1.5749 | 6.10 steps/s | VRAM=32.80 Go [2026-07-17 06:05:07] step 187700/1000000 | loss=2.4044 vocab=0.8904 token=1.4748 | 6.13 steps/s | VRAM=28.70 Go [2026-07-17 06:05:15] step 187750/1000000 | loss=2.4062 vocab=1.5273 token=1.6073 | 6.07 steps/s | VRAM=23.85 Go [2026-07-17 06:05:23] step 187800/1000000 | loss=2.3595 vocab=1.0749 token=1.6393 | 6.12 steps/s | VRAM=32.58 Go [2026-07-17 06:05:32] step 187850/1000000 | loss=2.5607 vocab=0.8449 token=1.6493 | 6.11 steps/s | VRAM=28.03 Go [2026-07-17 06:05:40] step 187900/1000000 | loss=2.3721 vocab=1.8508 token=1.5924 | 6.09 steps/s | VRAM=26.86 Go [2026-07-17 06:05:48] step 187950/1000000 | loss=2.4038 vocab=1.0144 token=1.6688 | 6.13 steps/s | VRAM=25.80 Go [2026-07-17 06:05:56] step 188000/1000000 | loss=2.4318 vocab=0.9332 token=0.8610 | 6.15 steps/s | VRAM=28.68 Go [2026-07-17 06:06:04] step 188050/1000000 | loss=2.3186 vocab=0.9895 token=1.5014 | 6.12 steps/s | VRAM=28.49 Go [2026-07-17 06:06:13] step 188100/1000000 | loss=2.3867 vocab=0.8914 token=1.4180 | 6.10 steps/s | VRAM=32.56 Go [2026-07-17 06:06:21] step 188150/1000000 | loss=2.4744 vocab=1.1399 token=1.1589 | 6.11 steps/s | VRAM=26.09 Go [2026-07-17 06:06:29] step 188200/1000000 | loss=2.5238 vocab=0.5571 token=1.4391 | 6.10 steps/s | VRAM=29.00 Go [2026-07-17 06:06:37] step 188250/1000000 | loss=2.4579 vocab=1.0752 token=1.3150 | 6.09 steps/s | VRAM=29.38 Go [2026-07-17 06:06:45] step 188300/1000000 | loss=2.4487 vocab=1.0741 token=1.7021 | 6.09 steps/s | VRAM=32.56 Go [2026-07-17 06:06:53] step 188350/1000000 | loss=2.4162 vocab=0.6299 token=1.4862 | 6.12 steps/s | VRAM=32.08 Go [2026-07-17 06:07:02] step 188400/1000000 | loss=2.3583 vocab=0.9656 token=1.6688 | 6.11 steps/s | VRAM=32.76 Go [2026-07-17 06:07:10] step 188450/1000000 | loss=2.4511 vocab=0.4737 token=1.6234 | 6.10 steps/s | VRAM=29.13 Go [2026-07-17 06:07:18] step 188500/1000000 | loss=2.4102 vocab=1.0562 token=1.2839 | 6.09 steps/s | VRAM=29.38 Go [2026-07-17 06:07:26] step 188550/1000000 | loss=2.4775 vocab=0.8889 token=1.8254 | 6.11 steps/s | VRAM=32.43 Go [2026-07-17 06:07:34] step 188600/1000000 | loss=2.4082 vocab=1.0493 token=1.5789 | 6.08 steps/s | VRAM=28.91 Go [2026-07-17 06:07:43] step 188650/1000000 | loss=2.3186 vocab=0.8986 token=1.2612 | 6.13 steps/s | VRAM=29.79 Go [2026-07-17 06:07:51] step 188700/1000000 | loss=2.3621 vocab=0.6454 token=1.1502 | 6.09 steps/s | VRAM=29.23 Go [2026-07-17 06:07:59] step 188750/1000000 | loss=2.4341 vocab=1.2398 token=1.5844 | 6.13 steps/s | VRAM=28.11 Go [2026-07-17 06:08:07] step 188800/1000000 | loss=2.4066 vocab=0.7377 token=1.3443 | 6.10 steps/s | VRAM=28.64 Go [2026-07-17 06:08:15] step 188850/1000000 | loss=2.4049 vocab=1.1184 token=1.8760 | 6.08 steps/s | VRAM=26.81 Go [2026-07-17 06:08:24] step 188900/1000000 | loss=2.4886 vocab=1.0377 token=1.6979 | 6.12 steps/s | VRAM=31.59 Go [2026-07-17 06:08:32] step 188950/1000000 | loss=2.3381 vocab=0.8546 token=1.5302 | 6.14 steps/s | VRAM=28.05 Go [2026-07-17 06:08:40] step 189000/1000000 | loss=2.2997 vocab=0.9092 token=1.6804 | 6.12 steps/s | VRAM=32.55 Go [2026-07-17 06:08:48] step 189050/1000000 | loss=2.3642 vocab=1.1241 token=1.2963 | 6.11 steps/s | VRAM=32.81 Go [2026-07-17 06:08:56] step 189100/1000000 | loss=2.4693 vocab=1.2715 token=1.6947 | 6.12 steps/s | VRAM=28.13 Go [2026-07-17 06:09:04] step 189150/1000000 | loss=2.4238 vocab=0.9364 token=1.3870 | 6.11 steps/s | VRAM=28.50 Go [2026-07-17 06:09:13] step 189200/1000000 | loss=2.3613 vocab=0.7497 token=1.5972 | 6.13 steps/s | VRAM=27.16 Go [2026-07-17 06:09:21] step 189250/1000000 | loss=2.3840 vocab=0.8577 token=1.1752 | 6.10 steps/s | VRAM=28.33 Go [2026-07-17 06:09:29] step 189300/1000000 | loss=2.3458 vocab=0.5910 token=1.0930 | 6.13 steps/s | VRAM=28.05 Go [2026-07-17 06:09:37] step 189350/1000000 | loss=2.4556 vocab=0.8927 token=1.6024 | 6.13 steps/s | VRAM=32.29 Go [2026-07-17 06:09:45] step 189400/1000000 | loss=2.3759 vocab=1.3845 token=1.2362 | 6.11 steps/s | VRAM=29.23 Go [2026-07-17 06:09:54] step 189450/1000000 | loss=2.3402 vocab=0.9081 token=1.0775 | 6.11 steps/s | VRAM=28.79 Go [2026-07-17 06:10:02] step 189500/1000000 | loss=2.3257 vocab=1.2118 token=1.7005 | 6.16 steps/s | VRAM=29.43 Go [2026-07-17 06:10:10] step 189550/1000000 | loss=2.3179 vocab=1.0744 token=1.3808 | 6.14 steps/s | VRAM=31.05 Go [2026-07-17 06:10:18] step 189600/1000000 | loss=2.4765 vocab=1.2534 token=1.2073 | 6.09 steps/s | VRAM=26.97 Go [2026-07-17 06:10:26] step 189650/1000000 | loss=2.4129 vocab=1.1215 token=1.7420 | 6.15 steps/s | VRAM=28.21 Go [2026-07-17 06:10:34] step 189700/1000000 | loss=2.2899 vocab=1.0153 token=1.1215 | 6.11 steps/s | VRAM=26.10 Go [2026-07-17 06:10:42] step 189750/1000000 | loss=2.4824 vocab=0.7037 token=1.5406 | 6.13 steps/s | VRAM=25.81 Go [2026-07-17 06:10:51] step 189800/1000000 | loss=2.2697 vocab=1.1143 token=1.6449 | 6.08 steps/s | VRAM=28.24 Go [2026-07-17 06:10:59] step 189850/1000000 | loss=2.4489 vocab=0.7077 token=1.2066 | 6.10 steps/s | VRAM=29.24 Go [2026-07-17 06:11:07] step 189900/1000000 | loss=2.3735 vocab=0.8630 token=1.3421 | 6.13 steps/s | VRAM=32.27 Go [2026-07-17 06:11:15] step 189950/1000000 | loss=2.3582 vocab=0.5014 token=1.2219 | 6.14 steps/s | VRAM=31.85 Go [2026-07-17 06:11:23] step 190000/1000000 | loss=2.3064 vocab=1.4724 token=1.2555 | 6.13 steps/s | VRAM=28.49 Go [2026-07-17 06:12:44] step 190050/1000000 | loss=2.4485 vocab=0.5575 token=1.3770 | 0.62 steps/s | VRAM=25.89 Go [2026-07-17 06:12:52] step 190100/1000000 | loss=2.4117 vocab=0.8210 token=1.2648 | 6.20 steps/s | VRAM=32.24 Go [2026-07-17 06:13:00] step 190150/1000000 | loss=2.3335 vocab=0.9172 token=1.3649 | 6.18 steps/s | VRAM=25.91 Go [2026-07-17 06:13:08] step 190200/1000000 | loss=2.5397 vocab=1.2114 token=1.4288 | 6.16 steps/s | VRAM=31.85 Go [2026-07-17 06:13:16] step 190250/1000000 | loss=2.4446 vocab=0.8678 token=1.3918 | 6.15 steps/s | VRAM=26.69 Go [2026-07-17 06:13:25] step 190300/1000000 | loss=2.4497 vocab=1.2563 token=1.7858 | 6.17 steps/s | VRAM=28.99 Go [2026-07-17 06:13:33] step 190350/1000000 | loss=2.4055 vocab=0.7698 token=1.2425 | 6.15 steps/s | VRAM=28.48 Go [2026-07-17 06:13:41] step 190400/1000000 | loss=2.3750 vocab=0.6861 token=1.1594 | 6.11 steps/s | VRAM=28.92 Go [2026-07-17 06:13:49] step 190450/1000000 | loss=2.3622 vocab=0.9705 token=1.2485 | 6.12 steps/s | VRAM=27.92 Go [2026-07-17 06:13:57] step 190500/1000000 | loss=2.3916 vocab=1.1767 token=1.6227 | 6.09 steps/s | VRAM=28.33 Go [2026-07-17 06:14:05] step 190550/1000000 | loss=2.3007 vocab=0.6777 token=1.4249 | 6.13 steps/s | VRAM=29.01 Go [2026-07-17 06:14:14] step 190600/1000000 | loss=2.2742 vocab=0.9469 token=1.1072 | 6.09 steps/s | VRAM=26.75 Go [2026-07-17 06:14:22] step 190650/1000000 | loss=2.4512 vocab=1.5799 token=1.6983 | 6.08 steps/s | VRAM=25.27 Go [2026-07-17 06:14:30] step 190700/1000000 | loss=2.3069 vocab=0.8415 token=1.4411 | 6.09 steps/s | VRAM=26.48 Go [2026-07-17 06:14:38] step 190750/1000000 | loss=2.4001 vocab=1.2036 token=1.4213 | 6.12 steps/s | VRAM=25.60 Go [2026-07-17 06:14:46] step 190800/1000000 | loss=2.3967 vocab=0.9974 token=1.2028 | 6.09 steps/s | VRAM=26.10 Go [2026-07-17 06:14:55] step 190850/1000000 | loss=2.3949 vocab=1.1512 token=1.4613 | 6.11 steps/s | VRAM=26.21 Go [2026-07-17 06:15:03] step 190900/1000000 | loss=2.4061 vocab=0.8364 token=1.3005 | 6.09 steps/s | VRAM=28.35 Go [2026-07-17 06:15:11] step 190950/1000000 | loss=2.3574 vocab=1.6407 token=1.5876 | 6.12 steps/s | VRAM=28.45 Go [2026-07-17 06:15:19] step 191000/1000000 | loss=2.3784 vocab=0.8487 token=1.2907 | 6.09 steps/s | VRAM=31.97 Go [2026-07-17 06:15:27] step 191050/1000000 | loss=2.4524 vocab=1.2370 token=1.4503 | 6.11 steps/s | VRAM=25.89 Go [2026-07-17 06:15:36] step 191100/1000000 | loss=2.3693 vocab=1.2146 token=1.4218 | 6.09 steps/s | VRAM=26.06 Go [2026-07-17 06:15:44] step 191150/1000000 | loss=2.3485 vocab=0.9723 token=1.0775 | 6.11 steps/s | VRAM=28.58 Go [2026-07-17 06:15:52] step 191200/1000000 | loss=2.3725 vocab=1.3123 token=1.3861 | 6.12 steps/s | VRAM=29.13 Go [2026-07-17 06:16:00] step 191250/1000000 | loss=2.3862 vocab=0.9310 token=1.6284 | 6.11 steps/s | VRAM=31.75 Go [2026-07-17 06:16:08] step 191300/1000000 | loss=2.3748 vocab=1.1630 token=1.3709 | 6.08 steps/s | VRAM=26.34 Go [2026-07-17 06:16:17] step 191350/1000000 | loss=2.4447 vocab=1.3982 token=1.8631 | 6.12 steps/s | VRAM=26.03 Go [2026-07-17 06:16:25] step 191400/1000000 | loss=2.4663 vocab=0.8984 token=1.1304 | 6.11 steps/s | VRAM=26.05 Go [2026-07-17 06:16:33] step 191450/1000000 | loss=2.2776 vocab=1.1087 token=1.2407 | 6.13 steps/s | VRAM=24.92 Go [2026-07-17 06:16:41] step 191500/1000000 | loss=2.4929 vocab=2.4159 token=1.6559 | 6.12 steps/s | VRAM=32.07 Go [2026-07-17 06:16:49] step 191550/1000000 | loss=2.4687 vocab=1.0725 token=1.1688 | 6.05 steps/s | VRAM=23.74 Go [2026-07-17 06:16:57] step 191600/1000000 | loss=2.4422 vocab=1.0082 token=1.1737 | 6.15 steps/s | VRAM=30.91 Go [2026-07-17 06:17:06] step 191650/1000000 | loss=2.4397 vocab=1.1712 token=1.2366 | 6.10 steps/s | VRAM=32.88 Go [2026-07-17 06:17:14] step 191700/1000000 | loss=2.3962 vocab=0.8728 token=1.3590 | 6.10 steps/s | VRAM=28.57 Go [2026-07-17 06:17:22] step 191750/1000000 | loss=2.4587 vocab=0.9740 token=1.4784 | 6.10 steps/s | VRAM=23.07 Go [2026-07-17 06:17:30] step 191800/1000000 | loss=2.2746 vocab=1.0656 token=1.3688 | 6.07 steps/s | VRAM=28.84 Go [2026-07-17 06:17:38] step 191850/1000000 | loss=2.5051 vocab=1.2363 token=1.4282 | 6.08 steps/s | VRAM=29.49 Go [2026-07-17 06:17:47] step 191900/1000000 | loss=2.4256 vocab=1.0702 token=1.5820 | 6.10 steps/s | VRAM=25.22 Go [2026-07-17 06:17:55] step 191950/1000000 | loss=2.3645 vocab=0.8400 token=1.6480 | 6.09 steps/s | VRAM=29.00 Go [2026-07-17 06:18:03] step 192000/1000000 | loss=2.3056 vocab=1.0106 token=1.7502 | 6.11 steps/s | VRAM=25.69 Go [2026-07-17 06:18:11] step 192050/1000000 | loss=2.3996 vocab=1.1623 token=1.4678 | 6.10 steps/s | VRAM=28.39 Go [2026-07-17 06:18:19] step 192100/1000000 | loss=2.4119 vocab=0.8707 token=1.3841 | 6.12 steps/s | VRAM=31.90 Go [2026-07-17 06:18:28] step 192150/1000000 | loss=2.3008 vocab=1.0995 token=1.8011 | 6.09 steps/s | VRAM=25.76 Go [2026-07-17 06:18:36] step 192200/1000000 | loss=2.4946 vocab=0.6401 token=0.9911 | 6.10 steps/s | VRAM=28.43 Go [2026-07-17 06:18:44] step 192250/1000000 | loss=2.3162 vocab=0.9354 token=1.4264 | 6.10 steps/s | VRAM=28.25 Go [2026-07-17 06:18:52] step 192300/1000000 | loss=2.4727 vocab=0.6319 token=1.0895 | 6.11 steps/s | VRAM=29.12 Go [2026-07-17 06:19:00] step 192350/1000000 | loss=2.2980 vocab=0.6329 token=1.3914 | 6.10 steps/s | VRAM=24.82 Go [2026-07-17 06:19:09] step 192400/1000000 | loss=2.3807 vocab=1.1684 token=1.1271 | 6.06 steps/s | VRAM=30.36 Go [2026-07-17 06:19:17] step 192450/1000000 | loss=2.4257 vocab=0.6341 token=1.0283 | 6.14 steps/s | VRAM=27.57 Go [2026-07-17 06:19:25] step 192500/1000000 | loss=2.3963 vocab=1.1284 token=1.8275 | 6.09 steps/s | VRAM=28.36 Go [2026-07-17 06:19:33] step 192550/1000000 | loss=2.4215 vocab=0.8891 token=1.2395 | 6.09 steps/s | VRAM=29.98 Go [2026-07-17 06:19:41] step 192600/1000000 | loss=2.4105 vocab=0.7361 token=1.3570 | 6.08 steps/s | VRAM=29.70 Go [2026-07-17 06:19:50] step 192650/1000000 | loss=2.4349 vocab=0.7892 token=1.4054 | 6.11 steps/s | VRAM=32.99 Go [2026-07-17 06:19:58] step 192700/1000000 | loss=2.4063 vocab=0.8028 token=1.5663 | 6.09 steps/s | VRAM=28.56 Go [2026-07-17 06:20:06] step 192750/1000000 | loss=2.4183 vocab=0.6948 token=1.8005 | 6.09 steps/s | VRAM=26.95 Go [2026-07-17 06:20:14] step 192800/1000000 | loss=2.4673 vocab=1.2641 token=1.5746 | 6.09 steps/s | VRAM=28.65 Go [2026-07-17 06:20:22] step 192850/1000000 | loss=2.3849 vocab=1.0574 token=1.3871 | 6.14 steps/s | VRAM=25.59 Go [2026-07-17 06:20:31] step 192900/1000000 | loss=2.3809 vocab=0.8478 token=1.5645 | 6.15 steps/s | VRAM=27.97 Go [2026-07-17 06:20:39] step 192950/1000000 | loss=2.3027 vocab=0.7899 token=1.4963 | 6.12 steps/s | VRAM=25.40 Go [2026-07-17 06:20:47] step 193000/1000000 | loss=2.3889 vocab=0.9053 token=1.4557 | 6.11 steps/s | VRAM=29.05 Go [2026-07-17 06:20:55] step 193050/1000000 | loss=2.4223 vocab=1.2577 token=1.8812 | 6.13 steps/s | VRAM=27.79 Go [2026-07-17 06:21:03] step 193100/1000000 | loss=2.3466 vocab=0.9585 token=1.5126 | 6.07 steps/s | VRAM=26.00 Go [2026-07-17 06:21:11] step 193150/1000000 | loss=2.3713 vocab=0.6525 token=1.1964 | 6.11 steps/s | VRAM=27.71 Go [2026-07-17 06:21:20] step 193200/1000000 | loss=2.4357 vocab=1.2218 token=1.6273 | 6.13 steps/s | VRAM=28.73 Go [2026-07-17 06:21:28] step 193250/1000000 | loss=2.3127 vocab=1.0598 token=1.4438 | 6.11 steps/s | VRAM=28.84 Go [2026-07-17 06:21:36] step 193300/1000000 | loss=2.4897 vocab=0.6383 token=1.2167 | 6.13 steps/s | VRAM=31.85 Go [2026-07-17 06:21:44] step 193350/1000000 | loss=2.3618 vocab=1.0294 token=1.5466 | 6.10 steps/s | VRAM=28.11 Go [2026-07-17 06:21:52] step 193400/1000000 | loss=2.4118 vocab=0.7123 token=1.9245 | 6.12 steps/s | VRAM=26.24 Go [2026-07-17 06:22:00] step 193450/1000000 | loss=2.4398 vocab=1.3669 token=1.5776 | 6.13 steps/s | VRAM=29.15 Go [2026-07-17 06:22:09] step 193500/1000000 | loss=2.3831 vocab=0.8675 token=1.2728 | 6.13 steps/s | VRAM=29.19 Go [2026-07-17 06:22:17] step 193550/1000000 | loss=2.3380 vocab=0.8242 token=1.4762 | 6.13 steps/s | VRAM=31.53 Go [2026-07-17 06:22:25] step 193600/1000000 | loss=2.3440 vocab=0.9794 token=1.2539 | 6.10 steps/s | VRAM=26.98 Go [2026-07-17 06:22:33] step 193650/1000000 | loss=2.3165 vocab=0.6025 token=1.2849 | 6.16 steps/s | VRAM=31.09 Go [2026-07-17 06:22:41] step 193700/1000000 | loss=2.4430 vocab=1.1122 token=1.3603 | 6.11 steps/s | VRAM=30.27 Go [2026-07-17 06:22:49] step 193750/1000000 | loss=2.3601 vocab=0.8887 token=1.2056 | 6.15 steps/s | VRAM=31.93 Go [2026-07-17 06:22:58] step 193800/1000000 | loss=2.4150 vocab=0.7814 token=1.8690 | 6.14 steps/s | VRAM=28.51 Go [2026-07-17 06:23:06] step 193850/1000000 | loss=2.4008 vocab=0.7399 token=1.8116 | 6.13 steps/s | VRAM=32.85 Go [2026-07-17 06:23:14] step 193900/1000000 | loss=2.3608 vocab=0.7606 token=1.5722 | 6.12 steps/s | VRAM=29.24 Go [2026-07-17 06:23:22] step 193950/1000000 | loss=2.3953 vocab=1.3281 token=1.3425 | 6.14 steps/s | VRAM=31.60 Go [2026-07-17 06:23:30] step 194000/1000000 | loss=2.4174 vocab=0.7170 token=1.2579 | 6.11 steps/s | VRAM=32.44 Go [2026-07-17 06:23:38] step 194050/1000000 | loss=2.4465 vocab=1.0866 token=1.3039 | 6.09 steps/s | VRAM=29.07 Go [2026-07-17 06:23:47] step 194100/1000000 | loss=2.5445 vocab=0.7647 token=1.3956 | 6.11 steps/s | VRAM=32.26 Go [2026-07-17 06:23:55] step 194150/1000000 | loss=2.3982 vocab=0.7072 token=1.5033 | 6.11 steps/s | VRAM=29.58 Go [2026-07-17 06:24:03] step 194200/1000000 | loss=2.3739 vocab=0.9452 token=1.1360 | 6.15 steps/s | VRAM=28.14 Go [2026-07-17 06:24:11] step 194250/1000000 | loss=2.3604 vocab=0.7480 token=1.2176 | 6.12 steps/s | VRAM=28.24 Go [2026-07-17 06:24:19] step 194300/1000000 | loss=2.4239 vocab=0.3175 token=0.9258 | 6.13 steps/s | VRAM=30.77 Go [2026-07-17 06:24:27] step 194350/1000000 | loss=2.3698 vocab=1.0044 token=1.5579 | 6.12 steps/s | VRAM=29.02 Go [2026-07-17 06:24:36] step 194400/1000000 | loss=2.4167 vocab=1.2858 token=1.3718 | 6.12 steps/s | VRAM=31.95 Go [2026-07-17 06:24:44] step 194450/1000000 | loss=2.4447 vocab=1.1438 token=1.5304 | 6.12 steps/s | VRAM=28.45 Go [2026-07-17 06:24:52] step 194500/1000000 | loss=2.3678 vocab=0.6875 token=1.3324 | 6.14 steps/s | VRAM=32.45 Go [2026-07-17 06:25:00] step 194550/1000000 | loss=2.3805 vocab=0.8265 token=1.1066 | 6.11 steps/s | VRAM=25.90 Go [2026-07-17 06:25:08] step 194600/1000000 | loss=2.4298 vocab=0.9978 token=1.7626 | 6.15 steps/s | VRAM=32.20 Go [2026-07-17 06:25:16] step 194650/1000000 | loss=2.4629 vocab=1.3907 token=1.3526 | 6.13 steps/s | VRAM=25.54 Go [2026-07-17 06:25:25] step 194700/1000000 | loss=2.3740 vocab=0.9378 token=1.7340 | 6.07 steps/s | VRAM=26.60 Go [2026-07-17 06:25:33] step 194750/1000000 | loss=2.3865 vocab=0.8422 token=1.3823 | 6.14 steps/s | VRAM=30.89 Go [2026-07-17 06:25:41] step 194800/1000000 | loss=2.3724 vocab=0.6576 token=1.0748 | 6.13 steps/s | VRAM=25.16 Go [2026-07-17 06:25:49] step 194850/1000000 | loss=2.4631 vocab=1.4949 token=1.4963 | 6.13 steps/s | VRAM=28.89 Go [2026-07-17 06:25:57] step 194900/1000000 | loss=2.4365 vocab=0.8891 token=1.3921 | 6.11 steps/s | VRAM=26.12 Go [2026-07-17 06:26:05] step 194950/1000000 | loss=2.3585 vocab=1.0146 token=1.2540 | 6.14 steps/s | VRAM=28.71 Go [2026-07-17 06:26:14] step 195000/1000000 | loss=2.3825 vocab=0.9285 token=1.5714 | 6.14 steps/s | VRAM=30.79 Go [2026-07-17 06:27:37] step 195050/1000000 | loss=2.3624 vocab=0.6654 token=1.3626 | 0.60 steps/s | VRAM=28.23 Go [2026-07-17 06:27:45] step 195100/1000000 | loss=2.3340 vocab=1.3073 token=1.4053 | 6.21 steps/s | VRAM=28.73 Go [2026-07-17 06:27:53] step 195150/1000000 | loss=2.4807 vocab=0.9945 token=1.1684 | 6.18 steps/s | VRAM=29.03 Go [2026-07-17 06:28:01] step 195200/1000000 | loss=2.2931 vocab=0.6958 token=1.3147 | 6.20 steps/s | VRAM=32.00 Go [2026-07-17 06:28:09] step 195250/1000000 | loss=2.3222 vocab=1.0842 token=1.2775 | 6.15 steps/s | VRAM=26.00 Go [2026-07-17 06:28:17] step 195300/1000000 | loss=2.4586 vocab=1.2264 token=1.6462 | 6.17 steps/s | VRAM=25.38 Go [2026-07-17 06:28:25] step 195350/1000000 | loss=2.3634 vocab=1.1495 token=1.5612 | 6.14 steps/s | VRAM=28.83 Go [2026-07-17 06:28:34] step 195400/1000000 | loss=2.3175 vocab=0.9147 token=1.6043 | 6.12 steps/s | VRAM=32.75 Go [2026-07-17 06:28:42] step 195450/1000000 | loss=2.2852 vocab=0.8671 token=1.2487 | 6.11 steps/s | VRAM=25.86 Go [2026-07-17 06:28:50] step 195500/1000000 | loss=2.3705 vocab=0.9949 token=1.0928 | 6.12 steps/s | VRAM=26.62 Go [2026-07-17 06:28:58] step 195550/1000000 | loss=2.3528 vocab=0.8994 token=1.3358 | 6.13 steps/s | VRAM=31.90 Go [2026-07-17 06:29:06] step 195600/1000000 | loss=2.4277 vocab=1.6494 token=1.8237 | 6.12 steps/s | VRAM=29.24 Go [2026-07-17 06:29:14] step 195650/1000000 | loss=2.3172 vocab=1.0025 token=1.1453 | 6.11 steps/s | VRAM=28.44 Go [2026-07-17 06:29:23] step 195700/1000000 | loss=2.5203 vocab=0.6285 token=1.4207 | 6.13 steps/s | VRAM=24.75 Go [2026-07-17 06:29:31] step 195750/1000000 | loss=2.3079 vocab=0.8268 token=1.3917 | 6.12 steps/s | VRAM=32.51 Go [2026-07-17 06:29:39] step 195800/1000000 | loss=2.3368 vocab=0.8000 token=1.3805 | 6.10 steps/s | VRAM=24.51 Go [2026-07-17 06:29:47] step 195850/1000000 | loss=2.4272 vocab=0.9086 token=1.7206 | 6.11 steps/s | VRAM=28.68 Go [2026-07-17 06:29:55] step 195900/1000000 | loss=2.3779 vocab=1.1249 token=1.3340 | 6.13 steps/s | VRAM=28.93 Go [2026-07-17 06:30:03] step 195950/1000000 | loss=2.4078 vocab=0.8358 token=1.4700 | 6.13 steps/s | VRAM=29.41 Go [2026-07-17 06:30:12] step 196000/1000000 | loss=2.4231 vocab=0.9368 token=2.1266 | 6.12 steps/s | VRAM=28.72 Go [2026-07-17 06:30:20] step 196050/1000000 | loss=2.3094 vocab=1.1494 token=1.2973 | 6.13 steps/s | VRAM=32.76 Go [2026-07-17 06:30:28] step 196100/1000000 | loss=2.3974 vocab=1.4222 token=1.4169 | 6.10 steps/s | VRAM=26.08 Go [2026-07-17 06:30:36] step 196150/1000000 | loss=2.3801 vocab=0.8448 token=1.3431 | 6.11 steps/s | VRAM=28.65 Go [2026-07-17 06:30:44] step 196200/1000000 | loss=2.3262 vocab=1.3083 token=1.7222 | 6.10 steps/s | VRAM=25.72 Go [2026-07-17 06:30:52] step 196250/1000000 | loss=2.4119 vocab=0.9088 token=1.8297 | 6.13 steps/s | VRAM=31.43 Go [2026-07-17 06:31:01] step 196300/1000000 | loss=2.3879 vocab=0.9478 token=1.0855 | 6.12 steps/s | VRAM=28.72 Go [2026-07-17 06:31:09] step 196350/1000000 | loss=2.3826 vocab=0.9336 token=1.5383 | 6.12 steps/s | VRAM=32.22 Go [2026-07-17 06:31:17] step 196400/1000000 | loss=2.4380 vocab=0.7894 token=1.3927 | 6.11 steps/s | VRAM=26.87 Go [2026-07-17 06:31:25] step 196450/1000000 | loss=2.3849 vocab=1.2029 token=1.6195 | 6.13 steps/s | VRAM=30.64 Go [2026-07-17 06:31:33] step 196500/1000000 | loss=2.3202 vocab=0.6687 token=1.3290 | 6.12 steps/s | VRAM=28.14 Go [2026-07-17 06:31:41] step 196550/1000000 | loss=2.4637 vocab=0.7038 token=1.3160 | 6.13 steps/s | VRAM=32.46 Go [2026-07-17 06:31:50] step 196600/1000000 | loss=2.3870 vocab=0.7451 token=1.3148 | 6.09 steps/s | VRAM=28.26 Go [2026-07-17 06:31:58] step 196650/1000000 | loss=2.3958 vocab=0.7090 token=1.3079 | 6.08 steps/s | VRAM=28.77 Go [2026-07-17 06:32:06] step 196700/1000000 | loss=2.4642 vocab=1.1130 token=1.8723 | 6.11 steps/s | VRAM=26.10 Go [2026-07-17 06:32:14] step 196750/1000000 | loss=2.4655 vocab=0.6842 token=0.9308 | 6.11 steps/s | VRAM=28.93 Go [2026-07-17 06:32:22] step 196800/1000000 | loss=2.3815 vocab=0.8233 token=1.1730 | 6.11 steps/s | VRAM=27.81 Go [2026-07-17 06:32:31] step 196850/1000000 | loss=2.4051 vocab=1.1555 token=1.3585 | 6.11 steps/s | VRAM=28.65 Go [2026-07-17 06:32:39] step 196900/1000000 | loss=2.4213 vocab=1.0766 token=1.2787 | 6.12 steps/s | VRAM=29.41 Go [2026-07-17 06:32:47] step 196950/1000000 | loss=2.3306 vocab=0.9719 token=1.1223 | 6.12 steps/s | VRAM=24.74 Go [2026-07-17 06:32:55] step 197000/1000000 | loss=2.3957 vocab=0.7285 token=1.2155 | 6.08 steps/s | VRAM=29.03 Go [2026-07-17 06:33:03] step 197050/1000000 | loss=2.3804 vocab=1.1804 token=1.3719 | 6.12 steps/s | VRAM=28.76 Go [2026-07-17 06:33:12] step 197100/1000000 | loss=2.4391 vocab=0.9065 token=1.5696 | 6.11 steps/s | VRAM=32.39 Go [2026-07-17 06:33:20] step 197150/1000000 | loss=2.4200 vocab=1.0306 token=1.3503 | 6.10 steps/s | VRAM=26.17 Go [2026-07-17 06:33:28] step 197200/1000000 | loss=2.3988 vocab=0.6715 token=1.1552 | 6.12 steps/s | VRAM=32.11 Go [2026-07-17 06:33:36] step 197250/1000000 | loss=2.2407 vocab=1.0699 token=1.2902 | 6.09 steps/s | VRAM=27.96 Go [2026-07-17 06:33:44] step 197300/1000000 | loss=2.3997 vocab=1.0026 token=1.5808 | 6.10 steps/s | VRAM=25.95 Go [2026-07-17 06:33:53] step 197350/1000000 | loss=2.4541 vocab=1.2184 token=1.7584 | 6.11 steps/s | VRAM=32.05 Go [2026-07-17 06:34:01] step 197400/1000000 | loss=2.5337 vocab=1.3133 token=1.5079 | 6.11 steps/s | VRAM=32.16 Go [2026-07-17 06:34:09] step 197450/1000000 | loss=2.4676 vocab=1.1130 token=1.7557 | 6.09 steps/s | VRAM=25.89 Go [2026-07-17 06:34:17] step 197500/1000000 | loss=2.5102 vocab=1.1784 token=1.4318 | 6.14 steps/s | VRAM=29.46 Go [2026-07-17 06:34:25] step 197550/1000000 | loss=2.4232 vocab=1.0499 token=1.5097 | 6.12 steps/s | VRAM=26.14 Go [2026-07-17 06:34:33] step 197600/1000000 | loss=2.3604 vocab=0.8086 token=1.4768 | 6.11 steps/s | VRAM=29.04 Go [2026-07-17 06:34:42] step 197650/1000000 | loss=2.3880 vocab=0.7930 token=1.0163 | 6.11 steps/s | VRAM=29.39 Go [2026-07-17 06:34:50] step 197700/1000000 | loss=2.3758 vocab=1.1595 token=1.6123 | 6.11 steps/s | VRAM=28.27 Go [2026-07-17 06:34:58] step 197750/1000000 | loss=2.4919 vocab=1.1165 token=1.5497 | 6.09 steps/s | VRAM=28.45 Go [2026-07-17 06:35:06] step 197800/1000000 | loss=2.3090 vocab=1.1863 token=1.7738 | 6.11 steps/s | VRAM=32.60 Go [2026-07-17 06:35:14] step 197850/1000000 | loss=2.3189 vocab=0.7895 token=1.4163 | 6.08 steps/s | VRAM=23.45 Go [2026-07-17 06:35:23] step 197900/1000000 | loss=2.3632 vocab=1.5017 token=1.7738 | 6.11 steps/s | VRAM=32.06 Go [2026-07-17 06:35:31] step 197950/1000000 | loss=2.4036 vocab=0.9444 token=1.4917 | 6.10 steps/s | VRAM=28.32 Go [2026-07-17 06:35:39] step 198000/1000000 | loss=2.3434 vocab=0.6452 token=1.0207 | 6.12 steps/s | VRAM=32.50 Go [2026-07-17 06:35:47] step 198050/1000000 | loss=2.4468 vocab=1.2133 token=1.3987 | 6.07 steps/s | VRAM=28.97 Go [2026-07-17 06:35:55] step 198100/1000000 | loss=2.3535 vocab=1.0865 token=1.2762 | 6.08 steps/s | VRAM=28.93 Go [2026-07-17 06:36:04] step 198150/1000000 | loss=2.3965 vocab=0.8779 token=1.5337 | 6.10 steps/s | VRAM=28.79 Go [2026-07-17 06:36:12] step 198200/1000000 | loss=2.4618 vocab=1.2308 token=1.8248 | 6.09 steps/s | VRAM=30.10 Go [2026-07-17 06:36:20] step 198250/1000000 | loss=2.3635 vocab=1.2272 token=1.4655 | 6.11 steps/s | VRAM=32.06 Go [2026-07-17 06:36:28] step 198300/1000000 | loss=2.4522 vocab=0.7511 token=1.2286 | 6.07 steps/s | VRAM=27.53 Go [2026-07-17 06:36:36] step 198350/1000000 | loss=2.4231 vocab=0.9188 token=1.1211 | 6.11 steps/s | VRAM=28.52 Go [2026-07-17 06:36:45] step 198400/1000000 | loss=2.3871 vocab=1.0160 token=1.3924 | 6.13 steps/s | VRAM=31.38 Go [2026-07-17 06:36:53] step 198450/1000000 | loss=2.4652 vocab=1.1144 token=1.1949 | 6.08 steps/s | VRAM=32.30 Go [2026-07-17 06:37:01] step 198500/1000000 | loss=2.4315 vocab=0.8949 token=1.6466 | 6.12 steps/s | VRAM=25.90 Go [2026-07-17 06:37:09] step 198550/1000000 | loss=2.3715 vocab=0.6566 token=1.3912 | 6.10 steps/s | VRAM=32.41 Go [2026-07-17 06:37:17] step 198600/1000000 | loss=2.4773 vocab=1.2143 token=1.5641 | 6.11 steps/s | VRAM=32.79 Go [2026-07-17 06:37:26] step 198650/1000000 | loss=2.3698 vocab=1.2280 token=1.5914 | 6.05 steps/s | VRAM=29.74 Go [2026-07-17 06:37:34] step 198700/1000000 | loss=2.4699 vocab=0.5714 token=1.1386 | 6.09 steps/s | VRAM=32.95 Go [2026-07-17 06:37:42] step 198750/1000000 | loss=2.4553 vocab=0.9810 token=1.3532 | 6.08 steps/s | VRAM=28.80 Go [2026-07-17 06:37:50] step 198800/1000000 | loss=2.3510 vocab=0.9888 token=1.3897 | 6.10 steps/s | VRAM=29.21 Go [2026-07-17 06:37:58] step 198850/1000000 | loss=2.4376 vocab=0.8321 token=1.7823 | 6.11 steps/s | VRAM=29.19 Go [2026-07-17 06:38:07] step 198900/1000000 | loss=2.3225 vocab=1.0303 token=1.4796 | 6.07 steps/s | VRAM=26.61 Go [2026-07-17 06:38:15] step 198950/1000000 | loss=2.4331 vocab=0.6280 token=1.2121 | 6.14 steps/s | VRAM=31.36 Go [2026-07-17 06:38:23] step 199000/1000000 | loss=2.3578 vocab=1.1372 token=1.7800 | 6.09 steps/s | VRAM=25.27 Go [2026-07-17 06:38:31] step 199050/1000000 | loss=2.3950 vocab=1.2193 token=1.6197 | 6.13 steps/s | VRAM=29.73 Go [2026-07-17 06:38:39] step 199100/1000000 | loss=2.4204 vocab=0.8549 token=1.4872 | 6.09 steps/s | VRAM=28.60 Go [2026-07-17 06:38:48] step 199150/1000000 | loss=2.4084 vocab=0.6968 token=1.3456 | 6.07 steps/s | VRAM=28.21 Go [2026-07-17 06:38:56] step 199200/1000000 | loss=2.3691 vocab=0.8789 token=0.8553 | 6.09 steps/s | VRAM=28.64 Go [2026-07-17 06:39:04] step 199250/1000000 | loss=2.4815 vocab=0.6888 token=1.4451 | 6.12 steps/s | VRAM=32.33 Go [2026-07-17 06:39:12] step 199300/1000000 | loss=2.4464 vocab=0.9393 token=1.4028 | 6.09 steps/s | VRAM=29.58 Go [2026-07-17 06:39:20] step 199350/1000000 | loss=2.4411 vocab=0.7156 token=1.2481 | 6.13 steps/s | VRAM=31.37 Go [2026-07-17 06:39:29] step 199400/1000000 | loss=2.3647 vocab=1.2867 token=1.8593 | 6.13 steps/s | VRAM=32.25 Go [2026-07-17 06:39:37] step 199450/1000000 | loss=2.3339 vocab=0.9490 token=1.4990 | 6.08 steps/s | VRAM=26.69 Go [2026-07-17 06:39:45] step 199500/1000000 | loss=2.4308 vocab=1.0185 token=1.3055 | 6.11 steps/s | VRAM=28.14 Go [2026-07-17 06:39:53] step 199550/1000000 | loss=2.4799 vocab=1.0403 token=1.6004 | 6.10 steps/s | VRAM=26.82 Go [2026-07-17 06:40:01] step 199600/1000000 | loss=2.4879 vocab=0.9345 token=1.7577 | 6.14 steps/s | VRAM=23.40 Go [2026-07-17 06:40:09] step 199650/1000000 | loss=2.3970 vocab=1.0508 token=1.4184 | 6.14 steps/s | VRAM=31.56 Go [2026-07-17 06:40:18] step 199700/1000000 | loss=2.4630 vocab=1.4993 token=1.9405 | 6.11 steps/s | VRAM=27.59 Go [2026-07-17 06:40:26] step 199750/1000000 | loss=2.4658 vocab=0.7906 token=1.5025 | 6.13 steps/s | VRAM=27.58 Go [2026-07-17 06:40:34] step 199800/1000000 | loss=2.4485 vocab=1.0403 token=1.3448 | 6.09 steps/s | VRAM=29.32 Go [2026-07-17 06:40:42] step 199850/1000000 | loss=2.3444 vocab=0.8561 token=1.4964 | 6.14 steps/s | VRAM=31.41 Go [2026-07-17 06:40:50] step 199900/1000000 | loss=2.2935 vocab=1.5001 token=1.5574 | 6.08 steps/s | VRAM=22.74 Go [2026-07-17 06:40:59] step 199950/1000000 | loss=2.3914 vocab=0.9212 token=1.8965 | 6.11 steps/s | VRAM=22.68 Go [2026-07-17 06:41:07] step 200000/1000000 | loss=2.3843 vocab=1.0304 token=1.7646 | 6.09 steps/s | VRAM=26.77 Go [2026-07-17 06:42:32] step 200050/1000000 | loss=2.5178 vocab=0.8235 token=1.4290 | 0.59 steps/s | VRAM=32.20 Go [2026-07-17 06:42:40] step 200100/1000000 | loss=2.4058 vocab=1.2173 token=1.6716 | 6.18 steps/s | VRAM=28.86 Go [2026-07-17 06:42:48] step 200150/1000000 | loss=2.4460 vocab=0.7994 token=1.3261 | 6.17 steps/s | VRAM=28.36 Go [2026-07-17 06:42:56] step 200200/1000000 | loss=2.3614 vocab=0.9645 token=1.4851 | 6.14 steps/s | VRAM=26.06 Go [2026-07-17 06:43:04] step 200250/1000000 | loss=2.3626 vocab=0.8286 token=1.2641 | 6.16 steps/s | VRAM=25.50 Go [2026-07-17 06:43:12] step 200300/1000000 | loss=2.4944 vocab=1.2449 token=1.2796 | 6.16 steps/s | VRAM=31.96 Go [2026-07-17 06:43:20] step 200350/1000000 | loss=2.2535 vocab=0.7397 token=1.3098 | 6.13 steps/s | VRAM=29.40 Go [2026-07-17 06:43:28] step 200400/1000000 | loss=2.5388 vocab=0.7466 token=1.4303 | 6.14 steps/s | VRAM=32.17 Go [2026-07-17 06:43:37] step 200450/1000000 | loss=2.3913 vocab=1.0183 token=1.4983 | 6.10 steps/s | VRAM=30.00 Go [2026-07-17 06:43:45] step 200500/1000000 | loss=2.3726 vocab=1.1219 token=1.3271 | 6.11 steps/s | VRAM=31.47 Go [2026-07-17 06:43:53] step 200550/1000000 | loss=2.3080 vocab=0.5263 token=1.3412 | 6.14 steps/s | VRAM=32.08 Go [2026-07-17 06:44:01] step 200600/1000000 | loss=2.3286 vocab=1.1092 token=1.3071 | 6.09 steps/s | VRAM=29.50 Go [2026-07-17 06:44:09] step 200650/1000000 | loss=2.4954 vocab=0.5418 token=1.1868 | 6.10 steps/s | VRAM=32.47 Go [2026-07-17 06:44:18] step 200700/1000000 | loss=2.4807 vocab=1.1577 token=1.2029 | 6.09 steps/s | VRAM=27.95 Go [2026-07-17 06:44:26] step 200750/1000000 | loss=2.5113 vocab=0.6344 token=1.0081 | 6.11 steps/s | VRAM=28.15 Go [2026-07-17 06:44:34] step 200800/1000000 | loss=2.3913 vocab=1.0900 token=1.8234 | 6.09 steps/s | VRAM=32.04 Go [2026-07-17 06:44:42] step 200850/1000000 | loss=2.3555 vocab=0.6300 token=0.8583 | 6.09 steps/s | VRAM=29.04 Go [2026-07-17 06:44:50] step 200900/1000000 | loss=2.4150 vocab=1.0057 token=1.2784 | 6.12 steps/s | VRAM=24.48 Go [2026-07-17 06:44:59] step 200950/1000000 | loss=2.3467 vocab=0.8446 token=1.1847 | 6.09 steps/s | VRAM=28.34 Go [2026-07-17 06:45:07] step 201000/1000000 | loss=2.4707 vocab=1.0784 token=1.3801 | 6.10 steps/s | VRAM=28.25 Go [2026-07-17 06:45:15] step 201050/1000000 | loss=2.5078 vocab=1.4164 token=1.7111 | 6.08 steps/s | VRAM=32.38 Go [2026-07-17 06:45:23] step 201100/1000000 | loss=2.4569 vocab=0.6897 token=1.4002 | 6.09 steps/s | VRAM=29.08 Go [2026-07-17 06:45:31] step 201150/1000000 | loss=2.3725 vocab=0.9502 token=1.4437 | 6.11 steps/s | VRAM=30.48 Go [2026-07-17 06:45:40] step 201200/1000000 | loss=2.4692 vocab=1.0280 token=1.5786 | 6.04 steps/s | VRAM=27.80 Go [2026-07-17 06:45:48] step 201250/1000000 | loss=2.4704 vocab=0.6588 token=1.5252 | 6.11 steps/s | VRAM=28.64 Go [2026-07-17 06:45:56] step 201300/1000000 | loss=2.4070 vocab=1.5349 token=1.3786 | 6.13 steps/s | VRAM=31.51 Go [2026-07-17 06:46:04] step 201350/1000000 | loss=2.3068 vocab=0.8680 token=1.3523 | 6.06 steps/s | VRAM=26.81 Go [2026-07-17 06:46:12] step 201400/1000000 | loss=2.4126 vocab=0.7355 token=1.1938 | 6.13 steps/s | VRAM=31.73 Go [2026-07-17 06:46:21] step 201450/1000000 | loss=2.3494 vocab=0.8936 token=1.6801 | 6.12 steps/s | VRAM=33.01 Go [2026-07-17 06:46:29] step 201500/1000000 | loss=2.4610 vocab=1.1609 token=1.8980 | 6.12 steps/s | VRAM=28.58 Go [2026-07-17 06:46:37] step 201550/1000000 | loss=2.3630 vocab=1.0095 token=1.4020 | 6.13 steps/s | VRAM=28.87 Go [2026-07-17 06:46:45] step 201600/1000000 | loss=2.3189 vocab=0.6767 token=0.8358 | 6.08 steps/s | VRAM=29.03 Go [2026-07-17 06:46:53] step 201650/1000000 | loss=2.4489 vocab=0.8869 token=1.4553 | 6.13 steps/s | VRAM=32.38 Go [2026-07-17 06:47:01] step 201700/1000000 | loss=2.3350 vocab=0.6870 token=1.4076 | 6.11 steps/s | VRAM=27.72 Go [2026-07-17 06:47:10] step 201750/1000000 | loss=2.5166 vocab=1.1202 token=1.7049 | 6.10 steps/s | VRAM=30.17 Go [2026-07-17 06:47:18] step 201800/1000000 | loss=2.2581 vocab=0.5777 token=1.5883 | 6.11 steps/s | VRAM=25.02 Go [2026-07-17 06:47:26] step 201850/1000000 | loss=2.3622 vocab=0.8517 token=0.8453 | 6.12 steps/s | VRAM=31.65 Go [2026-07-17 06:47:34] step 201900/1000000 | loss=2.3696 vocab=0.7935 token=1.5661 | 6.10 steps/s | VRAM=28.21 Go [2026-07-17 06:47:42] step 201950/1000000 | loss=2.4780 vocab=0.6329 token=1.1259 | 6.15 steps/s | VRAM=25.95 Go [2026-07-17 06:47:51] step 202000/1000000 | loss=2.3365 vocab=0.4267 token=0.7257 | 6.12 steps/s | VRAM=28.53 Go [2026-07-17 06:47:59] step 202050/1000000 | loss=2.3006 vocab=0.7177 token=1.6435 | 6.11 steps/s | VRAM=30.47 Go [2026-07-17 06:48:07] step 202100/1000000 | loss=2.3394 vocab=0.9315 token=1.4104 | 6.11 steps/s | VRAM=27.66 Go [2026-07-17 06:48:15] step 202150/1000000 | loss=2.3196 vocab=0.9231 token=1.3186 | 6.11 steps/s | VRAM=28.20 Go [2026-07-17 06:48:23] step 202200/1000000 | loss=2.3888 vocab=1.0260 token=1.5109 | 6.10 steps/s | VRAM=26.68 Go [2026-07-17 06:48:31] step 202250/1000000 | loss=2.3220 vocab=0.9270 token=1.3653 | 6.13 steps/s | VRAM=28.38 Go [2026-07-17 06:48:40] step 202300/1000000 | loss=2.5158 vocab=0.9941 token=1.1897 | 6.14 steps/s | VRAM=31.09 Go [2026-07-17 06:48:48] step 202350/1000000 | loss=2.4346 vocab=0.5486 token=1.2475 | 6.11 steps/s | VRAM=28.73 Go [2026-07-17 06:48:56] step 202400/1000000 | loss=2.3795 vocab=1.0504 token=1.2181 | 6.13 steps/s | VRAM=31.52 Go [2026-07-17 06:49:04] step 202450/1000000 | loss=2.3580 vocab=0.9008 token=0.9517 | 6.08 steps/s | VRAM=27.12 Go [2026-07-17 06:49:12] step 202500/1000000 | loss=2.4422 vocab=0.8908 token=1.7039 | 6.10 steps/s | VRAM=29.56 Go [2026-07-17 06:49:21] step 202550/1000000 | loss=2.3471 vocab=0.6456 token=1.0530 | 6.11 steps/s | VRAM=32.21 Go [2026-07-17 06:49:29] step 202600/1000000 | loss=2.3019 vocab=0.7605 token=1.2163 | 6.11 steps/s | VRAM=32.11 Go [2026-07-17 06:49:37] step 202650/1000000 | loss=2.3504 vocab=1.3085 token=1.3937 | 6.11 steps/s | VRAM=30.29 Go [2026-07-17 06:49:45] step 202700/1000000 | loss=2.4589 vocab=1.1738 token=1.6955 | 6.13 steps/s | VRAM=31.40 Go [2026-07-17 06:49:53] step 202750/1000000 | loss=2.4516 vocab=0.9183 token=1.2116 | 6.11 steps/s | VRAM=32.63 Go [2026-07-17 06:50:01] step 202800/1000000 | loss=2.2746 vocab=0.9046 token=1.5888 | 6.11 steps/s | VRAM=26.21 Go [2026-07-17 06:50:10] step 202850/1000000 | loss=2.3511 vocab=0.9351 token=1.4324 | 6.09 steps/s | VRAM=29.32 Go [2026-07-17 06:50:18] step 202900/1000000 | loss=2.3780 vocab=0.8925 token=1.4832 | 6.11 steps/s | VRAM=24.95 Go [2026-07-17 06:50:26] step 202950/1000000 | loss=2.3199 vocab=0.9492 token=1.6091 | 6.10 steps/s | VRAM=29.19 Go [2026-07-17 06:50:34] step 203000/1000000 | loss=2.3223 vocab=1.1256 token=1.6546 | 6.13 steps/s | VRAM=31.25 Go [2026-07-17 06:50:42] step 203050/1000000 | loss=2.4403 vocab=1.2306 token=1.4500 | 6.13 steps/s | VRAM=31.36 Go [2026-07-17 06:50:50] step 203100/1000000 | loss=2.3628 vocab=1.2188 token=1.9457 | 6.11 steps/s | VRAM=28.66 Go [2026-07-17 06:50:59] step 203150/1000000 | loss=2.3805 vocab=0.4772 token=1.2053 | 6.12 steps/s | VRAM=32.41 Go [2026-07-17 06:51:07] step 203200/1000000 | loss=2.4373 vocab=1.0763 token=1.5755 | 6.09 steps/s | VRAM=32.71 Go [2026-07-17 06:51:15] step 203250/1000000 | loss=2.2737 vocab=1.2742 token=1.3403 | 6.11 steps/s | VRAM=27.87 Go [2026-07-17 06:51:23] step 203300/1000000 | loss=2.4542 vocab=0.9079 token=1.4548 | 6.09 steps/s | VRAM=26.60 Go [2026-07-17 06:51:31] step 203350/1000000 | loss=2.5003 vocab=1.2867 token=1.5917 | 6.09 steps/s | VRAM=25.85 Go [2026-07-17 06:51:40] step 203400/1000000 | loss=2.3927 vocab=1.1792 token=1.5178 | 6.14 steps/s | VRAM=24.98 Go [2026-07-17 06:51:48] step 203450/1000000 | loss=2.4171 vocab=0.6199 token=1.2117 | 6.13 steps/s | VRAM=28.19 Go [2026-07-17 06:51:56] step 203500/1000000 | loss=2.4025 vocab=0.8956 token=1.6934 | 6.13 steps/s | VRAM=27.98 Go [2026-07-17 06:52:04] step 203550/1000000 | loss=2.4375 vocab=0.8615 token=1.7585 | 6.11 steps/s | VRAM=25.52 Go [2026-07-17 06:52:12] step 203600/1000000 | loss=2.4142 vocab=0.7697 token=1.4373 | 6.10 steps/s | VRAM=29.33 Go [2026-07-17 06:52:21] step 203650/1000000 | loss=2.4517 vocab=1.3235 token=1.2041 | 6.11 steps/s | VRAM=32.74 Go [2026-07-17 06:52:29] step 203700/1000000 | loss=2.4437 vocab=0.5235 token=1.3304 | 6.10 steps/s | VRAM=28.20 Go [2026-07-17 06:52:37] step 203750/1000000 | loss=2.4019 vocab=1.1718 token=1.5866 | 6.09 steps/s | VRAM=29.65 Go [2026-07-17 06:52:45] step 203800/1000000 | loss=2.3406 vocab=0.6952 token=1.2851 | 6.12 steps/s | VRAM=31.16 Go [2026-07-17 06:52:53] step 203850/1000000 | loss=2.3386 vocab=0.7387 token=1.1888 | 6.08 steps/s | VRAM=26.48 Go [2026-07-17 06:53:01] step 203900/1000000 | loss=2.3720 vocab=0.6297 token=1.6913 | 6.11 steps/s | VRAM=32.38 Go [2026-07-17 06:53:10] step 203950/1000000 | loss=2.3699 vocab=0.9528 token=1.5749 | 6.09 steps/s | VRAM=27.36 Go [2026-07-17 06:53:18] step 204000/1000000 | loss=2.3203 vocab=1.0094 token=1.7111 | 6.08 steps/s | VRAM=26.45 Go [2026-07-17 06:53:26] step 204050/1000000 | loss=2.3692 vocab=0.5486 token=1.1974 | 6.09 steps/s | VRAM=29.24 Go [2026-07-17 06:53:34] step 204100/1000000 | loss=2.2981 vocab=0.8724 token=1.6645 | 6.07 steps/s | VRAM=28.49 Go [2026-07-17 06:53:43] step 204150/1000000 | loss=2.3361 vocab=0.7188 token=1.2450 | 6.10 steps/s | VRAM=28.94 Go [2026-07-17 06:53:51] step 204200/1000000 | loss=2.3257 vocab=0.6704 token=1.2882 | 6.11 steps/s | VRAM=29.07 Go [2026-07-17 06:53:59] step 204250/1000000 | loss=2.3492 vocab=0.7577 token=1.5766 | 6.07 steps/s | VRAM=28.35 Go [2026-07-17 06:54:07] step 204300/1000000 | loss=2.4280 vocab=0.9415 token=1.3217 | 6.13 steps/s | VRAM=32.51 Go [2026-07-17 06:54:15] step 204350/1000000 | loss=2.4228 vocab=0.7187 token=0.8740 | 6.08 steps/s | VRAM=25.96 Go [2026-07-17 06:54:24] step 204400/1000000 | loss=2.3938 vocab=0.8331 token=1.2920 | 6.13 steps/s | VRAM=32.57 Go [2026-07-17 06:54:32] step 204450/1000000 | loss=2.3412 vocab=0.9581 token=1.6516 | 6.11 steps/s | VRAM=26.06 Go [2026-07-17 06:54:40] step 204500/1000000 | loss=2.4383 vocab=0.5628 token=1.1744 | 6.11 steps/s | VRAM=31.96 Go [2026-07-17 06:54:48] step 204550/1000000 | loss=2.3655 vocab=0.5702 token=1.1971 | 6.10 steps/s | VRAM=30.90 Go [2026-07-17 06:54:56] step 204600/1000000 | loss=2.3348 vocab=1.5411 token=1.4186 | 6.08 steps/s | VRAM=27.37 Go [2026-07-17 06:55:04] step 204650/1000000 | loss=2.2762 vocab=1.1183 token=1.4313 | 6.14 steps/s | VRAM=31.40 Go [2026-07-17 06:55:13] step 204700/1000000 | loss=2.3762 vocab=0.7410 token=1.5939 | 6.11 steps/s | VRAM=31.62 Go [2026-07-17 06:55:21] step 204750/1000000 | loss=2.4171 vocab=0.7595 token=1.2451 | 6.13 steps/s | VRAM=25.00 Go [2026-07-17 06:55:29] step 204800/1000000 | loss=2.3393 vocab=1.1840 token=1.6173 | 6.16 steps/s | VRAM=24.62 Go [2026-07-17 06:55:37] step 204850/1000000 | loss=2.3621 vocab=0.8959 token=1.4545 | 6.10 steps/s | VRAM=30.07 Go [2026-07-17 06:55:45] step 204900/1000000 | loss=2.5008 vocab=0.9613 token=1.5255 | 6.11 steps/s | VRAM=28.24 Go [2026-07-17 06:55:54] step 204950/1000000 | loss=2.3177 vocab=0.7202 token=1.5734 | 6.09 steps/s | VRAM=28.36 Go [2026-07-17 06:56:02] step 205000/1000000 | loss=2.3051 vocab=1.0872 token=1.2600 | 6.10 steps/s | VRAM=31.71 Go [2026-07-17 06:57:43] step 205050/1000000 | loss=2.3967 vocab=0.7007 token=1.3515 | 0.49 steps/s | VRAM=28.17 Go [2026-07-17 06:57:52] step 205100/1000000 | loss=2.3248 vocab=0.6442 token=1.2033 | 6.19 steps/s | VRAM=29.05 Go [2026-07-17 06:58:00] step 205150/1000000 | loss=2.4944 vocab=1.5979 token=1.8066 | 6.19 steps/s | VRAM=24.95 Go [2026-07-17 06:58:08] step 205200/1000000 | loss=2.4890 vocab=1.1183 token=1.5812 | 6.17 steps/s | VRAM=29.09 Go [2026-07-17 06:58:16] step 205250/1000000 | loss=2.3484 vocab=1.1258 token=1.4981 | 6.12 steps/s | VRAM=26.69 Go [2026-07-17 06:58:24] step 205300/1000000 | loss=2.4118 vocab=1.0624 token=1.5847 | 6.17 steps/s | VRAM=24.17 Go [2026-07-17 06:58:32] step 205350/1000000 | loss=2.4795 vocab=1.0010 token=1.1683 | 6.14 steps/s | VRAM=26.69 Go [2026-07-17 06:58:40] step 205400/1000000 | loss=2.5202 vocab=1.1384 token=1.4415 | 6.15 steps/s | VRAM=26.25 Go [2026-07-17 06:58:48] step 205450/1000000 | loss=2.3518 vocab=1.0124 token=1.1529 | 6.12 steps/s | VRAM=26.24 Go [2026-07-17 06:58:57] step 205500/1000000 | loss=2.5042 vocab=1.7474 token=1.3773 | 6.13 steps/s | VRAM=28.17 Go [2026-07-17 06:59:05] step 205550/1000000 | loss=2.3482 vocab=0.8162 token=1.1722 | 6.09 steps/s | VRAM=28.27 Go [2026-07-17 06:59:13] step 205600/1000000 | loss=2.3681 vocab=1.0674 token=1.7214 | 6.12 steps/s | VRAM=28.46 Go [2026-07-17 06:59:21] step 205650/1000000 | loss=2.4332 vocab=0.9788 token=0.8585 | 6.13 steps/s | VRAM=29.30 Go [2026-07-17 06:59:29] step 205700/1000000 | loss=2.4510 vocab=0.9438 token=1.1464 | 6.10 steps/s | VRAM=29.49 Go [2026-07-17 06:59:38] step 205750/1000000 | loss=2.4152 vocab=1.5744 token=1.2242 | 6.11 steps/s | VRAM=29.49 Go [2026-07-17 06:59:46] step 205800/1000000 | loss=2.4076 vocab=0.8833 token=1.4557 | 6.12 steps/s | VRAM=29.14 Go [2026-07-17 06:59:54] step 205850/1000000 | loss=2.4236 vocab=1.2981 token=1.1335 | 6.13 steps/s | VRAM=31.26 Go [2026-07-17 07:00:02] step 205900/1000000 | loss=2.2417 vocab=0.8735 token=1.0622 | 6.09 steps/s | VRAM=28.93 Go [2026-07-17 07:00:10] step 205950/1000000 | loss=2.3472 vocab=0.7608 token=1.7187 | 6.10 steps/s | VRAM=28.46 Go [2026-07-17 07:00:18] step 206000/1000000 | loss=2.2744 vocab=0.7520 token=1.1321 | 6.11 steps/s | VRAM=24.90 Go [2026-07-17 07:00:27] step 206050/1000000 | loss=2.4279 vocab=1.3241 token=1.4535 | 6.09 steps/s | VRAM=26.08 Go [2026-07-17 07:00:35] step 206100/1000000 | loss=2.4644 vocab=1.0493 token=1.3849 | 6.10 steps/s | VRAM=29.06 Go [2026-07-17 07:00:43] step 206150/1000000 | loss=2.3485 vocab=1.0087 token=1.5660 | 6.10 steps/s | VRAM=28.61 Go [2026-07-17 07:00:51] step 206200/1000000 | loss=2.4405 vocab=0.8155 token=1.2481 | 6.09 steps/s | VRAM=28.87 Go [2026-07-17 07:00:59] step 206250/1000000 | loss=2.5002 vocab=1.0186 token=1.3219 | 6.11 steps/s | VRAM=24.85 Go [2026-07-17 07:01:08] step 206300/1000000 | loss=2.3064 vocab=0.8839 token=1.5198 | 6.12 steps/s | VRAM=25.32 Go [2026-07-17 07:01:16] step 206350/1000000 | loss=2.3629 vocab=0.9516 token=1.7053 | 6.10 steps/s | VRAM=29.59 Go [2026-07-17 07:01:24] step 206400/1000000 | loss=2.3486 vocab=0.6983 token=1.4969 | 6.13 steps/s | VRAM=24.89 Go [2026-07-17 07:01:32] step 206450/1000000 | loss=2.3103 vocab=1.2816 token=1.6560 | 6.12 steps/s | VRAM=32.62 Go [2026-07-17 07:01:40] step 206500/1000000 | loss=2.3538 vocab=0.7456 token=1.4155 | 6.08 steps/s | VRAM=24.79 Go [2026-07-17 07:01:49] step 206550/1000000 | loss=2.3240 vocab=1.4473 token=1.7332 | 6.12 steps/s | VRAM=29.03 Go [2026-07-17 07:01:57] step 206600/1000000 | loss=2.2934 vocab=0.6790 token=1.5614 | 6.10 steps/s | VRAM=28.67 Go [2026-07-17 07:02:05] step 206650/1000000 | loss=2.2764 vocab=1.1723 token=1.2726 | 6.11 steps/s | VRAM=29.13 Go [2026-07-17 07:02:13] step 206700/1000000 | loss=2.3758 vocab=1.2296 token=1.8782 | 6.15 steps/s | VRAM=31.68 Go [2026-07-17 07:02:21] step 206750/1000000 | loss=2.4016 vocab=1.6684 token=1.6172 | 6.10 steps/s | VRAM=32.19 Go [2026-07-17 07:02:29] step 206800/1000000 | loss=2.3317 vocab=0.5020 token=0.9525 | 6.09 steps/s | VRAM=28.46 Go [2026-07-17 07:02:38] step 206850/1000000 | loss=2.5164 vocab=0.6780 token=1.4834 | 6.10 steps/s | VRAM=28.92 Go [2026-07-17 07:02:46] step 206900/1000000 | loss=2.5331 vocab=1.0416 token=1.0534 | 6.10 steps/s | VRAM=28.43 Go [2026-07-17 07:02:54] step 206950/1000000 | loss=2.3589 vocab=1.1960 token=1.2037 | 6.10 steps/s | VRAM=32.38 Go [2026-07-17 07:03:02] step 207000/1000000 | loss=2.4828 vocab=1.0884 token=1.9070 | 6.09 steps/s | VRAM=28.32 Go [2026-07-17 07:03:10] step 207050/1000000 | loss=2.3466 vocab=1.4336 token=1.6823 | 6.07 steps/s | VRAM=28.29 Go [2026-07-17 07:03:19] step 207100/1000000 | loss=2.4275 vocab=1.1759 token=1.3264 | 6.12 steps/s | VRAM=25.65 Go [2026-07-17 07:03:27] step 207150/1000000 | loss=2.3576 vocab=0.7064 token=1.5434 | 6.14 steps/s | VRAM=30.88 Go [2026-07-17 07:03:35] step 207200/1000000 | loss=2.3464 vocab=0.8155 token=1.8471 | 6.10 steps/s | VRAM=28.31 Go [2026-07-17 07:03:43] step 207250/1000000 | loss=2.3178 vocab=1.2702 token=1.7285 | 6.13 steps/s | VRAM=24.95 Go [2026-07-17 07:03:51] step 207300/1000000 | loss=2.3057 vocab=0.9983 token=1.8057 | 6.11 steps/s | VRAM=28.61 Go [2026-07-17 07:04:00] step 207350/1000000 | loss=2.4259 vocab=1.1837 token=1.1265 | 6.11 steps/s | VRAM=28.67 Go [2026-07-17 07:04:08] step 207400/1000000 | loss=2.3406 vocab=0.6261 token=1.3381 | 6.12 steps/s | VRAM=30.04 Go [2026-07-17 07:04:16] step 207450/1000000 | loss=2.3540 vocab=0.8797 token=1.2946 | 6.12 steps/s | VRAM=26.23 Go [2026-07-17 07:04:24] step 207500/1000000 | loss=2.2369 vocab=0.7350 token=1.3208 | 6.12 steps/s | VRAM=31.46 Go [2026-07-17 07:04:32] step 207550/1000000 | loss=2.3282 vocab=1.0401 token=1.5065 | 6.13 steps/s | VRAM=32.38 Go [2026-07-17 07:04:40] step 207600/1000000 | loss=2.4730 vocab=0.9046 token=1.5181 | 6.08 steps/s | VRAM=26.08 Go [2026-07-17 07:04:49] step 207650/1000000 | loss=2.4489 vocab=0.8891 token=1.4193 | 6.15 steps/s | VRAM=31.73 Go [2026-07-17 07:04:57] step 207700/1000000 | loss=2.4493 vocab=0.8218 token=1.4897 | 6.10 steps/s | VRAM=29.20 Go [2026-07-17 07:05:05] step 207750/1000000 | loss=2.3992 vocab=0.8091 token=1.5276 | 6.12 steps/s | VRAM=31.62 Go [2026-07-17 07:05:13] step 207800/1000000 | loss=2.3876 vocab=1.0996 token=1.3204 | 6.11 steps/s | VRAM=26.22 Go [2026-07-17 07:05:21] step 207850/1000000 | loss=2.4287 vocab=1.1829 token=1.1872 | 6.10 steps/s | VRAM=28.11 Go [2026-07-17 07:05:29] step 207900/1000000 | loss=2.4725 vocab=1.1753 token=1.3882 | 6.12 steps/s | VRAM=28.61 Go [2026-07-17 07:05:38] step 207950/1000000 | loss=2.4381 vocab=0.6163 token=1.5601 | 6.11 steps/s | VRAM=28.25 Go [2026-07-17 07:05:46] step 208000/1000000 | loss=2.3441 vocab=1.1566 token=1.0565 | 6.12 steps/s | VRAM=32.10 Go [2026-07-17 07:05:54] step 208050/1000000 | loss=2.3928 vocab=0.9209 token=1.2884 | 6.12 steps/s | VRAM=29.15 Go [2026-07-17 07:06:02] step 208100/1000000 | loss=2.3639 vocab=0.6814 token=1.2672 | 6.11 steps/s | VRAM=26.34 Go [2026-07-17 07:06:10] step 208150/1000000 | loss=2.3494 vocab=0.9303 token=1.3411 | 6.10 steps/s | VRAM=28.71 Go [2026-07-17 07:06:19] step 208200/1000000 | loss=2.4443 vocab=1.4902 token=1.5019 | 6.08 steps/s | VRAM=26.40 Go [2026-07-17 07:06:27] step 208250/1000000 | loss=2.3595 vocab=0.9290 token=1.3729 | 6.12 steps/s | VRAM=28.14 Go [2026-07-17 07:06:35] step 208300/1000000 | loss=2.4966 vocab=1.0063 token=1.5741 | 6.08 steps/s | VRAM=29.79 Go [2026-07-17 07:06:43] step 208350/1000000 | loss=2.3483 vocab=1.1213 token=1.1645 | 6.11 steps/s | VRAM=28.74 Go [2026-07-17 07:06:51] step 208400/1000000 | loss=2.3799 vocab=0.8872 token=1.1208 | 6.12 steps/s | VRAM=31.43 Go [2026-07-17 07:07:00] step 208450/1000000 | loss=2.4192 vocab=1.0847 token=1.5111 | 6.07 steps/s | VRAM=30.24 Go [2026-07-17 07:07:08] step 208500/1000000 | loss=2.3823 vocab=1.0673 token=1.1993 | 6.10 steps/s | VRAM=25.37 Go [2026-07-17 07:07:16] step 208550/1000000 | loss=2.3836 vocab=1.1149 token=1.2915 | 6.07 steps/s | VRAM=26.53 Go [2026-07-17 07:07:24] step 208600/1000000 | loss=2.4202 vocab=0.8079 token=1.1478 | 6.10 steps/s | VRAM=29.78 Go [2026-07-17 07:07:32] step 208650/1000000 | loss=2.3828 vocab=0.5082 token=1.2842 | 6.12 steps/s | VRAM=29.38 Go [2026-07-17 07:07:41] step 208700/1000000 | loss=2.4337 vocab=1.0679 token=1.3282 | 6.10 steps/s | VRAM=32.06 Go [2026-07-17 07:07:49] step 208750/1000000 | loss=2.3454 vocab=0.5515 token=1.5681 | 6.12 steps/s | VRAM=32.75 Go [2026-07-17 07:07:57] step 208800/1000000 | loss=2.3558 vocab=1.0049 token=1.3919 | 6.13 steps/s | VRAM=24.17 Go [2026-07-17 07:08:05] step 208850/1000000 | loss=2.3568 vocab=0.7052 token=1.4190 | 6.08 steps/s | VRAM=28.30 Go [2026-07-17 07:08:13] step 208900/1000000 | loss=2.3884 vocab=1.0126 token=1.3319 | 6.11 steps/s | VRAM=28.79 Go [2026-07-17 07:08:22] step 208950/1000000 | loss=2.4139 vocab=0.7743 token=1.0530 | 6.07 steps/s | VRAM=26.22 Go [2026-07-17 07:08:30] step 209000/1000000 | loss=2.4573 vocab=0.9698 token=1.7885 | 6.11 steps/s | VRAM=24.67 Go [2026-07-17 07:08:38] step 209050/1000000 | loss=2.3543 vocab=1.1882 token=1.1598 | 6.14 steps/s | VRAM=31.80 Go [2026-07-17 07:08:46] step 209100/1000000 | loss=2.2944 vocab=0.8088 token=1.3420 | 6.13 steps/s | VRAM=29.18 Go [2026-07-17 07:08:54] step 209150/1000000 | loss=2.4153 vocab=0.9509 token=1.5975 | 6.12 steps/s | VRAM=29.30 Go [2026-07-17 07:09:02] step 209200/1000000 | loss=2.2675 vocab=0.3522 token=1.4627 | 6.10 steps/s | VRAM=26.06 Go [2026-07-17 07:09:11] step 209250/1000000 | loss=2.3312 vocab=0.5731 token=1.0290 | 6.11 steps/s | VRAM=27.07 Go [2026-07-17 07:09:19] step 209300/1000000 | loss=2.3835 vocab=0.9931 token=1.5007 | 6.11 steps/s | VRAM=27.97 Go [2026-07-17 07:09:27] step 209350/1000000 | loss=2.3654 vocab=0.5895 token=0.9840 | 6.10 steps/s | VRAM=30.17 Go [2026-07-17 07:09:35] step 209400/1000000 | loss=2.4297 vocab=0.6174 token=1.4443 | 6.14 steps/s | VRAM=24.24 Go [2026-07-17 07:09:43] step 209450/1000000 | loss=2.3571 vocab=1.1452 token=1.4364 | 6.14 steps/s | VRAM=27.97 Go [2026-07-17 07:09:51] step 209500/1000000 | loss=2.4662 vocab=1.0007 token=1.1143 | 6.15 steps/s | VRAM=31.43 Go [2026-07-17 07:10:00] step 209550/1000000 | loss=2.4458 vocab=1.1095 token=1.4329 | 6.12 steps/s | VRAM=29.27 Go [2026-07-17 07:10:08] step 209600/1000000 | loss=2.4277 vocab=1.5000 token=1.6425 | 6.04 steps/s | VRAM=30.19 Go [2026-07-17 07:10:16] step 209650/1000000 | loss=2.3995 vocab=1.0648 token=1.8182 | 6.13 steps/s | VRAM=25.61 Go [2026-07-17 07:10:24] step 209700/1000000 | loss=2.3276 vocab=1.0401 token=1.2421 | 6.15 steps/s | VRAM=32.29 Go [2026-07-17 07:10:32] step 209750/1000000 | loss=2.4125 vocab=1.1907 token=1.3494 | 6.09 steps/s | VRAM=31.51 Go [2026-07-17 07:10:40] step 209800/1000000 | loss=2.4000 vocab=0.5318 token=1.1627 | 6.14 steps/s | VRAM=31.65 Go [2026-07-17 07:10:49] step 209850/1000000 | loss=2.3560 vocab=0.9655 token=1.4770 | 6.08 steps/s | VRAM=28.38 Go [2026-07-17 07:10:57] step 209900/1000000 | loss=2.3292 vocab=0.8899 token=1.3041 | 6.12 steps/s | VRAM=27.82 Go [2026-07-17 07:11:05] step 209950/1000000 | loss=2.4117 vocab=1.1050 token=1.5482 | 6.14 steps/s | VRAM=31.58 Go [2026-07-17 07:11:13] step 210000/1000000 | loss=2.4115 vocab=1.1771 token=1.7955 | 6.11 steps/s | VRAM=28.88 Go [2026-07-17 07:12:35] step 210050/1000000 | loss=2.3096 vocab=0.8904 token=1.5548 | 0.61 steps/s | VRAM=28.45 Go [2026-07-17 07:12:43] step 210100/1000000 | loss=2.4085 vocab=1.0437 token=1.2493 | 6.17 steps/s | VRAM=27.91 Go [2026-07-17 07:12:51] step 210150/1000000 | loss=2.4403 vocab=1.0653 token=1.5812 | 6.19 steps/s | VRAM=31.02 Go [2026-07-17 07:12:59] step 210200/1000000 | loss=2.4042 vocab=0.9749 token=1.2197 | 6.17 steps/s | VRAM=28.72 Go [2026-07-17 07:13:07] step 210250/1000000 | loss=2.4132 vocab=0.8806 token=1.3641 | 6.15 steps/s | VRAM=26.77 Go [2026-07-17 07:13:16] step 210300/1000000 | loss=2.2727 vocab=1.1145 token=1.5903 | 6.16 steps/s | VRAM=28.87 Go [2026-07-17 07:13:24] step 210350/1000000 | loss=2.3401 vocab=0.8182 token=1.4116 | 6.13 steps/s | VRAM=31.66 Go [2026-07-17 07:13:32] step 210400/1000000 | loss=2.3982 vocab=1.2323 token=1.9653 | 6.14 steps/s | VRAM=26.04 Go [2026-07-17 07:13:40] step 210450/1000000 | loss=2.3014 vocab=0.6505 token=1.3401 | 6.14 steps/s | VRAM=32.73 Go [2026-07-17 07:13:48] step 210500/1000000 | loss=2.4536 vocab=0.7144 token=1.5391 | 6.12 steps/s | VRAM=29.53 Go [2026-07-17 07:13:56] step 210550/1000000 | loss=2.4214 vocab=1.3344 token=1.3187 | 6.10 steps/s | VRAM=32.73 Go [2026-07-17 07:14:05] step 210600/1000000 | loss=2.4417 vocab=0.9535 token=1.6719 | 6.09 steps/s | VRAM=31.83 Go [2026-07-17 07:14:13] step 210650/1000000 | loss=2.4526 vocab=1.0796 token=1.4985 | 6.12 steps/s | VRAM=31.88 Go [2026-07-17 07:14:21] step 210700/1000000 | loss=2.4282 vocab=1.0752 token=1.9126 | 6.07 steps/s | VRAM=29.65 Go [2026-07-17 07:14:29] step 210750/1000000 | loss=2.3312 vocab=0.8845 token=1.3455 | 6.08 steps/s | VRAM=28.57 Go [2026-07-17 07:14:37] step 210800/1000000 | loss=2.3970 vocab=1.0458 token=1.7310 | 6.10 steps/s | VRAM=32.73 Go [2026-07-17 07:14:46] step 210850/1000000 | loss=2.4632 vocab=0.2555 token=1.1338 | 6.09 steps/s | VRAM=25.87 Go [2026-07-17 07:14:55] step 210900/1000000 | loss=2.4190 vocab=1.1711 token=1.1170 | 5.27 steps/s | VRAM=25.21 Go [2026-07-17 07:15:03] step 210950/1000000 | loss=2.3848 vocab=0.8623 token=1.1322 | 6.11 steps/s | VRAM=26.93 Go [2026-07-17 07:15:11] step 211000/1000000 | loss=2.4665 vocab=0.6774 token=1.1602 | 6.08 steps/s | VRAM=32.69 Go [2026-07-17 07:15:20] step 211050/1000000 | loss=2.3620 vocab=0.6863 token=0.8627 | 6.09 steps/s | VRAM=31.92 Go [2026-07-17 07:15:28] step 211100/1000000 | loss=2.3355 vocab=1.3436 token=1.4063 | 6.07 steps/s | VRAM=28.30 Go [2026-07-17 07:15:36] step 211150/1000000 | loss=2.4317 vocab=1.2436 token=1.5075 | 6.10 steps/s | VRAM=25.47 Go [2026-07-17 07:15:44] step 211200/1000000 | loss=2.3358 vocab=0.9326 token=1.3152 | 6.11 steps/s | VRAM=27.81 Go [2026-07-17 07:15:53] step 211250/1000000 | loss=2.3114 vocab=0.7352 token=1.3564 | 6.09 steps/s | VRAM=32.53 Go [2026-07-17 07:16:01] step 211300/1000000 | loss=2.4419 vocab=1.0108 token=1.4192 | 6.07 steps/s | VRAM=28.14 Go [2026-07-17 07:16:09] step 211350/1000000 | loss=2.3275 vocab=0.6933 token=1.2306 | 6.09 steps/s | VRAM=30.37 Go [2026-07-17 07:16:17] step 211400/1000000 | loss=2.3296 vocab=1.0584 token=1.3578 | 6.12 steps/s | VRAM=32.18 Go [2026-07-17 07:16:25] step 211450/1000000 | loss=2.2608 vocab=0.7061 token=1.4192 | 6.10 steps/s | VRAM=31.56 Go [2026-07-17 07:16:34] step 211500/1000000 | loss=2.4234 vocab=1.2204 token=1.6564 | 6.11 steps/s | VRAM=28.46 Go [2026-07-17 07:16:42] step 211550/1000000 | loss=2.4227 vocab=1.1567 token=1.3315 | 6.08 steps/s | VRAM=32.72 Go [2026-07-17 07:16:50] step 211600/1000000 | loss=2.4348 vocab=0.8766 token=1.3245 | 6.07 steps/s | VRAM=29.52 Go [2026-07-17 07:16:58] step 211650/1000000 | loss=2.3885 vocab=1.2050 token=1.2709 | 6.08 steps/s | VRAM=28.91 Go [2026-07-17 07:17:06] step 211700/1000000 | loss=2.4416 vocab=1.4511 token=1.4733 | 6.14 steps/s | VRAM=28.89 Go [2026-07-17 07:17:15] step 211750/1000000 | loss=2.3488 vocab=1.1055 token=1.3319 | 6.09 steps/s | VRAM=31.42 Go [2026-07-17 07:17:23] step 211800/1000000 | loss=2.3578 vocab=1.3555 token=1.4990 | 6.06 steps/s | VRAM=28.09 Go [2026-07-17 07:17:31] step 211850/1000000 | loss=2.3425 vocab=1.1499 token=1.0059 | 6.10 steps/s | VRAM=28.52 Go [2026-07-17 07:17:39] step 211900/1000000 | loss=2.3865 vocab=1.1961 token=1.4136 | 6.09 steps/s | VRAM=29.28 Go [2026-07-17 07:17:47] step 211950/1000000 | loss=2.4771 vocab=0.7525 token=1.3615 | 6.09 steps/s | VRAM=26.33 Go [2026-07-17 07:17:56] step 212000/1000000 | loss=2.3662 vocab=1.0167 token=1.5228 | 6.10 steps/s | VRAM=28.47 Go [2026-07-17 07:18:04] step 212050/1000000 | loss=2.3657 vocab=0.8554 token=1.5712 | 6.05 steps/s | VRAM=28.28 Go [2026-07-17 07:18:12] step 212100/1000000 | loss=2.4092 vocab=0.9098 token=1.4039 | 6.09 steps/s | VRAM=29.35 Go [2026-07-17 07:18:20] step 212150/1000000 | loss=2.4493 vocab=1.0401 token=1.4814 | 6.07 steps/s | VRAM=26.43 Go [2026-07-17 07:18:29] step 212200/1000000 | loss=2.4497 vocab=0.9399 token=1.5756 | 6.07 steps/s | VRAM=28.64 Go [2026-07-17 07:18:37] step 212250/1000000 | loss=2.3872 vocab=0.7997 token=1.2643 | 6.07 steps/s | VRAM=29.42 Go [2026-07-17 07:18:45] step 212300/1000000 | loss=2.3131 vocab=0.8286 token=1.4857 | 6.11 steps/s | VRAM=24.71 Go [2026-07-17 07:18:53] step 212350/1000000 | loss=2.2856 vocab=1.2719 token=1.2453 | 6.09 steps/s | VRAM=28.40 Go [2026-07-17 07:19:01] step 212400/1000000 | loss=2.3779 vocab=1.2492 token=1.5821 | 6.09 steps/s | VRAM=25.94 Go [2026-07-17 07:19:10] step 212450/1000000 | loss=2.4462 vocab=1.0381 token=1.4106 | 6.06 steps/s | VRAM=28.44 Go [2026-07-17 07:19:18] step 212500/1000000 | loss=2.4299 vocab=1.5292 token=1.4500 | 6.12 steps/s | VRAM=24.74 Go [2026-07-17 07:19:26] step 212550/1000000 | loss=2.3553 vocab=1.0208 token=1.5137 | 6.10 steps/s | VRAM=32.22 Go [2026-07-17 07:19:34] step 212600/1000000 | loss=2.3635 vocab=0.8856 token=1.3728 | 6.08 steps/s | VRAM=28.32 Go [2026-07-17 07:19:42] step 212650/1000000 | loss=2.3030 vocab=0.7809 token=1.5595 | 6.09 steps/s | VRAM=29.20 Go [2026-07-17 07:19:51] step 212700/1000000 | loss=2.4457 vocab=0.9945 token=1.9892 | 6.11 steps/s | VRAM=32.75 Go [2026-07-17 07:19:59] step 212750/1000000 | loss=2.3578 vocab=0.7550 token=1.6599 | 6.06 steps/s | VRAM=25.79 Go [2026-07-17 07:20:07] step 212800/1000000 | loss=2.5428 vocab=1.2782 token=1.5531 | 6.08 steps/s | VRAM=27.64 Go [2026-07-17 07:20:15] step 212850/1000000 | loss=2.3348 vocab=0.6539 token=1.2531 | 6.13 steps/s | VRAM=31.83 Go [2026-07-17 07:20:23] step 212900/1000000 | loss=2.3913 vocab=0.9393 token=1.2291 | 6.07 steps/s | VRAM=28.79 Go [2026-07-17 07:20:32] step 212950/1000000 | loss=2.4787 vocab=1.1302 token=1.5019 | 6.05 steps/s | VRAM=26.56 Go [2026-07-17 07:20:40] step 213000/1000000 | loss=2.3664 vocab=0.7970 token=1.6611 | 6.11 steps/s | VRAM=28.19 Go [2026-07-17 07:20:48] step 213050/1000000 | loss=2.3727 vocab=1.4905 token=1.4020 | 6.06 steps/s | VRAM=30.12 Go [2026-07-17 07:20:56] step 213100/1000000 | loss=2.3398 vocab=0.6687 token=0.8930 | 6.09 steps/s | VRAM=32.41 Go [2026-07-17 07:21:05] step 213150/1000000 | loss=2.4315 vocab=0.6562 token=1.0295 | 6.09 steps/s | VRAM=28.15 Go [2026-07-17 07:21:13] step 213200/1000000 | loss=2.3504 vocab=0.4840 token=1.0872 | 6.09 steps/s | VRAM=28.91 Go [2026-07-17 07:21:21] step 213250/1000000 | loss=2.3833 vocab=0.6630 token=1.1371 | 6.08 steps/s | VRAM=26.98 Go [2026-07-17 07:21:29] step 213300/1000000 | loss=2.2913 vocab=0.5207 token=1.3043 | 6.10 steps/s | VRAM=31.83 Go [2026-07-17 07:21:37] step 213350/1000000 | loss=2.4036 vocab=0.7129 token=1.3860 | 6.09 steps/s | VRAM=29.20 Go [2026-07-17 07:21:46] step 213400/1000000 | loss=2.4113 vocab=1.3862 token=1.8243 | 6.10 steps/s | VRAM=28.41 Go [2026-07-17 07:21:54] step 213450/1000000 | loss=2.3092 vocab=0.9174 token=1.3457 | 6.10 steps/s | VRAM=29.20 Go [2026-07-17 07:22:02] step 213500/1000000 | loss=2.2399 vocab=0.9536 token=1.4304 | 6.12 steps/s | VRAM=27.87 Go [2026-07-17 07:22:10] step 213550/1000000 | loss=2.3842 vocab=1.2606 token=1.5422 | 6.09 steps/s | VRAM=31.96 Go [2026-07-17 07:22:19] step 213600/1000000 | loss=2.4480 vocab=0.5319 token=0.8685 | 6.05 steps/s | VRAM=23.40 Go [2026-07-17 07:22:27] step 213650/1000000 | loss=2.3872 vocab=0.7825 token=1.0679 | 6.12 steps/s | VRAM=32.13 Go [2026-07-17 07:22:35] step 213700/1000000 | loss=2.4263 vocab=1.5936 token=1.8057 | 6.07 steps/s | VRAM=26.45 Go [2026-07-17 07:22:43] step 213750/1000000 | loss=2.4089 vocab=1.1272 token=1.6420 | 6.07 steps/s | VRAM=26.22 Go [2026-07-17 07:22:51] step 213800/1000000 | loss=2.4930 vocab=1.1491 token=1.3349 | 6.09 steps/s | VRAM=29.53 Go [2026-07-17 07:23:00] step 213850/1000000 | loss=2.4039 vocab=0.7294 token=1.1673 | 6.10 steps/s | VRAM=32.55 Go [2026-07-17 07:23:08] step 213900/1000000 | loss=2.3442 vocab=0.8182 token=1.2699 | 6.10 steps/s | VRAM=30.26 Go [2026-07-17 07:23:16] step 213950/1000000 | loss=2.4406 vocab=0.9894 token=1.8478 | 6.10 steps/s | VRAM=29.36 Go [2026-07-17 07:23:24] step 214000/1000000 | loss=2.4284 vocab=1.3276 token=1.7709 | 6.12 steps/s | VRAM=31.78 Go [2026-07-17 07:23:32] step 214050/1000000 | loss=2.3781 vocab=1.3724 token=1.5672 | 6.10 steps/s | VRAM=28.86 Go [2026-07-17 07:23:41] step 214100/1000000 | loss=2.2738 vocab=0.6561 token=1.0515 | 6.09 steps/s | VRAM=28.95 Go [2026-07-17 07:23:49] step 214150/1000000 | loss=2.4160 vocab=1.0778 token=1.3731 | 6.12 steps/s | VRAM=28.05 Go [2026-07-17 07:23:57] step 214200/1000000 | loss=2.3510 vocab=0.9826 token=1.5008 | 6.13 steps/s | VRAM=31.62 Go [2026-07-17 07:24:05] step 214250/1000000 | loss=2.3162 vocab=0.7385 token=1.1994 | 6.09 steps/s | VRAM=28.84 Go [2026-07-17 07:24:13] step 214300/1000000 | loss=2.3235 vocab=1.0122 token=1.5831 | 6.11 steps/s | VRAM=28.71 Go [2026-07-17 07:24:21] step 214350/1000000 | loss=2.3699 vocab=0.9875 token=1.4034 | 6.10 steps/s | VRAM=29.40 Go [2026-07-17 07:24:30] step 214400/1000000 | loss=2.3631 vocab=0.8446 token=1.2385 | 6.11 steps/s | VRAM=32.51 Go [2026-07-17 07:24:38] step 214450/1000000 | loss=2.2938 vocab=0.6687 token=1.2813 | 6.12 steps/s | VRAM=31.48 Go [2026-07-17 07:24:46] step 214500/1000000 | loss=2.3734 vocab=1.0629 token=1.3429 | 6.10 steps/s | VRAM=28.97 Go [2026-07-17 07:24:54] step 214550/1000000 | loss=2.3827 vocab=1.2014 token=1.6049 | 6.11 steps/s | VRAM=24.44 Go [2026-07-17 07:25:02] step 214600/1000000 | loss=2.3610 vocab=0.8868 token=1.3374 | 6.14 steps/s | VRAM=32.03 Go [2026-07-17 07:25:11] step 214650/1000000 | loss=2.3974 vocab=0.7292 token=1.2580 | 6.10 steps/s | VRAM=24.58 Go [2026-07-17 07:25:19] step 214700/1000000 | loss=2.2397 vocab=0.8211 token=1.2832 | 6.09 steps/s | VRAM=28.40 Go [2026-07-17 07:25:27] step 214750/1000000 | loss=2.3076 vocab=1.4146 token=1.2483 | 6.10 steps/s | VRAM=26.29 Go [2026-07-17 07:25:35] step 214800/1000000 | loss=2.2449 vocab=0.8080 token=1.1581 | 6.10 steps/s | VRAM=27.75 Go [2026-07-17 07:25:43] step 214850/1000000 | loss=2.3585 vocab=1.5487 token=1.6108 | 6.12 steps/s | VRAM=25.46 Go [2026-07-17 07:25:51] step 214900/1000000 | loss=2.4423 vocab=1.4574 token=1.6432 | 6.12 steps/s | VRAM=32.14 Go [2026-07-17 07:26:00] step 214950/1000000 | loss=2.2255 vocab=0.7022 token=1.2349 | 6.08 steps/s | VRAM=30.48 Go [2026-07-17 07:26:08] step 215000/1000000 | loss=2.4095 vocab=0.7929 token=1.5404 | 6.09 steps/s | VRAM=28.57 Go [2026-07-17 07:27:36] step 215050/1000000 | loss=2.4279 vocab=0.7975 token=1.5364 | 0.57 steps/s | VRAM=31.47 Go [2026-07-17 07:27:44] step 215100/1000000 | loss=2.3969 vocab=1.3270 token=1.5732 | 6.15 steps/s | VRAM=28.45 Go [2026-07-17 07:27:52] step 215150/1000000 | loss=2.4572 vocab=0.9638 token=1.5364 | 6.13 steps/s | VRAM=26.80 Go [2026-07-17 07:28:00] step 215200/1000000 | loss=2.4964 vocab=0.8130 token=1.5226 | 6.08 steps/s | VRAM=28.74 Go [2026-07-17 07:28:08] step 215250/1000000 | loss=2.3277 vocab=1.0399 token=1.5673 | 6.13 steps/s | VRAM=28.00 Go [2026-07-17 07:28:16] step 215300/1000000 | loss=2.3545 vocab=0.7959 token=1.5376 | 6.14 steps/s | VRAM=24.39 Go [2026-07-17 07:28:25] step 215350/1000000 | loss=2.3917 vocab=1.5186 token=1.5021 | 6.14 steps/s | VRAM=32.32 Go [2026-07-17 07:28:33] step 215400/1000000 | loss=2.3932 vocab=0.6400 token=1.2774 | 6.10 steps/s | VRAM=26.24 Go [2026-07-17 07:28:41] step 215450/1000000 | loss=2.3696 vocab=0.5099 token=1.2127 | 6.12 steps/s | VRAM=24.88 Go [2026-07-17 07:28:49] step 215500/1000000 | loss=2.2800 vocab=1.0434 token=1.1969 | 6.10 steps/s | VRAM=28.90 Go [2026-07-17 07:28:57] step 215550/1000000 | loss=2.3024 vocab=0.9384 token=1.2847 | 6.10 steps/s | VRAM=28.36 Go [2026-07-17 07:29:06] step 215600/1000000 | loss=2.4694 vocab=0.8364 token=1.4999 | 6.10 steps/s | VRAM=28.83 Go [2026-07-17 07:29:14] step 215650/1000000 | loss=2.2543 vocab=0.8974 token=1.4194 | 6.08 steps/s | VRAM=25.03 Go [2026-07-17 07:29:22] step 215700/1000000 | loss=2.4274 vocab=0.8635 token=1.4005 | 6.08 steps/s | VRAM=31.38 Go [2026-07-17 07:29:30] step 215750/1000000 | loss=2.2969 vocab=0.6492 token=1.3206 | 6.07 steps/s | VRAM=26.55 Go [2026-07-17 07:29:38] step 215800/1000000 | loss=2.3760 vocab=1.3537 token=1.5577 | 6.07 steps/s | VRAM=29.20 Go [2026-07-17 07:29:47] step 215850/1000000 | loss=2.3135 vocab=0.6064 token=1.3956 | 6.04 steps/s | VRAM=28.86 Go [2026-07-17 07:29:55] step 215900/1000000 | loss=2.4483 vocab=0.9132 token=1.5838 | 6.08 steps/s | VRAM=28.46 Go [2026-07-17 07:30:03] step 215950/1000000 | loss=2.3387 vocab=1.2005 token=1.4452 | 6.08 steps/s | VRAM=28.60 Go [2026-07-17 07:30:11] step 216000/1000000 | loss=2.4456 vocab=0.9580 token=1.6442 | 6.10 steps/s | VRAM=29.33 Go [2026-07-17 07:30:20] step 216050/1000000 | loss=2.2888 vocab=0.9486 token=1.1522 | 6.08 steps/s | VRAM=26.14 Go [2026-07-17 07:30:28] step 216100/1000000 | loss=2.5701 vocab=0.8572 token=1.9595 | 6.04 steps/s | VRAM=29.47 Go [2026-07-17 07:30:36] step 216150/1000000 | loss=2.3874 vocab=0.9005 token=1.1210 | 6.06 steps/s | VRAM=31.75 Go [2026-07-17 07:30:44] step 216200/1000000 | loss=2.3024 vocab=1.2429 token=1.5637 | 6.08 steps/s | VRAM=28.51 Go [2026-07-17 07:30:53] step 216250/1000000 | loss=2.4112 vocab=0.9372 token=1.3420 | 6.10 steps/s | VRAM=24.18 Go [2026-07-17 07:31:01] step 216300/1000000 | loss=2.2968 vocab=1.1134 token=1.7186 | 6.08 steps/s | VRAM=31.80 Go [2026-07-17 07:31:09] step 216350/1000000 | loss=2.3061 vocab=0.7860 token=1.4421 | 6.07 steps/s | VRAM=31.92 Go [2026-07-17 07:31:17] step 216400/1000000 | loss=2.3672 vocab=0.6339 token=1.1047 | 6.09 steps/s | VRAM=30.97 Go [2026-07-17 07:31:25] step 216450/1000000 | loss=2.3825 vocab=1.0793 token=1.5716 | 6.09 steps/s | VRAM=27.00 Go [2026-07-17 07:31:34] step 216500/1000000 | loss=2.4792 vocab=1.0317 token=1.4674 | 6.08 steps/s | VRAM=28.84 Go [2026-07-17 07:31:42] step 216550/1000000 | loss=2.3572 vocab=1.2958 token=1.3552 | 6.09 steps/s | VRAM=28.07 Go [2026-07-17 07:31:50] step 216600/1000000 | loss=2.3541 vocab=0.8552 token=1.3506 | 6.11 steps/s | VRAM=32.06 Go [2026-07-17 07:31:58] step 216650/1000000 | loss=2.3262 vocab=1.7144 token=1.5371 | 6.08 steps/s | VRAM=32.78 Go [2026-07-17 07:32:07] step 216700/1000000 | loss=2.3034 vocab=0.6125 token=1.2519 | 6.10 steps/s | VRAM=28.66 Go [2026-07-17 07:32:15] step 216750/1000000 | loss=2.4077 vocab=1.3901 token=1.6259 | 6.07 steps/s | VRAM=27.19 Go [2026-07-17 07:32:23] step 216800/1000000 | loss=2.4332 vocab=1.2060 token=1.7015 | 6.07 steps/s | VRAM=28.86 Go [2026-07-17 07:32:31] step 216850/1000000 | loss=2.4103 vocab=1.5093 token=1.4742 | 6.09 steps/s | VRAM=25.74 Go [2026-07-17 07:32:39] step 216900/1000000 | loss=2.4308 vocab=0.7666 token=1.3651 | 6.10 steps/s | VRAM=31.78 Go [2026-07-17 07:32:48] step 216950/1000000 | loss=2.4058 vocab=0.9347 token=1.2898 | 6.06 steps/s | VRAM=26.53 Go [2026-07-17 07:32:56] step 217000/1000000 | loss=2.4655 vocab=0.7763 token=1.4073 | 6.07 steps/s | VRAM=30.61 Go [2026-07-17 07:33:04] step 217050/1000000 | loss=2.3778 vocab=1.0489 token=1.5599 | 6.08 steps/s | VRAM=28.44 Go [2026-07-17 07:33:12] step 217100/1000000 | loss=2.4302 vocab=0.6913 token=1.2902 | 6.09 steps/s | VRAM=29.03 Go [2026-07-17 07:33:21] step 217150/1000000 | loss=2.2274 vocab=0.9938 token=1.3353 | 6.09 steps/s | VRAM=28.44 Go [2026-07-17 07:33:29] step 217200/1000000 | loss=2.3642 vocab=0.8709 token=1.2361 | 6.08 steps/s | VRAM=31.87 Go [2026-07-17 07:33:37] step 217250/1000000 | loss=2.3008 vocab=1.1458 token=1.7196 | 6.09 steps/s | VRAM=25.69 Go [2026-07-17 07:33:45] step 217300/1000000 | loss=2.3154 vocab=0.4716 token=1.2578 | 6.11 steps/s | VRAM=25.32 Go [2026-07-17 07:33:53] step 217350/1000000 | loss=2.3946 vocab=0.8830 token=1.4372 | 6.11 steps/s | VRAM=31.82 Go [2026-07-17 07:34:02] step 217400/1000000 | loss=2.3667 vocab=1.0304 token=1.0799 | 6.09 steps/s | VRAM=29.41 Go [2026-07-17 07:34:10] step 217450/1000000 | loss=2.4429 vocab=0.8215 token=1.3456 | 6.09 steps/s | VRAM=24.90 Go [2026-07-17 07:34:18] step 217500/1000000 | loss=2.2329 vocab=0.5735 token=1.5389 | 6.10 steps/s | VRAM=32.06 Go [2026-07-17 07:34:26] step 217550/1000000 | loss=2.3626 vocab=0.6812 token=1.2289 | 6.07 steps/s | VRAM=29.87 Go [2026-07-17 07:34:34] step 217600/1000000 | loss=2.3859 vocab=0.8483 token=1.4214 | 6.11 steps/s | VRAM=24.87 Go [2026-07-17 07:34:43] step 217650/1000000 | loss=2.3869 vocab=0.7911 token=1.5002 | 6.11 steps/s | VRAM=31.56 Go [2026-07-17 07:34:51] step 217700/1000000 | loss=2.3296 vocab=0.9962 token=1.3947 | 6.07 steps/s | VRAM=31.99 Go [2026-07-17 07:34:59] step 217750/1000000 | loss=2.2740 vocab=1.1412 token=1.7812 | 6.08 steps/s | VRAM=26.55 Go [2026-07-17 07:35:07] step 217800/1000000 | loss=2.4448 vocab=1.2332 token=1.3959 | 6.11 steps/s | VRAM=29.18 Go [2026-07-17 07:35:15] step 217850/1000000 | loss=2.4288 vocab=1.0098 token=1.2503 | 6.13 steps/s | VRAM=26.48 Go [2026-07-17 07:35:24] step 217900/1000000 | loss=2.3944 vocab=0.3866 token=1.1485 | 6.09 steps/s | VRAM=26.50 Go [2026-07-17 07:35:32] step 217950/1000000 | loss=2.3819 vocab=1.0779 token=1.4083 | 6.13 steps/s | VRAM=29.67 Go [2026-07-17 07:35:40] step 218000/1000000 | loss=2.3814 vocab=0.7582 token=1.3125 | 6.07 steps/s | VRAM=28.99 Go [2026-07-17 07:35:48] step 218050/1000000 | loss=2.3922 vocab=0.8903 token=1.3619 | 6.09 steps/s | VRAM=26.28 Go [2026-07-17 07:35:56] step 218100/1000000 | loss=2.2928 vocab=0.7760 token=1.3186 | 6.06 steps/s | VRAM=29.20 Go [2026-07-17 07:36:05] step 218150/1000000 | loss=2.3994 vocab=0.9200 token=1.3428 | 6.08 steps/s | VRAM=28.43 Go [2026-07-17 07:36:13] step 218200/1000000 | loss=2.3119 vocab=1.3267 token=1.6937 | 6.11 steps/s | VRAM=29.60 Go [2026-07-17 07:36:21] step 218250/1000000 | loss=2.3022 vocab=1.3112 token=1.2878 | 6.09 steps/s | VRAM=31.97 Go [2026-07-17 07:36:29] step 218300/1000000 | loss=2.4917 vocab=0.8824 token=1.6622 | 6.11 steps/s | VRAM=32.33 Go [2026-07-17 07:36:37] step 218350/1000000 | loss=2.4258 vocab=0.9494 token=1.3223 | 6.07 steps/s | VRAM=29.37 Go [2026-07-17 07:36:46] step 218400/1000000 | loss=2.2566 vocab=1.2978 token=1.4589 | 6.09 steps/s | VRAM=32.38 Go [2026-07-17 07:36:54] step 218450/1000000 | loss=2.4133 vocab=1.1976 token=1.1255 | 6.08 steps/s | VRAM=30.22 Go [2026-07-17 07:37:02] step 218500/1000000 | loss=2.4493 vocab=0.5757 token=1.2094 | 6.10 steps/s | VRAM=29.09 Go [2026-07-17 07:37:10] step 218550/1000000 | loss=2.3638 vocab=1.5367 token=1.4694 | 6.09 steps/s | VRAM=29.10 Go [2026-07-17 07:37:19] step 218600/1000000 | loss=2.4389 vocab=1.3686 token=1.1059 | 6.09 steps/s | VRAM=28.17 Go [2026-07-17 07:37:27] step 218650/1000000 | loss=2.4485 vocab=1.0137 token=1.8960 | 6.11 steps/s | VRAM=28.59 Go [2026-07-17 07:37:35] step 218700/1000000 | loss=2.5474 vocab=1.1642 token=1.4185 | 6.10 steps/s | VRAM=32.27 Go [2026-07-17 07:37:43] step 218750/1000000 | loss=2.2979 vocab=0.5018 token=1.1813 | 6.11 steps/s | VRAM=31.83 Go [2026-07-17 07:37:51] step 218800/1000000 | loss=2.3571 vocab=0.9871 token=1.2465 | 6.11 steps/s | VRAM=32.46 Go [2026-07-17 07:37:59] step 218850/1000000 | loss=2.4013 vocab=0.9044 token=1.2983 | 6.09 steps/s | VRAM=23.69 Go [2026-07-17 07:38:08] step 218900/1000000 | loss=2.3249 vocab=1.0497 token=1.8096 | 6.10 steps/s | VRAM=28.27 Go [2026-07-17 07:38:16] step 218950/1000000 | loss=2.3130 vocab=1.0806 token=1.2076 | 6.10 steps/s | VRAM=29.96 Go [2026-07-17 07:38:24] step 219000/1000000 | loss=2.4066 vocab=0.6207 token=1.4841 | 6.09 steps/s | VRAM=28.59 Go [2026-07-17 07:38:32] step 219050/1000000 | loss=2.3490 vocab=1.0785 token=1.3389 | 6.06 steps/s | VRAM=30.39 Go [2026-07-17 07:38:41] step 219100/1000000 | loss=2.3591 vocab=1.1425 token=1.5015 | 6.12 steps/s | VRAM=31.94 Go [2026-07-17 07:38:49] step 219150/1000000 | loss=2.4155 vocab=0.5514 token=1.2865 | 6.09 steps/s | VRAM=28.44 Go [2026-07-17 07:38:57] step 219200/1000000 | loss=2.4337 vocab=0.8864 token=1.2656 | 6.10 steps/s | VRAM=29.03 Go [2026-07-17 07:39:05] step 219250/1000000 | loss=2.3605 vocab=1.8758 token=1.9383 | 6.10 steps/s | VRAM=31.80 Go [2026-07-17 07:39:13] step 219300/1000000 | loss=2.4337 vocab=0.7656 token=1.1002 | 6.07 steps/s | VRAM=29.58 Go [2026-07-17 07:39:22] step 219350/1000000 | loss=2.4636 vocab=1.1244 token=1.5696 | 6.10 steps/s | VRAM=26.09 Go [2026-07-17 07:39:30] step 219400/1000000 | loss=2.3173 vocab=0.7983 token=1.1806 | 6.10 steps/s | VRAM=26.40 Go [2026-07-17 07:39:38] step 219450/1000000 | loss=2.3079 vocab=0.7668 token=0.9154 | 6.07 steps/s | VRAM=29.92 Go [2026-07-17 07:39:46] step 219500/1000000 | loss=2.3091 vocab=1.0957 token=1.4881 | 6.09 steps/s | VRAM=28.13 Go [2026-07-17 07:39:54] step 219550/1000000 | loss=2.3403 vocab=0.9592 token=1.2693 | 6.10 steps/s | VRAM=32.45 Go [2026-07-17 07:40:03] step 219600/1000000 | loss=2.3291 vocab=1.1208 token=1.3278 | 6.07 steps/s | VRAM=26.34 Go [2026-07-17 07:40:11] step 219650/1000000 | loss=2.2846 vocab=1.2992 token=1.5211 | 6.12 steps/s | VRAM=28.94 Go [2026-07-17 07:40:19] step 219700/1000000 | loss=2.3203 vocab=0.8127 token=1.5548 | 6.14 steps/s | VRAM=28.29 Go [2026-07-17 07:40:27] step 219750/1000000 | loss=2.5185 vocab=0.9183 token=1.3976 | 6.08 steps/s | VRAM=26.49 Go [2026-07-17 07:40:35] step 219800/1000000 | loss=2.3336 vocab=1.1829 token=1.3307 | 6.10 steps/s | VRAM=30.06 Go [2026-07-17 07:40:44] step 219850/1000000 | loss=2.3425 vocab=0.6702 token=1.1895 | 6.10 steps/s | VRAM=28.61 Go [2026-07-17 07:40:52] step 219900/1000000 | loss=2.3339 vocab=0.9265 token=1.3271 | 6.11 steps/s | VRAM=29.21 Go [2026-07-17 07:41:00] step 219950/1000000 | loss=2.4639 vocab=0.6317 token=1.1937 | 6.08 steps/s | VRAM=28.96 Go [2026-07-17 07:41:08] step 220000/1000000 | loss=2.4487 vocab=0.9231 token=1.2870 | 6.13 steps/s | VRAM=31.78 Go [2026-07-17 07:42:32] step 220050/1000000 | loss=2.3575 vocab=0.8465 token=1.2699 | 0.60 steps/s | VRAM=26.14 Go [2026-07-17 07:42:40] step 220100/1000000 | loss=2.4610 vocab=1.2461 token=1.7003 | 6.18 steps/s | VRAM=29.23 Go [2026-07-17 07:42:48] step 220150/1000000 | loss=2.3218 vocab=1.1823 token=1.7632 | 6.19 steps/s | VRAM=31.57 Go [2026-07-17 07:42:56] step 220200/1000000 | loss=2.3276 vocab=1.3557 token=1.8736 | 6.18 steps/s | VRAM=24.94 Go [2026-07-17 07:43:04] step 220250/1000000 | loss=2.3532 vocab=1.1312 token=1.5552 | 6.15 steps/s | VRAM=32.14 Go [2026-07-17 07:43:12] step 220300/1000000 | loss=2.3902 vocab=0.8521 token=1.5019 | 6.16 steps/s | VRAM=28.56 Go [2026-07-17 07:43:21] step 220350/1000000 | loss=2.4209 vocab=1.1364 token=1.4087 | 6.12 steps/s | VRAM=31.49 Go [2026-07-17 07:43:29] step 220400/1000000 | loss=2.5322 vocab=1.7712 token=1.6818 | 6.13 steps/s | VRAM=31.66 Go [2026-07-17 07:43:37] step 220450/1000000 | loss=2.4023 vocab=0.7008 token=1.3646 | 6.10 steps/s | VRAM=32.59 Go [2026-07-17 07:43:45] step 220500/1000000 | loss=2.3983 vocab=1.6426 token=2.2030 | 6.15 steps/s | VRAM=32.52 Go [2026-07-17 07:43:53] step 220550/1000000 | loss=2.3574 vocab=1.2742 token=1.4971 | 6.11 steps/s | VRAM=32.54 Go [2026-07-17 07:44:01] step 220600/1000000 | loss=2.3923 vocab=1.0528 token=1.4582 | 6.12 steps/s | VRAM=25.61 Go [2026-07-17 07:44:10] step 220650/1000000 | loss=2.4336 vocab=0.7705 token=1.4276 | 6.11 steps/s | VRAM=26.01 Go [2026-07-17 07:44:18] step 220700/1000000 | loss=2.3705 vocab=0.5889 token=1.2876 | 6.11 steps/s | VRAM=26.43 Go [2026-07-17 07:44:26] step 220750/1000000 | loss=2.3497 vocab=1.4849 token=1.8769 | 6.12 steps/s | VRAM=28.12 Go [2026-07-17 07:44:34] step 220800/1000000 | loss=2.3655 vocab=1.0830 token=1.4068 | 6.10 steps/s | VRAM=29.98 Go [2026-07-17 07:44:42] step 220850/1000000 | loss=2.2930 vocab=0.9020 token=1.6591 | 6.11 steps/s | VRAM=28.88 Go [2026-07-17 07:44:51] step 220900/1000000 | loss=2.4104 vocab=0.7006 token=1.4028 | 6.11 steps/s | VRAM=31.09 Go [2026-07-17 07:44:59] step 220950/1000000 | loss=2.3925 vocab=0.6248 token=1.4266 | 6.08 steps/s | VRAM=26.38 Go [2026-07-17 07:45:07] step 221000/1000000 | loss=2.3979 vocab=1.3339 token=1.3396 | 6.11 steps/s | VRAM=27.13 Go [2026-07-17 07:45:15] step 221050/1000000 | loss=2.3659 vocab=0.9475 token=1.3739 | 6.09 steps/s | VRAM=29.74 Go [2026-07-17 07:45:23] step 221100/1000000 | loss=2.3443 vocab=1.1724 token=1.4153 | 6.12 steps/s | VRAM=31.82 Go [2026-07-17 07:45:32] step 221150/1000000 | loss=2.3096 vocab=0.9493 token=1.6218 | 6.10 steps/s | VRAM=24.12 Go [2026-07-17 07:45:40] step 221200/1000000 | loss=2.3107 vocab=1.0687 token=1.5718 | 6.07 steps/s | VRAM=27.03 Go [2026-07-17 07:45:48] step 221250/1000000 | loss=2.3149 vocab=0.8994 token=1.1534 | 6.10 steps/s | VRAM=28.49 Go [2026-07-17 07:45:56] step 221300/1000000 | loss=2.4761 vocab=1.3762 token=1.6714 | 6.11 steps/s | VRAM=30.84 Go [2026-07-17 07:46:04] step 221350/1000000 | loss=2.5021 vocab=0.5869 token=1.3609 | 6.11 steps/s | VRAM=31.61 Go [2026-07-17 07:46:13] step 221400/1000000 | loss=2.4155 vocab=0.8784 token=1.5160 | 6.10 steps/s | VRAM=32.03 Go [2026-07-17 07:46:21] step 221450/1000000 | loss=2.3770 vocab=1.1787 token=1.4642 | 6.11 steps/s | VRAM=28.83 Go [2026-07-17 07:46:29] step 221500/1000000 | loss=2.4123 vocab=1.0568 token=1.6788 | 6.09 steps/s | VRAM=29.44 Go [2026-07-17 07:46:37] step 221550/1000000 | loss=2.3602 vocab=1.0899 token=1.5913 | 6.11 steps/s | VRAM=24.38 Go [2026-07-17 07:46:45] step 221600/1000000 | loss=2.3885 vocab=1.0981 token=1.6746 | 6.10 steps/s | VRAM=32.28 Go [2026-07-17 07:46:53] step 221650/1000000 | loss=2.2973 vocab=1.1647 token=0.9263 | 6.10 steps/s | VRAM=28.53 Go [2026-07-17 07:47:02] step 221700/1000000 | loss=2.4021 vocab=1.1147 token=1.7215 | 6.10 steps/s | VRAM=32.45 Go [2026-07-17 07:47:10] step 221750/1000000 | loss=2.4000 vocab=1.1327 token=1.2717 | 6.11 steps/s | VRAM=32.36 Go [2026-07-17 07:47:18] step 221800/1000000 | loss=2.3521 vocab=1.0956 token=1.4874 | 6.11 steps/s | VRAM=28.79 Go [2026-07-17 07:47:26] step 221850/1000000 | loss=2.4689 vocab=1.2128 token=1.5027 | 6.10 steps/s | VRAM=26.17 Go [2026-07-17 07:47:34] step 221900/1000000 | loss=2.4461 vocab=1.1674 token=1.3408 | 6.13 steps/s | VRAM=29.07 Go [2026-07-17 07:47:43] step 221950/1000000 | loss=2.4419 vocab=1.2211 token=1.6423 | 6.11 steps/s | VRAM=28.03 Go [2026-07-17 07:47:51] step 222000/1000000 | loss=2.4661 vocab=1.0196 token=1.5548 | 6.09 steps/s | VRAM=24.92 Go [2026-07-17 07:47:59] step 222050/1000000 | loss=2.3552 vocab=0.7427 token=1.1273 | 6.09 steps/s | VRAM=28.66 Go [2026-07-17 07:48:07] step 222100/1000000 | loss=2.3369 vocab=1.2584 token=1.5278 | 6.13 steps/s | VRAM=31.25 Go [2026-07-17 07:48:15] step 222150/1000000 | loss=2.3881 vocab=1.2650 token=1.3045 | 6.10 steps/s | VRAM=32.73 Go [2026-07-17 07:48:24] step 222200/1000000 | loss=2.3873 vocab=0.7861 token=1.3099 | 6.08 steps/s | VRAM=26.41 Go [2026-07-17 07:48:32] step 222250/1000000 | loss=2.4049 vocab=1.5221 token=1.5702 | 6.12 steps/s | VRAM=28.34 Go [2026-07-17 07:48:40] step 222300/1000000 | loss=2.3382 vocab=0.7481 token=1.5158 | 6.14 steps/s | VRAM=28.63 Go [2026-07-17 07:48:48] step 222350/1000000 | loss=2.2970 vocab=0.8366 token=1.1997 | 6.12 steps/s | VRAM=29.80 Go [2026-07-17 07:48:56] step 222400/1000000 | loss=2.2604 vocab=0.9407 token=1.6187 | 6.10 steps/s | VRAM=29.41 Go [2026-07-17 07:49:04] step 222450/1000000 | loss=2.3476 vocab=0.9119 token=1.0418 | 6.14 steps/s | VRAM=28.01 Go [2026-07-17 07:49:13] step 222500/1000000 | loss=2.3298 vocab=0.8510 token=1.2681 | 6.07 steps/s | VRAM=26.91 Go [2026-07-17 07:49:21] step 222550/1000000 | loss=2.3567 vocab=0.7048 token=1.3490 | 6.13 steps/s | VRAM=28.96 Go [2026-07-17 07:49:29] step 222600/1000000 | loss=2.4676 vocab=1.4721 token=1.5604 | 6.08 steps/s | VRAM=29.16 Go [2026-07-17 07:49:37] step 222650/1000000 | loss=2.4153 vocab=0.5637 token=1.5266 | 6.14 steps/s | VRAM=31.37 Go [2026-07-17 07:49:45] step 222700/1000000 | loss=2.4520 vocab=1.3295 token=1.4636 | 6.09 steps/s | VRAM=29.22 Go [2026-07-17 07:49:54] step 222750/1000000 | loss=2.3877 vocab=0.7879 token=1.4964 | 6.07 steps/s | VRAM=26.88 Go [2026-07-17 07:50:02] step 222800/1000000 | loss=2.2991 vocab=0.7376 token=1.0702 | 6.14 steps/s | VRAM=27.61 Go [2026-07-17 07:50:10] step 222850/1000000 | loss=2.4345 vocab=0.8815 token=1.1780 | 6.09 steps/s | VRAM=26.71 Go [2026-07-17 07:50:18] step 222900/1000000 | loss=2.3843 vocab=1.4078 token=1.3313 | 6.12 steps/s | VRAM=29.04 Go [2026-07-17 07:50:26] step 222950/1000000 | loss=2.3046 vocab=1.4298 token=1.3281 | 6.13 steps/s | VRAM=27.33 Go [2026-07-17 07:50:34] step 223000/1000000 | loss=2.3793 vocab=0.7821 token=1.4079 | 6.13 steps/s | VRAM=27.72 Go [2026-07-17 07:50:43] step 223050/1000000 | loss=2.4032 vocab=0.8361 token=1.5393 | 6.10 steps/s | VRAM=27.07 Go [2026-07-17 07:50:51] step 223100/1000000 | loss=2.3891 vocab=0.7513 token=1.2025 | 6.13 steps/s | VRAM=27.41 Go [2026-07-17 07:50:59] step 223150/1000000 | loss=2.4373 vocab=1.0422 token=1.7158 | 6.14 steps/s | VRAM=25.91 Go [2026-07-17 07:51:07] step 223200/1000000 | loss=2.4285 vocab=1.0294 token=1.5185 | 6.11 steps/s | VRAM=32.20 Go [2026-07-17 07:51:15] step 223250/1000000 | loss=2.4121 vocab=0.7567 token=1.3305 | 6.10 steps/s | VRAM=28.10 Go [2026-07-17 07:51:24] step 223300/1000000 | loss=2.3539 vocab=1.1193 token=1.6043 | 6.08 steps/s | VRAM=25.88 Go [2026-07-17 07:51:32] step 223350/1000000 | loss=2.3841 vocab=1.0267 token=1.4662 | 6.14 steps/s | VRAM=25.42 Go [2026-07-17 07:51:40] step 223400/1000000 | loss=2.3285 vocab=0.6630 token=1.1571 | 6.16 steps/s | VRAM=28.53 Go [2026-07-17 07:51:48] step 223450/1000000 | loss=2.3363 vocab=0.6912 token=1.2605 | 6.11 steps/s | VRAM=31.97 Go [2026-07-17 07:51:56] step 223500/1000000 | loss=2.3505 vocab=1.1010 token=1.6148 | 6.12 steps/s | VRAM=25.36 Go [2026-07-17 07:52:04] step 223550/1000000 | loss=2.4988 vocab=0.8581 token=1.3987 | 6.09 steps/s | VRAM=26.40 Go [2026-07-17 07:52:13] step 223600/1000000 | loss=2.5528 vocab=0.9523 token=1.4423 | 6.12 steps/s | VRAM=26.08 Go [2026-07-17 07:52:21] step 223650/1000000 | loss=2.4135 vocab=1.0491 token=0.9758 | 6.10 steps/s | VRAM=26.36 Go [2026-07-17 07:52:29] step 223700/1000000 | loss=2.3753 vocab=1.0505 token=1.6365 | 6.11 steps/s | VRAM=28.17 Go [2026-07-17 07:52:37] step 223750/1000000 | loss=2.3082 vocab=0.9635 token=1.6356 | 6.11 steps/s | VRAM=29.05 Go [2026-07-17 07:52:45] step 223800/1000000 | loss=2.3447 vocab=0.8371 token=1.4312 | 6.09 steps/s | VRAM=25.84 Go [2026-07-17 07:52:54] step 223850/1000000 | loss=2.4548 vocab=0.9977 token=1.6621 | 6.10 steps/s | VRAM=32.46 Go [2026-07-17 07:53:02] step 223900/1000000 | loss=2.4855 vocab=1.0251 token=1.4964 | 6.08 steps/s | VRAM=26.22 Go [2026-07-17 07:53:10] step 223950/1000000 | loss=2.3799 vocab=0.8992 token=1.3436 | 6.11 steps/s | VRAM=24.74 Go [2026-07-17 07:53:18] step 224000/1000000 | loss=2.2891 vocab=0.9704 token=1.9064 | 6.10 steps/s | VRAM=32.99 Go [2026-07-17 07:53:26] step 224050/1000000 | loss=2.3537 vocab=1.2129 token=1.7624 | 6.12 steps/s | VRAM=27.58 Go [2026-07-17 07:53:35] step 224100/1000000 | loss=2.3590 vocab=1.3573 token=1.3379 | 6.11 steps/s | VRAM=28.89 Go [2026-07-17 07:53:43] step 224150/1000000 | loss=2.3252 vocab=1.1825 token=0.9872 | 6.11 steps/s | VRAM=32.32 Go [2026-07-17 07:53:51] step 224200/1000000 | loss=2.3920 vocab=0.8517 token=1.3258 | 6.10 steps/s | VRAM=27.66 Go [2026-07-17 07:53:59] step 224250/1000000 | loss=2.3376 vocab=0.8179 token=1.2498 | 6.09 steps/s | VRAM=26.10 Go [2026-07-17 07:54:07] step 224300/1000000 | loss=2.2270 vocab=1.0125 token=1.2623 | 6.13 steps/s | VRAM=32.73 Go [2026-07-17 07:54:15] step 224350/1000000 | loss=2.3548 vocab=1.1904 token=1.7804 | 6.09 steps/s | VRAM=26.37 Go [2026-07-17 07:54:24] step 224400/1000000 | loss=2.3920 vocab=1.3299 token=1.4602 | 6.13 steps/s | VRAM=31.47 Go [2026-07-17 07:54:32] step 224450/1000000 | loss=2.4121 vocab=0.7857 token=1.0384 | 6.08 steps/s | VRAM=28.41 Go [2026-07-17 07:54:40] step 224500/1000000 | loss=2.2662 vocab=1.5121 token=1.4702 | 6.12 steps/s | VRAM=26.28 Go [2026-07-17 07:54:48] step 224550/1000000 | loss=2.4020 vocab=0.8464 token=1.1431 | 6.11 steps/s | VRAM=31.61 Go [2026-07-17 07:54:56] step 224600/1000000 | loss=2.4388 vocab=0.8108 token=1.4928 | 6.07 steps/s | VRAM=26.32 Go [2026-07-17 07:55:05] step 224650/1000000 | loss=2.3783 vocab=0.9391 token=1.3523 | 6.11 steps/s | VRAM=32.60 Go [2026-07-17 07:55:13] step 224700/1000000 | loss=2.5629 vocab=1.3375 token=1.5685 | 6.10 steps/s | VRAM=28.75 Go [2026-07-17 07:55:21] step 224750/1000000 | loss=2.3957 vocab=1.0758 token=1.5507 | 6.09 steps/s | VRAM=29.46 Go [2026-07-17 07:55:29] step 224800/1000000 | loss=2.3798 vocab=1.0761 token=1.2208 | 6.10 steps/s | VRAM=29.53 Go [2026-07-17 07:55:37] step 224850/1000000 | loss=2.3472 vocab=1.0465 token=1.5041 | 6.10 steps/s | VRAM=27.17 Go [2026-07-17 07:55:46] step 224900/1000000 | loss=2.3482 vocab=0.5739 token=1.0488 | 6.11 steps/s | VRAM=28.23 Go [2026-07-17 07:55:54] step 224950/1000000 | loss=2.3493 vocab=0.7658 token=1.4547 | 6.10 steps/s | VRAM=28.69 Go [2026-07-17 07:56:02] step 225000/1000000 | loss=2.3678 vocab=0.9593 token=1.0957 | 6.10 steps/s | VRAM=26.11 Go [2026-07-17 07:57:24] step 225050/1000000 | loss=2.3641 vocab=1.3009 token=1.6684 | 0.61 steps/s | VRAM=28.66 Go [2026-07-17 07:57:32] step 225100/1000000 | loss=2.3754 vocab=0.6005 token=1.3059 | 6.17 steps/s | VRAM=26.38 Go [2026-07-17 07:57:40] step 225150/1000000 | loss=2.3410 vocab=1.0192 token=1.3305 | 6.16 steps/s | VRAM=29.15 Go [2026-07-17 07:57:48] step 225200/1000000 | loss=2.3785 vocab=1.0052 token=1.2173 | 6.18 steps/s | VRAM=29.46 Go [2026-07-17 07:57:56] step 225250/1000000 | loss=2.4158 vocab=0.7424 token=1.1582 | 6.15 steps/s | VRAM=28.21 Go [2026-07-17 07:58:04] step 225300/1000000 | loss=2.4510 vocab=1.3148 token=1.3649 | 6.15 steps/s | VRAM=28.96 Go [2026-07-17 07:58:12] step 225350/1000000 | loss=2.3586 vocab=1.3427 token=1.8073 | 6.17 steps/s | VRAM=32.39 Go [2026-07-17 07:58:21] step 225400/1000000 | loss=2.3748 vocab=1.0922 token=1.7231 | 6.13 steps/s | VRAM=31.23 Go [2026-07-17 07:58:29] step 225450/1000000 | loss=2.4194 vocab=1.0837 token=1.3861 | 6.10 steps/s | VRAM=28.13 Go [2026-07-17 07:58:37] step 225500/1000000 | loss=2.3758 vocab=1.1819 token=0.9329 | 6.13 steps/s | VRAM=32.00 Go [2026-07-17 07:58:45] step 225550/1000000 | loss=2.3184 vocab=0.6042 token=1.4761 | 6.06 steps/s | VRAM=28.99 Go [2026-07-17 07:58:53] step 225600/1000000 | loss=2.3013 vocab=1.1299 token=0.8766 | 6.13 steps/s | VRAM=28.11 Go [2026-07-17 07:59:02] step 225650/1000000 | loss=2.2984 vocab=0.9372 token=1.2484 | 6.10 steps/s | VRAM=28.43 Go [2026-07-17 07:59:10] step 225700/1000000 | loss=2.3360 vocab=0.7621 token=1.4856 | 6.12 steps/s | VRAM=26.05 Go [2026-07-17 07:59:18] step 225750/1000000 | loss=2.4110 vocab=1.2522 token=1.3384 | 6.11 steps/s | VRAM=32.27 Go [2026-07-17 07:59:28] step 225800/1000000 | loss=2.3573 vocab=1.2767 token=1.4599 | 5.10 steps/s | VRAM=28.32 Go [2026-07-17 07:59:36] step 225850/1000000 | loss=2.4117 vocab=0.8258 token=1.1360 | 6.13 steps/s | VRAM=32.82 Go [2026-07-17 07:59:44] step 225900/1000000 | loss=2.3726 vocab=1.1344 token=1.6681 | 6.13 steps/s | VRAM=32.43 Go [2026-07-17 07:59:52] step 225950/1000000 | loss=2.3730 vocab=0.7515 token=1.4384 | 6.10 steps/s | VRAM=28.61 Go [2026-07-17 08:00:00] step 226000/1000000 | loss=2.3911 vocab=1.3578 token=1.7514 | 6.13 steps/s | VRAM=25.55 Go [2026-07-17 08:00:09] step 226050/1000000 | loss=2.4026 vocab=1.0537 token=1.7527 | 6.12 steps/s | VRAM=31.26 Go [2026-07-17 08:00:17] step 226100/1000000 | loss=2.4057 vocab=1.1146 token=1.7513 | 6.08 steps/s | VRAM=29.28 Go [2026-07-17 08:00:25] step 226150/1000000 | loss=2.3557 vocab=1.4127 token=1.4885 | 6.12 steps/s | VRAM=29.36 Go [2026-07-17 08:00:33] step 226200/1000000 | loss=2.4108 vocab=1.0770 token=2.0015 | 6.11 steps/s | VRAM=28.73 Go [2026-07-17 08:00:41] step 226250/1000000 | loss=2.4250 vocab=1.7313 token=1.7520 | 6.12 steps/s | VRAM=31.84 Go [2026-07-17 08:00:50] step 226300/1000000 | loss=2.3187 vocab=0.5513 token=1.2694 | 6.10 steps/s | VRAM=31.57 Go [2026-07-17 08:00:58] step 226350/1000000 | loss=2.3475 vocab=0.8244 token=1.2890 | 6.07 steps/s | VRAM=27.10 Go [2026-07-17 08:01:06] step 226400/1000000 | loss=2.3739 vocab=0.8122 token=1.2708 | 6.14 steps/s | VRAM=28.03 Go [2026-07-17 08:01:14] step 226450/1000000 | loss=2.3321 vocab=0.9205 token=1.2619 | 6.11 steps/s | VRAM=28.92 Go [2026-07-17 08:01:22] step 226500/1000000 | loss=2.2705 vocab=1.0047 token=1.5132 | 6.11 steps/s | VRAM=32.04 Go [2026-07-17 08:01:31] step 226550/1000000 | loss=2.4204 vocab=0.6510 token=1.2302 | 6.07 steps/s | VRAM=25.90 Go [2026-07-17 08:01:39] step 226600/1000000 | loss=2.3919 vocab=1.1179 token=2.0094 | 6.10 steps/s | VRAM=31.73 Go [2026-07-17 08:01:47] step 226650/1000000 | loss=2.3318 vocab=0.7457 token=1.3537 | 6.07 steps/s | VRAM=26.26 Go [2026-07-17 08:01:55] step 226700/1000000 | loss=2.3992 vocab=0.7169 token=1.1107 | 6.10 steps/s | VRAM=32.30 Go [2026-07-17 08:02:03] step 226750/1000000 | loss=2.4375 vocab=1.5357 token=1.5963 | 6.14 steps/s | VRAM=24.34 Go [2026-07-17 08:02:11] step 226800/1000000 | loss=2.3714 vocab=1.2162 token=1.3152 | 6.12 steps/s | VRAM=26.69 Go [2026-07-17 08:02:20] step 226850/1000000 | loss=2.4288 vocab=0.9797 token=1.5131 | 6.04 steps/s | VRAM=29.00 Go [2026-07-17 08:02:28] step 226900/1000000 | loss=2.3524 vocab=0.8436 token=1.4101 | 6.11 steps/s | VRAM=24.52 Go [2026-07-17 08:02:36] step 226950/1000000 | loss=2.5111 vocab=1.4311 token=1.5759 | 6.14 steps/s | VRAM=24.93 Go [2026-07-17 08:02:44] step 227000/1000000 | loss=2.4389 vocab=0.6880 token=1.4717 | 6.11 steps/s | VRAM=28.49 Go [2026-07-17 08:02:52] step 227050/1000000 | loss=2.3506 vocab=1.1160 token=1.4378 | 6.10 steps/s | VRAM=28.64 Go [2026-07-17 08:03:01] step 227100/1000000 | loss=2.5015 vocab=0.7967 token=1.0922 | 6.08 steps/s | VRAM=31.86 Go [2026-07-17 08:03:09] step 227150/1000000 | loss=2.3335 vocab=0.8519 token=1.4697 | 6.08 steps/s | VRAM=26.49 Go [2026-07-17 08:03:17] step 227200/1000000 | loss=2.4045 vocab=0.7124 token=1.0586 | 6.10 steps/s | VRAM=25.69 Go [2026-07-17 08:03:25] step 227250/1000000 | loss=2.2302 vocab=0.9974 token=1.1282 | 6.10 steps/s | VRAM=32.71 Go [2026-07-17 08:03:33] step 227300/1000000 | loss=2.4376 vocab=1.4145 token=1.7091 | 6.11 steps/s | VRAM=27.69 Go [2026-07-17 08:03:42] step 227350/1000000 | loss=2.3701 vocab=0.9043 token=1.7624 | 6.12 steps/s | VRAM=31.35 Go [2026-07-17 08:03:50] step 227400/1000000 | loss=2.3738 vocab=0.9726 token=1.2422 | 6.08 steps/s | VRAM=28.96 Go [2026-07-17 08:03:58] step 227450/1000000 | loss=2.3732 vocab=0.7388 token=1.3710 | 6.09 steps/s | VRAM=28.48 Go [2026-07-17 08:04:06] step 227500/1000000 | loss=2.3930 vocab=1.0154 token=1.5253 | 6.09 steps/s | VRAM=28.80 Go [2026-07-17 08:04:15] step 227550/1000000 | loss=2.4075 vocab=0.9829 token=1.4058 | 6.10 steps/s | VRAM=32.34 Go [2026-07-17 08:04:23] step 227600/1000000 | loss=2.4055 vocab=1.1236 token=1.3944 | 6.06 steps/s | VRAM=27.85 Go [2026-07-17 08:04:31] step 227650/1000000 | loss=2.4267 vocab=0.8229 token=1.4689 | 6.11 steps/s | VRAM=28.35 Go [2026-07-17 08:04:39] step 227700/1000000 | loss=2.3309 vocab=0.7038 token=1.4373 | 6.11 steps/s | VRAM=25.54 Go [2026-07-17 08:04:47] step 227750/1000000 | loss=2.2772 vocab=0.9071 token=1.5409 | 6.11 steps/s | VRAM=26.12 Go [2026-07-17 08:04:56] step 227800/1000000 | loss=2.3604 vocab=1.1482 token=1.4579 | 6.10 steps/s | VRAM=28.44 Go [2026-07-17 08:05:04] step 227850/1000000 | loss=2.3787 vocab=0.8441 token=1.4987 | 6.11 steps/s | VRAM=29.34 Go [2026-07-17 08:05:12] step 227900/1000000 | loss=2.1887 vocab=0.6292 token=1.0328 | 6.12 steps/s | VRAM=27.68 Go [2026-07-17 08:05:20] step 227950/1000000 | loss=2.3766 vocab=1.1140 token=1.2768 | 6.14 steps/s | VRAM=26.57 Go [2026-07-17 08:05:28] step 228000/1000000 | loss=2.3332 vocab=1.0718 token=1.2589 | 6.12 steps/s | VRAM=28.43 Go [2026-07-17 08:05:36] step 228050/1000000 | loss=2.3356 vocab=1.0084 token=1.4658 | 6.09 steps/s | VRAM=28.37 Go [2026-07-17 08:05:45] step 228100/1000000 | loss=2.4850 vocab=0.8786 token=1.4662 | 6.12 steps/s | VRAM=28.71 Go [2026-07-17 08:05:53] step 228150/1000000 | loss=2.3415 vocab=1.1746 token=1.6407 | 6.12 steps/s | VRAM=26.08 Go [2026-07-17 08:06:01] step 228200/1000000 | loss=2.2632 vocab=0.7034 token=1.2268 | 6.12 steps/s | VRAM=28.46 Go [2026-07-17 08:06:09] step 228250/1000000 | loss=2.2771 vocab=0.8973 token=1.5456 | 6.10 steps/s | VRAM=26.16 Go [2026-07-17 08:06:17] step 228300/1000000 | loss=2.3897 vocab=0.8785 token=1.4802 | 6.13 steps/s | VRAM=27.84 Go [2026-07-17 08:06:25] step 228350/1000000 | loss=2.2454 vocab=1.2524 token=1.3081 | 6.13 steps/s | VRAM=32.77 Go [2026-07-17 08:06:34] step 228400/1000000 | loss=2.5114 vocab=0.8958 token=1.2640 | 6.13 steps/s | VRAM=31.51 Go [2026-07-17 08:06:42] step 228450/1000000 | loss=2.4523 vocab=0.8099 token=1.2017 | 6.11 steps/s | VRAM=26.13 Go [2026-07-17 08:06:50] step 228500/1000000 | loss=2.2432 vocab=0.8101 token=1.4355 | 6.09 steps/s | VRAM=28.43 Go [2026-07-17 08:06:58] step 228550/1000000 | loss=2.4689 vocab=0.9985 token=1.7014 | 6.09 steps/s | VRAM=29.02 Go [2026-07-17 08:07:06] step 228600/1000000 | loss=2.2818 vocab=0.8080 token=1.4663 | 6.12 steps/s | VRAM=31.89 Go [2026-07-17 08:07:15] step 228650/1000000 | loss=2.2995 vocab=1.2500 token=1.1341 | 6.11 steps/s | VRAM=28.76 Go [2026-07-17 08:07:23] step 228700/1000000 | loss=2.4186 vocab=1.4712 token=1.4652 | 6.11 steps/s | VRAM=28.76 Go [2026-07-17 08:07:31] step 228750/1000000 | loss=2.1849 vocab=1.2335 token=1.6004 | 6.11 steps/s | VRAM=31.26 Go [2026-07-17 08:07:39] step 228800/1000000 | loss=2.3200 vocab=1.3080 token=1.6659 | 6.10 steps/s | VRAM=26.70 Go [2026-07-17 08:07:47] step 228850/1000000 | loss=2.4736 vocab=0.6855 token=1.0485 | 6.10 steps/s | VRAM=28.21 Go [2026-07-17 08:07:55] step 228900/1000000 | loss=2.3028 vocab=0.8489 token=1.3122 | 6.09 steps/s | VRAM=28.76 Go [2026-07-17 08:08:04] step 228950/1000000 | loss=2.4461 vocab=0.9444 token=1.5236 | 6.11 steps/s | VRAM=26.15 Go [2026-07-17 08:08:12] step 229000/1000000 | loss=2.3462 vocab=1.7229 token=1.9028 | 6.11 steps/s | VRAM=26.04 Go [2026-07-17 08:08:20] step 229050/1000000 | loss=2.3087 vocab=1.0082 token=1.1303 | 6.13 steps/s | VRAM=26.22 Go [2026-07-17 08:08:28] step 229100/1000000 | loss=2.4024 vocab=0.9038 token=1.4272 | 6.14 steps/s | VRAM=26.04 Go [2026-07-17 08:08:36] step 229150/1000000 | loss=2.3614 vocab=0.8775 token=1.2777 | 6.11 steps/s | VRAM=27.79 Go [2026-07-17 08:08:45] step 229200/1000000 | loss=2.3290 vocab=0.6384 token=1.2705 | 6.10 steps/s | VRAM=28.66 Go [2026-07-17 08:08:53] step 229250/1000000 | loss=2.3042 vocab=0.7124 token=1.4229 | 6.09 steps/s | VRAM=25.94 Go [2026-07-17 08:09:01] step 229300/1000000 | loss=2.3802 vocab=0.5826 token=1.1659 | 6.13 steps/s | VRAM=31.39 Go [2026-07-17 08:09:09] step 229350/1000000 | loss=2.3543 vocab=1.0539 token=1.8087 | 6.11 steps/s | VRAM=28.40 Go [2026-07-17 08:09:17] step 229400/1000000 | loss=2.2773 vocab=1.1007 token=1.9048 | 6.13 steps/s | VRAM=32.82 Go [2026-07-17 08:09:25] step 229450/1000000 | loss=2.4510 vocab=1.1830 token=1.7336 | 6.13 steps/s | VRAM=28.08 Go [2026-07-17 08:09:34] step 229500/1000000 | loss=2.4364 vocab=1.2411 token=1.2548 | 6.12 steps/s | VRAM=29.40 Go [2026-07-17 08:09:42] step 229550/1000000 | loss=2.2918 vocab=0.8881 token=1.3136 | 6.12 steps/s | VRAM=32.32 Go [2026-07-17 08:09:50] step 229600/1000000 | loss=2.3767 vocab=1.0785 token=1.5662 | 6.14 steps/s | VRAM=28.80 Go [2026-07-17 08:09:58] step 229650/1000000 | loss=2.3938 vocab=0.8964 token=1.5574 | 6.11 steps/s | VRAM=28.83 Go [2026-07-17 08:10:06] step 229700/1000000 | loss=2.3609 vocab=0.5887 token=1.4726 | 6.11 steps/s | VRAM=29.41 Go [2026-07-17 08:10:14] step 229750/1000000 | loss=2.4126 vocab=0.8213 token=1.3821 | 6.09 steps/s | VRAM=26.21 Go [2026-07-17 08:10:23] step 229800/1000000 | loss=2.3818 vocab=0.8886 token=1.4425 | 6.13 steps/s | VRAM=28.29 Go [2026-07-17 08:10:31] step 229850/1000000 | loss=2.3944 vocab=1.0904 token=1.2522 | 6.09 steps/s | VRAM=27.04 Go [2026-07-17 08:10:39] step 229900/1000000 | loss=2.3315 vocab=0.5390 token=0.8236 | 6.13 steps/s | VRAM=28.50 Go [2026-07-17 08:10:47] step 229950/1000000 | loss=2.3433 vocab=0.7157 token=1.3823 | 6.12 steps/s | VRAM=27.80 Go [2026-07-17 08:10:55] step 230000/1000000 | loss=2.3571 vocab=1.2902 token=1.4493 | 6.15 steps/s | VRAM=29.29 Go [2026-07-17 08:12:17] step 230050/1000000 | loss=2.4138 vocab=0.7889 token=1.5981 | 0.61 steps/s | VRAM=28.87 Go [2026-07-17 08:12:25] step 230100/1000000 | loss=2.3951 vocab=0.9528 token=1.6640 | 6.13 steps/s | VRAM=29.97 Go [2026-07-17 08:12:33] step 230150/1000000 | loss=2.2514 vocab=0.7443 token=1.1658 | 6.16 steps/s | VRAM=31.77 Go [2026-07-17 08:12:41] step 230200/1000000 | loss=2.3437 vocab=0.9825 token=1.3694 | 6.18 steps/s | VRAM=25.30 Go [2026-07-17 08:12:49] step 230250/1000000 | loss=2.4269 vocab=1.1192 token=1.4904 | 6.14 steps/s | VRAM=24.65 Go [2026-07-17 08:12:58] step 230300/1000000 | loss=2.3541 vocab=1.0006 token=1.2115 | 6.10 steps/s | VRAM=26.44 Go [2026-07-17 08:13:06] step 230350/1000000 | loss=2.4322 vocab=1.1692 token=1.1579 | 6.12 steps/s | VRAM=26.51 Go [2026-07-17 08:13:14] step 230400/1000000 | loss=2.3045 vocab=1.2719 token=1.5328 | 6.14 steps/s | VRAM=32.58 Go [2026-07-17 08:13:22] step 230450/1000000 | loss=2.3223 vocab=0.8439 token=1.2104 | 6.13 steps/s | VRAM=28.86 Go [2026-07-17 08:13:30] step 230500/1000000 | loss=2.3561 vocab=0.7123 token=1.3928 | 6.11 steps/s | VRAM=29.03 Go [2026-07-17 08:13:38] step 230550/1000000 | loss=2.3378 vocab=1.1355 token=1.4891 | 6.12 steps/s | VRAM=28.24 Go [2026-07-17 08:13:47] step 230600/1000000 | loss=2.3216 vocab=1.2680 token=1.4771 | 6.10 steps/s | VRAM=28.59 Go [2026-07-17 08:13:55] step 230650/1000000 | loss=2.1550 vocab=0.9253 token=1.4154 | 6.11 steps/s | VRAM=28.88 Go [2026-07-17 08:14:03] step 230700/1000000 | loss=2.4386 vocab=0.8378 token=1.5099 | 6.10 steps/s | VRAM=32.25 Go [2026-07-17 08:14:11] step 230750/1000000 | loss=2.3585 vocab=0.5397 token=1.0473 | 6.10 steps/s | VRAM=29.78 Go [2026-07-17 08:14:19] step 230800/1000000 | loss=2.4617 vocab=0.9374 token=1.5805 | 6.09 steps/s | VRAM=28.87 Go [2026-07-17 08:14:28] step 230850/1000000 | loss=2.3776 vocab=0.9229 token=1.2642 | 6.10 steps/s | VRAM=29.24 Go [2026-07-17 08:14:36] step 230900/1000000 | loss=2.3237 vocab=1.2394 token=1.5632 | 6.08 steps/s | VRAM=31.43 Go [2026-07-17 08:14:44] step 230950/1000000 | loss=2.3262 vocab=1.0423 token=1.3907 | 6.11 steps/s | VRAM=32.28 Go [2026-07-17 08:14:52] step 231000/1000000 | loss=2.3382 vocab=1.0622 token=1.4333 | 6.11 steps/s | VRAM=27.96 Go [2026-07-17 08:15:00] step 231050/1000000 | loss=2.4013 vocab=1.2146 token=1.0581 | 6.11 steps/s | VRAM=28.73 Go [2026-07-17 08:15:09] step 231100/1000000 | loss=2.2715 vocab=1.4400 token=1.8445 | 6.12 steps/s | VRAM=26.43 Go [2026-07-17 08:15:17] step 231150/1000000 | loss=2.4824 vocab=0.8634 token=1.6396 | 6.14 steps/s | VRAM=27.94 Go [2026-07-17 08:15:25] step 231200/1000000 | loss=2.3913 vocab=1.0080 token=1.2020 | 6.12 steps/s | VRAM=26.04 Go [2026-07-17 08:15:33] step 231250/1000000 | loss=2.4511 vocab=0.8766 token=1.5185 | 6.11 steps/s | VRAM=28.85 Go [2026-07-17 08:15:41] step 231300/1000000 | loss=2.3737 vocab=0.9164 token=1.3376 | 6.13 steps/s | VRAM=31.59 Go [2026-07-17 08:15:49] step 231350/1000000 | loss=2.4053 vocab=1.5799 token=1.7018 | 6.09 steps/s | VRAM=27.00 Go [2026-07-17 08:15:58] step 231400/1000000 | loss=2.3763 vocab=0.6715 token=0.9917 | 6.13 steps/s | VRAM=31.31 Go [2026-07-17 08:16:06] step 231450/1000000 | loss=2.3843 vocab=0.5307 token=1.0815 | 6.11 steps/s | VRAM=25.10 Go [2026-07-17 08:16:14] step 231500/1000000 | loss=2.3359 vocab=0.6772 token=1.3357 | 6.13 steps/s | VRAM=32.05 Go [2026-07-17 08:16:22] step 231550/1000000 | loss=2.4263 vocab=0.6418 token=1.4037 | 6.10 steps/s | VRAM=26.75 Go [2026-07-17 08:16:30] step 231600/1000000 | loss=2.4492 vocab=1.0517 token=1.5892 | 6.13 steps/s | VRAM=30.19 Go [2026-07-17 08:16:39] step 231650/1000000 | loss=2.3990 vocab=0.8837 token=1.5393 | 6.11 steps/s | VRAM=28.17 Go [2026-07-17 08:16:47] step 231700/1000000 | loss=2.4260 vocab=0.8846 token=1.5346 | 6.10 steps/s | VRAM=26.91 Go [2026-07-17 08:16:55] step 231750/1000000 | loss=2.4078 vocab=0.7890 token=1.4043 | 6.12 steps/s | VRAM=32.10 Go [2026-07-17 08:17:03] step 231800/1000000 | loss=2.3264 vocab=0.9803 token=1.2072 | 6.10 steps/s | VRAM=29.18 Go [2026-07-17 08:17:11] step 231850/1000000 | loss=2.3499 vocab=0.5850 token=1.2884 | 6.07 steps/s | VRAM=26.16 Go [2026-07-17 08:17:20] step 231900/1000000 | loss=2.3861 vocab=1.0589 token=1.6730 | 6.12 steps/s | VRAM=32.16 Go [2026-07-17 08:17:28] step 231950/1000000 | loss=2.4193 vocab=0.8966 token=1.1887 | 6.09 steps/s | VRAM=28.15 Go [2026-07-17 08:17:36] step 232000/1000000 | loss=2.3501 vocab=0.6879 token=1.0152 | 6.06 steps/s | VRAM=26.45 Go [2026-07-17 08:17:44] step 232050/1000000 | loss=2.4035 vocab=1.2725 token=1.2895 | 6.11 steps/s | VRAM=29.63 Go [2026-07-17 08:17:52] step 232100/1000000 | loss=2.3800 vocab=1.0807 token=1.4373 | 6.13 steps/s | VRAM=31.74 Go [2026-07-17 08:18:00] step 232150/1000000 | loss=2.2670 vocab=0.5899 token=1.4780 | 6.12 steps/s | VRAM=32.19 Go [2026-07-17 08:18:09] step 232200/1000000 | loss=2.3970 vocab=0.7393 token=1.4067 | 6.10 steps/s | VRAM=28.40 Go [2026-07-17 08:18:17] step 232250/1000000 | loss=2.2837 vocab=0.7927 token=1.1850 | 6.13 steps/s | VRAM=27.97 Go [2026-07-17 08:18:25] step 232300/1000000 | loss=2.5604 vocab=1.1330 token=1.6811 | 6.10 steps/s | VRAM=25.88 Go [2026-07-17 08:18:33] step 232350/1000000 | loss=2.4774 vocab=1.7686 token=1.4826 | 6.12 steps/s | VRAM=25.28 Go [2026-07-17 08:18:41] step 232400/1000000 | loss=2.4047 vocab=0.8878 token=1.3456 | 6.11 steps/s | VRAM=26.24 Go [2026-07-17 08:18:49] step 232450/1000000 | loss=2.3141 vocab=1.0208 token=1.6285 | 6.15 steps/s | VRAM=30.97 Go [2026-07-17 08:18:58] step 232500/1000000 | loss=2.2701 vocab=1.0573 token=1.5035 | 6.11 steps/s | VRAM=26.80 Go [2026-07-17 08:19:06] step 232550/1000000 | loss=2.3406 vocab=0.7647 token=1.2678 | 6.13 steps/s | VRAM=28.60 Go [2026-07-17 08:19:14] step 232600/1000000 | loss=2.3510 vocab=0.8591 token=1.6291 | 6.13 steps/s | VRAM=27.94 Go [2026-07-17 08:19:22] step 232650/1000000 | loss=2.3851 vocab=0.9916 token=1.3432 | 6.11 steps/s | VRAM=32.86 Go [2026-07-17 08:19:30] step 232700/1000000 | loss=2.4522 vocab=1.3469 token=1.4986 | 6.12 steps/s | VRAM=26.03 Go [2026-07-17 08:19:39] step 232750/1000000 | loss=2.3303 vocab=0.8870 token=1.4457 | 6.12 steps/s | VRAM=28.57 Go [2026-07-17 08:19:47] step 232800/1000000 | loss=2.3900 vocab=0.6886 token=1.5035 | 6.13 steps/s | VRAM=31.64 Go [2026-07-17 08:19:55] step 232850/1000000 | loss=2.3704 vocab=1.0407 token=1.4215 | 6.12 steps/s | VRAM=28.96 Go [2026-07-17 08:20:03] step 232900/1000000 | loss=2.3550 vocab=0.7720 token=1.2817 | 6.15 steps/s | VRAM=31.96 Go [2026-07-17 08:20:11] step 232950/1000000 | loss=2.4214 vocab=0.9834 token=1.1780 | 6.12 steps/s | VRAM=25.53 Go [2026-07-17 08:20:19] step 233000/1000000 | loss=2.3529 vocab=1.3218 token=1.3201 | 6.14 steps/s | VRAM=28.88 Go [2026-07-17 08:20:27] step 233050/1000000 | loss=2.3030 vocab=1.0554 token=1.3817 | 6.15 steps/s | VRAM=28.57 Go [2026-07-17 08:20:36] step 233100/1000000 | loss=2.3001 vocab=0.8911 token=1.3843 | 6.11 steps/s | VRAM=29.23 Go [2026-07-17 08:20:44] step 233150/1000000 | loss=2.2817 vocab=1.1077 token=1.2242 | 6.13 steps/s | VRAM=28.56 Go [2026-07-17 08:20:52] step 233200/1000000 | loss=2.4217 vocab=1.0645 token=1.3250 | 6.11 steps/s | VRAM=28.48 Go [2026-07-17 08:21:00] step 233250/1000000 | loss=2.3104 vocab=1.2286 token=1.5530 | 6.09 steps/s | VRAM=26.36 Go [2026-07-17 08:21:08] step 233300/1000000 | loss=2.3733 vocab=0.6375 token=1.5498 | 6.16 steps/s | VRAM=28.54 Go [2026-07-17 08:21:16] step 233350/1000000 | loss=2.3452 vocab=0.8198 token=1.2058 | 6.12 steps/s | VRAM=24.62 Go [2026-07-17 08:21:25] step 233400/1000000 | loss=2.3588 vocab=0.7842 token=1.4087 | 6.09 steps/s | VRAM=28.29 Go [2026-07-17 08:21:33] step 233450/1000000 | loss=2.4058 vocab=1.6225 token=1.7466 | 6.14 steps/s | VRAM=32.18 Go [2026-07-17 08:21:41] step 233500/1000000 | loss=2.3080 vocab=0.6367 token=1.2321 | 6.10 steps/s | VRAM=29.91 Go [2026-07-17 08:21:49] step 233550/1000000 | loss=2.3317 vocab=1.0794 token=1.3595 | 6.11 steps/s | VRAM=24.67 Go [2026-07-17 08:21:57] step 233600/1000000 | loss=2.3400 vocab=0.3878 token=1.0675 | 6.12 steps/s | VRAM=29.27 Go [2026-07-17 08:22:06] step 233650/1000000 | loss=2.2688 vocab=0.7021 token=1.2127 | 6.10 steps/s | VRAM=27.27 Go [2026-07-17 08:22:14] step 233700/1000000 | loss=2.3714 vocab=1.0220 token=1.2099 | 6.16 steps/s | VRAM=31.80 Go [2026-07-17 08:22:22] step 233750/1000000 | loss=2.2617 vocab=0.9051 token=1.2171 | 6.13 steps/s | VRAM=29.42 Go [2026-07-17 08:22:30] step 233800/1000000 | loss=2.3924 vocab=1.2123 token=1.6741 | 6.15 steps/s | VRAM=25.03 Go [2026-07-17 08:22:38] step 233850/1000000 | loss=2.4479 vocab=1.3069 token=1.6969 | 6.14 steps/s | VRAM=30.00 Go [2026-07-17 08:22:46] step 233900/1000000 | loss=2.4065 vocab=0.7320 token=1.6841 | 6.14 steps/s | VRAM=27.84 Go [2026-07-17 08:22:54] step 233950/1000000 | loss=2.4343 vocab=0.8682 token=1.4126 | 6.15 steps/s | VRAM=31.42 Go [2026-07-17 08:23:03] step 234000/1000000 | loss=2.3262 vocab=0.9613 token=1.0158 | 6.13 steps/s | VRAM=29.17 Go [2026-07-17 08:23:11] step 234050/1000000 | loss=2.4235 vocab=1.5579 token=1.2985 | 6.11 steps/s | VRAM=31.82 Go [2026-07-17 08:23:19] step 234100/1000000 | loss=2.4341 vocab=0.5951 token=1.5424 | 6.12 steps/s | VRAM=26.60 Go [2026-07-17 08:23:27] step 234150/1000000 | loss=2.3543 vocab=0.9769 token=1.3974 | 6.14 steps/s | VRAM=27.60 Go [2026-07-17 08:23:35] step 234200/1000000 | loss=2.4065 vocab=1.1869 token=1.5811 | 6.13 steps/s | VRAM=30.95 Go [2026-07-17 08:23:43] step 234250/1000000 | loss=2.4800 vocab=0.8917 token=1.2043 | 6.14 steps/s | VRAM=29.08 Go [2026-07-17 08:23:51] step 234300/1000000 | loss=2.4200 vocab=0.8861 token=1.1508 | 6.14 steps/s | VRAM=31.10 Go [2026-07-17 08:24:00] step 234350/1000000 | loss=2.3700 vocab=1.6185 token=1.2859 | 6.13 steps/s | VRAM=31.06 Go [2026-07-17 08:24:08] step 234400/1000000 | loss=2.3193 vocab=0.8732 token=2.0139 | 6.12 steps/s | VRAM=28.45 Go [2026-07-17 08:24:16] step 234450/1000000 | loss=2.3393 vocab=1.0013 token=1.1848 | 6.10 steps/s | VRAM=26.59 Go [2026-07-17 08:24:24] step 234500/1000000 | loss=2.3526 vocab=2.0525 token=1.7817 | 6.11 steps/s | VRAM=27.87 Go [2026-07-17 08:24:32] step 234550/1000000 | loss=2.3581 vocab=1.0667 token=1.4749 | 6.13 steps/s | VRAM=25.59 Go [2026-07-17 08:24:41] step 234600/1000000 | loss=2.4024 vocab=0.9023 token=1.3467 | 6.08 steps/s | VRAM=25.23 Go [2026-07-17 08:24:49] step 234650/1000000 | loss=2.3332 vocab=0.6219 token=1.2831 | 6.12 steps/s | VRAM=28.95 Go [2026-07-17 08:24:57] step 234700/1000000 | loss=2.3883 vocab=0.6777 token=1.4689 | 6.11 steps/s | VRAM=29.22 Go [2026-07-17 08:25:05] step 234750/1000000 | loss=2.4328 vocab=1.0010 token=1.6548 | 6.11 steps/s | VRAM=26.24 Go [2026-07-17 08:25:13] step 234800/1000000 | loss=2.4309 vocab=0.9389 token=1.4459 | 6.13 steps/s | VRAM=32.51 Go [2026-07-17 08:25:21] step 234850/1000000 | loss=2.3329 vocab=1.0842 token=1.3669 | 6.08 steps/s | VRAM=26.39 Go [2026-07-17 08:25:30] step 234900/1000000 | loss=2.3334 vocab=0.9306 token=1.5276 | 6.10 steps/s | VRAM=28.03 Go [2026-07-17 08:25:38] step 234950/1000000 | loss=2.3379 vocab=1.2675 token=1.4841 | 6.06 steps/s | VRAM=28.90 Go [2026-07-17 08:25:46] step 235000/1000000 | loss=2.3689 vocab=0.8803 token=1.1052 | 6.14 steps/s | VRAM=31.23 Go [2026-07-17 08:27:16] step 235050/1000000 | loss=2.3439 vocab=0.7565 token=1.3342 | 0.55 steps/s | VRAM=28.45 Go [2026-07-17 08:27:25] step 235100/1000000 | loss=2.4481 vocab=1.2644 token=1.7016 | 6.18 steps/s | VRAM=28.40 Go [2026-07-17 08:27:33] step 235150/1000000 | loss=2.3775 vocab=0.8613 token=1.4089 | 6.19 steps/s | VRAM=32.75 Go [2026-07-17 08:27:41] step 235200/1000000 | loss=2.4140 vocab=0.7216 token=1.1739 | 6.18 steps/s | VRAM=32.20 Go [2026-07-17 08:27:49] step 235250/1000000 | loss=2.4305 vocab=0.8928 token=1.4782 | 6.14 steps/s | VRAM=32.92 Go [2026-07-17 08:27:57] step 235300/1000000 | loss=2.3704 vocab=1.4178 token=1.3185 | 6.14 steps/s | VRAM=31.60 Go [2026-07-17 08:28:05] step 235350/1000000 | loss=2.2992 vocab=0.8709 token=1.3897 | 6.14 steps/s | VRAM=26.09 Go [2026-07-17 08:28:13] step 235400/1000000 | loss=2.3195 vocab=0.8532 token=1.2618 | 6.15 steps/s | VRAM=31.60 Go [2026-07-17 08:28:21] step 235450/1000000 | loss=2.4387 vocab=0.9893 token=1.4702 | 6.14 steps/s | VRAM=28.34 Go [2026-07-17 08:28:30] step 235500/1000000 | loss=2.4548 vocab=0.9219 token=1.3997 | 6.12 steps/s | VRAM=28.35 Go [2026-07-17 08:28:38] step 235550/1000000 | loss=2.3080 vocab=0.5302 token=0.8570 | 6.13 steps/s | VRAM=31.95 Go [2026-07-17 08:28:46] step 235600/1000000 | loss=2.4136 vocab=1.5797 token=1.1275 | 6.06 steps/s | VRAM=28.40 Go [2026-07-17 08:28:54] step 235650/1000000 | loss=2.4472 vocab=1.2162 token=1.2748 | 6.07 steps/s | VRAM=28.74 Go [2026-07-17 08:29:02] step 235700/1000000 | loss=2.3096 vocab=0.8031 token=1.3152 | 6.10 steps/s | VRAM=31.64 Go [2026-07-17 08:29:11] step 235750/1000000 | loss=2.3173 vocab=1.0120 token=1.6708 | 6.11 steps/s | VRAM=29.44 Go [2026-07-17 08:29:19] step 235800/1000000 | loss=2.3765 vocab=0.8392 token=1.6056 | 6.12 steps/s | VRAM=28.73 Go [2026-07-17 08:29:27] step 235850/1000000 | loss=2.3344 vocab=0.5812 token=1.6733 | 6.10 steps/s | VRAM=32.56 Go [2026-07-17 08:29:35] step 235900/1000000 | loss=2.3820 vocab=1.0252 token=1.4172 | 6.07 steps/s | VRAM=26.28 Go [2026-07-17 08:29:43] step 235950/1000000 | loss=2.4223 vocab=0.8351 token=1.3634 | 6.07 steps/s | VRAM=32.38 Go [2026-07-17 08:29:52] step 236000/1000000 | loss=2.3671 vocab=0.9917 token=1.5020 | 6.11 steps/s | VRAM=28.82 Go [2026-07-17 08:30:00] step 236050/1000000 | loss=2.4558 vocab=0.6477 token=1.1831 | 6.15 steps/s | VRAM=28.24 Go [2026-07-17 08:30:08] step 236100/1000000 | loss=2.3762 vocab=0.7500 token=1.5210 | 6.11 steps/s | VRAM=28.61 Go [2026-07-17 08:30:16] step 236150/1000000 | loss=2.3422 vocab=0.8093 token=1.4550 | 6.14 steps/s | VRAM=31.97 Go [2026-07-17 08:30:24] step 236200/1000000 | loss=2.2855 vocab=0.8484 token=0.8690 | 6.09 steps/s | VRAM=29.53 Go [2026-07-17 08:30:33] step 236250/1000000 | loss=2.2666 vocab=0.7172 token=1.6686 | 6.12 steps/s | VRAM=32.25 Go [2026-07-17 08:30:41] step 236300/1000000 | loss=2.3929 vocab=0.7177 token=1.4407 | 6.12 steps/s | VRAM=28.32 Go [2026-07-17 08:30:49] step 236350/1000000 | loss=2.3925 vocab=1.4225 token=1.5718 | 6.12 steps/s | VRAM=31.10 Go [2026-07-17 08:30:57] step 236400/1000000 | loss=2.3347 vocab=0.6481 token=1.7415 | 6.10 steps/s | VRAM=29.68 Go [2026-07-17 08:31:05] step 236450/1000000 | loss=2.3349 vocab=0.5743 token=1.3458 | 6.12 steps/s | VRAM=29.07 Go [2026-07-17 08:31:13] step 236500/1000000 | loss=2.3839 vocab=1.4899 token=1.3648 | 6.11 steps/s | VRAM=28.52 Go [2026-07-17 08:31:22] step 236550/1000000 | loss=2.3129 vocab=2.0501 token=1.3821 | 6.07 steps/s | VRAM=26.46 Go [2026-07-17 08:31:30] step 236600/1000000 | loss=2.4167 vocab=0.3975 token=1.1748 | 6.13 steps/s | VRAM=25.70 Go [2026-07-17 08:31:38] step 236650/1000000 | loss=2.4356 vocab=1.0292 token=1.2949 | 6.14 steps/s | VRAM=29.18 Go [2026-07-17 08:31:46] step 236700/1000000 | loss=2.2965 vocab=1.2163 token=1.7110 | 6.08 steps/s | VRAM=26.01 Go [2026-07-17 08:31:54] step 236750/1000000 | loss=2.4743 vocab=0.8290 token=1.3811 | 6.14 steps/s | VRAM=32.02 Go [2026-07-17 08:32:02] step 236800/1000000 | loss=2.3304 vocab=1.2926 token=1.3829 | 6.14 steps/s | VRAM=32.36 Go [2026-07-17 08:32:11] step 236850/1000000 | loss=2.3435 vocab=1.5271 token=1.4877 | 6.12 steps/s | VRAM=28.45 Go [2026-07-17 08:32:19] step 236900/1000000 | loss=2.3583 vocab=1.3542 token=1.7053 | 6.13 steps/s | VRAM=28.53 Go [2026-07-17 08:32:27] step 236950/1000000 | loss=2.3504 vocab=0.6530 token=1.2209 | 6.10 steps/s | VRAM=32.07 Go [2026-07-17 08:32:35] step 237000/1000000 | loss=2.3736 vocab=0.8281 token=1.1761 | 6.13 steps/s | VRAM=25.96 Go [2026-07-17 08:32:43] step 237050/1000000 | loss=2.4605 vocab=1.1629 token=1.7678 | 6.11 steps/s | VRAM=28.92 Go [2026-07-17 08:32:52] step 237100/1000000 | loss=2.4337 vocab=0.6198 token=1.4481 | 6.10 steps/s | VRAM=28.98 Go [2026-07-17 08:33:00] step 237150/1000000 | loss=2.3414 vocab=1.2273 token=1.4417 | 6.11 steps/s | VRAM=32.79 Go [2026-07-17 08:33:08] step 237200/1000000 | loss=2.3533 vocab=0.9096 token=1.4076 | 6.07 steps/s | VRAM=29.06 Go [2026-07-17 08:33:16] step 237250/1000000 | loss=2.3228 vocab=0.8561 token=1.6729 | 6.10 steps/s | VRAM=28.49 Go [2026-07-17 08:33:24] step 237300/1000000 | loss=2.3159 vocab=1.2622 token=1.9081 | 6.10 steps/s | VRAM=28.51 Go [2026-07-17 08:33:32] step 237350/1000000 | loss=2.3887 vocab=0.6524 token=1.0177 | 6.13 steps/s | VRAM=31.20 Go [2026-07-17 08:33:41] step 237400/1000000 | loss=2.3105 vocab=1.5353 token=1.4640 | 6.11 steps/s | VRAM=32.27 Go [2026-07-17 08:33:49] step 237450/1000000 | loss=2.3536 vocab=0.6304 token=1.4214 | 6.12 steps/s | VRAM=31.65 Go [2026-07-17 08:33:57] step 237500/1000000 | loss=2.3223 vocab=0.9515 token=1.2587 | 6.11 steps/s | VRAM=28.45 Go [2026-07-17 08:34:05] step 237550/1000000 | loss=2.2453 vocab=0.5242 token=1.3452 | 6.12 steps/s | VRAM=26.13 Go [2026-07-17 08:34:13] step 237600/1000000 | loss=2.3207 vocab=0.9431 token=1.0976 | 6.13 steps/s | VRAM=29.02 Go [2026-07-17 08:34:21] step 237650/1000000 | loss=2.3317 vocab=0.8494 token=1.5827 | 6.13 steps/s | VRAM=31.51 Go [2026-07-17 08:34:30] step 237700/1000000 | loss=2.2530 vocab=1.1525 token=1.4989 | 6.10 steps/s | VRAM=27.85 Go [2026-07-17 08:34:38] step 237750/1000000 | loss=2.4236 vocab=1.1083 token=1.2690 | 6.12 steps/s | VRAM=31.67 Go [2026-07-17 08:34:46] step 237800/1000000 | loss=2.3930 vocab=1.0814 token=1.4730 | 6.09 steps/s | VRAM=26.84 Go [2026-07-17 08:34:54] step 237850/1000000 | loss=2.4182 vocab=0.9911 token=1.2593 | 6.11 steps/s | VRAM=28.64 Go [2026-07-17 08:35:02] step 237900/1000000 | loss=2.3137 vocab=1.1635 token=1.1689 | 6.11 steps/s | VRAM=26.60 Go [2026-07-17 08:35:11] step 237950/1000000 | loss=2.3396 vocab=0.7214 token=1.5399 | 6.11 steps/s | VRAM=29.07 Go [2026-07-17 08:35:19] step 238000/1000000 | loss=2.3648 vocab=0.5041 token=1.1509 | 6.14 steps/s | VRAM=32.28 Go [2026-07-17 08:35:27] step 238050/1000000 | loss=2.3378 vocab=1.1414 token=1.2322 | 6.09 steps/s | VRAM=32.57 Go [2026-07-17 08:35:35] step 238100/1000000 | loss=2.2745 vocab=1.1152 token=1.3718 | 6.14 steps/s | VRAM=31.56 Go [2026-07-17 08:35:43] step 238150/1000000 | loss=2.3722 vocab=0.6597 token=1.2429 | 6.14 steps/s | VRAM=32.38 Go [2026-07-17 08:35:51] step 238200/1000000 | loss=2.2731 vocab=0.7242 token=1.3505 | 6.11 steps/s | VRAM=26.30 Go [2026-07-17 08:36:00] step 238250/1000000 | loss=2.3616 vocab=0.6966 token=0.9770 | 6.12 steps/s | VRAM=28.80 Go [2026-07-17 08:36:08] step 238300/1000000 | loss=2.3772 vocab=0.7656 token=1.4432 | 6.07 steps/s | VRAM=27.26 Go [2026-07-17 08:36:16] step 238350/1000000 | loss=2.2539 vocab=0.7516 token=1.0933 | 6.11 steps/s | VRAM=31.79 Go [2026-07-17 08:36:24] step 238400/1000000 | loss=2.3862 vocab=0.4828 token=1.1202 | 6.14 steps/s | VRAM=28.93 Go [2026-07-17 08:36:32] step 238450/1000000 | loss=2.4417 vocab=1.0756 token=1.5169 | 6.13 steps/s | VRAM=28.92 Go [2026-07-17 08:36:40] step 238500/1000000 | loss=2.4243 vocab=1.2280 token=1.6334 | 6.12 steps/s | VRAM=28.51 Go [2026-07-17 08:36:49] step 238550/1000000 | loss=2.3163 vocab=0.5739 token=1.0799 | 6.11 steps/s | VRAM=27.27 Go [2026-07-17 08:36:57] step 238600/1000000 | loss=2.3474 vocab=0.8637 token=1.3422 | 6.14 steps/s | VRAM=28.91 Go [2026-07-17 08:37:05] step 238650/1000000 | loss=2.2625 vocab=0.8019 token=1.4036 | 6.13 steps/s | VRAM=29.03 Go [2026-07-17 08:37:13] step 238700/1000000 | loss=2.4395 vocab=0.5617 token=1.4517 | 6.15 steps/s | VRAM=29.19 Go [2026-07-17 08:37:21] step 238750/1000000 | loss=2.2316 vocab=0.8860 token=1.1543 | 6.12 steps/s | VRAM=24.49 Go [2026-07-17 08:37:29] step 238800/1000000 | loss=2.4298 vocab=0.8193 token=1.2615 | 6.13 steps/s | VRAM=28.86 Go [2026-07-17 08:37:38] step 238850/1000000 | loss=2.3898 vocab=0.6544 token=0.9995 | 6.12 steps/s | VRAM=32.52 Go [2026-07-17 08:37:46] step 238900/1000000 | loss=2.3828 vocab=1.0436 token=1.1904 | 6.15 steps/s | VRAM=28.71 Go [2026-07-17 08:37:54] step 238950/1000000 | loss=2.3325 vocab=0.8899 token=1.3161 | 6.10 steps/s | VRAM=32.56 Go [2026-07-17 08:38:02] step 239000/1000000 | loss=2.2768 vocab=1.2371 token=1.2784 | 6.12 steps/s | VRAM=29.29 Go [2026-07-17 08:38:10] step 239050/1000000 | loss=2.3216 vocab=0.9016 token=1.5513 | 6.11 steps/s | VRAM=28.31 Go [2026-07-17 08:38:18] step 239100/1000000 | loss=2.3505 vocab=0.7391 token=1.4280 | 6.13 steps/s | VRAM=28.72 Go [2026-07-17 08:38:27] step 239150/1000000 | loss=2.4323 vocab=1.2367 token=1.5136 | 6.11 steps/s | VRAM=29.14 Go [2026-07-17 08:38:35] step 239200/1000000 | loss=2.4051 vocab=1.3877 token=1.2304 | 6.11 steps/s | VRAM=26.21 Go [2026-07-17 08:38:43] step 239250/1000000 | loss=2.3334 vocab=1.2679 token=1.8770 | 6.12 steps/s | VRAM=27.42 Go [2026-07-17 08:38:51] step 239300/1000000 | loss=2.3664 vocab=1.0916 token=1.4637 | 6.14 steps/s | VRAM=32.26 Go [2026-07-17 08:38:59] step 239350/1000000 | loss=2.2846 vocab=0.9666 token=0.9255 | 6.12 steps/s | VRAM=27.98 Go [2026-07-17 08:39:07] step 239400/1000000 | loss=2.4238 vocab=0.8790 token=1.2853 | 6.14 steps/s | VRAM=28.16 Go [2026-07-17 08:39:16] step 239450/1000000 | loss=2.3726 vocab=1.2116 token=1.2436 | 6.14 steps/s | VRAM=31.79 Go [2026-07-17 08:39:24] step 239500/1000000 | loss=2.3573 vocab=0.9941 token=1.1816 | 6.15 steps/s | VRAM=29.02 Go [2026-07-17 08:39:32] step 239550/1000000 | loss=2.3713 vocab=0.5248 token=1.1837 | 6.12 steps/s | VRAM=29.60 Go [2026-07-17 08:39:40] step 239600/1000000 | loss=2.4252 vocab=0.8826 token=1.0316 | 6.13 steps/s | VRAM=28.84 Go [2026-07-17 08:39:48] step 239650/1000000 | loss=2.2830 vocab=1.0186 token=1.6853 | 6.14 steps/s | VRAM=29.28 Go [2026-07-17 08:39:56] step 239700/1000000 | loss=2.3121 vocab=0.8416 token=1.2119 | 6.13 steps/s | VRAM=25.95 Go [2026-07-17 08:40:04] step 239750/1000000 | loss=2.1683 vocab=1.0103 token=1.4590 | 6.16 steps/s | VRAM=32.40 Go [2026-07-17 08:40:13] step 239800/1000000 | loss=2.2686 vocab=0.5073 token=1.4439 | 6.10 steps/s | VRAM=29.91 Go [2026-07-17 08:40:21] step 239850/1000000 | loss=2.3817 vocab=1.1796 token=1.3972 | 6.10 steps/s | VRAM=28.72 Go [2026-07-17 08:40:29] step 239900/1000000 | loss=2.5335 vocab=1.3294 token=1.2769 | 6.10 steps/s | VRAM=26.13 Go [2026-07-17 08:40:37] step 239950/1000000 | loss=2.3850 vocab=1.3526 token=1.8412 | 6.16 steps/s | VRAM=28.96 Go [2026-07-17 08:40:45] step 240000/1000000 | loss=2.4068 vocab=0.6589 token=1.6328 | 6.13 steps/s | VRAM=28.22 Go [2026-07-17 08:42:07] step 240050/1000000 | loss=2.3548 vocab=0.4898 token=1.3738 | 0.61 steps/s | VRAM=31.51 Go [2026-07-17 08:42:15] step 240100/1000000 | loss=2.2715 vocab=1.1192 token=1.3533 | 6.20 steps/s | VRAM=27.92 Go [2026-07-17 08:42:23] step 240150/1000000 | loss=2.3536 vocab=1.0383 token=1.5498 | 6.15 steps/s | VRAM=32.66 Go [2026-07-17 08:42:31] step 240200/1000000 | loss=2.4604 vocab=1.0015 token=1.7502 | 6.18 steps/s | VRAM=28.60 Go [2026-07-17 08:42:40] step 240250/1000000 | loss=2.2748 vocab=1.1709 token=1.5888 | 6.15 steps/s | VRAM=26.47 Go [2026-07-17 08:42:48] step 240300/1000000 | loss=2.3973 vocab=0.8335 token=1.2273 | 6.17 steps/s | VRAM=29.09 Go [2026-07-17 08:42:56] step 240350/1000000 | loss=2.3759 vocab=0.9012 token=1.3743 | 6.18 steps/s | VRAM=30.09 Go [2026-07-17 08:43:04] step 240400/1000000 | loss=2.3811 vocab=1.3729 token=1.5204 | 6.15 steps/s | VRAM=25.18 Go [2026-07-17 08:43:12] step 240450/1000000 | loss=2.3220 vocab=1.2422 token=1.4243 | 6.14 steps/s | VRAM=27.85 Go [2026-07-17 08:43:20] step 240500/1000000 | loss=2.3295 vocab=1.0661 token=1.3414 | 6.18 steps/s | VRAM=31.65 Go [2026-07-17 08:43:28] step 240550/1000000 | loss=2.3441 vocab=0.7953 token=0.9328 | 6.10 steps/s | VRAM=30.39 Go [2026-07-17 08:43:36] step 240600/1000000 | loss=2.2775 vocab=1.2412 token=1.5815 | 6.15 steps/s | VRAM=31.98 Go [2026-07-17 08:43:45] step 240650/1000000 | loss=2.3492 vocab=0.6245 token=1.2534 | 6.16 steps/s | VRAM=32.20 Go [2026-07-17 08:43:53] step 240700/1000000 | loss=2.3280 vocab=1.0262 token=1.6324 | 6.13 steps/s | VRAM=29.79 Go [2026-07-17 08:44:01] step 240750/1000000 | loss=2.3274 vocab=0.6796 token=1.2528 | 6.13 steps/s | VRAM=25.73 Go [2026-07-17 08:44:09] step 240800/1000000 | loss=2.3040 vocab=0.8554 token=1.4636 | 6.12 steps/s | VRAM=29.14 Go [2026-07-17 08:44:17] step 240850/1000000 | loss=2.3564 vocab=0.7216 token=1.3070 | 6.14 steps/s | VRAM=29.43 Go [2026-07-17 08:44:25] step 240900/1000000 | loss=2.3692 vocab=0.7714 token=0.8538 | 6.13 steps/s | VRAM=27.98 Go [2026-07-17 08:44:34] step 240950/1000000 | loss=2.2661 vocab=1.1601 token=1.6853 | 6.11 steps/s | VRAM=32.72 Go [2026-07-17 08:44:42] step 241000/1000000 | loss=2.3256 vocab=0.9223 token=1.3803 | 6.14 steps/s | VRAM=28.01 Go [2026-07-17 08:44:50] step 241050/1000000 | loss=2.3597 vocab=0.7713 token=1.2366 | 6.13 steps/s | VRAM=25.58 Go [2026-07-17 08:44:58] step 241100/1000000 | loss=2.2946 vocab=0.8053 token=0.9782 | 6.15 steps/s | VRAM=31.92 Go [2026-07-17 08:45:06] step 241150/1000000 | loss=2.4131 vocab=1.4344 token=1.5036 | 6.11 steps/s | VRAM=25.96 Go [2026-07-17 08:45:14] step 241200/1000000 | loss=2.3398 vocab=1.4593 token=1.6938 | 6.18 steps/s | VRAM=31.25 Go [2026-07-17 08:45:22] step 241250/1000000 | loss=2.4080 vocab=1.4920 token=1.4772 | 6.14 steps/s | VRAM=28.78 Go [2026-07-17 08:45:31] step 241300/1000000 | loss=2.3634 vocab=0.7644 token=1.6048 | 6.15 steps/s | VRAM=28.93 Go [2026-07-17 08:45:39] step 241350/1000000 | loss=2.4119 vocab=1.2842 token=1.8228 | 6.13 steps/s | VRAM=28.71 Go [2026-07-17 08:45:47] step 241400/1000000 | loss=2.3648 vocab=1.0020 token=1.3539 | 6.14 steps/s | VRAM=31.49 Go [2026-07-17 08:45:55] step 241450/1000000 | loss=2.3570 vocab=1.2083 token=1.5469 | 6.08 steps/s | VRAM=27.53 Go [2026-07-17 08:46:03] step 241500/1000000 | loss=2.3887 vocab=0.7923 token=1.3722 | 6.16 steps/s | VRAM=28.29 Go [2026-07-17 08:46:11] step 241550/1000000 | loss=2.4203 vocab=0.9836 token=1.4422 | 6.15 steps/s | VRAM=30.61 Go [2026-07-17 08:46:19] step 241600/1000000 | loss=2.3428 vocab=0.8049 token=1.9402 | 6.17 steps/s | VRAM=30.71 Go [2026-07-17 08:46:28] step 241650/1000000 | loss=2.2868 vocab=0.9414 token=1.4997 | 6.12 steps/s | VRAM=27.49 Go [2026-07-17 08:46:36] step 241700/1000000 | loss=2.3803 vocab=0.6971 token=1.4189 | 6.14 steps/s | VRAM=31.54 Go [2026-07-17 08:46:44] step 241750/1000000 | loss=2.4186 vocab=0.6915 token=1.3069 | 6.13 steps/s | VRAM=28.78 Go [2026-07-17 08:46:52] step 241800/1000000 | loss=2.4792 vocab=0.6479 token=1.0491 | 6.09 steps/s | VRAM=29.15 Go [2026-07-17 08:47:00] step 241850/1000000 | loss=2.4188 vocab=0.9704 token=1.7819 | 6.11 steps/s | VRAM=29.07 Go [2026-07-17 08:47:08] step 241900/1000000 | loss=2.3786 vocab=0.7797 token=1.9926 | 6.08 steps/s | VRAM=27.10 Go [2026-07-17 08:47:17] step 241950/1000000 | loss=2.4129 vocab=0.9991 token=1.5767 | 6.14 steps/s | VRAM=28.25 Go [2026-07-17 08:47:25] step 242000/1000000 | loss=2.3659 vocab=0.9173 token=1.2263 | 6.14 steps/s | VRAM=32.18 Go [2026-07-17 08:47:33] step 242050/1000000 | loss=2.3274 vocab=0.8178 token=1.4659 | 6.11 steps/s | VRAM=31.11 Go [2026-07-17 08:47:41] step 242100/1000000 | loss=2.3756 vocab=1.6417 token=1.7497 | 6.09 steps/s | VRAM=26.37 Go [2026-07-17 08:47:49] step 242150/1000000 | loss=2.4196 vocab=1.0392 token=1.9245 | 6.16 steps/s | VRAM=27.79 Go [2026-07-17 08:47:57] step 242200/1000000 | loss=2.3052 vocab=1.1366 token=1.3961 | 6.10 steps/s | VRAM=28.67 Go [2026-07-17 08:48:06] step 242250/1000000 | loss=2.4474 vocab=1.3086 token=1.3552 | 6.12 steps/s | VRAM=29.49 Go [2026-07-17 08:48:14] step 242300/1000000 | loss=2.3972 vocab=1.1028 token=1.1702 | 6.15 steps/s | VRAM=29.41 Go [2026-07-17 08:48:22] step 242350/1000000 | loss=2.3219 vocab=0.9925 token=1.4968 | 6.17 steps/s | VRAM=28.97 Go [2026-07-17 08:48:30] step 242400/1000000 | loss=2.4544 vocab=1.1182 token=1.3589 | 6.12 steps/s | VRAM=26.17 Go [2026-07-17 08:48:38] step 242450/1000000 | loss=2.3825 vocab=1.0916 token=1.5914 | 6.13 steps/s | VRAM=28.44 Go [2026-07-17 08:48:46] step 242500/1000000 | loss=2.3892 vocab=1.2649 token=1.7638 | 6.13 steps/s | VRAM=28.02 Go [2026-07-17 08:48:55] step 242550/1000000 | loss=2.3920 vocab=0.9826 token=1.5120 | 6.13 steps/s | VRAM=26.08 Go [2026-07-17 08:49:03] step 242600/1000000 | loss=2.3715 vocab=1.1774 token=1.0824 | 6.13 steps/s | VRAM=29.26 Go [2026-07-17 08:49:11] step 242650/1000000 | loss=2.4201 vocab=0.8095 token=1.2065 | 6.14 steps/s | VRAM=27.19 Go [2026-07-17 08:49:19] step 242700/1000000 | loss=2.3864 vocab=1.1808 token=1.5557 | 6.11 steps/s | VRAM=29.13 Go [2026-07-17 08:49:27] step 242750/1000000 | loss=2.3901 vocab=1.1646 token=1.4357 | 6.09 steps/s | VRAM=31.97 Go [2026-07-17 08:49:35] step 242800/1000000 | loss=2.3849 vocab=1.2008 token=1.6606 | 6.11 steps/s | VRAM=31.76 Go [2026-07-17 08:49:44] step 242850/1000000 | loss=2.3920 vocab=1.0034 token=1.5125 | 6.12 steps/s | VRAM=26.18 Go [2026-07-17 08:49:52] step 242900/1000000 | loss=2.4392 vocab=0.8429 token=1.3318 | 6.07 steps/s | VRAM=25.65 Go [2026-07-17 08:50:00] step 242950/1000000 | loss=2.3315 vocab=1.1092 token=1.4185 | 6.11 steps/s | VRAM=31.50 Go [2026-07-17 08:50:08] step 243000/1000000 | loss=2.2881 vocab=0.6762 token=1.5282 | 6.14 steps/s | VRAM=32.54 Go [2026-07-17 08:50:16] step 243050/1000000 | loss=2.3263 vocab=0.3512 token=0.8194 | 6.12 steps/s | VRAM=25.69 Go [2026-07-17 08:50:24] step 243100/1000000 | loss=2.3976 vocab=0.9511 token=1.2961 | 6.15 steps/s | VRAM=25.06 Go [2026-07-17 08:50:33] step 243150/1000000 | loss=2.4024 vocab=1.1468 token=1.0960 | 6.15 steps/s | VRAM=24.88 Go [2026-07-17 08:50:41] step 243200/1000000 | loss=2.4142 vocab=0.6277 token=1.3523 | 6.13 steps/s | VRAM=25.48 Go [2026-07-17 08:50:49] step 243250/1000000 | loss=2.4613 vocab=1.5704 token=1.5884 | 6.10 steps/s | VRAM=26.40 Go [2026-07-17 08:50:57] step 243300/1000000 | loss=2.3736 vocab=1.1960 token=1.3698 | 6.11 steps/s | VRAM=26.93 Go [2026-07-17 08:51:05] step 243350/1000000 | loss=2.3408 vocab=1.0548 token=1.4807 | 6.09 steps/s | VRAM=24.75 Go [2026-07-17 08:51:14] step 243400/1000000 | loss=2.4085 vocab=1.4043 token=1.6230 | 6.11 steps/s | VRAM=28.25 Go [2026-07-17 08:51:22] step 243450/1000000 | loss=2.4421 vocab=0.8479 token=1.3558 | 6.16 steps/s | VRAM=31.52 Go [2026-07-17 08:51:30] step 243500/1000000 | loss=2.3486 vocab=0.9984 token=1.4851 | 6.14 steps/s | VRAM=26.02 Go [2026-07-17 08:51:38] step 243550/1000000 | loss=2.3399 vocab=0.9329 token=1.5867 | 6.14 steps/s | VRAM=28.19 Go [2026-07-17 08:51:46] step 243600/1000000 | loss=2.4495 vocab=1.0548 token=1.5462 | 6.13 steps/s | VRAM=27.86 Go [2026-07-17 08:51:54] step 243650/1000000 | loss=2.3010 vocab=0.7852 token=1.3884 | 6.15 steps/s | VRAM=29.03 Go [2026-07-17 08:52:02] step 243700/1000000 | loss=2.3664 vocab=0.8700 token=1.4225 | 6.13 steps/s | VRAM=26.42 Go [2026-07-17 08:52:11] step 243750/1000000 | loss=2.4418 vocab=0.7694 token=1.3167 | 6.12 steps/s | VRAM=30.06 Go [2026-07-17 08:52:19] step 243800/1000000 | loss=2.3173 vocab=0.6819 token=1.1980 | 6.13 steps/s | VRAM=28.39 Go [2026-07-17 08:52:27] step 243850/1000000 | loss=2.2945 vocab=0.8859 token=1.2347 | 6.15 steps/s | VRAM=31.67 Go [2026-07-17 08:52:35] step 243900/1000000 | loss=2.4741 vocab=1.1611 token=1.4562 | 6.12 steps/s | VRAM=28.50 Go [2026-07-17 08:52:43] step 243950/1000000 | loss=2.3806 vocab=1.4414 token=1.4853 | 6.13 steps/s | VRAM=25.91 Go [2026-07-17 08:52:51] step 244000/1000000 | loss=2.3756 vocab=0.7791 token=1.2530 | 6.16 steps/s | VRAM=28.77 Go [2026-07-17 08:52:59] step 244050/1000000 | loss=2.4393 vocab=1.1202 token=1.3148 | 6.14 steps/s | VRAM=28.28 Go [2026-07-17 08:53:08] step 244100/1000000 | loss=2.4296 vocab=0.9817 token=1.0771 | 6.11 steps/s | VRAM=30.48 Go [2026-07-17 08:53:16] step 244150/1000000 | loss=2.3228 vocab=0.7752 token=1.3017 | 6.15 steps/s | VRAM=28.30 Go [2026-07-17 08:53:24] step 244200/1000000 | loss=2.3998 vocab=0.7283 token=1.4190 | 6.12 steps/s | VRAM=32.60 Go [2026-07-17 08:53:32] step 244250/1000000 | loss=2.2755 vocab=0.7941 token=0.9600 | 6.12 steps/s | VRAM=31.51 Go [2026-07-17 08:53:40] step 244300/1000000 | loss=2.4343 vocab=1.0719 token=1.6725 | 6.14 steps/s | VRAM=28.78 Go [2026-07-17 08:53:48] step 244350/1000000 | loss=2.4774 vocab=1.4804 token=1.5341 | 6.11 steps/s | VRAM=26.30 Go [2026-07-17 08:53:57] step 244400/1000000 | loss=2.3017 vocab=1.0349 token=1.3473 | 6.16 steps/s | VRAM=31.20 Go [2026-07-17 08:54:05] step 244450/1000000 | loss=2.3227 vocab=1.1917 token=1.6587 | 6.11 steps/s | VRAM=28.62 Go [2026-07-17 08:54:13] step 244500/1000000 | loss=2.3908 vocab=1.3638 token=1.5903 | 6.15 steps/s | VRAM=27.67 Go [2026-07-17 08:54:21] step 244550/1000000 | loss=2.3420 vocab=1.2210 token=1.4716 | 6.13 steps/s | VRAM=26.32 Go [2026-07-17 08:54:29] step 244600/1000000 | loss=2.3259 vocab=0.8638 token=1.6859 | 6.10 steps/s | VRAM=28.54 Go [2026-07-17 08:54:37] step 244650/1000000 | loss=2.3231 vocab=0.8949 token=1.2972 | 6.10 steps/s | VRAM=29.99 Go [2026-07-17 08:54:46] step 244700/1000000 | loss=2.3570 vocab=1.2885 token=1.3027 | 6.16 steps/s | VRAM=29.81 Go [2026-07-17 08:54:54] step 244750/1000000 | loss=2.2795 vocab=1.1003 token=1.5686 | 6.12 steps/s | VRAM=27.95 Go [2026-07-17 08:55:02] step 244800/1000000 | loss=2.3376 vocab=0.6635 token=1.4269 | 6.10 steps/s | VRAM=27.24 Go [2026-07-17 08:55:10] step 244850/1000000 | loss=2.3620 vocab=1.4373 token=1.3485 | 6.12 steps/s | VRAM=27.42 Go [2026-07-17 08:55:18] step 244900/1000000 | loss=2.4504 vocab=0.9543 token=1.7895 | 6.14 steps/s | VRAM=31.91 Go [2026-07-17 08:55:26] step 244950/1000000 | loss=2.3827 vocab=0.7928 token=1.2225 | 6.08 steps/s | VRAM=27.55 Go [2026-07-17 08:55:35] step 245000/1000000 | loss=2.3855 vocab=0.9812 token=1.3887 | 6.12 steps/s | VRAM=32.15 Go [2026-07-17 08:56:57] step 245050/1000000 | loss=2.3215 vocab=0.7411 token=1.3572 | 0.60 steps/s | VRAM=25.83 Go [2026-07-17 08:57:05] step 245100/1000000 | loss=2.3959 vocab=0.7817 token=1.4965 | 6.17 steps/s | VRAM=28.33 Go [2026-07-17 08:57:14] step 245150/1000000 | loss=2.3655 vocab=1.0168 token=1.7347 | 6.18 steps/s | VRAM=26.28 Go [2026-07-17 08:57:22] step 245200/1000000 | loss=2.3855 vocab=0.8498 token=1.8035 | 6.20 steps/s | VRAM=29.40 Go [2026-07-17 08:57:30] step 245250/1000000 | loss=2.2541 vocab=0.9654 token=1.2880 | 6.10 steps/s | VRAM=28.33 Go [2026-07-17 08:57:38] step 245300/1000000 | loss=2.3261 vocab=0.7625 token=1.5508 | 6.17 steps/s | VRAM=31.77 Go [2026-07-17 08:57:46] step 245350/1000000 | loss=2.2418 vocab=0.4327 token=1.0690 | 6.16 steps/s | VRAM=28.75 Go [2026-07-17 08:57:54] step 245400/1000000 | loss=2.4484 vocab=0.9720 token=2.0146 | 6.14 steps/s | VRAM=27.84 Go [2026-07-17 08:58:02] step 245450/1000000 | loss=2.2914 vocab=0.9341 token=1.3738 | 6.13 steps/s | VRAM=31.76 Go [2026-07-17 08:58:11] step 245500/1000000 | loss=2.4662 vocab=0.9092 token=1.1757 | 6.12 steps/s | VRAM=26.40 Go [2026-07-17 08:58:19] step 245550/1000000 | loss=2.5090 vocab=1.1621 token=1.4556 | 6.09 steps/s | VRAM=26.27 Go [2026-07-17 08:58:27] step 245600/1000000 | loss=2.2603 vocab=0.8102 token=1.1935 | 6.10 steps/s | VRAM=29.46 Go [2026-07-17 08:58:35] step 245650/1000000 | loss=2.2971 vocab=1.5212 token=1.8738 | 6.08 steps/s | VRAM=28.81 Go [2026-07-17 08:58:43] step 245700/1000000 | loss=2.3802 vocab=0.7848 token=1.2686 | 6.13 steps/s | VRAM=31.74 Go [2026-07-17 08:58:51] step 245750/1000000 | loss=2.3410 vocab=0.8185 token=1.4045 | 6.13 steps/s | VRAM=29.41 Go [2026-07-17 08:59:00] step 245800/1000000 | loss=2.3906 vocab=0.9821 token=1.5498 | 6.17 steps/s | VRAM=30.46 Go [2026-07-17 08:59:08] step 245850/1000000 | loss=2.2805 vocab=0.5048 token=1.4427 | 6.10 steps/s | VRAM=28.53 Go [2026-07-17 08:59:16] step 245900/1000000 | loss=2.5005 vocab=0.7821 token=1.3106 | 6.13 steps/s | VRAM=26.37 Go [2026-07-17 08:59:24] step 245950/1000000 | loss=2.2222 vocab=0.8063 token=1.5371 | 6.13 steps/s | VRAM=32.74 Go [2026-07-17 08:59:32] step 246000/1000000 | loss=2.4014 vocab=0.3866 token=1.1875 | 6.11 steps/s | VRAM=32.22 Go [2026-07-17 08:59:40] step 246050/1000000 | loss=2.4314 vocab=1.0132 token=1.5723 | 6.13 steps/s | VRAM=31.44 Go [2026-07-17 08:59:49] step 246100/1000000 | loss=2.2711 vocab=1.8411 token=1.5554 | 6.05 steps/s | VRAM=26.66 Go [2026-07-17 08:59:57] step 246150/1000000 | loss=2.3768 vocab=0.6147 token=0.9897 | 6.12 steps/s | VRAM=25.36 Go [2026-07-17 09:00:05] step 246200/1000000 | loss=2.3751 vocab=0.8155 token=1.1342 | 6.09 steps/s | VRAM=27.81 Go [2026-07-17 09:00:13] step 246250/1000000 | loss=2.3694 vocab=1.1547 token=1.6013 | 6.09 steps/s | VRAM=26.24 Go [2026-07-17 09:00:21] step 246300/1000000 | loss=2.4301 vocab=1.0543 token=1.6767 | 6.12 steps/s | VRAM=32.08 Go [2026-07-17 09:00:30] step 246350/1000000 | loss=2.4134 vocab=0.5805 token=1.1468 | 6.09 steps/s | VRAM=28.85 Go [2026-07-17 09:00:38] step 246400/1000000 | loss=2.3791 vocab=1.3691 token=1.6313 | 6.11 steps/s | VRAM=25.68 Go [2026-07-17 09:00:46] step 246450/1000000 | loss=2.3897 vocab=1.0512 token=1.3448 | 6.09 steps/s | VRAM=27.94 Go [2026-07-17 09:00:54] step 246500/1000000 | loss=2.4220 vocab=0.8399 token=1.6745 | 6.10 steps/s | VRAM=28.89 Go [2026-07-17 09:01:02] step 246550/1000000 | loss=2.3221 vocab=0.7287 token=1.2548 | 6.10 steps/s | VRAM=29.29 Go [2026-07-17 09:01:11] step 246600/1000000 | loss=2.3102 vocab=1.1743 token=1.2731 | 6.10 steps/s | VRAM=32.13 Go [2026-07-17 09:01:19] step 246650/1000000 | loss=2.4300 vocab=0.6730 token=1.5960 | 6.10 steps/s | VRAM=31.51 Go [2026-07-17 09:01:27] step 246700/1000000 | loss=2.4022 vocab=0.9011 token=1.5521 | 6.07 steps/s | VRAM=28.51 Go [2026-07-17 09:01:35] step 246750/1000000 | loss=2.3700 vocab=1.1066 token=1.5431 | 6.08 steps/s | VRAM=29.53 Go [2026-07-17 09:01:44] step 246800/1000000 | loss=2.2726 vocab=0.9825 token=1.2472 | 6.08 steps/s | VRAM=28.96 Go [2026-07-17 09:01:52] step 246850/1000000 | loss=2.3744 vocab=1.1541 token=1.5773 | 6.12 steps/s | VRAM=28.61 Go [2026-07-17 09:02:00] step 246900/1000000 | loss=2.4182 vocab=1.9886 token=2.0153 | 6.07 steps/s | VRAM=26.95 Go [2026-07-17 09:02:08] step 246950/1000000 | loss=2.2767 vocab=0.9509 token=1.5095 | 6.10 steps/s | VRAM=28.75 Go [2026-07-17 09:02:16] step 247000/1000000 | loss=2.3442 vocab=0.8073 token=1.6159 | 6.11 steps/s | VRAM=30.83 Go [2026-07-17 09:02:25] step 247050/1000000 | loss=2.2462 vocab=0.9043 token=1.2394 | 6.10 steps/s | VRAM=29.17 Go [2026-07-17 09:02:33] step 247100/1000000 | loss=2.3700 vocab=0.7179 token=1.2432 | 6.09 steps/s | VRAM=28.60 Go [2026-07-17 09:02:41] step 247150/1000000 | loss=2.4185 vocab=0.9292 token=1.2841 | 6.06 steps/s | VRAM=28.76 Go [2026-07-17 09:02:49] step 247200/1000000 | loss=2.3888 vocab=1.3341 token=1.7866 | 6.12 steps/s | VRAM=28.27 Go [2026-07-17 09:02:57] step 247250/1000000 | loss=2.3924 vocab=0.8354 token=1.2528 | 6.12 steps/s | VRAM=26.87 Go [2026-07-17 09:03:05] step 247300/1000000 | loss=2.3703 vocab=0.7897 token=1.1287 | 6.13 steps/s | VRAM=31.16 Go [2026-07-17 09:03:14] step 247350/1000000 | loss=2.2870 vocab=0.6925 token=0.9205 | 6.12 steps/s | VRAM=29.34 Go [2026-07-17 09:03:22] step 247400/1000000 | loss=2.3260 vocab=1.1443 token=1.6568 | 6.12 steps/s | VRAM=31.08 Go [2026-07-17 09:03:30] step 247450/1000000 | loss=2.3631 vocab=0.7021 token=1.3458 | 6.11 steps/s | VRAM=32.45 Go [2026-07-17 09:03:38] step 247500/1000000 | loss=2.3678 vocab=1.0331 token=1.4426 | 6.13 steps/s | VRAM=25.54 Go [2026-07-17 09:03:46] step 247550/1000000 | loss=2.3199 vocab=1.0446 token=1.4602 | 6.11 steps/s | VRAM=28.63 Go [2026-07-17 09:03:55] step 247600/1000000 | loss=2.3902 vocab=0.4043 token=1.0342 | 6.12 steps/s | VRAM=27.45 Go [2026-07-17 09:04:03] step 247650/1000000 | loss=2.2935 vocab=0.6386 token=1.1675 | 6.13 steps/s | VRAM=27.30 Go [2026-07-17 09:04:11] step 247700/1000000 | loss=2.4483 vocab=1.1786 token=1.6708 | 6.14 steps/s | VRAM=30.87 Go [2026-07-17 09:04:19] step 247750/1000000 | loss=2.3909 vocab=0.7180 token=1.0267 | 6.12 steps/s | VRAM=30.16 Go [2026-07-17 09:04:27] step 247800/1000000 | loss=2.3875 vocab=1.0939 token=1.6864 | 6.12 steps/s | VRAM=32.40 Go [2026-07-17 09:04:35] step 247850/1000000 | loss=2.3862 vocab=0.9674 token=1.5435 | 6.08 steps/s | VRAM=28.55 Go [2026-07-17 09:04:44] step 247900/1000000 | loss=2.3051 vocab=1.1939 token=1.4252 | 6.11 steps/s | VRAM=32.81 Go [2026-07-17 09:04:52] step 247950/1000000 | loss=2.3953 vocab=0.8469 token=0.9922 | 6.14 steps/s | VRAM=28.37 Go [2026-07-17 09:05:00] step 248000/1000000 | loss=2.3711 vocab=0.9605 token=1.1327 | 6.12 steps/s | VRAM=31.98 Go [2026-07-17 09:05:08] step 248050/1000000 | loss=2.3974 vocab=1.0257 token=1.4340 | 6.13 steps/s | VRAM=27.61 Go [2026-07-17 09:05:16] step 248100/1000000 | loss=2.3025 vocab=0.7010 token=1.2627 | 6.13 steps/s | VRAM=25.71 Go [2026-07-17 09:05:24] step 248150/1000000 | loss=2.3214 vocab=0.9971 token=1.6173 | 6.15 steps/s | VRAM=31.49 Go [2026-07-17 09:05:32] step 248200/1000000 | loss=2.3983 vocab=1.5525 token=1.3656 | 6.12 steps/s | VRAM=32.94 Go [2026-07-17 09:05:41] step 248250/1000000 | loss=2.2248 vocab=0.7720 token=1.7063 | 6.15 steps/s | VRAM=32.36 Go [2026-07-17 09:05:49] step 248300/1000000 | loss=2.2685 vocab=0.5433 token=1.1230 | 6.13 steps/s | VRAM=31.68 Go [2026-07-17 09:05:57] step 248350/1000000 | loss=2.2228 vocab=0.8434 token=1.3974 | 6.11 steps/s | VRAM=28.42 Go [2026-07-17 09:06:05] step 248400/1000000 | loss=2.3470 vocab=1.0561 token=1.5155 | 6.13 steps/s | VRAM=27.99 Go [2026-07-17 09:06:13] step 248450/1000000 | loss=2.3645 vocab=0.7638 token=1.1861 | 6.12 steps/s | VRAM=25.03 Go [2026-07-17 09:06:21] step 248500/1000000 | loss=2.2895 vocab=1.2818 token=1.7050 | 6.15 steps/s | VRAM=27.44 Go [2026-07-17 09:06:30] step 248550/1000000 | loss=2.3698 vocab=0.9729 token=1.1745 | 6.16 steps/s | VRAM=31.43 Go [2026-07-17 09:06:38] step 248600/1000000 | loss=2.4143 vocab=1.4245 token=1.3493 | 6.13 steps/s | VRAM=28.88 Go [2026-07-17 09:06:46] step 248650/1000000 | loss=2.4357 vocab=1.3045 token=1.6552 | 6.14 steps/s | VRAM=26.28 Go [2026-07-17 09:06:54] step 248700/1000000 | loss=2.3855 vocab=0.9913 token=1.1472 | 6.14 steps/s | VRAM=28.18 Go [2026-07-17 09:07:02] step 248750/1000000 | loss=2.3974 vocab=0.8171 token=1.3819 | 6.13 steps/s | VRAM=26.47 Go [2026-07-17 09:07:10] step 248800/1000000 | loss=2.3778 vocab=1.2186 token=1.8259 | 6.11 steps/s | VRAM=32.22 Go [2026-07-17 09:07:19] step 248850/1000000 | loss=2.3231 vocab=0.6653 token=1.1492 | 6.11 steps/s | VRAM=32.75 Go [2026-07-17 09:07:27] step 248900/1000000 | loss=2.3542 vocab=0.6876 token=1.2228 | 6.13 steps/s | VRAM=29.08 Go [2026-07-17 09:07:35] step 248950/1000000 | loss=2.2867 vocab=0.8793 token=1.5491 | 6.15 steps/s | VRAM=32.18 Go [2026-07-17 09:07:43] step 249000/1000000 | loss=2.3740 vocab=1.6347 token=1.4758 | 6.13 steps/s | VRAM=29.47 Go [2026-07-17 09:07:51] step 249050/1000000 | loss=2.3377 vocab=0.9044 token=1.0756 | 6.13 steps/s | VRAM=26.77 Go [2026-07-17 09:07:59] step 249100/1000000 | loss=2.3237 vocab=0.4843 token=1.1511 | 6.15 steps/s | VRAM=28.96 Go [2026-07-17 09:08:07] step 249150/1000000 | loss=2.4192 vocab=1.0277 token=1.2823 | 6.08 steps/s | VRAM=28.59 Go [2026-07-17 09:08:16] step 249200/1000000 | loss=2.3267 vocab=1.3667 token=1.5276 | 6.11 steps/s | VRAM=24.75 Go [2026-07-17 09:08:24] step 249250/1000000 | loss=2.4232 vocab=0.9319 token=1.3617 | 6.11 steps/s | VRAM=25.83 Go [2026-07-17 09:08:32] step 249300/1000000 | loss=2.1953 vocab=0.7043 token=1.1164 | 6.14 steps/s | VRAM=26.27 Go [2026-07-17 09:08:40] step 249350/1000000 | loss=2.3904 vocab=0.9778 token=0.9762 | 6.12 steps/s | VRAM=23.44 Go [2026-07-17 09:08:48] step 249400/1000000 | loss=2.3543 vocab=0.5908 token=1.3379 | 6.15 steps/s | VRAM=31.40 Go [2026-07-17 09:08:56] step 249450/1000000 | loss=2.3406 vocab=1.2497 token=1.5179 | 6.13 steps/s | VRAM=32.63 Go [2026-07-17 09:09:05] step 249500/1000000 | loss=2.3677 vocab=0.7319 token=1.4429 | 6.14 steps/s | VRAM=22.54 Go [2026-07-17 09:09:13] step 249550/1000000 | loss=2.3412 vocab=0.8808 token=1.2046 | 6.16 steps/s | VRAM=28.64 Go [2026-07-17 09:09:21] step 249600/1000000 | loss=2.3390 vocab=0.6548 token=1.1066 | 6.11 steps/s | VRAM=29.56 Go [2026-07-17 09:09:29] step 249650/1000000 | loss=2.2711 vocab=0.7767 token=0.8768 | 6.11 steps/s | VRAM=28.56 Go [2026-07-17 09:09:37] step 249700/1000000 | loss=2.3298 vocab=0.8192 token=1.3076 | 6.13 steps/s | VRAM=32.65 Go [2026-07-17 09:09:45] step 249750/1000000 | loss=2.4233 vocab=0.9179 token=1.2925 | 6.10 steps/s | VRAM=23.26 Go [2026-07-17 09:09:54] step 249800/1000000 | loss=2.3232 vocab=0.8175 token=1.2346 | 6.14 steps/s | VRAM=28.54 Go [2026-07-17 09:10:02] step 249850/1000000 | loss=2.4537 vocab=0.8951 token=0.9463 | 6.10 steps/s | VRAM=28.81 Go [2026-07-17 09:10:10] step 249900/1000000 | loss=2.3606 vocab=1.3635 token=1.4054 | 6.10 steps/s | VRAM=31.92 Go [2026-07-17 09:10:18] step 249950/1000000 | loss=2.2857 vocab=0.7997 token=1.4405 | 6.11 steps/s | VRAM=26.72 Go [2026-07-17 09:10:26] step 250000/1000000 | loss=2.2682 vocab=1.3256 token=1.3657 | 6.13 steps/s | VRAM=28.77 Go [2026-07-17 09:11:51] step 250050/1000000 | loss=2.3559 vocab=1.0976 token=1.4013 | 0.59 steps/s | VRAM=28.78 Go [2026-07-17 09:12:00] step 250100/1000000 | loss=2.3465 vocab=0.4604 token=1.1429 | 6.16 steps/s | VRAM=30.38 Go [2026-07-17 09:12:08] step 250150/1000000 | loss=2.3664 vocab=0.6955 token=1.3098 | 6.18 steps/s | VRAM=28.26 Go [2026-07-17 09:12:16] step 250200/1000000 | loss=2.4176 vocab=0.8470 token=1.1925 | 6.12 steps/s | VRAM=26.44 Go [2026-07-17 09:12:24] step 250250/1000000 | loss=2.3261 vocab=0.7985 token=1.1213 | 6.13 steps/s | VRAM=26.19 Go [2026-07-17 09:12:32] step 250300/1000000 | loss=2.4161 vocab=0.9302 token=1.3250 | 6.17 steps/s | VRAM=29.00 Go [2026-07-17 09:12:40] step 250350/1000000 | loss=2.3284 vocab=0.5609 token=1.4037 | 6.17 steps/s | VRAM=24.73 Go [2026-07-17 09:12:48] step 250400/1000000 | loss=2.3654 vocab=0.6586 token=1.1491 | 6.17 steps/s | VRAM=32.58 Go [2026-07-17 09:12:56] step 250450/1000000 | loss=2.3191 vocab=0.9142 token=1.3216 | 6.14 steps/s | VRAM=32.69 Go [2026-07-17 09:13:05] step 250500/1000000 | loss=2.3445 vocab=0.7336 token=1.3655 | 6.13 steps/s | VRAM=28.54 Go [2026-07-17 09:13:13] step 250550/1000000 | loss=2.3893 vocab=0.9195 token=1.0459 | 6.14 steps/s | VRAM=29.54 Go [2026-07-17 09:13:21] step 250600/1000000 | loss=2.3358 vocab=1.3623 token=1.1211 | 6.14 steps/s | VRAM=31.73 Go [2026-07-17 09:13:29] step 250650/1000000 | loss=2.3654 vocab=1.5350 token=1.8995 | 6.10 steps/s | VRAM=29.12 Go [2026-07-17 09:13:37] step 250700/1000000 | loss=2.4524 vocab=0.6116 token=1.1360 | 6.13 steps/s | VRAM=25.23 Go [2026-07-17 09:13:45] step 250750/1000000 | loss=2.3115 vocab=1.1059 token=1.3588 | 6.14 steps/s | VRAM=27.74 Go [2026-07-17 09:13:54] step 250800/1000000 | loss=2.3695 vocab=0.9593 token=1.3627 | 6.12 steps/s | VRAM=26.08 Go [2026-07-17 09:14:02] step 250850/1000000 | loss=2.4514 vocab=1.5270 token=1.4968 | 6.13 steps/s | VRAM=23.95 Go [2026-07-17 09:14:10] step 250900/1000000 | loss=2.4222 vocab=0.9559 token=1.4743 | 6.12 steps/s | VRAM=31.56 Go [2026-07-17 09:14:18] step 250950/1000000 | loss=2.2734 vocab=0.5653 token=1.1006 | 6.11 steps/s | VRAM=32.18 Go [2026-07-17 09:14:26] step 251000/1000000 | loss=2.3432 vocab=1.0843 token=1.6739 | 6.10 steps/s | VRAM=28.79 Go [2026-07-17 09:14:34] step 251050/1000000 | loss=2.3255 vocab=1.1575 token=1.4767 | 6.10 steps/s | VRAM=31.83 Go [2026-07-17 09:14:43] step 251100/1000000 | loss=2.4258 vocab=1.2600 token=1.2508 | 6.08 steps/s | VRAM=27.21 Go [2026-07-17 09:14:51] step 251150/1000000 | loss=2.3807 vocab=0.6729 token=1.0398 | 6.14 steps/s | VRAM=32.13 Go [2026-07-17 09:14:59] step 251200/1000000 | loss=2.1505 vocab=1.0389 token=1.5552 | 6.10 steps/s | VRAM=31.57 Go [2026-07-17 09:15:07] step 251250/1000000 | loss=2.2608 vocab=1.1980 token=1.1492 | 6.10 steps/s | VRAM=28.46 Go [2026-07-17 09:15:15] step 251300/1000000 | loss=2.4863 vocab=0.6868 token=1.4034 | 6.11 steps/s | VRAM=28.65 Go [2026-07-17 09:15:24] step 251350/1000000 | loss=2.3095 vocab=0.5388 token=1.2611 | 6.09 steps/s | VRAM=29.83 Go [2026-07-17 09:15:32] step 251400/1000000 | loss=2.3756 vocab=0.5975 token=1.1456 | 6.14 steps/s | VRAM=24.44 Go [2026-07-17 09:15:40] step 251450/1000000 | loss=2.2755 vocab=0.5932 token=1.0563 | 6.10 steps/s | VRAM=32.66 Go [2026-07-17 09:15:48] step 251500/1000000 | loss=2.4107 vocab=1.3358 token=1.6151 | 6.13 steps/s | VRAM=31.47 Go [2026-07-17 09:15:56] step 251550/1000000 | loss=2.3979 vocab=0.9245 token=1.0014 | 6.10 steps/s | VRAM=32.33 Go [2026-07-17 09:16:05] step 251600/1000000 | loss=2.3290 vocab=1.4475 token=1.2858 | 6.12 steps/s | VRAM=29.83 Go [2026-07-17 09:16:13] step 251650/1000000 | loss=2.3685 vocab=1.0390 token=1.1213 | 6.13 steps/s | VRAM=25.39 Go [2026-07-17 09:16:21] step 251700/1000000 | loss=2.3261 vocab=0.7510 token=1.1125 | 6.13 steps/s | VRAM=28.83 Go [2026-07-17 09:16:29] step 251750/1000000 | loss=2.4315 vocab=1.2461 token=1.9493 | 6.09 steps/s | VRAM=26.74 Go [2026-07-17 09:16:37] step 251800/1000000 | loss=2.4086 vocab=1.8947 token=1.7661 | 6.09 steps/s | VRAM=29.66 Go [2026-07-17 09:16:45] step 251850/1000000 | loss=2.3938 vocab=1.5481 token=1.4764 | 6.11 steps/s | VRAM=31.77 Go [2026-07-17 09:16:54] step 251900/1000000 | loss=2.3719 vocab=0.8713 token=1.2966 | 6.11 steps/s | VRAM=31.70 Go [2026-07-17 09:17:02] step 251950/1000000 | loss=2.3718 vocab=0.7140 token=1.7348 | 6.06 steps/s | VRAM=30.29 Go [2026-07-17 09:17:10] step 252000/1000000 | loss=2.4139 vocab=1.2273 token=1.5243 | 6.12 steps/s | VRAM=28.25 Go [2026-07-17 09:17:18] step 252050/1000000 | loss=2.3282 vocab=0.6916 token=1.1414 | 6.12 steps/s | VRAM=31.37 Go [2026-07-17 09:17:26] step 252100/1000000 | loss=2.3513 vocab=0.7826 token=1.3414 | 6.11 steps/s | VRAM=28.59 Go [2026-07-17 09:17:35] step 252150/1000000 | loss=2.3528 vocab=0.7701 token=1.2375 | 6.13 steps/s | VRAM=28.57 Go [2026-07-17 09:17:43] step 252200/1000000 | loss=2.2945 vocab=0.9727 token=1.2466 | 6.13 steps/s | VRAM=32.26 Go [2026-07-17 09:17:51] step 252250/1000000 | loss=2.4556 vocab=0.7459 token=1.2843 | 6.10 steps/s | VRAM=29.37 Go [2026-07-17 09:17:59] step 252300/1000000 | loss=2.3269 vocab=0.9021 token=1.2789 | 6.11 steps/s | VRAM=32.49 Go [2026-07-17 09:18:07] step 252350/1000000 | loss=2.4595 vocab=0.9494 token=1.0928 | 6.10 steps/s | VRAM=29.59 Go [2026-07-17 09:18:15] step 252400/1000000 | loss=2.2082 vocab=1.0176 token=1.3101 | 6.13 steps/s | VRAM=31.30 Go [2026-07-17 09:18:24] step 252450/1000000 | loss=2.3801 vocab=0.8664 token=1.4218 | 6.11 steps/s | VRAM=24.91 Go [2026-07-17 09:18:32] step 252500/1000000 | loss=2.2623 vocab=0.8949 token=1.4891 | 6.11 steps/s | VRAM=32.41 Go [2026-07-17 09:18:40] step 252550/1000000 | loss=2.4204 vocab=1.0050 token=1.6373 | 6.13 steps/s | VRAM=31.94 Go [2026-07-17 09:18:48] step 252600/1000000 | loss=2.3495 vocab=1.0259 token=1.3140 | 6.11 steps/s | VRAM=25.38 Go [2026-07-17 09:18:56] step 252650/1000000 | loss=2.3415 vocab=1.1599 token=1.7499 | 6.08 steps/s | VRAM=30.35 Go [2026-07-17 09:19:04] step 252700/1000000 | loss=2.2961 vocab=0.7053 token=0.8792 | 6.13 steps/s | VRAM=31.61 Go [2026-07-17 09:19:13] step 252750/1000000 | loss=2.2878 vocab=1.2094 token=1.6427 | 6.12 steps/s | VRAM=26.43 Go [2026-07-17 09:19:21] step 252800/1000000 | loss=2.2654 vocab=1.1296 token=1.3507 | 6.11 steps/s | VRAM=32.66 Go [2026-07-17 09:19:29] step 252850/1000000 | loss=2.3502 vocab=1.1191 token=1.3535 | 6.05 steps/s | VRAM=28.43 Go [2026-07-17 09:19:37] step 252900/1000000 | loss=2.3308 vocab=1.0051 token=1.3522 | 6.07 steps/s | VRAM=27.05 Go [2026-07-17 09:19:45] step 252950/1000000 | loss=2.3184 vocab=0.9448 token=1.3935 | 6.14 steps/s | VRAM=25.31 Go [2026-07-17 09:19:54] step 253000/1000000 | loss=2.3391 vocab=1.4611 token=1.3378 | 6.14 steps/s | VRAM=32.36 Go [2026-07-17 09:20:02] step 253050/1000000 | loss=2.2079 vocab=0.6894 token=1.1913 | 6.13 steps/s | VRAM=31.57 Go [2026-07-17 09:20:10] step 253100/1000000 | loss=2.3494 vocab=0.6597 token=1.4904 | 6.13 steps/s | VRAM=27.45 Go [2026-07-17 09:20:18] step 253150/1000000 | loss=2.3721 vocab=1.0119 token=1.1986 | 6.13 steps/s | VRAM=31.55 Go [2026-07-17 09:20:26] step 253200/1000000 | loss=2.3560 vocab=1.0462 token=1.3415 | 6.14 steps/s | VRAM=32.58 Go [2026-07-17 09:20:34] step 253250/1000000 | loss=2.3712 vocab=0.9530 token=1.4220 | 6.13 steps/s | VRAM=31.85 Go [2026-07-17 09:20:43] step 253300/1000000 | loss=2.3755 vocab=0.8348 token=1.4318 | 6.14 steps/s | VRAM=29.12 Go [2026-07-17 09:20:51] step 253350/1000000 | loss=2.4708 vocab=1.6415 token=1.7356 | 6.09 steps/s | VRAM=26.71 Go [2026-07-17 09:20:59] step 253400/1000000 | loss=2.3969 vocab=0.5610 token=1.5712 | 6.11 steps/s | VRAM=32.50 Go [2026-07-17 09:21:07] step 253450/1000000 | loss=2.3927 vocab=0.7372 token=1.4055 | 6.14 steps/s | VRAM=24.99 Go [2026-07-17 09:21:15] step 253500/1000000 | loss=2.3996 vocab=1.0214 token=1.5079 | 6.11 steps/s | VRAM=26.81 Go [2026-07-17 09:21:23] step 253550/1000000 | loss=2.2825 vocab=1.5067 token=1.3307 | 6.11 steps/s | VRAM=28.16 Go [2026-07-17 09:21:32] step 253600/1000000 | loss=2.2810 vocab=0.5204 token=1.1234 | 6.09 steps/s | VRAM=23.59 Go [2026-07-17 09:21:40] step 253650/1000000 | loss=2.3747 vocab=0.9203 token=1.4626 | 6.14 steps/s | VRAM=28.89 Go [2026-07-17 09:21:48] step 253700/1000000 | loss=2.3807 vocab=1.0609 token=1.1726 | 6.12 steps/s | VRAM=28.57 Go [2026-07-17 09:21:56] step 253750/1000000 | loss=2.3984 vocab=0.5550 token=0.9153 | 6.17 steps/s | VRAM=24.62 Go [2026-07-17 09:22:04] step 253800/1000000 | loss=2.3048 vocab=1.1227 token=1.3689 | 6.18 steps/s | VRAM=31.86 Go [2026-07-17 09:22:12] step 253850/1000000 | loss=2.3043 vocab=0.8722 token=1.6526 | 6.09 steps/s | VRAM=28.61 Go [2026-07-17 09:22:21] step 253900/1000000 | loss=2.3407 vocab=0.9593 token=1.2939 | 6.15 steps/s | VRAM=32.23 Go [2026-07-17 09:22:29] step 253950/1000000 | loss=2.5101 vocab=0.9877 token=1.2859 | 6.14 steps/s | VRAM=28.95 Go [2026-07-17 09:22:37] step 254000/1000000 | loss=2.3239 vocab=0.9281 token=1.1094 | 6.12 steps/s | VRAM=28.33 Go [2026-07-17 09:22:45] step 254050/1000000 | loss=2.3032 vocab=0.7943 token=1.5347 | 6.13 steps/s | VRAM=31.66 Go [2026-07-17 09:22:53] step 254100/1000000 | loss=2.4180 vocab=1.3121 token=0.9575 | 6.11 steps/s | VRAM=26.84 Go [2026-07-17 09:23:01] step 254150/1000000 | loss=2.2744 vocab=0.8869 token=1.3810 | 6.15 steps/s | VRAM=30.34 Go [2026-07-17 09:23:10] step 254200/1000000 | loss=2.3616 vocab=0.7738 token=1.5910 | 6.10 steps/s | VRAM=26.55 Go [2026-07-17 09:23:18] step 254250/1000000 | loss=2.3427 vocab=0.7516 token=1.3256 | 6.10 steps/s | VRAM=26.92 Go [2026-07-17 09:23:26] step 254300/1000000 | loss=2.2477 vocab=0.6231 token=1.1991 | 6.13 steps/s | VRAM=31.92 Go [2026-07-17 09:23:34] step 254350/1000000 | loss=2.3149 vocab=1.2070 token=1.3755 | 6.12 steps/s | VRAM=32.79 Go [2026-07-17 09:23:42] step 254400/1000000 | loss=2.3325 vocab=0.7349 token=1.4865 | 6.09 steps/s | VRAM=28.34 Go [2026-07-17 09:23:50] step 254450/1000000 | loss=2.4162 vocab=1.6581 token=1.7365 | 6.14 steps/s | VRAM=31.80 Go [2026-07-17 09:23:59] step 254500/1000000 | loss=2.3636 vocab=0.8965 token=1.4940 | 6.12 steps/s | VRAM=25.88 Go [2026-07-17 09:24:07] step 254550/1000000 | loss=2.4561 vocab=0.9139 token=1.3419 | 6.11 steps/s | VRAM=25.91 Go [2026-07-17 09:24:15] step 254600/1000000 | loss=2.4634 vocab=1.0239 token=1.4017 | 6.10 steps/s | VRAM=28.37 Go [2026-07-17 09:24:23] step 254650/1000000 | loss=2.2869 vocab=1.1298 token=1.5444 | 6.09 steps/s | VRAM=25.62 Go [2026-07-17 09:24:31] step 254700/1000000 | loss=2.4128 vocab=1.1796 token=1.7183 | 6.12 steps/s | VRAM=32.41 Go [2026-07-17 09:24:40] step 254750/1000000 | loss=2.3205 vocab=1.2158 token=1.2668 | 6.11 steps/s | VRAM=28.77 Go [2026-07-17 09:24:48] step 254800/1000000 | loss=2.4860 vocab=0.5533 token=1.0473 | 6.09 steps/s | VRAM=28.44 Go [2026-07-17 09:24:56] step 254850/1000000 | loss=2.2286 vocab=1.0455 token=1.1805 | 6.09 steps/s | VRAM=28.71 Go [2026-07-17 09:25:04] step 254900/1000000 | loss=2.3456 vocab=0.6467 token=1.2069 | 6.10 steps/s | VRAM=28.24 Go [2026-07-17 09:25:12] step 254950/1000000 | loss=2.3027 vocab=1.3305 token=1.8196 | 6.12 steps/s | VRAM=29.71 Go [2026-07-17 09:25:20] step 255000/1000000 | loss=2.4499 vocab=0.8123 token=1.2250 | 6.12 steps/s | VRAM=25.42 Go [2026-07-17 09:26:45] step 255050/1000000 | loss=2.2779 vocab=0.6821 token=1.2604 | 0.59 steps/s | VRAM=29.26 Go [2026-07-17 09:26:53] step 255100/1000000 | loss=2.3945 vocab=0.8191 token=1.0410 | 6.19 steps/s | VRAM=28.11 Go [2026-07-17 09:27:01] step 255150/1000000 | loss=2.3766 vocab=0.7697 token=1.5944 | 6.18 steps/s | VRAM=31.38 Go [2026-07-17 09:27:09] step 255200/1000000 | loss=2.4503 vocab=0.9383 token=1.4734 | 6.17 steps/s | VRAM=32.37 Go [2026-07-17 09:27:17] step 255250/1000000 | loss=2.3819 vocab=0.9337 token=1.9763 | 6.20 steps/s | VRAM=24.77 Go [2026-07-17 09:27:25] step 255300/1000000 | loss=2.3572 vocab=0.6016 token=0.9483 | 6.18 steps/s | VRAM=31.21 Go [2026-07-17 09:27:33] step 255350/1000000 | loss=2.2711 vocab=0.6469 token=1.2933 | 6.16 steps/s | VRAM=31.30 Go [2026-07-17 09:27:41] step 255400/1000000 | loss=2.3203 vocab=0.7497 token=0.7915 | 6.12 steps/s | VRAM=23.75 Go [2026-07-17 09:27:49] step 255450/1000000 | loss=2.3196 vocab=0.4853 token=1.5302 | 6.15 steps/s | VRAM=28.27 Go [2026-07-17 09:27:58] step 255500/1000000 | loss=2.2850 vocab=1.1198 token=1.5333 | 6.14 steps/s | VRAM=31.55 Go [2026-07-17 09:28:06] step 255550/1000000 | loss=2.3274 vocab=0.9166 token=1.5787 | 6.16 steps/s | VRAM=30.79 Go [2026-07-17 09:34:47] step 255600/1000000 | loss=1.7185 vocab=0.6420 token=1.7077 | 7.74 steps/s | VRAM=28.91 Go [2026-07-17 09:34:55] step 255650/1000000 | loss=2.3417 vocab=0.5193 token=1.2142 | 6.24 steps/s | VRAM=32.46 Go [2026-07-17 09:35:03] step 255700/1000000 | loss=2.3841 vocab=0.6740 token=1.6618 | 6.20 steps/s | VRAM=29.23 Go [2026-07-17 09:35:11] step 255750/1000000 | loss=2.3584 vocab=0.9696 token=1.3366 | 6.20 steps/s | VRAM=28.49 Go [2026-07-17 09:35:19] step 255800/1000000 | loss=2.2332 vocab=0.4700 token=0.7963 | 6.17 steps/s | VRAM=28.00 Go [2026-07-17 09:35:27] step 255850/1000000 | loss=2.4004 vocab=0.7116 token=1.2694 | 6.19 steps/s | VRAM=31.91 Go [2026-07-17 09:35:35] step 255900/1000000 | loss=2.3485 vocab=0.8976 token=1.3167 | 6.16 steps/s | VRAM=25.88 Go [2026-07-17 09:35:43] step 255950/1000000 | loss=2.2925 vocab=0.5679 token=1.0561 | 6.17 steps/s | VRAM=28.54 Go [2026-07-17 09:35:51] step 256000/1000000 | loss=2.3327 vocab=1.5033 token=1.4295 | 6.15 steps/s | VRAM=28.51 Go [2026-07-17 09:35:59] step 256050/1000000 | loss=2.3262 vocab=1.2037 token=1.8564 | 6.17 steps/s | VRAM=29.90 Go [2026-07-17 09:36:08] step 256100/1000000 | loss=2.3670 vocab=0.9748 token=1.6662 | 6.13 steps/s | VRAM=32.67 Go [2026-07-17 09:36:16] step 256150/1000000 | loss=2.4721 vocab=0.9123 token=1.4806 | 6.15 steps/s | VRAM=27.64 Go [2026-07-17 09:36:24] step 256200/1000000 | loss=2.2712 vocab=0.7891 token=1.3785 | 6.13 steps/s | VRAM=31.98 Go [2026-07-17 09:36:32] step 256250/1000000 | loss=2.3504 vocab=0.7907 token=1.1464 | 6.12 steps/s | VRAM=28.28 Go [2026-07-17 09:36:40] step 256300/1000000 | loss=2.4310 vocab=1.3819 token=1.6019 | 6.12 steps/s | VRAM=26.06 Go [2026-07-17 09:36:48] step 256350/1000000 | loss=2.4245 vocab=1.0507 token=1.5420 | 6.12 steps/s | VRAM=28.08 Go [2026-07-17 09:36:57] step 256400/1000000 | loss=2.2843 vocab=1.2142 token=1.6543 | 6.09 steps/s | VRAM=26.63 Go [2026-07-17 09:37:05] step 256450/1000000 | loss=2.3650 vocab=1.2467 token=1.8415 | 6.12 steps/s | VRAM=28.66 Go [2026-07-17 09:37:13] step 256500/1000000 | loss=2.3517 vocab=0.4936 token=1.1092 | 6.09 steps/s | VRAM=30.40 Go [2026-07-17 09:37:21] step 256550/1000000 | loss=2.4466 vocab=0.6580 token=1.2221 | 6.08 steps/s | VRAM=28.80 Go [2026-07-17 09:37:29] step 256600/1000000 | loss=2.3539 vocab=1.3102 token=1.6483 | 6.12 steps/s | VRAM=31.73 Go [2026-07-17 09:37:38] step 256650/1000000 | loss=2.3068 vocab=0.9635 token=1.6154 | 6.12 steps/s | VRAM=28.22 Go [2026-07-17 09:37:46] step 256700/1000000 | loss=2.2099 vocab=1.3257 token=1.5598 | 6.11 steps/s | VRAM=28.49 Go [2026-07-17 09:37:54] step 256750/1000000 | loss=2.3262 vocab=1.0365 token=1.4032 | 6.09 steps/s | VRAM=32.95 Go [2026-07-17 09:38:02] step 256800/1000000 | loss=2.4195 vocab=0.9386 token=1.2814 | 6.09 steps/s | VRAM=28.42 Go [2026-07-17 09:38:10] step 256850/1000000 | loss=2.2713 vocab=1.1937 token=1.4635 | 6.12 steps/s | VRAM=32.03 Go [2026-07-17 09:38:19] step 256900/1000000 | loss=2.4836 vocab=0.6862 token=0.9858 | 6.11 steps/s | VRAM=30.00 Go [2026-07-17 09:38:27] step 256950/1000000 | loss=2.2989 vocab=0.6184 token=1.3898 | 6.10 steps/s | VRAM=28.62 Go [2026-07-17 09:38:35] step 257000/1000000 | loss=2.2728 vocab=0.9579 token=1.5401 | 6.10 steps/s | VRAM=28.18 Go [2026-07-17 09:38:43] step 257050/1000000 | loss=2.3587 vocab=1.0833 token=1.3593 | 6.10 steps/s | VRAM=28.81 Go [2026-07-17 09:38:51] step 257100/1000000 | loss=2.3676 vocab=0.8240 token=1.1581 | 6.11 steps/s | VRAM=28.65 Go [2026-07-17 09:39:00] step 257150/1000000 | loss=2.2973 vocab=0.7709 token=0.9579 | 6.08 steps/s | VRAM=28.43 Go [2026-07-17 09:39:08] step 257200/1000000 | loss=2.3383 vocab=1.4076 token=1.7353 | 6.11 steps/s | VRAM=32.28 Go [2026-07-17 09:39:16] step 257250/1000000 | loss=2.3956 vocab=1.0139 token=1.1656 | 6.08 steps/s | VRAM=29.05 Go [2026-07-17 09:39:24] step 257300/1000000 | loss=2.3510 vocab=1.2874 token=1.1517 | 6.11 steps/s | VRAM=28.37 Go [2026-07-17 09:39:32] step 257350/1000000 | loss=2.3486 vocab=0.7667 token=1.2890 | 6.07 steps/s | VRAM=25.99 Go [2026-07-17 09:39:41] step 257400/1000000 | loss=2.3408 vocab=1.1208 token=1.6963 | 6.10 steps/s | VRAM=28.23 Go [2026-07-17 09:39:49] step 257450/1000000 | loss=2.2458 vocab=0.8388 token=1.1199 | 6.12 steps/s | VRAM=32.46 Go [2026-07-17 09:39:57] step 257500/1000000 | loss=2.3745 vocab=1.0675 token=1.3654 | 6.10 steps/s | VRAM=30.29 Go [2026-07-17 09:40:05] step 257550/1000000 | loss=2.2916 vocab=1.0125 token=1.4271 | 6.11 steps/s | VRAM=29.82 Go [2026-07-17 09:40:13] step 257600/1000000 | loss=2.3297 vocab=0.7591 token=1.5940 | 6.14 steps/s | VRAM=31.46 Go [2026-07-17 09:40:21] step 257650/1000000 | loss=2.1972 vocab=0.6035 token=0.8092 | 6.12 steps/s | VRAM=29.93 Go [2026-07-17 09:40:30] step 257700/1000000 | loss=2.3698 vocab=0.8313 token=1.3178 | 6.11 steps/s | VRAM=28.57 Go [2026-07-17 09:40:40] step 257750/1000000 | loss=2.4132 vocab=0.9189 token=1.9051 | 4.78 steps/s | VRAM=25.18 Go [2026-07-17 09:40:48] step 257800/1000000 | loss=2.3817 vocab=0.8827 token=1.4263 | 6.14 steps/s | VRAM=29.24 Go [2026-07-17 09:40:56] step 257850/1000000 | loss=2.1907 vocab=1.2578 token=1.8311 | 6.12 steps/s | VRAM=29.14 Go [2026-07-17 09:41:05] step 257900/1000000 | loss=2.1833 vocab=0.7947 token=1.2932 | 6.12 steps/s | VRAM=28.55 Go [2026-07-17 09:41:13] step 257950/1000000 | loss=2.3849 vocab=0.7205 token=1.4727 | 6.10 steps/s | VRAM=29.33 Go [2026-07-17 09:41:21] step 258000/1000000 | loss=2.3486 vocab=0.9534 token=1.1519 | 6.10 steps/s | VRAM=29.11 Go [2026-07-17 09:41:29] step 258050/1000000 | loss=2.2034 vocab=0.5859 token=1.4472 | 6.11 steps/s | VRAM=32.40 Go [2026-07-17 09:41:37] step 258100/1000000 | loss=2.3898 vocab=0.8372 token=1.2901 | 6.13 steps/s | VRAM=31.86 Go [2026-07-17 09:41:45] step 258150/1000000 | loss=2.3538 vocab=0.9169 token=1.6464 | 6.13 steps/s | VRAM=25.49 Go [2026-07-17 09:41:54] step 258200/1000000 | loss=2.3180 vocab=1.2019 token=1.2598 | 6.12 steps/s | VRAM=26.50 Go [2026-07-17 09:42:02] step 258250/1000000 | loss=2.2617 vocab=0.8773 token=0.9440 | 6.13 steps/s | VRAM=24.78 Go [2026-07-17 09:42:10] step 258300/1000000 | loss=2.4284 vocab=0.7152 token=1.1311 | 6.09 steps/s | VRAM=27.26 Go [2026-07-17 09:42:18] step 258350/1000000 | loss=2.2740 vocab=0.8533 token=1.2636 | 6.12 steps/s | VRAM=28.48 Go [2026-07-17 09:42:26] step 258400/1000000 | loss=2.4721 vocab=0.9362 token=1.6447 | 6.08 steps/s | VRAM=26.26 Go [2026-07-17 09:42:35] step 258450/1000000 | loss=2.3191 vocab=0.4602 token=1.1240 | 6.13 steps/s | VRAM=30.96 Go [2026-07-17 09:42:43] step 258500/1000000 | loss=2.4422 vocab=0.9943 token=1.4580 | 6.06 steps/s | VRAM=29.50 Go [2026-07-17 09:42:51] step 258550/1000000 | loss=2.3011 vocab=0.7736 token=1.4889 | 6.09 steps/s | VRAM=26.31 Go [2026-07-17 09:42:59] step 258600/1000000 | loss=2.3422 vocab=0.4057 token=0.9271 | 6.15 steps/s | VRAM=25.05 Go [2026-07-17 09:43:07] step 258650/1000000 | loss=2.2666 vocab=1.1996 token=1.5185 | 6.11 steps/s | VRAM=28.76 Go [2026-07-17 09:43:16] step 258700/1000000 | loss=2.2943 vocab=1.4026 token=1.7272 | 6.08 steps/s | VRAM=28.61 Go [2026-07-17 09:43:24] step 258750/1000000 | loss=2.3748 vocab=1.4875 token=1.7670 | 6.11 steps/s | VRAM=28.02 Go [2026-07-17 09:43:32] step 258800/1000000 | loss=2.4094 vocab=1.0724 token=1.6233 | 6.14 steps/s | VRAM=26.64 Go [2026-07-17 09:43:40] step 258850/1000000 | loss=2.4149 vocab=0.9892 token=1.3520 | 6.13 steps/s | VRAM=27.46 Go [2026-07-17 09:43:48] step 258900/1000000 | loss=2.3371 vocab=1.0619 token=1.2662 | 6.10 steps/s | VRAM=28.43 Go [2026-07-17 09:43:56] step 258950/1000000 | loss=2.3074 vocab=0.6865 token=1.1855 | 6.12 steps/s | VRAM=32.09 Go [2026-07-17 09:44:05] step 259000/1000000 | loss=2.4278 vocab=1.4139 token=1.2815 | 6.09 steps/s | VRAM=29.37 Go [2026-07-17 09:44:13] step 259050/1000000 | loss=2.3884 vocab=1.3131 token=1.4435 | 6.12 steps/s | VRAM=31.65 Go [2026-07-17 09:44:21] step 259100/1000000 | loss=2.3636 vocab=1.0697 token=1.7791 | 6.10 steps/s | VRAM=32.63 Go [2026-07-17 09:44:29] step 259150/1000000 | loss=2.2721 vocab=1.0322 token=1.2004 | 6.08 steps/s | VRAM=26.59 Go [2026-07-17 09:44:37] step 259200/1000000 | loss=2.4264 vocab=1.0478 token=1.1892 | 6.13 steps/s | VRAM=32.10 Go [2026-07-17 09:44:46] step 259250/1000000 | loss=2.3182 vocab=0.9653 token=1.3661 | 6.08 steps/s | VRAM=23.41 Go [2026-07-17 09:44:54] step 259300/1000000 | loss=2.2589 vocab=1.1856 token=1.7684 | 6.15 steps/s | VRAM=31.90 Go [2026-07-17 09:45:02] step 259350/1000000 | loss=2.3569 vocab=0.8769 token=1.6090 | 6.10 steps/s | VRAM=32.32 Go [2026-07-17 09:45:10] step 259400/1000000 | loss=2.3088 vocab=0.9355 token=1.4671 | 6.14 steps/s | VRAM=32.37 Go [2026-07-17 09:45:18] step 259450/1000000 | loss=2.2887 vocab=0.7773 token=0.9897 | 6.06 steps/s | VRAM=29.62 Go [2026-07-17 09:45:26] step 259500/1000000 | loss=2.3395 vocab=1.1911 token=1.0621 | 6.11 steps/s | VRAM=29.75 Go [2026-07-17 09:45:35] step 259550/1000000 | loss=2.2834 vocab=0.6798 token=1.7798 | 6.15 steps/s | VRAM=31.34 Go [2026-07-17 09:45:43] step 259600/1000000 | loss=2.3216 vocab=1.1303 token=1.5135 | 6.07 steps/s | VRAM=32.62 Go [2026-07-17 09:45:51] step 259650/1000000 | loss=2.3072 vocab=0.9867 token=1.4353 | 6.11 steps/s | VRAM=32.71 Go [2026-07-17 09:45:59] step 259700/1000000 | loss=2.3723 vocab=1.4388 token=1.2173 | 6.07 steps/s | VRAM=29.54 Go [2026-07-17 09:46:07] step 259750/1000000 | loss=2.3363 vocab=0.9418 token=1.5487 | 6.12 steps/s | VRAM=27.67 Go [2026-07-17 09:46:16] step 259800/1000000 | loss=2.3722 vocab=0.9997 token=1.3994 | 6.11 steps/s | VRAM=27.98 Go [2026-07-17 09:46:24] step 259850/1000000 | loss=2.3637 vocab=0.6320 token=0.8566 | 6.11 steps/s | VRAM=25.61 Go [2026-07-17 09:46:32] step 259900/1000000 | loss=2.2683 vocab=0.8236 token=1.2289 | 6.11 steps/s | VRAM=27.91 Go [2026-07-17 09:46:40] step 259950/1000000 | loss=2.4191 vocab=0.5861 token=1.0021 | 6.10 steps/s | VRAM=29.59 Go [2026-07-17 09:46:48] step 260000/1000000 | loss=2.4008 vocab=0.7403 token=1.4308 | 6.10 steps/s | VRAM=32.77 Go [push] ECHEC (HF injoignable ?) a step 260000 -- entrainement CONTINUE : RuntimeError('[enforce fail at inline_container.cc:672] . unexpected pos 2530337792 vs 2530337680') [2026-07-17 09:47:01] step 260050/1000000 | loss=2.3881 vocab=1.3382 token=1.6342 | 3.82 steps/s | VRAM=26.22 Go [2026-07-17 09:47:10] step 260100/1000000 | loss=2.3873 vocab=1.6125 token=1.5085 | 6.13 steps/s | VRAM=26.19 Go [2026-07-17 09:47:18] step 260150/1000000 | loss=2.3170 vocab=0.7752 token=1.4829 | 6.14 steps/s | VRAM=29.46 Go [2026-07-17 09:47:26] step 260200/1000000 | loss=2.3258 vocab=0.7158 token=1.4166 | 6.12 steps/s | VRAM=29.18 Go [2026-07-17 09:47:34] step 260250/1000000 | loss=2.3537 vocab=0.7666 token=1.3624 | 6.13 steps/s | VRAM=29.66 Go [2026-07-17 09:47:42] step 260300/1000000 | loss=2.3255 vocab=1.0518 token=1.5551 | 6.13 steps/s | VRAM=25.01 Go [2026-07-17 09:47:50] step 260350/1000000 | loss=2.3122 vocab=1.1316 token=1.5467 | 6.13 steps/s | VRAM=28.37 Go [2026-07-17 09:47:59] step 260400/1000000 | loss=2.3824 vocab=1.0399 token=1.4025 | 6.12 steps/s | VRAM=32.24 Go [2026-07-17 09:48:07] step 260450/1000000 | loss=2.3915 vocab=1.0232 token=1.6700 | 6.14 steps/s | VRAM=32.09 Go [2026-07-17 09:48:15] step 260500/1000000 | loss=2.3082 vocab=1.0873 token=1.5141 | 6.11 steps/s | VRAM=28.51 Go [2026-07-17 09:48:23] step 260550/1000000 | loss=2.3097 vocab=0.5135 token=1.5429 | 6.11 steps/s | VRAM=29.25 Go [2026-07-17 10:31:31] step 255600/1000000 | loss=1.6779 vocab=1.2680 token=1.0295 | 8.45 steps/s | VRAM=32.27 Go [2026-07-17 10:31:39] step 255650/1000000 | loss=2.3520 vocab=0.7899 token=1.2159 | 6.17 steps/s | VRAM=28.49 Go [2026-07-17 10:31:48] step 255700/1000000 | loss=2.3346 vocab=0.9684 token=1.3660 | 6.17 steps/s | VRAM=28.44 Go [2026-07-17 10:31:56] step 255750/1000000 | loss=2.2709 vocab=1.0005 token=1.2479 | 6.20 steps/s | VRAM=31.90 Go [2026-07-17 10:32:04] step 255800/1000000 | loss=2.2905 vocab=1.3995 token=1.4532 | 6.19 steps/s | VRAM=28.43 Go [2026-07-17 10:32:12] step 255850/1000000 | loss=2.3821 vocab=0.8528 token=1.5933 | 6.14 steps/s | VRAM=31.86 Go [2026-07-17 10:32:20] step 255900/1000000 | loss=2.3141 vocab=1.2663 token=1.4970 | 6.14 steps/s | VRAM=25.21 Go [2026-07-17 10:32:28] step 255950/1000000 | loss=2.3974 vocab=0.9262 token=1.5009 | 6.15 steps/s | VRAM=28.10 Go [2026-07-17 10:32:36] step 256000/1000000 | loss=2.3448 vocab=0.5615 token=1.1396 | 6.11 steps/s | VRAM=25.04 Go [2026-07-17 10:32:45] step 256050/1000000 | loss=2.3936 vocab=0.9908 token=1.2427 | 6.09 steps/s | VRAM=25.45 Go [2026-07-17 10:32:53] step 256100/1000000 | loss=2.2925 vocab=1.4937 token=1.8951 | 6.12 steps/s | VRAM=28.10 Go [2026-07-17 10:33:01] step 256150/1000000 | loss=2.2842 vocab=1.2081 token=1.3401 | 6.11 steps/s | VRAM=32.07 Go [2026-07-17 10:33:09] step 256200/1000000 | loss=2.2477 vocab=0.9168 token=1.2682 | 6.07 steps/s | VRAM=27.87 Go [2026-07-17 10:33:17] step 256250/1000000 | loss=2.3681 vocab=1.1755 token=1.3399 | 6.10 steps/s | VRAM=27.87 Go [2026-07-17 10:33:26] step 256300/1000000 | loss=2.2541 vocab=0.7557 token=1.1028 | 6.08 steps/s | VRAM=27.73 Go [2026-07-17 10:33:34] step 256350/1000000 | loss=2.3287 vocab=1.6622 token=1.7336 | 6.12 steps/s | VRAM=32.79 Go [2026-07-17 10:33:42] step 256400/1000000 | loss=2.3492 vocab=1.0455 token=1.6540 | 6.07 steps/s | VRAM=27.96 Go [2026-07-17 10:33:50] step 256450/1000000 | loss=2.3299 vocab=0.6072 token=1.2218 | 6.07 steps/s | VRAM=27.85 Go [2026-07-17 10:33:58] step 256500/1000000 | loss=2.3523 vocab=1.0086 token=1.4475 | 6.07 steps/s | VRAM=31.15 Go [2026-07-17 10:34:07] step 256550/1000000 | loss=2.3338 vocab=0.5234 token=1.1908 | 6.10 steps/s | VRAM=31.18 Go [2026-07-17 10:34:15] step 256600/1000000 | loss=2.4357 vocab=1.1001 token=1.8107 | 6.06 steps/s | VRAM=28.31 Go [2026-07-17 10:34:23] step 256650/1000000 | loss=2.3990 vocab=1.2387 token=1.3327 | 6.08 steps/s | VRAM=28.38 Go [2026-07-17 10:34:31] step 256700/1000000 | loss=2.3623 vocab=0.5030 token=1.4369 | 6.09 steps/s | VRAM=27.70 Go [2026-07-17 10:34:39] step 256750/1000000 | loss=2.4466 vocab=0.3673 token=1.3400 | 6.12 steps/s | VRAM=28.14 Go [2026-07-17 10:34:48] step 256800/1000000 | loss=2.2849 vocab=0.8269 token=1.0202 | 6.10 steps/s | VRAM=28.26 Go [2026-07-17 10:34:56] step 256850/1000000 | loss=2.2901 vocab=0.7334 token=1.5715 | 6.09 steps/s | VRAM=28.49 Go [2026-07-17 10:35:04] step 256900/1000000 | loss=2.2849 vocab=0.5465 token=1.2695 | 6.10 steps/s | VRAM=27.49 Go [2026-07-17 10:35:12] step 256950/1000000 | loss=2.3524 vocab=0.9277 token=1.2639 | 6.10 steps/s | VRAM=31.93 Go [2026-07-17 10:35:20] step 257000/1000000 | loss=2.3974 vocab=0.9352 token=1.4096 | 6.11 steps/s | VRAM=24.80 Go [2026-07-17 10:35:29] step 257050/1000000 | loss=2.1687 vocab=0.7417 token=1.2554 | 6.08 steps/s | VRAM=27.85 Go [2026-07-17 10:35:37] step 257100/1000000 | loss=2.3642 vocab=0.9335 token=1.3199 | 6.12 steps/s | VRAM=31.23 Go [2026-07-17 10:35:45] step 257150/1000000 | loss=2.3556 vocab=0.6140 token=1.1773 | 6.07 steps/s | VRAM=24.49 Go [2026-07-17 10:35:53] step 257200/1000000 | loss=2.3344 vocab=0.7614 token=1.0248 | 6.08 steps/s | VRAM=28.39 Go [2026-07-17 10:36:02] step 257250/1000000 | loss=2.3562 vocab=0.6885 token=1.4402 | 6.09 steps/s | VRAM=31.38 Go [2026-07-17 10:36:10] step 257300/1000000 | loss=2.3267 vocab=1.0627 token=1.7161 | 6.10 steps/s | VRAM=27.75 Go [2026-07-17 10:36:18] step 257350/1000000 | loss=2.2944 vocab=0.7812 token=0.8635 | 6.11 steps/s | VRAM=28.13 Go [2026-07-17 10:36:26] step 257400/1000000 | loss=2.2667 vocab=1.1939 token=0.9085 | 6.08 steps/s | VRAM=27.50 Go [2026-07-17 10:36:34] step 257450/1000000 | loss=2.3016 vocab=1.2752 token=1.3223 | 6.07 steps/s | VRAM=27.59 Go [2026-07-17 10:36:43] step 257500/1000000 | loss=2.2593 vocab=0.7530 token=1.4795 | 6.07 steps/s | VRAM=27.96 Go [2026-07-17 10:36:51] step 257550/1000000 | loss=2.3390 vocab=1.1039 token=1.4260 | 6.06 steps/s | VRAM=28.54 Go [2026-07-17 10:50:18] step 255600/1000000 | loss=1.7163 vocab=0.8833 token=1.5154 | 7.68 steps/s | VRAM=30.02 Go [2026-07-17 10:50:26] step 255650/1000000 | loss=2.3344 vocab=0.9816 token=1.5163 | 6.19 steps/s | VRAM=29.99 Go [2026-07-17 10:50:35] step 255700/1000000 | loss=2.3136 vocab=0.9224 token=1.4002 | 6.20 steps/s | VRAM=32.02 Go [2026-07-17 10:50:43] step 255750/1000000 | loss=2.2571 vocab=1.0387 token=1.6263 | 6.17 steps/s | VRAM=32.75 Go [2026-07-17 10:50:51] step 255800/1000000 | loss=2.4165 vocab=0.6331 token=1.6472 | 6.16 steps/s | VRAM=25.42 Go [2026-07-17 10:50:59] step 255850/1000000 | loss=2.3453 vocab=0.7119 token=1.1085 | 6.12 steps/s | VRAM=25.80 Go [2026-07-17 10:51:07] step 255900/1000000 | loss=2.2965 vocab=1.0566 token=1.2633 | 6.17 steps/s | VRAM=28.98 Go [2026-07-17 10:51:15] step 255950/1000000 | loss=2.3026 vocab=1.0244 token=1.6503 | 6.19 steps/s | VRAM=29.27 Go [2026-07-17 10:51:23] step 256000/1000000 | loss=2.3922 vocab=0.7774 token=1.4906 | 6.12 steps/s | VRAM=30.93 Go [2026-07-17 10:51:31] step 256050/1000000 | loss=2.3306 vocab=1.1886 token=1.6059 | 6.10 steps/s | VRAM=28.49 Go [2026-07-17 10:51:40] step 256100/1000000 | loss=2.3410 vocab=0.8556 token=1.4736 | 6.08 steps/s | VRAM=26.89 Go [2026-07-17 10:51:48] step 256150/1000000 | loss=2.2576 vocab=0.4687 token=1.3923 | 6.11 steps/s | VRAM=29.15 Go [2026-07-17 10:51:56] step 256200/1000000 | loss=2.4598 vocab=0.6874 token=1.0631 | 6.10 steps/s | VRAM=31.04 Go [2026-07-17 10:52:04] step 256250/1000000 | loss=2.3306 vocab=1.1744 token=1.1400 | 6.06 steps/s | VRAM=26.29 Go [2026-07-17 10:52:13] step 256300/1000000 | loss=2.3504 vocab=0.9997 token=1.3573 | 6.08 steps/s | VRAM=30.97 Go [2026-07-17 10:52:21] step 256350/1000000 | loss=2.2499 vocab=0.9320 token=1.4500 | 6.09 steps/s | VRAM=28.84 Go [2026-07-17 10:52:29] step 256400/1000000 | loss=2.4608 vocab=0.7598 token=0.9748 | 6.10 steps/s | VRAM=28.37 Go [2026-07-17 10:52:37] step 256450/1000000 | loss=2.3760 vocab=0.7394 token=1.6696 | 6.04 steps/s | VRAM=32.02 Go [2026-07-17 10:52:45] step 256500/1000000 | loss=2.3179 vocab=0.9956 token=1.5173 | 6.07 steps/s | VRAM=28.12 Go [2026-07-17 10:52:54] step 256550/1000000 | loss=2.2663 vocab=0.5825 token=1.7860 | 6.07 steps/s | VRAM=29.70 Go [2026-07-17 10:53:02] step 256600/1000000 | loss=2.4131 vocab=1.0548 token=1.5396 | 6.07 steps/s | VRAM=28.43 Go [2026-07-17 10:53:10] step 256650/1000000 | loss=2.3637 vocab=0.8251 token=1.2485 | 6.08 steps/s | VRAM=27.69 Go [2026-07-17 10:53:18] step 256700/1000000 | loss=2.3168 vocab=0.7784 token=1.1002 | 6.11 steps/s | VRAM=32.87 Go [2026-07-17 10:53:27] step 256750/1000000 | loss=2.3255 vocab=0.7242 token=1.2773 | 6.03 steps/s | VRAM=24.66 Go [2026-07-17 10:53:35] step 256800/1000000 | loss=2.3148 vocab=1.2360 token=1.6449 | 6.05 steps/s | VRAM=29.41 Go [2026-07-17 10:53:43] step 256850/1000000 | loss=2.3713 vocab=0.8251 token=1.1912 | 6.06 steps/s | VRAM=28.73 Go [2026-07-17 10:53:51] step 256900/1000000 | loss=2.4040 vocab=1.1768 token=1.5813 | 6.07 steps/s | VRAM=28.98 Go [2026-07-17 10:54:00] step 256950/1000000 | loss=2.2671 vocab=1.0370 token=0.9120 | 6.08 steps/s | VRAM=26.55 Go [2026-07-17 10:54:08] step 257000/1000000 | loss=2.2707 vocab=0.9842 token=1.2092 | 6.12 steps/s | VRAM=32.59 Go [2026-07-17 10:54:16] step 257050/1000000 | loss=2.2476 vocab=1.0518 token=1.2946 | 6.09 steps/s | VRAM=28.41 Go [2026-07-17 10:54:24] step 257100/1000000 | loss=2.3221 vocab=0.8076 token=1.5983 | 6.08 steps/s | VRAM=26.52 Go [2026-07-17 10:54:32] step 257150/1000000 | loss=2.2937 vocab=0.5711 token=1.3144 | 6.09 steps/s | VRAM=31.41 Go [2026-07-17 10:54:41] step 257200/1000000 | loss=2.3797 vocab=0.8785 token=1.2090 | 6.09 steps/s | VRAM=28.65 Go [2026-07-17 10:54:49] step 257250/1000000 | loss=2.3009 vocab=0.6472 token=1.2138 | 6.09 steps/s | VRAM=26.04 Go [2026-07-17 10:54:57] step 257300/1000000 | loss=2.3410 vocab=1.2343 token=1.7609 | 6.09 steps/s | VRAM=27.67 Go [2026-07-17 10:55:05] step 257350/1000000 | loss=2.3767 vocab=0.7102 token=1.2351 | 6.08 steps/s | VRAM=32.89 Go [2026-07-17 10:55:13] step 257400/1000000 | loss=2.3344 vocab=0.5779 token=1.4867 | 6.10 steps/s | VRAM=30.76 Go [2026-07-17 10:55:22] step 257450/1000000 | loss=2.2344 vocab=0.6699 token=0.8627 | 6.07 steps/s | VRAM=26.62 Go [2026-07-17 10:55:30] step 257500/1000000 | loss=2.3683 vocab=0.7937 token=1.5327 | 6.11 steps/s | VRAM=28.42 Go [2026-07-17 10:55:38] step 257550/1000000 | loss=2.2963 vocab=0.9795 token=1.2185 | 6.09 steps/s | VRAM=31.41 Go [2026-07-17 10:55:46] step 257600/1000000 | loss=2.4233 vocab=1.9302 token=1.6317 | 6.11 steps/s | VRAM=28.83 Go [2026-07-17 10:55:55] step 257650/1000000 | loss=2.2727 vocab=1.0919 token=1.8833 | 6.09 steps/s | VRAM=26.19 Go [2026-07-17 10:56:03] step 257700/1000000 | loss=2.3406 vocab=1.1751 token=1.3295 | 6.11 steps/s | VRAM=31.46 Go [2026-07-17 10:56:11] step 257750/1000000 | loss=2.3430 vocab=0.7548 token=1.3056 | 6.08 steps/s | VRAM=26.69 Go [2026-07-17 10:56:19] step 257800/1000000 | loss=2.4368 vocab=1.1032 token=1.1696 | 6.12 steps/s | VRAM=32.75 Go [2026-07-17 10:56:27] step 257850/1000000 | loss=2.3432 vocab=1.0890 token=1.7019 | 6.09 steps/s | VRAM=27.73 Go [2026-07-17 10:56:35] step 257900/1000000 | loss=2.3777 vocab=0.9542 token=1.7076 | 6.13 steps/s | VRAM=29.04 Go [2026-07-17 10:56:44] step 257950/1000000 | loss=2.3572 vocab=0.4559 token=1.3039 | 6.10 steps/s | VRAM=28.87 Go [2026-07-17 10:56:52] step 258000/1000000 | loss=2.3881 vocab=1.4768 token=1.3507 | 6.09 steps/s | VRAM=28.18 Go [2026-07-17 10:57:00] step 258050/1000000 | loss=2.2392 vocab=0.6734 token=1.3210 | 6.09 steps/s | VRAM=28.56 Go [2026-07-17 10:57:08] step 258100/1000000 | loss=2.4040 vocab=0.9916 token=1.2691 | 6.10 steps/s | VRAM=28.20 Go [2026-07-17 10:57:16] step 258150/1000000 | loss=2.2834 vocab=1.0190 token=1.2724 | 6.09 steps/s | VRAM=32.66 Go [2026-07-17 10:57:25] step 258200/1000000 | loss=2.3366 vocab=0.8538 token=1.1703 | 6.09 steps/s | VRAM=29.56 Go [2026-07-17 10:57:33] step 258250/1000000 | loss=2.3796 vocab=1.3225 token=1.7256 | 6.06 steps/s | VRAM=26.75 Go [2026-07-17 10:57:41] step 258300/1000000 | loss=2.3222 vocab=1.0087 token=1.6084 | 6.12 steps/s | VRAM=32.00 Go [2026-07-17 10:57:49] step 258350/1000000 | loss=2.3247 vocab=1.0325 token=1.4413 | 6.10 steps/s | VRAM=31.63 Go [2026-07-17 10:57:57] step 258400/1000000 | loss=2.3724 vocab=0.8309 token=1.1354 | 6.11 steps/s | VRAM=27.83 Go [2026-07-17 10:58:06] step 258450/1000000 | loss=2.3157 vocab=0.7635 token=1.3817 | 6.06 steps/s | VRAM=26.58 Go [2026-07-17 10:58:14] step 258500/1000000 | loss=2.3945 vocab=1.4842 token=1.5742 | 6.10 steps/s | VRAM=31.91 Go [2026-07-17 10:58:22] step 258550/1000000 | loss=2.3215 vocab=1.1056 token=1.6576 | 6.11 steps/s | VRAM=26.22 Go [2026-07-17 10:58:30] step 258600/1000000 | loss=2.2949 vocab=1.0839 token=1.2846 | 6.14 steps/s | VRAM=29.14 Go [2026-07-17 10:58:38] step 258650/1000000 | loss=2.4228 vocab=0.6332 token=1.4760 | 6.09 steps/s | VRAM=29.10 Go [2026-07-17 10:58:47] step 258700/1000000 | loss=2.3885 vocab=0.6713 token=1.4171 | 6.10 steps/s | VRAM=29.05 Go [2026-07-17 10:58:55] step 258750/1000000 | loss=2.4075 vocab=1.2916 token=1.3559 | 6.13 steps/s | VRAM=29.02 Go [2026-07-17 10:59:03] step 258800/1000000 | loss=2.3359 vocab=1.1032 token=1.3311 | 6.11 steps/s | VRAM=32.88 Go [2026-07-17 10:59:11] step 258850/1000000 | loss=2.3584 vocab=0.8129 token=1.4078 | 6.08 steps/s | VRAM=26.63 Go [2026-07-17 10:59:19] step 258900/1000000 | loss=2.4164 vocab=1.1376 token=1.6226 | 6.10 steps/s | VRAM=28.75 Go [2026-07-17 10:59:28] step 258950/1000000 | loss=2.3018 vocab=1.1265 token=1.5062 | 6.10 steps/s | VRAM=28.23 Go [2026-07-17 10:59:36] step 259000/1000000 | loss=2.3062 vocab=1.0130 token=1.2512 | 6.08 steps/s | VRAM=29.54 Go [2026-07-17 10:59:44] step 259050/1000000 | loss=2.2622 vocab=0.4808 token=1.1427 | 6.11 steps/s | VRAM=32.21 Go [2026-07-17 10:59:52] step 259100/1000000 | loss=2.3504 vocab=0.8983 token=1.0902 | 6.10 steps/s | VRAM=28.06 Go [2026-07-17 11:00:00] step 259150/1000000 | loss=2.2483 vocab=0.7596 token=1.6995 | 6.10 steps/s | VRAM=31.99 Go [2026-07-17 11:00:09] step 259200/1000000 | loss=2.3024 vocab=0.5715 token=1.1610 | 6.10 steps/s | VRAM=29.10 Go [2026-07-17 11:00:17] step 259250/1000000 | loss=2.3861 vocab=1.5250 token=1.6176 | 6.10 steps/s | VRAM=26.73 Go [2026-07-17 11:00:25] step 259300/1000000 | loss=2.2407 vocab=0.8027 token=1.2287 | 6.13 steps/s | VRAM=31.32 Go [2026-07-17 11:00:33] step 259350/1000000 | loss=2.4957 vocab=0.6549 token=1.4431 | 6.06 steps/s | VRAM=25.58 Go [2026-07-17 11:00:41] step 259400/1000000 | loss=2.4329 vocab=1.4748 token=1.3153 | 6.11 steps/s | VRAM=31.26 Go [2026-07-17 11:00:50] step 259450/1000000 | loss=2.2605 vocab=0.7922 token=1.6557 | 6.10 steps/s | VRAM=29.18 Go [2026-07-17 11:00:58] step 259500/1000000 | loss=2.4305 vocab=1.4531 token=1.4350 | 6.11 steps/s | VRAM=32.04 Go [2026-07-17 11:01:06] step 259550/1000000 | loss=2.4106 vocab=0.7344 token=1.5654 | 6.10 steps/s | VRAM=32.95 Go [2026-07-17 11:01:14] step 259600/1000000 | loss=2.3322 vocab=0.8579 token=1.5857 | 6.12 steps/s | VRAM=28.29 Go [2026-07-17 11:01:22] step 259650/1000000 | loss=2.2444 vocab=0.6839 token=1.1222 | 6.11 steps/s | VRAM=32.79 Go [2026-07-17 11:01:31] step 259700/1000000 | loss=2.2562 vocab=0.9092 token=1.3574 | 6.10 steps/s | VRAM=26.62 Go [2026-07-17 11:01:39] step 259750/1000000 | loss=2.3348 vocab=0.8847 token=1.0025 | 6.14 steps/s | VRAM=30.33 Go [2026-07-17 11:01:47] step 259800/1000000 | loss=2.3510 vocab=1.2608 token=1.6580 | 6.10 steps/s | VRAM=29.05 Go [2026-07-17 11:01:55] step 259850/1000000 | loss=2.3673 vocab=0.7771 token=1.3669 | 6.14 steps/s | VRAM=25.28 Go [2026-07-17 11:02:03] step 259900/1000000 | loss=2.3447 vocab=1.0217 token=1.4619 | 6.10 steps/s | VRAM=27.86 Go [2026-07-17 11:02:11] step 259950/1000000 | loss=2.3611 vocab=0.8360 token=1.4810 | 6.10 steps/s | VRAM=26.80 Go [2026-07-17 11:02:20] step 260000/1000000 | loss=2.3073 vocab=0.3984 token=1.2498 | 6.13 steps/s | VRAM=28.11 Go [push] step 260000 OK (best-of-2, loss_token=1.2498) [2026-07-17 11:03:51] step 260050/1000000 | loss=2.3815 vocab=0.7779 token=1.3037 | 0.54 steps/s | VRAM=26.17 Go [2026-07-17 11:04:00] step 260100/1000000 | loss=2.2807 vocab=0.5809 token=0.9619 | 6.17 steps/s | VRAM=29.32 Go [2026-07-17 11:04:08] step 260150/1000000 | loss=2.3548 vocab=1.3948 token=1.7119 | 6.18 steps/s | VRAM=28.85 Go [2026-07-17 11:04:16] step 260200/1000000 | loss=2.3660 vocab=0.8495 token=1.5064 | 6.17 steps/s | VRAM=29.05 Go [2026-07-17 11:04:24] step 260250/1000000 | loss=2.4444 vocab=1.1747 token=1.4926 | 6.13 steps/s | VRAM=28.96 Go [2026-07-17 11:04:32] step 260300/1000000 | loss=2.2105 vocab=0.6313 token=1.1128 | 6.16 steps/s | VRAM=30.99 Go [2026-07-17 11:04:40] step 260350/1000000 | loss=2.3302 vocab=0.9882 token=1.5363 | 6.11 steps/s | VRAM=26.12 Go [2026-07-17 11:04:48] step 260400/1000000 | loss=2.3662 vocab=1.1361 token=1.5367 | 6.13 steps/s | VRAM=27.88 Go [2026-07-17 11:04:57] step 260450/1000000 | loss=2.2169 vocab=0.9402 token=1.3456 | 6.13 steps/s | VRAM=32.50 Go [2026-07-17 11:05:05] step 260500/1000000 | loss=2.3335 vocab=0.6860 token=1.0472 | 6.12 steps/s | VRAM=29.36 Go [2026-07-17 11:05:13] step 260550/1000000 | loss=2.2646 vocab=0.7660 token=1.2896 | 6.11 steps/s | VRAM=29.84 Go [2026-07-17 11:05:21] step 260600/1000000 | loss=2.3145 vocab=0.7242 token=1.3305 | 6.11 steps/s | VRAM=28.45 Go [2026-07-17 11:05:29] step 260650/1000000 | loss=2.3946 vocab=0.5315 token=1.0346 | 6.10 steps/s | VRAM=25.97 Go [2026-07-17 11:05:37] step 260700/1000000 | loss=2.2671 vocab=0.6279 token=1.5533 | 6.10 steps/s | VRAM=25.19 Go [2026-07-17 11:05:46] step 260750/1000000 | loss=2.2795 vocab=0.8597 token=1.5089 | 6.07 steps/s | VRAM=31.82 Go [2026-07-17 11:05:54] step 260800/1000000 | loss=2.3385 vocab=0.5898 token=1.3430 | 6.12 steps/s | VRAM=32.06 Go [2026-07-17 11:06:02] step 260850/1000000 | loss=2.3080 vocab=0.9470 token=1.6465 | 6.07 steps/s | VRAM=32.47 Go [2026-07-17 11:06:10] step 260900/1000000 | loss=2.3122 vocab=1.3160 token=1.4310 | 6.09 steps/s | VRAM=23.34 Go [2026-07-17 11:06:19] step 260950/1000000 | loss=2.3305 vocab=0.7996 token=1.3852 | 6.09 steps/s | VRAM=25.77 Go [2026-07-17 11:06:27] step 261000/1000000 | loss=2.3486 vocab=1.0911 token=1.3872 | 6.07 steps/s | VRAM=26.28 Go [2026-07-17 11:06:35] step 261050/1000000 | loss=2.4050 vocab=0.6747 token=1.2474 | 6.07 steps/s | VRAM=28.81 Go [2026-07-17 11:06:43] step 261100/1000000 | loss=2.2539 vocab=0.8217 token=1.4907 | 6.10 steps/s | VRAM=32.31 Go [2026-07-17 11:06:51] step 261150/1000000 | loss=2.3399 vocab=1.1607 token=1.1388 | 6.07 steps/s | VRAM=29.73 Go [2026-07-17 11:07:00] step 261200/1000000 | loss=2.4063 vocab=1.0112 token=1.9389 | 6.15 steps/s | VRAM=24.85 Go [2026-07-17 11:07:08] step 261250/1000000 | loss=2.2952 vocab=1.0032 token=1.5214 | 6.09 steps/s | VRAM=26.02 Go [2026-07-17 11:07:16] step 261300/1000000 | loss=2.3991 vocab=1.4388 token=1.5076 | 6.14 steps/s | VRAM=32.01 Go [2026-07-17 11:07:24] step 261350/1000000 | loss=2.2830 vocab=1.0430 token=1.2237 | 6.12 steps/s | VRAM=28.87 Go [2026-07-17 11:07:32] step 261400/1000000 | loss=2.3426 vocab=0.8323 token=1.3087 | 6.08 steps/s | VRAM=28.61 Go [2026-07-17 11:07:40] step 261450/1000000 | loss=2.2514 vocab=1.1477 token=1.6048 | 6.12 steps/s | VRAM=31.77 Go [2026-07-17 11:07:49] step 261500/1000000 | loss=2.2980 vocab=1.0403 token=1.3623 | 6.09 steps/s | VRAM=28.52 Go [2026-07-17 11:07:57] step 261550/1000000 | loss=2.3619 vocab=1.0018 token=1.2215 | 6.09 steps/s | VRAM=26.07 Go [2026-07-17 11:08:05] step 261600/1000000 | loss=2.2239 vocab=1.2302 token=1.3214 | 6.08 steps/s | VRAM=29.11 Go [2026-07-17 11:08:13] step 261650/1000000 | loss=2.3125 vocab=0.8353 token=1.5776 | 6.09 steps/s | VRAM=28.72 Go [2026-07-17 11:08:22] step 261700/1000000 | loss=2.3686 vocab=1.0628 token=1.4667 | 6.09 steps/s | VRAM=28.59 Go [2026-07-17 11:08:30] step 261750/1000000 | loss=2.3609 vocab=0.7582 token=1.5871 | 6.12 steps/s | VRAM=29.03 Go [2026-07-17 11:08:38] step 261800/1000000 | loss=2.3772 vocab=0.5513 token=1.2261 | 6.11 steps/s | VRAM=28.80 Go [2026-07-17 11:08:46] step 261850/1000000 | loss=2.3330 vocab=1.0126 token=1.5693 | 6.10 steps/s | VRAM=26.13 Go [2026-07-17 11:08:54] step 261900/1000000 | loss=2.3370 vocab=0.6745 token=1.3978 | 6.13 steps/s | VRAM=28.34 Go [2026-07-17 11:09:02] step 261950/1000000 | loss=2.4425 vocab=0.7162 token=1.2269 | 6.12 steps/s | VRAM=29.41 Go [2026-07-17 11:09:11] step 262000/1000000 | loss=2.2369 vocab=1.1031 token=1.5872 | 6.10 steps/s | VRAM=32.44 Go [2026-07-17 11:09:19] step 262050/1000000 | loss=2.3283 vocab=1.0396 token=1.3903 | 6.12 steps/s | VRAM=29.43 Go [2026-07-17 11:09:27] step 262100/1000000 | loss=2.2407 vocab=1.1669 token=1.3590 | 6.12 steps/s | VRAM=29.70 Go [2026-07-17 11:09:35] step 262150/1000000 | loss=2.2544 vocab=0.9794 token=1.6845 | 6.08 steps/s | VRAM=28.91 Go [2026-07-17 11:09:43] step 262200/1000000 | loss=2.4077 vocab=0.7023 token=1.0781 | 6.08 steps/s | VRAM=28.97 Go [2026-07-17 11:09:52] step 262250/1000000 | loss=2.3512 vocab=0.4805 token=1.1406 | 6.09 steps/s | VRAM=29.12 Go [2026-07-17 11:10:00] step 262300/1000000 | loss=2.3139 vocab=1.3444 token=1.7146 | 6.10 steps/s | VRAM=26.50 Go [2026-07-17 11:10:08] step 262350/1000000 | loss=2.5109 vocab=0.6868 token=1.3851 | 6.12 steps/s | VRAM=26.49 Go [2026-07-17 11:10:16] step 262400/1000000 | loss=2.3461 vocab=1.1802 token=1.4388 | 6.09 steps/s | VRAM=28.49 Go [2026-07-17 11:10:24] step 262450/1000000 | loss=2.3511 vocab=0.7830 token=1.3186 | 6.10 steps/s | VRAM=26.30 Go [2026-07-17 11:10:33] step 262500/1000000 | loss=2.3064 vocab=0.9158 token=1.6662 | 6.11 steps/s | VRAM=28.67 Go [2026-07-17 11:10:41] step 262550/1000000 | loss=2.3008 vocab=0.7871 token=1.6170 | 6.12 steps/s | VRAM=32.57 Go [2026-07-17 11:10:49] step 262600/1000000 | loss=2.3037 vocab=1.1562 token=1.1546 | 6.10 steps/s | VRAM=30.99 Go [2026-07-17 11:10:57] step 262650/1000000 | loss=2.3178 vocab=0.8510 token=1.3335 | 6.11 steps/s | VRAM=24.54 Go [2026-07-17 11:11:05] step 262700/1000000 | loss=2.3052 vocab=0.9919 token=1.1453 | 6.13 steps/s | VRAM=29.08 Go [2026-07-17 11:11:13] step 262750/1000000 | loss=2.4917 vocab=1.5814 token=1.4670 | 6.11 steps/s | VRAM=29.81 Go [2026-07-17 11:11:22] step 262800/1000000 | loss=2.3710 vocab=0.8776 token=1.3359 | 6.12 steps/s | VRAM=28.43 Go [2026-07-17 11:11:30] step 262850/1000000 | loss=2.3543 vocab=1.1263 token=1.7674 | 6.12 steps/s | VRAM=32.31 Go [2026-07-17 11:11:38] step 262900/1000000 | loss=2.2826 vocab=1.0264 token=1.5940 | 6.09 steps/s | VRAM=23.59 Go [2026-07-17 11:11:46] step 262950/1000000 | loss=2.3641 vocab=1.1687 token=1.6591 | 6.16 steps/s | VRAM=30.80 Go [2026-07-17 11:11:54] step 263000/1000000 | loss=2.2163 vocab=0.7055 token=1.3632 | 6.09 steps/s | VRAM=26.05 Go [2026-07-17 11:12:03] step 263050/1000000 | loss=2.3224 vocab=1.0892 token=1.3434 | 6.10 steps/s | VRAM=25.72 Go [2026-07-17 11:12:11] step 263100/1000000 | loss=2.2369 vocab=1.2394 token=1.6405 | 6.14 steps/s | VRAM=32.26 Go [2026-07-17 11:12:19] step 263150/1000000 | loss=2.4385 vocab=1.1351 token=1.4046 | 6.09 steps/s | VRAM=32.66 Go [2026-07-17 11:12:27] step 263200/1000000 | loss=2.2884 vocab=0.7998 token=1.4403 | 6.11 steps/s | VRAM=32.98 Go [2026-07-17 11:12:35] step 263250/1000000 | loss=2.2655 vocab=0.7596 token=1.0762 | 6.11 steps/s | VRAM=25.87 Go [2026-07-17 11:12:43] step 263300/1000000 | loss=2.3417 vocab=0.9542 token=1.1253 | 6.10 steps/s | VRAM=32.27 Go [2026-07-17 11:12:52] step 263350/1000000 | loss=2.2961 vocab=1.2426 token=1.7383 | 6.06 steps/s | VRAM=29.06 Go [2026-07-17 11:13:00] step 263400/1000000 | loss=2.3159 vocab=0.9080 token=1.1897 | 6.12 steps/s | VRAM=29.00 Go [2026-07-17 11:13:08] step 263450/1000000 | loss=2.3037 vocab=0.6498 token=1.0512 | 6.13 steps/s | VRAM=31.78 Go [2026-07-17 11:13:16] step 263500/1000000 | loss=2.4616 vocab=0.9978 token=1.4071 | 6.13 steps/s | VRAM=25.96 Go [2026-07-17 11:13:24] step 263550/1000000 | loss=2.2863 vocab=0.6155 token=1.3672 | 6.12 steps/s | VRAM=29.57 Go [2026-07-17 11:13:33] step 263600/1000000 | loss=2.4257 vocab=0.7047 token=1.5153 | 6.12 steps/s | VRAM=26.14 Go [2026-07-17 11:13:41] step 263650/1000000 | loss=2.2867 vocab=1.2341 token=1.3671 | 6.13 steps/s | VRAM=28.30 Go [2026-07-17 11:13:49] step 263700/1000000 | loss=2.3671 vocab=1.1514 token=1.2459 | 6.11 steps/s | VRAM=30.82 Go [2026-07-17 11:13:57] step 263750/1000000 | loss=2.3911 vocab=0.8800 token=1.0040 | 6.10 steps/s | VRAM=29.98 Go [2026-07-17 11:14:05] step 263800/1000000 | loss=2.2406 vocab=0.9962 token=1.0934 | 6.10 steps/s | VRAM=28.62 Go [2026-07-17 11:14:13] step 263850/1000000 | loss=2.2874 vocab=0.7080 token=1.1056 | 6.10 steps/s | VRAM=32.93 Go [2026-07-17 11:14:22] step 263900/1000000 | loss=2.3960 vocab=1.1358 token=1.4961 | 6.13 steps/s | VRAM=29.36 Go [2026-07-17 11:14:30] step 263950/1000000 | loss=2.2785 vocab=0.6722 token=1.3026 | 6.13 steps/s | VRAM=31.58 Go [2026-07-17 11:14:38] step 264000/1000000 | loss=2.3116 vocab=0.8318 token=1.4756 | 6.13 steps/s | VRAM=26.81 Go [2026-07-17 11:14:46] step 264050/1000000 | loss=2.2626 vocab=0.5996 token=1.0554 | 6.10 steps/s | VRAM=26.45 Go [2026-07-17 11:14:54] step 264100/1000000 | loss=2.3968 vocab=0.5465 token=1.4745 | 6.12 steps/s | VRAM=25.68 Go [2026-07-17 11:15:02] step 264150/1000000 | loss=2.3695 vocab=1.2729 token=1.4178 | 6.12 steps/s | VRAM=31.54 Go [2026-07-17 11:15:11] step 264200/1000000 | loss=2.3152 vocab=1.0267 token=1.0372 | 6.11 steps/s | VRAM=29.24 Go [2026-07-17 11:15:19] step 264250/1000000 | loss=2.4669 vocab=1.5635 token=1.8389 | 6.12 steps/s | VRAM=28.91 Go [2026-07-17 11:15:27] step 264300/1000000 | loss=2.3567 vocab=1.4336 token=1.6501 | 6.12 steps/s | VRAM=28.72 Go [2026-07-17 11:15:35] step 264350/1000000 | loss=2.3763 vocab=1.3645 token=1.2481 | 6.09 steps/s | VRAM=24.95 Go [2026-07-17 11:15:43] step 264400/1000000 | loss=2.2651 vocab=0.9671 token=1.4195 | 6.08 steps/s | VRAM=23.73 Go [2026-07-17 11:15:52] step 264450/1000000 | loss=2.3910 vocab=0.8574 token=1.3158 | 6.13 steps/s | VRAM=25.13 Go [2026-07-17 11:16:00] step 264500/1000000 | loss=2.4115 vocab=1.0619 token=1.5819 | 6.11 steps/s | VRAM=29.88 Go [2026-07-17 11:16:08] step 264550/1000000 | loss=2.3560 vocab=0.7190 token=1.0406 | 6.13 steps/s | VRAM=24.49 Go [2026-07-17 11:16:16] step 264600/1000000 | loss=2.3206 vocab=0.9058 token=1.3544 | 6.12 steps/s | VRAM=32.08 Go [2026-07-17 11:16:24] step 264650/1000000 | loss=2.3574 vocab=0.6109 token=1.0985 | 6.09 steps/s | VRAM=28.25 Go [2026-07-17 11:16:33] step 264700/1000000 | loss=2.3095 vocab=0.7421 token=1.6395 | 6.09 steps/s | VRAM=25.70 Go [2026-07-17 11:16:41] step 264750/1000000 | loss=2.3427 vocab=0.7746 token=1.6146 | 6.12 steps/s | VRAM=32.10 Go [2026-07-17 11:16:49] step 264800/1000000 | loss=2.2945 vocab=0.7523 token=1.3910 | 6.12 steps/s | VRAM=29.15 Go [2026-07-17 11:16:57] step 264850/1000000 | loss=2.2793 vocab=0.6308 token=1.1747 | 6.12 steps/s | VRAM=26.52 Go [2026-07-17 11:17:05] step 264900/1000000 | loss=2.2073 vocab=0.7744 token=1.3093 | 6.10 steps/s | VRAM=28.57 Go [2026-07-17 11:17:13] step 264950/1000000 | loss=2.3440 vocab=1.4370 token=1.4580 | 6.11 steps/s | VRAM=29.85 Go [2026-07-17 11:17:22] step 265000/1000000 | loss=2.3620 vocab=0.5731 token=1.3292 | 6.09 steps/s | VRAM=28.76 Go [step 265000] loss_token=1.3292 > min recent (1.2498) -> push saute [2026-07-17 11:17:30] step 265050/1000000 | loss=2.3508 vocab=0.6990 token=1.3196 | 6.11 steps/s | VRAM=28.63 Go [2026-07-17 11:17:38] step 265100/1000000 | loss=2.2932 vocab=1.6489 token=1.7167 | 6.08 steps/s | VRAM=29.06 Go [2026-07-17 11:17:46] step 265150/1000000 | loss=2.2724 vocab=1.2419 token=1.0883 | 6.12 steps/s | VRAM=29.11 Go [2026-07-17 11:17:54] step 265200/1000000 | loss=2.3723 vocab=1.1245 token=1.0907 | 6.10 steps/s | VRAM=25.82 Go [2026-07-17 11:18:03] step 265250/1000000 | loss=2.4444 vocab=0.6186 token=1.3841 | 6.13 steps/s | VRAM=28.55 Go [2026-07-17 11:18:11] step 265300/1000000 | loss=2.3365 vocab=0.7303 token=1.3527 | 6.11 steps/s | VRAM=26.60 Go [2026-07-17 11:18:19] step 265350/1000000 | loss=2.3133 vocab=0.9358 token=1.4028 | 6.12 steps/s | VRAM=28.28 Go [2026-07-17 11:18:27] step 265400/1000000 | loss=2.2960 vocab=1.0898 token=1.3759 | 6.10 steps/s | VRAM=32.31 Go [2026-07-17 11:18:35] step 265450/1000000 | loss=2.3215 vocab=1.3901 token=1.3766 | 6.12 steps/s | VRAM=29.48 Go [2026-07-17 11:18:44] step 265500/1000000 | loss=2.3055 vocab=0.6249 token=1.3176 | 6.08 steps/s | VRAM=26.09 Go [2026-07-17 11:18:52] step 265550/1000000 | loss=2.3153 vocab=1.1769 token=1.4678 | 6.12 steps/s | VRAM=29.09 Go [2026-07-17 11:19:00] step 265600/1000000 | loss=2.3550 vocab=0.8720 token=1.3006 | 6.11 steps/s | VRAM=32.77 Go [2026-07-17 11:19:08] step 265650/1000000 | loss=2.3181 vocab=0.7973 token=1.2235 | 6.12 steps/s | VRAM=25.96 Go [2026-07-17 11:19:16] step 265700/1000000 | loss=2.3785 vocab=1.1528 token=1.3050 | 6.11 steps/s | VRAM=31.09 Go [2026-07-17 11:19:24] step 265750/1000000 | loss=2.3247 vocab=0.9584 token=1.3039 | 6.07 steps/s | VRAM=29.95 Go [2026-07-17 11:19:33] step 265800/1000000 | loss=2.3411 vocab=1.1631 token=1.4888 | 6.09 steps/s | VRAM=25.70 Go [2026-07-17 11:19:41] step 265850/1000000 | loss=2.3270 vocab=0.6434 token=1.2637 | 6.10 steps/s | VRAM=28.24 Go [2026-07-17 11:19:49] step 265900/1000000 | loss=2.3697 vocab=1.2607 token=1.4369 | 6.11 steps/s | VRAM=28.42 Go [2026-07-17 11:19:57] step 265950/1000000 | loss=2.3094 vocab=0.7844 token=1.3336 | 6.12 steps/s | VRAM=25.86 Go [2026-07-17 11:20:05] step 266000/1000000 | loss=2.3217 vocab=0.9412 token=1.4680 | 6.10 steps/s | VRAM=29.50 Go [2026-07-17 11:20:14] step 266050/1000000 | loss=2.3853 vocab=1.2809 token=1.4043 | 6.13 steps/s | VRAM=29.10 Go [2026-07-17 11:20:22] step 266100/1000000 | loss=2.4418 vocab=0.9033 token=1.0014 | 6.10 steps/s | VRAM=31.67 Go [2026-07-17 11:20:30] step 266150/1000000 | loss=2.3158 vocab=1.2061 token=1.3840 | 6.09 steps/s | VRAM=29.03 Go [2026-07-17 11:20:38] step 266200/1000000 | loss=2.4038 vocab=0.7520 token=1.3847 | 6.11 steps/s | VRAM=29.78 Go [2026-07-17 11:20:46] step 266250/1000000 | loss=2.3600 vocab=1.0465 token=1.2183 | 6.09 steps/s | VRAM=28.81 Go [2026-07-17 11:20:55] step 266300/1000000 | loss=2.3451 vocab=1.1612 token=1.6219 | 6.07 steps/s | VRAM=29.00 Go [2026-07-17 11:21:03] step 266350/1000000 | loss=2.3179 vocab=1.1001 token=1.3912 | 6.11 steps/s | VRAM=32.12 Go [2026-07-17 11:21:11] step 266400/1000000 | loss=2.2517 vocab=0.6891 token=0.9917 | 6.10 steps/s | VRAM=32.54 Go [2026-07-17 11:21:19] step 266450/1000000 | loss=2.3138 vocab=0.4916 token=1.0335 | 6.12 steps/s | VRAM=28.35 Go [2026-07-17 11:21:27] step 266500/1000000 | loss=2.3926 vocab=0.6818 token=1.2755 | 6.12 steps/s | VRAM=28.90 Go [2026-07-17 11:21:36] step 266550/1000000 | loss=2.3205 vocab=0.8742 token=1.2326 | 6.10 steps/s | VRAM=28.03 Go [2026-07-17 11:21:44] step 266600/1000000 | loss=2.3570 vocab=0.4644 token=1.1362 | 6.05 steps/s | VRAM=28.20 Go [2026-07-17 11:21:52] step 266650/1000000 | loss=2.2682 vocab=0.9263 token=1.1723 | 6.09 steps/s | VRAM=29.41 Go [2026-07-17 11:22:00] step 266700/1000000 | loss=2.3902 vocab=0.7139 token=1.6211 | 6.14 steps/s | VRAM=29.31 Go [2026-07-17 11:22:08] step 266750/1000000 | loss=2.2894 vocab=1.1395 token=1.4237 | 6.14 steps/s | VRAM=32.45 Go [2026-07-17 11:22:16] step 266800/1000000 | loss=2.2435 vocab=0.8475 token=1.1388 | 6.09 steps/s | VRAM=28.23 Go [2026-07-17 11:22:25] step 266850/1000000 | loss=2.3231 vocab=0.9551 token=1.6180 | 6.14 steps/s | VRAM=25.46 Go [2026-07-17 11:22:33] step 266900/1000000 | loss=2.4179 vocab=1.2820 token=1.3883 | 6.09 steps/s | VRAM=28.51 Go [2026-07-17 11:22:41] step 266950/1000000 | loss=2.3831 vocab=1.1100 token=1.7297 | 6.11 steps/s | VRAM=28.52 Go [2026-07-17 11:22:49] step 267000/1000000 | loss=2.2707 vocab=0.6036 token=1.2212 | 6.10 steps/s | VRAM=32.57 Go [2026-07-17 11:22:57] step 267050/1000000 | loss=2.2975 vocab=1.2642 token=1.5946 | 6.09 steps/s | VRAM=23.11 Go [2026-07-17 11:23:06] step 267100/1000000 | loss=2.3343 vocab=0.6093 token=1.2093 | 6.12 steps/s | VRAM=25.24 Go [2026-07-17 11:23:14] step 267150/1000000 | loss=2.3793 vocab=0.5180 token=0.9823 | 6.11 steps/s | VRAM=26.01 Go [2026-07-17 11:23:22] step 267200/1000000 | loss=2.3724 vocab=1.0927 token=1.2200 | 6.13 steps/s | VRAM=28.17 Go [2026-07-17 11:23:30] step 267250/1000000 | loss=2.2918 vocab=1.3209 token=1.1722 | 6.09 steps/s | VRAM=29.97 Go [2026-07-17 11:23:38] step 267300/1000000 | loss=2.4805 vocab=0.6304 token=1.2001 | 6.13 steps/s | VRAM=29.04 Go [2026-07-17 11:23:46] step 267350/1000000 | loss=2.4494 vocab=0.8136 token=1.5289 | 6.12 steps/s | VRAM=27.32 Go [2026-07-17 11:23:55] step 267400/1000000 | loss=2.3400 vocab=0.8105 token=1.3957 | 6.13 steps/s | VRAM=32.17 Go [2026-07-17 11:24:03] step 267450/1000000 | loss=2.3256 vocab=1.2396 token=1.4642 | 6.11 steps/s | VRAM=26.76 Go [2026-07-17 11:24:11] step 267500/1000000 | loss=2.3370 vocab=1.1710 token=1.2819 | 6.10 steps/s | VRAM=27.98 Go [2026-07-17 11:24:19] step 267550/1000000 | loss=2.2521 vocab=1.1823 token=1.6831 | 6.05 steps/s | VRAM=23.32 Go [2026-07-17 11:24:27] step 267600/1000000 | loss=2.3124 vocab=1.2643 token=1.4744 | 6.11 steps/s | VRAM=31.67 Go [2026-07-17 11:24:36] step 267650/1000000 | loss=2.4100 vocab=0.8265 token=1.5963 | 6.12 steps/s | VRAM=31.64 Go [2026-07-17 11:24:44] step 267700/1000000 | loss=2.3708 vocab=0.5679 token=1.2237 | 6.10 steps/s | VRAM=28.94 Go [2026-07-17 11:24:52] step 267750/1000000 | loss=2.3852 vocab=1.4345 token=1.7904 | 6.09 steps/s | VRAM=32.37 Go [2026-07-17 11:25:00] step 267800/1000000 | loss=2.3247 vocab=0.6625 token=1.2404 | 6.11 steps/s | VRAM=32.44 Go [2026-07-17 11:25:08] step 267850/1000000 | loss=2.3700 vocab=0.6707 token=1.2659 | 6.09 steps/s | VRAM=29.38 Go [2026-07-17 11:25:17] step 267900/1000000 | loss=2.3128 vocab=1.0043 token=1.7518 | 6.13 steps/s | VRAM=25.98 Go [2026-07-17 11:25:25] step 267950/1000000 | loss=2.3950 vocab=0.6690 token=1.1447 | 6.09 steps/s | VRAM=28.14 Go [2026-07-17 11:25:33] step 268000/1000000 | loss=2.3334 vocab=1.0731 token=1.4231 | 6.12 steps/s | VRAM=28.27 Go [2026-07-17 11:25:41] step 268050/1000000 | loss=2.4380 vocab=0.7330 token=1.1558 | 6.09 steps/s | VRAM=28.86 Go [2026-07-17 11:25:49] step 268100/1000000 | loss=2.2753 vocab=1.3330 token=1.2572 | 6.08 steps/s | VRAM=28.16 Go [2026-07-17 11:25:58] step 268150/1000000 | loss=2.3204 vocab=1.1237 token=1.1581 | 6.13 steps/s | VRAM=25.40 Go [2026-07-17 11:26:06] step 268200/1000000 | loss=2.2417 vocab=1.0345 token=1.3005 | 6.11 steps/s | VRAM=28.95 Go [2026-07-17 11:26:14] step 268250/1000000 | loss=2.3516 vocab=0.9124 token=1.1545 | 6.12 steps/s | VRAM=31.69 Go [2026-07-17 11:26:22] step 268300/1000000 | loss=2.3687 vocab=1.0567 token=1.3295 | 6.11 steps/s | VRAM=29.62 Go [2026-07-17 11:26:30] step 268350/1000000 | loss=2.3479 vocab=0.7661 token=0.7454 | 6.14 steps/s | VRAM=31.33 Go [2026-07-17 11:26:38] step 268400/1000000 | loss=2.2941 vocab=0.9030 token=1.1834 | 6.11 steps/s | VRAM=29.71 Go [2026-07-17 11:26:47] step 268450/1000000 | loss=2.2779 vocab=1.2415 token=1.4268 | 6.14 steps/s | VRAM=28.59 Go [2026-07-17 11:26:55] step 268500/1000000 | loss=2.3295 vocab=0.5415 token=1.2290 | 6.14 steps/s | VRAM=26.19 Go [2026-07-17 11:27:03] step 268550/1000000 | loss=2.4021 vocab=0.9919 token=1.4288 | 6.11 steps/s | VRAM=31.92 Go [2026-07-17 11:27:11] step 268600/1000000 | loss=2.3621 vocab=1.1110 token=1.8258 | 6.10 steps/s | VRAM=28.54 Go [2026-07-17 11:27:19] step 268650/1000000 | loss=2.4662 vocab=1.0565 token=1.3138 | 6.10 steps/s | VRAM=28.37 Go [2026-07-17 11:27:27] step 268700/1000000 | loss=2.2641 vocab=0.9535 token=1.0861 | 6.14 steps/s | VRAM=32.62 Go [2026-07-17 11:27:36] step 268750/1000000 | loss=2.3137 vocab=1.4700 token=1.8575 | 6.09 steps/s | VRAM=32.45 Go [2026-07-17 11:27:44] step 268800/1000000 | loss=2.2691 vocab=1.7147 token=1.9061 | 6.09 steps/s | VRAM=29.91 Go [2026-07-17 11:27:52] step 268850/1000000 | loss=2.2516 vocab=0.6921 token=1.1531 | 6.10 steps/s | VRAM=25.80 Go [2026-07-17 11:28:00] step 268900/1000000 | loss=2.3657 vocab=1.2307 token=1.5491 | 6.11 steps/s | VRAM=28.92 Go [2026-07-17 11:28:08] step 268950/1000000 | loss=2.3337 vocab=0.9716 token=1.8568 | 6.11 steps/s | VRAM=26.51 Go [2026-07-17 11:28:17] step 269000/1000000 | loss=2.3091 vocab=0.7573 token=1.4056 | 6.10 steps/s | VRAM=28.72 Go [2026-07-17 11:28:25] step 269050/1000000 | loss=2.2101 vocab=0.8795 token=1.4293 | 6.11 steps/s | VRAM=29.92 Go [2026-07-17 11:28:33] step 269100/1000000 | loss=2.3702 vocab=0.6740 token=0.9607 | 6.12 steps/s | VRAM=32.30 Go [2026-07-17 11:28:41] step 269150/1000000 | loss=2.2986 vocab=0.8519 token=1.3351 | 6.14 steps/s | VRAM=31.32 Go [2026-07-17 11:28:49] step 269200/1000000 | loss=2.2941 vocab=1.0438 token=1.5815 | 6.10 steps/s | VRAM=28.76 Go [2026-07-17 11:28:57] step 269250/1000000 | loss=2.3865 vocab=1.2953 token=1.4953 | 6.11 steps/s | VRAM=32.47 Go [2026-07-17 11:29:06] step 269300/1000000 | loss=2.3518 vocab=0.9917 token=1.4809 | 6.09 steps/s | VRAM=28.41 Go [2026-07-17 11:29:14] step 269350/1000000 | loss=2.4135 vocab=1.4366 token=1.1910 | 6.12 steps/s | VRAM=29.08 Go [2026-07-17 11:29:22] step 269400/1000000 | loss=2.3269 vocab=1.1916 token=1.3371 | 6.13 steps/s | VRAM=25.79 Go [2026-07-17 11:29:30] step 269450/1000000 | loss=2.2724 vocab=0.8628 token=0.9966 | 6.11 steps/s | VRAM=31.92 Go [2026-07-17 11:29:38] step 269500/1000000 | loss=2.4226 vocab=1.0570 token=1.2726 | 6.13 steps/s | VRAM=31.10 Go [2026-07-17 11:29:47] step 269550/1000000 | loss=2.4439 vocab=0.8796 token=1.8032 | 6.09 steps/s | VRAM=24.30 Go [2026-07-17 11:29:55] step 269600/1000000 | loss=2.2607 vocab=1.1791 token=1.2152 | 6.13 steps/s | VRAM=32.61 Go [2026-07-17 11:30:03] step 269650/1000000 | loss=2.3346 vocab=1.1394 token=1.5894 | 6.13 steps/s | VRAM=32.22 Go [2026-07-17 11:30:11] step 269700/1000000 | loss=2.2840 vocab=0.9582 token=1.5750 | 6.12 steps/s | VRAM=29.46 Go [2026-07-17 11:30:19] step 269750/1000000 | loss=2.3003 vocab=0.5540 token=1.0670 | 6.14 steps/s | VRAM=31.36 Go [2026-07-17 11:30:27] step 269800/1000000 | loss=2.2795 vocab=0.4317 token=1.5109 | 6.12 steps/s | VRAM=29.85 Go [2026-07-17 11:30:36] step 269850/1000000 | loss=2.2942 vocab=0.8247 token=1.3335 | 6.10 steps/s | VRAM=30.03 Go [2026-07-17 11:30:44] step 269900/1000000 | loss=2.2569 vocab=1.0127 token=1.5842 | 6.14 steps/s | VRAM=25.58 Go [2026-07-17 11:30:52] step 269950/1000000 | loss=2.2611 vocab=0.8225 token=1.1685 | 6.16 steps/s | VRAM=29.65 Go [2026-07-17 11:31:00] step 270000/1000000 | loss=2.3954 vocab=1.1110 token=1.3628 | 6.14 steps/s | VRAM=23.70 Go [step 270000] loss_token=1.3628 > min recent (1.3292) -> push saute [2026-07-17 11:31:08] step 270050/1000000 | loss=2.4054 vocab=1.0119 token=1.5182 | 6.15 steps/s | VRAM=31.55 Go [2026-07-17 11:31:16] step 270100/1000000 | loss=2.3298 vocab=1.3068 token=1.2264 | 6.13 steps/s | VRAM=26.20 Go [2026-07-17 11:31:24] step 270150/1000000 | loss=2.2949 vocab=0.7695 token=1.1903 | 6.15 steps/s | VRAM=25.48 Go [2026-07-17 11:31:33] step 270200/1000000 | loss=2.3690 vocab=1.3118 token=1.5378 | 6.14 steps/s | VRAM=33.02 Go [2026-07-17 11:31:41] step 270250/1000000 | loss=2.3563 vocab=0.9836 token=1.2682 | 6.13 steps/s | VRAM=29.86 Go [2026-07-17 11:31:49] step 270300/1000000 | loss=2.2612 vocab=1.1651 token=1.5619 | 6.12 steps/s | VRAM=28.83 Go [2026-07-17 11:31:57] step 270350/1000000 | loss=2.4471 vocab=0.6053 token=1.4303 | 6.08 steps/s | VRAM=26.42 Go [2026-07-17 11:32:05] step 270400/1000000 | loss=2.4120 vocab=0.9397 token=1.6182 | 6.11 steps/s | VRAM=29.09 Go [2026-07-17 11:32:13] step 270450/1000000 | loss=2.3849 vocab=0.7478 token=1.3580 | 6.13 steps/s | VRAM=29.65 Go [2026-07-17 11:32:22] step 270500/1000000 | loss=2.2849 vocab=1.0808 token=1.1882 | 6.10 steps/s | VRAM=28.49 Go [2026-07-17 11:32:30] step 270550/1000000 | loss=2.3524 vocab=0.4961 token=1.3191 | 6.13 steps/s | VRAM=28.59 Go [2026-07-17 11:32:38] step 270600/1000000 | loss=2.3128 vocab=0.6483 token=1.4888 | 6.14 steps/s | VRAM=32.13 Go [2026-07-17 11:32:46] step 270650/1000000 | loss=2.3032 vocab=1.0967 token=1.3628 | 6.12 steps/s | VRAM=29.32 Go [2026-07-17 11:32:54] step 270700/1000000 | loss=2.3242 vocab=1.0350 token=1.5351 | 6.12 steps/s | VRAM=28.95 Go [2026-07-17 11:33:02] step 270750/1000000 | loss=2.3089 vocab=1.1386 token=1.9392 | 6.12 steps/s | VRAM=32.50 Go [2026-07-17 11:33:11] step 270800/1000000 | loss=2.4283 vocab=0.8564 token=1.0611 | 6.08 steps/s | VRAM=28.66 Go [2026-07-17 11:33:19] step 270850/1000000 | loss=2.3134 vocab=0.9180 token=1.2019 | 6.15 steps/s | VRAM=31.73 Go [2026-07-17 11:33:27] step 270900/1000000 | loss=2.2093 vocab=0.8475 token=0.8486 | 6.13 steps/s | VRAM=29.20 Go [2026-07-17 11:33:35] step 270950/1000000 | loss=2.4044 vocab=0.5801 token=1.4499 | 6.13 steps/s | VRAM=29.28 Go [2026-07-17 11:33:43] step 271000/1000000 | loss=2.2795 vocab=1.0659 token=1.2232 | 6.12 steps/s | VRAM=29.59 Go [2026-07-17 11:33:51] step 271050/1000000 | loss=2.2500 vocab=0.9254 token=1.3410 | 6.13 steps/s | VRAM=25.77 Go [2026-07-17 11:34:00] step 271100/1000000 | loss=2.2736 vocab=0.6627 token=1.1725 | 6.13 steps/s | VRAM=26.26 Go [2026-07-17 11:34:08] step 271150/1000000 | loss=2.2766 vocab=1.0000 token=1.4380 | 6.14 steps/s | VRAM=28.85 Go [2026-07-17 11:34:16] step 271200/1000000 | loss=2.3101 vocab=0.8552 token=1.1766 | 6.14 steps/s | VRAM=31.93 Go [2026-07-17 11:34:24] step 271250/1000000 | loss=2.3478 vocab=1.4017 token=1.8417 | 6.14 steps/s | VRAM=31.29 Go [2026-07-17 11:34:32] step 271300/1000000 | loss=2.3536 vocab=0.7459 token=1.0986 | 6.15 steps/s | VRAM=29.05 Go [2026-07-17 11:34:40] step 271350/1000000 | loss=2.3724 vocab=0.7941 token=0.8764 | 6.14 steps/s | VRAM=26.02 Go [2026-07-17 11:34:48] step 271400/1000000 | loss=2.3126 vocab=1.2248 token=1.5042 | 6.12 steps/s | VRAM=32.93 Go [2026-07-17 11:34:57] step 271450/1000000 | loss=2.2144 vocab=0.9040 token=1.1028 | 6.13 steps/s | VRAM=28.70 Go [2026-07-17 11:35:05] step 271500/1000000 | loss=2.3554 vocab=0.7170 token=1.0486 | 6.14 steps/s | VRAM=29.17 Go [2026-07-17 11:35:13] step 271550/1000000 | loss=2.3070 vocab=0.5628 token=1.1099 | 6.12 steps/s | VRAM=27.91 Go [2026-07-17 11:35:21] step 271600/1000000 | loss=2.4241 vocab=1.3334 token=1.6639 | 6.14 steps/s | VRAM=32.34 Go [2026-07-17 11:35:29] step 271650/1000000 | loss=2.3297 vocab=0.8530 token=1.2116 | 6.13 steps/s | VRAM=32.32 Go [2026-07-17 11:35:37] step 271700/1000000 | loss=2.3692 vocab=0.7520 token=1.3178 | 6.11 steps/s | VRAM=26.06 Go [2026-07-17 11:35:46] step 271750/1000000 | loss=2.3441 vocab=0.9616 token=1.4212 | 6.14 steps/s | VRAM=31.39 Go [2026-07-17 11:35:54] step 271800/1000000 | loss=2.4077 vocab=0.8152 token=1.2660 | 6.09 steps/s | VRAM=26.63 Go [2026-07-17 11:36:02] step 271850/1000000 | loss=2.3070 vocab=1.0809 token=1.3009 | 6.11 steps/s | VRAM=26.49 Go [2026-07-17 11:36:10] step 271900/1000000 | loss=2.3148 vocab=0.8436 token=1.5739 | 6.13 steps/s | VRAM=28.15 Go [2026-07-17 11:36:18] step 271950/1000000 | loss=2.2988 vocab=0.6227 token=1.3640 | 6.12 steps/s | VRAM=25.78 Go [2026-07-17 11:36:26] step 272000/1000000 | loss=2.3066 vocab=0.9502 token=1.8190 | 6.10 steps/s | VRAM=28.12 Go [2026-07-17 11:36:35] step 272050/1000000 | loss=2.3703 vocab=1.6328 token=1.8037 | 6.16 steps/s | VRAM=32.79 Go [2026-07-17 11:36:43] step 272100/1000000 | loss=2.2977 vocab=1.3970 token=1.5953 | 6.13 steps/s | VRAM=31.78 Go [2026-07-17 11:36:51] step 272150/1000000 | loss=2.4059 vocab=1.0714 token=1.4804 | 6.12 steps/s | VRAM=29.60 Go [2026-07-17 11:36:59] step 272200/1000000 | loss=2.3174 vocab=0.6310 token=1.3396 | 6.09 steps/s | VRAM=32.89 Go [2026-07-17 11:37:07] step 272250/1000000 | loss=2.3814 vocab=1.0318 token=1.5571 | 6.09 steps/s | VRAM=29.21 Go [2026-07-17 11:37:15] step 272300/1000000 | loss=2.3571 vocab=1.2126 token=1.7189 | 6.12 steps/s | VRAM=27.98 Go [2026-07-17 11:37:24] step 272350/1000000 | loss=2.2318 vocab=1.3450 token=1.8227 | 6.12 steps/s | VRAM=28.64 Go [2026-07-17 11:37:32] step 272400/1000000 | loss=2.4135 vocab=1.1693 token=1.6085 | 6.11 steps/s | VRAM=28.48 Go [2026-07-17 11:37:40] step 272450/1000000 | loss=2.3170 vocab=0.7922 token=1.3756 | 6.12 steps/s | VRAM=32.11 Go [2026-07-17 11:37:48] step 272500/1000000 | loss=2.3907 vocab=0.6385 token=1.4640 | 6.15 steps/s | VRAM=31.35 Go [2026-07-17 11:37:56] step 272550/1000000 | loss=2.2663 vocab=0.9279 token=1.4282 | 6.08 steps/s | VRAM=26.52 Go [2026-07-17 11:38:04] step 272600/1000000 | loss=2.4238 vocab=0.8601 token=1.4134 | 6.13 steps/s | VRAM=27.59 Go [2026-07-17 11:38:13] step 272650/1000000 | loss=2.2522 vocab=1.1296 token=1.5446 | 6.11 steps/s | VRAM=32.70 Go [2026-07-17 11:38:21] step 272700/1000000 | loss=2.2667 vocab=0.6573 token=1.3094 | 6.13 steps/s | VRAM=32.89 Go [2026-07-17 11:38:29] step 272750/1000000 | loss=2.2491 vocab=0.7874 token=1.4201 | 6.15 steps/s | VRAM=32.33 Go [2026-07-17 11:38:37] step 272800/1000000 | loss=2.3086 vocab=1.2777 token=1.4738 | 6.12 steps/s | VRAM=26.50 Go [2026-07-17 11:38:45] step 272850/1000000 | loss=2.3306 vocab=0.9731 token=1.5205 | 6.11 steps/s | VRAM=32.14 Go [2026-07-17 11:38:53] step 272900/1000000 | loss=2.3126 vocab=0.7366 token=0.9623 | 6.15 steps/s | VRAM=28.22 Go [2026-07-17 11:39:02] step 272950/1000000 | loss=2.3103 vocab=1.0432 token=1.4585 | 6.12 steps/s | VRAM=29.35 Go [2026-07-17 11:39:10] step 273000/1000000 | loss=2.3518 vocab=0.9990 token=1.2768 | 6.14 steps/s | VRAM=29.04 Go [2026-07-17 11:39:18] step 273050/1000000 | loss=2.2942 vocab=0.9567 token=1.3180 | 6.14 steps/s | VRAM=24.56 Go [2026-07-17 11:39:26] step 273100/1000000 | loss=2.3603 vocab=0.8750 token=1.6149 | 6.13 steps/s | VRAM=28.94 Go [2026-07-17 11:39:34] step 273150/1000000 | loss=2.1919 vocab=0.5212 token=1.0582 | 6.14 steps/s | VRAM=31.60 Go [2026-07-17 11:39:42] step 273200/1000000 | loss=2.2397 vocab=1.0796 token=1.5912 | 6.10 steps/s | VRAM=27.44 Go [2026-07-17 11:39:51] step 273250/1000000 | loss=2.2987 vocab=0.7749 token=1.0945 | 6.12 steps/s | VRAM=27.96 Go [2026-07-17 11:39:59] step 273300/1000000 | loss=2.3445 vocab=1.9882 token=1.5965 | 6.14 steps/s | VRAM=27.45 Go [2026-07-17 11:40:07] step 273350/1000000 | loss=2.3470 vocab=0.6351 token=1.3296 | 6.11 steps/s | VRAM=29.09 Go [2026-07-17 11:40:15] step 273400/1000000 | loss=2.3580 vocab=0.9199 token=1.3856 | 6.10 steps/s | VRAM=28.01 Go [2026-07-17 11:40:23] step 273450/1000000 | loss=2.2593 vocab=1.3892 token=1.1856 | 6.12 steps/s | VRAM=27.93 Go [2026-07-17 11:40:31] step 273500/1000000 | loss=2.4135 vocab=1.0496 token=1.2290 | 6.12 steps/s | VRAM=26.76 Go [2026-07-17 11:40:40] step 273550/1000000 | loss=2.3427 vocab=0.9481 token=1.4906 | 6.11 steps/s | VRAM=29.38 Go [2026-07-17 11:40:48] step 273600/1000000 | loss=2.4111 vocab=0.9614 token=1.2629 | 6.15 steps/s | VRAM=28.41 Go [2026-07-17 11:40:56] step 273650/1000000 | loss=2.2184 vocab=0.8481 token=1.5152 | 6.14 steps/s | VRAM=25.67 Go [2026-07-17 11:41:04] step 273700/1000000 | loss=2.3204 vocab=0.8800 token=1.6071 | 6.13 steps/s | VRAM=29.86 Go [2026-07-17 11:41:12] step 273750/1000000 | loss=2.4540 vocab=0.7789 token=1.2108 | 6.13 steps/s | VRAM=31.55 Go [2026-07-17 11:41:20] step 273800/1000000 | loss=2.4392 vocab=1.2240 token=1.5214 | 6.12 steps/s | VRAM=29.76 Go [2026-07-17 11:41:29] step 273850/1000000 | loss=2.4234 vocab=0.9324 token=1.3778 | 6.08 steps/s | VRAM=29.33 Go [2026-07-17 11:41:37] step 273900/1000000 | loss=2.3214 vocab=0.9337 token=1.3840 | 6.13 steps/s | VRAM=28.22 Go [2026-07-17 11:41:45] step 273950/1000000 | loss=2.2408 vocab=0.4874 token=1.0559 | 6.09 steps/s | VRAM=32.89 Go [2026-07-17 11:41:53] step 274000/1000000 | loss=2.3491 vocab=1.2798 token=1.8282 | 6.14 steps/s | VRAM=28.52 Go [2026-07-17 11:42:01] step 274050/1000000 | loss=2.2782 vocab=0.8268 token=1.5236 | 6.11 steps/s | VRAM=32.80 Go [2026-07-17 11:42:09] step 274100/1000000 | loss=2.2676 vocab=1.3285 token=1.4715 | 6.13 steps/s | VRAM=28.11 Go [2026-07-17 11:42:18] step 274150/1000000 | loss=2.2859 vocab=1.0370 token=1.5725 | 6.09 steps/s | VRAM=29.13 Go [2026-07-17 11:42:26] step 274200/1000000 | loss=2.3068 vocab=1.4377 token=1.4222 | 6.15 steps/s | VRAM=31.58 Go [2026-07-17 11:42:34] step 274250/1000000 | loss=2.3553 vocab=1.2210 token=1.3055 | 6.11 steps/s | VRAM=29.09 Go [2026-07-17 11:42:42] step 274300/1000000 | loss=2.2350 vocab=0.7472 token=1.5812 | 6.13 steps/s | VRAM=28.75 Go [2026-07-17 11:42:50] step 274350/1000000 | loss=2.3625 vocab=0.5608 token=1.1136 | 6.13 steps/s | VRAM=29.41 Go [2026-07-17 11:42:59] step 274400/1000000 | loss=2.3276 vocab=0.9747 token=1.1355 | 6.11 steps/s | VRAM=26.02 Go [2026-07-17 11:43:07] step 274450/1000000 | loss=2.2819 vocab=0.6819 token=1.4190 | 6.11 steps/s | VRAM=28.98 Go [2026-07-17 11:43:15] step 274500/1000000 | loss=2.3135 vocab=0.5580 token=1.4555 | 6.09 steps/s | VRAM=26.51 Go [2026-07-17 11:43:23] step 274550/1000000 | loss=2.4161 vocab=1.3482 token=1.0644 | 6.11 steps/s | VRAM=32.63 Go [2026-07-17 11:43:31] step 274600/1000000 | loss=2.3220 vocab=0.5710 token=1.5585 | 6.14 steps/s | VRAM=30.93 Go [2026-07-17 11:43:39] step 274650/1000000 | loss=2.2986 vocab=0.9535 token=1.3643 | 6.11 steps/s | VRAM=28.89 Go [2026-07-17 11:43:48] step 274700/1000000 | loss=2.3712 vocab=0.8226 token=1.2149 | 6.13 steps/s | VRAM=28.57 Go [2026-07-17 11:43:56] step 274750/1000000 | loss=2.4159 vocab=0.6662 token=1.0705 | 6.13 steps/s | VRAM=31.52 Go [2026-07-17 11:44:04] step 274800/1000000 | loss=2.2816 vocab=1.3418 token=1.4095 | 6.12 steps/s | VRAM=32.85 Go [2026-07-17 11:44:12] step 274850/1000000 | loss=2.3870 vocab=1.6046 token=1.6184 | 6.10 steps/s | VRAM=32.26 Go [2026-07-17 11:44:20] step 274900/1000000 | loss=2.2117 vocab=0.9290 token=1.2769 | 6.11 steps/s | VRAM=32.30 Go [2026-07-17 11:44:29] step 274950/1000000 | loss=2.2607 vocab=0.9502 token=1.1316 | 6.11 steps/s | VRAM=30.04 Go [2026-07-17 11:44:37] step 275000/1000000 | loss=2.1775 vocab=0.6780 token=1.0222 | 6.08 steps/s | VRAM=28.41 Go [push] step 275000 OK (best-of-2, loss_token=1.0222) [2026-07-17 11:46:07] step 275050/1000000 | loss=2.3437 vocab=0.3399 token=1.3018 | 0.55 steps/s | VRAM=29.24 Go [2026-07-17 11:46:15] step 275100/1000000 | loss=2.4086 vocab=0.8872 token=1.5226 | 6.19 steps/s | VRAM=24.95 Go [2026-07-17 11:46:23] step 275150/1000000 | loss=2.4100 vocab=0.8642 token=1.3186 | 6.19 steps/s | VRAM=24.40 Go [2026-07-17 11:46:31] step 275200/1000000 | loss=2.3502 vocab=0.8210 token=1.6327 | 6.16 steps/s | VRAM=28.82 Go [2026-07-17 11:46:40] step 275250/1000000 | loss=2.2651 vocab=0.3661 token=0.9206 | 6.16 steps/s | VRAM=31.74 Go [2026-07-17 11:46:48] step 275300/1000000 | loss=2.3638 vocab=1.0473 token=1.5187 | 6.14 steps/s | VRAM=29.21 Go [2026-07-17 11:46:56] step 275350/1000000 | loss=2.2766 vocab=1.0620 token=1.3771 | 6.11 steps/s | VRAM=29.26 Go [2026-07-17 11:47:04] step 275400/1000000 | loss=2.3514 vocab=0.6753 token=1.5049 | 6.14 steps/s | VRAM=30.10 Go [2026-07-17 11:47:12] step 275450/1000000 | loss=2.3302 vocab=0.7650 token=1.5119 | 6.16 steps/s | VRAM=32.08 Go [2026-07-17 11:47:20] step 275500/1000000 | loss=2.3607 vocab=0.7870 token=1.1632 | 6.11 steps/s | VRAM=29.59 Go [2026-07-17 11:47:28] step 275550/1000000 | loss=2.2809 vocab=1.1039 token=1.5019 | 6.14 steps/s | VRAM=32.34 Go [2026-07-17 11:47:37] step 275600/1000000 | loss=2.2136 vocab=1.0774 token=1.7717 | 6.14 steps/s | VRAM=24.94 Go [2026-07-17 11:47:45] step 275650/1000000 | loss=2.2779 vocab=0.5662 token=1.0647 | 6.12 steps/s | VRAM=25.98 Go [2026-07-17 11:47:53] step 275700/1000000 | loss=2.2957 vocab=0.8684 token=1.2953 | 6.14 steps/s | VRAM=29.19 Go [2026-07-17 11:48:01] step 275750/1000000 | loss=2.2993 vocab=0.7385 token=1.5476 | 6.13 steps/s | VRAM=25.86 Go [2026-07-17 11:48:09] step 275800/1000000 | loss=2.3789 vocab=1.0730 token=1.5063 | 6.11 steps/s | VRAM=29.05 Go [2026-07-17 11:48:17] step 275850/1000000 | loss=2.3636 vocab=1.5165 token=1.7131 | 6.09 steps/s | VRAM=29.12 Go [2026-07-17 11:48:26] step 275900/1000000 | loss=2.3578 vocab=0.8642 token=1.2997 | 6.11 steps/s | VRAM=28.88 Go [2026-07-17 11:48:34] step 275950/1000000 | loss=2.3589 vocab=1.0959 token=1.7042 | 6.09 steps/s | VRAM=28.54 Go [2026-07-17 11:48:42] step 276000/1000000 | loss=2.4574 vocab=0.8230 token=0.9926 | 6.13 steps/s | VRAM=29.04 Go [2026-07-17 11:48:50] step 276050/1000000 | loss=2.3094 vocab=0.6410 token=1.5712 | 6.12 steps/s | VRAM=26.19 Go [2026-07-17 11:48:58] step 276100/1000000 | loss=2.3531 vocab=1.2151 token=1.3424 | 6.11 steps/s | VRAM=28.48 Go [2026-07-17 11:49:07] step 276150/1000000 | loss=2.3640 vocab=0.6097 token=1.7775 | 6.10 steps/s | VRAM=25.76 Go [2026-07-17 11:49:15] step 276200/1000000 | loss=2.3096 vocab=0.6442 token=0.9617 | 6.13 steps/s | VRAM=29.03 Go [2026-07-17 11:49:23] step 276250/1000000 | loss=2.2941 vocab=1.7505 token=1.5347 | 6.15 steps/s | VRAM=26.52 Go [2026-07-17 11:49:31] step 276300/1000000 | loss=2.3650 vocab=0.9698 token=1.2502 | 6.13 steps/s | VRAM=32.72 Go [2026-07-17 11:49:39] step 276350/1000000 | loss=2.3685 vocab=1.3617 token=1.1638 | 6.11 steps/s | VRAM=26.76 Go [2026-07-17 11:49:47] step 276400/1000000 | loss=2.3944 vocab=0.9660 token=1.5058 | 6.12 steps/s | VRAM=31.28 Go [2026-07-17 11:49:56] step 276450/1000000 | loss=2.3877 vocab=1.6997 token=2.0958 | 6.12 steps/s | VRAM=28.61 Go [2026-07-17 11:50:04] step 276500/1000000 | loss=2.1965 vocab=0.6410 token=0.9114 | 6.10 steps/s | VRAM=29.39 Go [2026-07-17 11:50:12] step 276550/1000000 | loss=2.3819 vocab=1.1914 token=1.4691 | 6.11 steps/s | VRAM=28.97 Go [2026-07-17 11:50:20] step 276600/1000000 | loss=2.4267 vocab=0.9897 token=1.3282 | 6.12 steps/s | VRAM=31.80 Go [2026-07-17 11:50:28] step 276650/1000000 | loss=2.3196 vocab=0.6613 token=1.1945 | 6.06 steps/s | VRAM=26.96 Go [2026-07-17 11:50:37] step 276700/1000000 | loss=2.2959 vocab=0.9782 token=1.1873 | 6.13 steps/s | VRAM=28.93 Go [2026-07-17 11:50:45] step 276750/1000000 | loss=2.2932 vocab=0.9206 token=1.7814 | 6.14 steps/s | VRAM=28.55 Go [2026-07-17 11:50:53] step 276800/1000000 | loss=2.2119 vocab=0.6496 token=0.9946 | 6.10 steps/s | VRAM=29.11 Go [2026-07-17 11:51:01] step 276850/1000000 | loss=2.2575 vocab=0.7885 token=0.9116 | 6.13 steps/s | VRAM=32.18 Go [2026-07-17 11:51:09] step 276900/1000000 | loss=2.3510 vocab=1.1176 token=1.5700 | 6.09 steps/s | VRAM=28.32 Go [2026-07-17 11:51:17] step 276950/1000000 | loss=2.2846 vocab=0.8299 token=1.3371 | 6.11 steps/s | VRAM=29.29 Go [2026-07-17 11:51:26] step 277000/1000000 | loss=2.2188 vocab=0.7536 token=1.3325 | 6.14 steps/s | VRAM=28.09 Go [2026-07-17 11:51:34] step 277050/1000000 | loss=2.3099 vocab=1.1418 token=1.4997 | 6.13 steps/s | VRAM=32.10 Go [2026-07-17 11:51:42] step 277100/1000000 | loss=2.2291 vocab=2.3478 token=1.6308 | 6.12 steps/s | VRAM=28.75 Go [2026-07-17 11:51:50] step 277150/1000000 | loss=2.3008 vocab=0.6938 token=1.4750 | 6.11 steps/s | VRAM=29.14 Go [2026-07-17 11:51:58] step 277200/1000000 | loss=2.3272 vocab=0.9173 token=1.4790 | 6.11 steps/s | VRAM=29.37 Go [2026-07-17 11:52:06] step 277250/1000000 | loss=2.3385 vocab=1.0018 token=1.3682 | 6.13 steps/s | VRAM=28.86 Go [2026-07-17 11:52:15] step 277300/1000000 | loss=2.2249 vocab=1.0094 token=1.3065 | 6.14 steps/s | VRAM=29.40 Go [2026-07-17 11:52:23] step 277350/1000000 | loss=2.4159 vocab=1.0573 token=1.2017 | 6.12 steps/s | VRAM=31.94 Go [2026-07-17 11:52:31] step 277400/1000000 | loss=2.3008 vocab=0.8516 token=1.3377 | 6.12 steps/s | VRAM=29.34 Go [2026-07-17 11:52:39] step 277450/1000000 | loss=2.2288 vocab=0.7876 token=1.3184 | 6.16 steps/s | VRAM=31.64 Go [2026-07-17 11:52:47] step 277500/1000000 | loss=2.3784 vocab=1.0633 token=1.0364 | 6.16 steps/s | VRAM=28.07 Go [2026-07-17 11:52:55] step 277550/1000000 | loss=2.2551 vocab=1.1718 token=1.5697 | 6.10 steps/s | VRAM=28.45 Go [2026-07-17 11:53:03] step 277600/1000000 | loss=2.2916 vocab=0.6592 token=1.3385 | 6.15 steps/s | VRAM=27.39 Go [2026-07-17 11:53:12] step 277650/1000000 | loss=2.3215 vocab=1.1933 token=1.5030 | 6.10 steps/s | VRAM=29.71 Go [2026-07-17 11:53:20] step 277700/1000000 | loss=2.3623 vocab=0.4388 token=1.2595 | 6.13 steps/s | VRAM=27.79 Go [2026-07-17 11:53:28] step 277750/1000000 | loss=2.2470 vocab=1.8543 token=1.5364 | 6.10 steps/s | VRAM=26.69 Go [2026-07-17 11:53:36] step 277800/1000000 | loss=2.2475 vocab=1.2881 token=1.4485 | 6.12 steps/s | VRAM=32.85 Go [2026-07-17 11:53:44] step 277850/1000000 | loss=2.3153 vocab=0.3843 token=1.0673 | 6.10 steps/s | VRAM=27.26 Go [2026-07-17 11:53:53] step 277900/1000000 | loss=2.3761 vocab=1.2049 token=1.6723 | 6.14 steps/s | VRAM=32.23 Go [2026-07-17 11:54:01] step 277950/1000000 | loss=2.2564 vocab=1.1851 token=1.4327 | 6.09 steps/s | VRAM=28.95 Go [2026-07-17 11:54:09] step 278000/1000000 | loss=2.1751 vocab=0.5842 token=1.2822 | 6.11 steps/s | VRAM=29.08 Go [2026-07-17 11:54:17] step 278050/1000000 | loss=2.2687 vocab=0.5628 token=1.0013 | 6.10 steps/s | VRAM=27.89 Go [2026-07-17 11:54:25] step 278100/1000000 | loss=2.3956 vocab=0.9659 token=1.1449 | 6.10 steps/s | VRAM=25.84 Go [2026-07-17 11:54:34] step 278150/1000000 | loss=2.4241 vocab=0.9519 token=1.2422 | 6.09 steps/s | VRAM=29.21 Go [2026-07-17 11:54:42] step 278200/1000000 | loss=2.2641 vocab=0.7316 token=1.0895 | 6.09 steps/s | VRAM=28.52 Go [2026-07-17 11:54:50] step 278250/1000000 | loss=2.3688 vocab=0.9075 token=1.1718 | 6.09 steps/s | VRAM=29.07 Go [2026-07-17 11:54:58] step 278300/1000000 | loss=2.3258 vocab=0.8699 token=1.2093 | 6.09 steps/s | VRAM=28.45 Go [2026-07-17 11:55:06] step 278350/1000000 | loss=2.2760 vocab=0.8297 token=1.2512 | 6.11 steps/s | VRAM=28.48 Go [2026-07-17 11:55:14] step 278400/1000000 | loss=2.4045 vocab=0.9649 token=1.5703 | 6.12 steps/s | VRAM=29.65 Go [2026-07-17 11:55:23] step 278450/1000000 | loss=2.2911 vocab=1.0605 token=1.5591 | 6.13 steps/s | VRAM=28.29 Go [2026-07-17 11:55:31] step 278500/1000000 | loss=2.3020 vocab=0.9865 token=1.5960 | 6.13 steps/s | VRAM=28.57 Go [2026-07-17 11:55:39] step 278550/1000000 | loss=2.3529 vocab=0.7960 token=1.6067 | 6.12 steps/s | VRAM=28.85 Go [2026-07-17 11:55:47] step 278600/1000000 | loss=2.3893 vocab=0.9408 token=1.3667 | 6.12 steps/s | VRAM=31.70 Go [2026-07-17 11:55:55] step 278650/1000000 | loss=2.3923 vocab=1.6665 token=1.6781 | 6.11 steps/s | VRAM=29.56 Go [2026-07-17 11:56:04] step 278700/1000000 | loss=2.3056 vocab=0.8785 token=1.5718 | 6.12 steps/s | VRAM=28.64 Go [2026-07-17 11:56:12] step 278750/1000000 | loss=2.3958 vocab=1.0716 token=1.3711 | 6.11 steps/s | VRAM=28.33 Go [2026-07-17 11:56:20] step 278800/1000000 | loss=2.3528 vocab=0.9555 token=1.2441 | 6.11 steps/s | VRAM=32.33 Go [2026-07-17 11:56:28] step 278850/1000000 | loss=2.2197 vocab=0.7289 token=1.2342 | 6.12 steps/s | VRAM=29.16 Go [2026-07-17 11:56:36] step 278900/1000000 | loss=2.2598 vocab=0.5901 token=0.9555 | 6.11 steps/s | VRAM=32.41 Go [2026-07-17 11:56:44] step 278950/1000000 | loss=2.4843 vocab=0.9150 token=1.2630 | 6.09 steps/s | VRAM=29.17 Go [2026-07-17 11:56:53] step 279000/1000000 | loss=2.2715 vocab=0.8112 token=1.4914 | 6.14 steps/s | VRAM=29.00 Go [2026-07-17 11:57:01] step 279050/1000000 | loss=2.2971 vocab=1.0911 token=1.6421 | 6.11 steps/s | VRAM=28.35 Go [2026-07-17 11:57:09] step 279100/1000000 | loss=2.4263 vocab=0.8682 token=1.3139 | 6.15 steps/s | VRAM=27.83 Go [2026-07-17 11:57:17] step 279150/1000000 | loss=2.2515 vocab=1.0628 token=1.4018 | 6.15 steps/s | VRAM=30.60 Go [2026-07-17 11:57:25] step 279200/1000000 | loss=2.4651 vocab=0.7329 token=1.3615 | 6.14 steps/s | VRAM=28.45 Go [2026-07-17 11:57:33] step 279250/1000000 | loss=2.3645 vocab=1.4429 token=1.7282 | 6.08 steps/s | VRAM=26.64 Go [2026-07-17 11:57:42] step 279300/1000000 | loss=2.3966 vocab=0.7179 token=0.9512 | 6.16 steps/s | VRAM=28.05 Go [2026-07-17 11:57:50] step 279350/1000000 | loss=2.5469 vocab=1.6071 token=1.5305 | 6.12 steps/s | VRAM=26.84 Go [2026-07-17 11:57:58] step 279400/1000000 | loss=2.4025 vocab=0.5268 token=1.2275 | 6.13 steps/s | VRAM=29.81 Go [2026-07-17 11:58:06] step 279450/1000000 | loss=2.3329 vocab=1.4491 token=1.6335 | 6.13 steps/s | VRAM=32.31 Go [2026-07-17 11:58:14] step 279500/1000000 | loss=2.2773 vocab=0.9036 token=1.0834 | 6.12 steps/s | VRAM=25.76 Go [2026-07-17 11:58:22] step 279550/1000000 | loss=2.3326 vocab=1.2906 token=1.5388 | 6.10 steps/s | VRAM=31.31 Go [2026-07-17 11:58:31] step 279600/1000000 | loss=2.1943 vocab=0.6343 token=0.9551 | 6.14 steps/s | VRAM=28.11 Go [2026-07-17 11:58:39] step 279650/1000000 | loss=2.2755 vocab=0.9350 token=1.5514 | 6.11 steps/s | VRAM=32.57 Go [2026-07-17 11:58:47] step 279700/1000000 | loss=2.3885 vocab=0.9232 token=1.6972 | 6.11 steps/s | VRAM=26.33 Go [2026-07-17 11:58:55] step 279750/1000000 | loss=2.3690 vocab=0.6787 token=1.1940 | 6.10 steps/s | VRAM=32.54 Go [2026-07-17 11:59:03] step 279800/1000000 | loss=2.4069 vocab=0.6726 token=1.3946 | 6.12 steps/s | VRAM=28.11 Go [2026-07-17 11:59:11] step 279850/1000000 | loss=2.3594 vocab=1.4856 token=1.3684 | 6.12 steps/s | VRAM=28.74 Go [2026-07-17 11:59:20] step 279900/1000000 | loss=2.2480 vocab=1.6667 token=1.7247 | 6.10 steps/s | VRAM=32.08 Go [2026-07-17 11:59:28] step 279950/1000000 | loss=2.2940 vocab=0.9798 token=1.5531 | 6.13 steps/s | VRAM=25.20 Go [2026-07-17 11:59:36] step 280000/1000000 | loss=2.3221 vocab=0.7915 token=1.3890 | 6.15 steps/s | VRAM=28.40 Go [step 280000] loss_token=1.3890 > min recent (1.0222) -> push saute [2026-07-17 11:59:44] step 280050/1000000 | loss=2.3573 vocab=0.9369 token=1.4608 | 6.11 steps/s | VRAM=27.93 Go [2026-07-17 11:59:52] step 280100/1000000 | loss=2.3481 vocab=1.3047 token=1.4078 | 6.11 steps/s | VRAM=32.30 Go [2026-07-17 12:00:00] step 280150/1000000 | loss=2.3500 vocab=1.1292 token=1.1312 | 6.08 steps/s | VRAM=26.31 Go [2026-07-17 12:00:09] step 280200/1000000 | loss=2.2882 vocab=0.9074 token=1.1772 | 6.11 steps/s | VRAM=29.39 Go [2026-07-17 12:00:17] step 280250/1000000 | loss=2.3927 vocab=0.4092 token=1.5398 | 6.09 steps/s | VRAM=26.12 Go [2026-07-17 12:00:25] step 280300/1000000 | loss=2.2619 vocab=0.6584 token=1.1714 | 6.14 steps/s | VRAM=31.80 Go [2026-07-17 12:00:33] step 280350/1000000 | loss=2.2986 vocab=0.5328 token=1.2346 | 6.11 steps/s | VRAM=32.70 Go [2026-07-17 12:00:41] step 280400/1000000 | loss=2.3597 vocab=1.2991 token=1.5062 | 6.10 steps/s | VRAM=26.72 Go [2026-07-17 12:00:50] step 280450/1000000 | loss=2.3113 vocab=1.0497 token=1.7994 | 6.12 steps/s | VRAM=32.51 Go [2026-07-17 12:00:58] step 280500/1000000 | loss=2.2943 vocab=1.0503 token=1.5798 | 6.13 steps/s | VRAM=29.20 Go [2026-07-17 12:01:06] step 280550/1000000 | loss=2.3718 vocab=0.9888 token=1.0777 | 6.11 steps/s | VRAM=32.17 Go [2026-07-17 12:01:14] step 280600/1000000 | loss=2.3558 vocab=1.4028 token=1.2871 | 6.13 steps/s | VRAM=24.75 Go [2026-07-17 12:01:22] step 280650/1000000 | loss=2.3671 vocab=1.6704 token=1.6069 | 6.08 steps/s | VRAM=27.39 Go [2026-07-17 12:01:30] step 280700/1000000 | loss=2.4352 vocab=0.7309 token=1.6576 | 6.12 steps/s | VRAM=27.84 Go [2026-07-17 12:01:39] step 280750/1000000 | loss=2.3070 vocab=1.0039 token=1.4712 | 6.11 steps/s | VRAM=32.31 Go [2026-07-17 12:01:47] step 280800/1000000 | loss=2.3135 vocab=1.1471 token=1.1720 | 6.13 steps/s | VRAM=31.80 Go [2026-07-17 12:01:55] step 280850/1000000 | loss=2.3205 vocab=1.2046 token=1.6317 | 6.10 steps/s | VRAM=28.87 Go [2026-07-17 12:02:03] step 280900/1000000 | loss=2.2901 vocab=0.9092 token=1.3191 | 6.13 steps/s | VRAM=29.03 Go [2026-07-17 12:02:11] step 280950/1000000 | loss=2.2990 vocab=0.6301 token=1.1694 | 6.14 steps/s | VRAM=31.83 Go [2026-07-17 12:02:19] step 281000/1000000 | loss=2.3445 vocab=1.2011 token=1.2640 | 6.13 steps/s | VRAM=28.00 Go [2026-07-17 12:02:28] step 281050/1000000 | loss=2.3171 vocab=1.0828 token=1.3083 | 6.11 steps/s | VRAM=28.40 Go [2026-07-17 12:02:36] step 281100/1000000 | loss=2.2858 vocab=0.7652 token=1.0707 | 6.09 steps/s | VRAM=29.50 Go [2026-07-17 12:02:44] step 281150/1000000 | loss=2.3616 vocab=0.7858 token=1.9438 | 6.14 steps/s | VRAM=31.62 Go [2026-07-17 12:02:52] step 281200/1000000 | loss=2.3550 vocab=1.7972 token=1.8420 | 6.11 steps/s | VRAM=32.46 Go [2026-07-17 12:03:00] step 281250/1000000 | loss=2.3410 vocab=1.4798 token=1.4545 | 6.07 steps/s | VRAM=29.54 Go [2026-07-17 12:03:09] step 281300/1000000 | loss=2.3280 vocab=1.1205 token=1.4988 | 6.08 steps/s | VRAM=29.39 Go [2026-07-17 12:03:17] step 281350/1000000 | loss=2.2231 vocab=1.4430 token=1.2958 | 6.10 steps/s | VRAM=29.17 Go [2026-07-17 12:03:25] step 281400/1000000 | loss=2.2656 vocab=0.6964 token=1.3102 | 6.12 steps/s | VRAM=32.30 Go [2026-07-17 12:03:33] step 281450/1000000 | loss=2.3617 vocab=1.3148 token=1.5589 | 6.12 steps/s | VRAM=26.22 Go [2026-07-17 12:03:41] step 281500/1000000 | loss=2.3055 vocab=1.2036 token=1.2925 | 6.13 steps/s | VRAM=29.01 Go [2026-07-17 12:03:49] step 281550/1000000 | loss=2.3919 vocab=1.2423 token=1.4561 | 6.12 steps/s | VRAM=27.85 Go [2026-07-17 12:03:58] step 281600/1000000 | loss=2.3355 vocab=1.1975 token=1.8338 | 6.09 steps/s | VRAM=27.02 Go [2026-07-17 12:04:06] step 281650/1000000 | loss=2.3383 vocab=1.0069 token=1.3822 | 6.17 steps/s | VRAM=31.06 Go [2026-07-17 12:04:14] step 281700/1000000 | loss=2.3795 vocab=1.2101 token=0.8359 | 6.11 steps/s | VRAM=29.97 Go [2026-07-17 12:04:22] step 281750/1000000 | loss=2.3785 vocab=1.0395 token=1.0359 | 6.13 steps/s | VRAM=28.71 Go [2026-07-17 12:04:30] step 281800/1000000 | loss=2.2682 vocab=0.9114 token=1.1336 | 6.14 steps/s | VRAM=31.91 Go [2026-07-17 12:04:38] step 281850/1000000 | loss=2.3497 vocab=0.9459 token=1.1566 | 6.11 steps/s | VRAM=29.19 Go [2026-07-17 12:04:47] step 281900/1000000 | loss=2.3132 vocab=1.0937 token=1.5748 | 6.11 steps/s | VRAM=31.59 Go [2026-07-17 12:04:55] step 281950/1000000 | loss=2.3296 vocab=0.8284 token=1.1710 | 6.14 steps/s | VRAM=31.41 Go [2026-07-17 12:05:03] step 282000/1000000 | loss=2.3071 vocab=1.2812 token=1.6734 | 6.10 steps/s | VRAM=26.64 Go [2026-07-17 12:05:11] step 282050/1000000 | loss=2.3117 vocab=1.1048 token=1.2662 | 6.11 steps/s | VRAM=29.54 Go [2026-07-17 12:05:19] step 282100/1000000 | loss=2.2749 vocab=0.7690 token=1.5056 | 6.16 steps/s | VRAM=28.42 Go [2026-07-17 12:05:27] step 282150/1000000 | loss=2.3116 vocab=0.7744 token=1.5657 | 6.15 steps/s | VRAM=32.15 Go [2026-07-17 12:05:36] step 282200/1000000 | loss=2.2859 vocab=0.6605 token=1.2585 | 6.15 steps/s | VRAM=31.26 Go [2026-07-17 12:05:44] step 282250/1000000 | loss=2.3252 vocab=1.4721 token=1.4865 | 6.14 steps/s | VRAM=24.85 Go [2026-07-17 12:05:52] step 282300/1000000 | loss=2.4246 vocab=1.0366 token=1.6052 | 6.11 steps/s | VRAM=28.47 Go [2026-07-17 12:06:00] step 282350/1000000 | loss=2.3337 vocab=1.2825 token=1.7182 | 6.12 steps/s | VRAM=23.79 Go [2026-07-17 12:06:08] step 282400/1000000 | loss=2.2358 vocab=0.7856 token=1.3431 | 6.13 steps/s | VRAM=24.72 Go [2026-07-17 12:06:16] step 282450/1000000 | loss=2.4216 vocab=1.2896 token=1.3700 | 6.11 steps/s | VRAM=30.35 Go [2026-07-17 12:06:24] step 282500/1000000 | loss=2.3664 vocab=1.0540 token=1.2926 | 6.15 steps/s | VRAM=32.53 Go [2026-07-17 12:06:33] step 282550/1000000 | loss=2.2846 vocab=0.6359 token=1.5240 | 6.18 steps/s | VRAM=32.00 Go [2026-07-17 12:06:41] step 282600/1000000 | loss=2.3291 vocab=0.6629 token=1.0708 | 6.14 steps/s | VRAM=28.98 Go [2026-07-17 12:06:49] step 282650/1000000 | loss=2.3358 vocab=1.0901 token=1.2918 | 6.14 steps/s | VRAM=29.59 Go [2026-07-17 12:06:57] step 282700/1000000 | loss=2.3835 vocab=0.6082 token=1.4094 | 6.12 steps/s | VRAM=32.93 Go [2026-07-17 12:07:05] step 282750/1000000 | loss=2.3761 vocab=1.1011 token=1.8683 | 6.16 steps/s | VRAM=30.55 Go [2026-07-17 12:07:13] step 282800/1000000 | loss=2.3495 vocab=0.7215 token=1.1090 | 6.13 steps/s | VRAM=32.38 Go [2026-07-17 12:07:21] step 282850/1000000 | loss=2.2976 vocab=1.0568 token=1.5290 | 6.15 steps/s | VRAM=29.32 Go [2026-07-17 12:07:30] step 282900/1000000 | loss=2.3583 vocab=0.6025 token=1.5683 | 6.13 steps/s | VRAM=26.72 Go [2026-07-17 12:07:38] step 282950/1000000 | loss=2.1921 vocab=0.9016 token=1.1062 | 6.10 steps/s | VRAM=28.98 Go [2026-07-17 12:07:46] step 283000/1000000 | loss=2.3724 vocab=1.1285 token=1.5036 | 6.11 steps/s | VRAM=29.26 Go [2026-07-17 12:07:54] step 283050/1000000 | loss=2.2990 vocab=0.8784 token=1.1161 | 6.15 steps/s | VRAM=32.50 Go [2026-07-17 12:08:02] step 283100/1000000 | loss=2.3887 vocab=0.6996 token=1.3444 | 6.13 steps/s | VRAM=29.40 Go [2026-07-17 12:08:10] step 283150/1000000 | loss=2.3159 vocab=0.8761 token=1.7120 | 6.13 steps/s | VRAM=25.85 Go [2026-07-17 12:08:18] step 283200/1000000 | loss=2.3781 vocab=0.6731 token=1.4409 | 6.17 steps/s | VRAM=28.10 Go [2026-07-17 12:08:27] step 283250/1000000 | loss=2.4434 vocab=1.2751 token=1.5537 | 6.13 steps/s | VRAM=27.92 Go [2026-07-17 12:08:35] step 283300/1000000 | loss=2.3321 vocab=0.6235 token=0.9961 | 6.13 steps/s | VRAM=27.41 Go [2026-07-17 12:08:43] step 283350/1000000 | loss=2.3623 vocab=0.4422 token=1.2259 | 6.16 steps/s | VRAM=27.49 Go [2026-07-17 12:08:51] step 283400/1000000 | loss=2.3090 vocab=0.6004 token=1.4900 | 6.15 steps/s | VRAM=28.86 Go [2026-07-17 12:08:59] step 283450/1000000 | loss=2.2750 vocab=1.2570 token=1.6281 | 6.15 steps/s | VRAM=29.39 Go [2026-07-17 12:09:07] step 283500/1000000 | loss=2.3660 vocab=0.7682 token=1.3965 | 6.11 steps/s | VRAM=29.26 Go [2026-07-17 12:09:16] step 283550/1000000 | loss=2.3019 vocab=0.6994 token=1.1882 | 6.13 steps/s | VRAM=29.20 Go [2026-07-17 12:09:24] step 283600/1000000 | loss=2.3781 vocab=0.6492 token=1.3510 | 6.10 steps/s | VRAM=28.25 Go [2026-07-17 12:09:32] step 283650/1000000 | loss=2.2561 vocab=1.6647 token=1.3099 | 6.12 steps/s | VRAM=28.59 Go [2026-07-17 12:09:40] step 283700/1000000 | loss=2.3544 vocab=0.8181 token=1.4220 | 6.15 steps/s | VRAM=29.68 Go [2026-07-17 12:09:48] step 283750/1000000 | loss=2.2654 vocab=1.0501 token=1.4608 | 6.12 steps/s | VRAM=26.34 Go [2026-07-17 12:09:56] step 283800/1000000 | loss=2.3876 vocab=0.8615 token=1.6412 | 6.16 steps/s | VRAM=31.55 Go [2026-07-17 12:10:05] step 283850/1000000 | loss=2.3461 vocab=1.1030 token=1.5295 | 6.12 steps/s | VRAM=25.52 Go [2026-07-17 12:10:13] step 283900/1000000 | loss=2.3167 vocab=1.0182 token=1.1430 | 6.15 steps/s | VRAM=31.34 Go [2026-07-17 12:10:21] step 283950/1000000 | loss=2.3521 vocab=1.1443 token=1.3578 | 6.11 steps/s | VRAM=26.17 Go [2026-07-17 12:10:29] step 284000/1000000 | loss=2.4118 vocab=1.7478 token=1.6261 | 6.12 steps/s | VRAM=28.33 Go [2026-07-17 12:10:37] step 284050/1000000 | loss=2.2622 vocab=0.8895 token=1.3087 | 6.14 steps/s | VRAM=25.77 Go [2026-07-17 12:10:45] step 284100/1000000 | loss=2.2531 vocab=0.9787 token=1.1927 | 6.12 steps/s | VRAM=29.32 Go [2026-07-17 12:10:53] step 284150/1000000 | loss=2.3652 vocab=0.7355 token=1.2807 | 6.13 steps/s | VRAM=25.29 Go [2026-07-17 12:11:02] step 284200/1000000 | loss=2.2947 vocab=0.8236 token=1.3067 | 6.14 steps/s | VRAM=28.21 Go [2026-07-17 12:11:10] step 284250/1000000 | loss=2.2010 vocab=0.6917 token=1.2753 | 6.13 steps/s | VRAM=32.81 Go [2026-07-17 12:11:18] step 284300/1000000 | loss=2.2922 vocab=1.1781 token=1.3890 | 6.12 steps/s | VRAM=32.87 Go [2026-07-17 12:11:26] step 284350/1000000 | loss=2.3262 vocab=0.6507 token=1.1969 | 6.15 steps/s | VRAM=30.54 Go [2026-07-17 12:11:34] step 284400/1000000 | loss=2.2014 vocab=0.9006 token=1.2991 | 6.12 steps/s | VRAM=31.81 Go [2026-07-17 12:11:42] step 284450/1000000 | loss=2.3108 vocab=1.0905 token=1.2161 | 6.10 steps/s | VRAM=23.57 Go [2026-07-17 12:11:51] step 284500/1000000 | loss=2.3265 vocab=1.0162 token=1.1371 | 6.18 steps/s | VRAM=30.85 Go [2026-07-17 12:11:59] step 284550/1000000 | loss=2.2469 vocab=1.3242 token=1.7456 | 6.14 steps/s | VRAM=32.19 Go [2026-07-17 12:12:07] step 284600/1000000 | loss=2.3033 vocab=0.5412 token=1.1394 | 6.09 steps/s | VRAM=26.94 Go [2026-07-17 12:12:15] step 284650/1000000 | loss=2.3635 vocab=0.7897 token=1.3226 | 6.17 steps/s | VRAM=26.69 Go [2026-07-17 12:12:23] step 284700/1000000 | loss=2.4118 vocab=1.0667 token=1.1437 | 6.14 steps/s | VRAM=31.37 Go [2026-07-17 12:12:31] step 284750/1000000 | loss=2.3603 vocab=0.9007 token=1.5516 | 6.11 steps/s | VRAM=28.98 Go [2026-07-17 12:12:40] step 284800/1000000 | loss=2.3311 vocab=1.1947 token=1.5424 | 6.10 steps/s | VRAM=29.24 Go [2026-07-17 12:12:48] step 284850/1000000 | loss=2.2688 vocab=1.2505 token=1.9187 | 6.10 steps/s | VRAM=31.89 Go [2026-07-17 12:12:56] step 284900/1000000 | loss=2.2505 vocab=0.6790 token=1.5584 | 6.12 steps/s | VRAM=31.28 Go [2026-07-17 12:13:04] step 284950/1000000 | loss=2.3713 vocab=0.9867 token=1.3039 | 6.13 steps/s | VRAM=28.35 Go [2026-07-17 12:13:12] step 285000/1000000 | loss=2.3188 vocab=0.7243 token=1.3013 | 6.12 steps/s | VRAM=29.39 Go [push] step 285000 OK (best-of-2, loss_token=1.3013) [2026-07-17 12:14:54] step 285050/1000000 | loss=2.3040 vocab=0.7094 token=1.2046 | 0.49 steps/s | VRAM=26.35 Go [2026-07-17 12:15:02] step 285100/1000000 | loss=2.2985 vocab=0.6016 token=1.2024 | 6.21 steps/s | VRAM=32.03 Go [2026-07-17 12:15:11] step 285150/1000000 | loss=2.3677 vocab=1.3602 token=1.4335 | 6.19 steps/s | VRAM=28.11 Go [2026-07-17 12:15:19] step 285200/1000000 | loss=2.2902 vocab=0.8180 token=1.5621 | 6.19 steps/s | VRAM=30.15 Go [2026-07-17 12:15:27] step 285250/1000000 | loss=2.3372 vocab=0.9638 token=1.2490 | 6.17 steps/s | VRAM=29.35 Go [2026-07-17 12:15:35] step 285300/1000000 | loss=2.4750 vocab=1.2060 token=1.3627 | 6.18 steps/s | VRAM=27.84 Go [2026-07-17 12:15:43] step 285350/1000000 | loss=2.3674 vocab=0.8513 token=1.5783 | 6.15 steps/s | VRAM=24.31 Go [2026-07-17 12:15:51] step 285400/1000000 | loss=2.3233 vocab=0.7927 token=1.0771 | 6.14 steps/s | VRAM=31.93 Go [2026-07-17 12:15:59] step 285450/1000000 | loss=2.4553 vocab=0.8844 token=1.3017 | 6.11 steps/s | VRAM=32.92 Go [2026-07-17 12:16:07] step 285500/1000000 | loss=2.3292 vocab=0.6774 token=0.8909 | 6.12 steps/s | VRAM=28.82 Go [2026-07-17 12:16:16] step 285550/1000000 | loss=2.3153 vocab=0.8384 token=1.4277 | 6.10 steps/s | VRAM=25.92 Go [2026-07-17 12:16:24] step 285600/1000000 | loss=2.3301 vocab=0.8167 token=1.0008 | 6.09 steps/s | VRAM=27.01 Go [2026-07-17 12:16:32] step 285650/1000000 | loss=2.3305 vocab=0.7859 token=1.2766 | 6.10 steps/s | VRAM=28.32 Go [2026-07-17 12:16:40] step 285700/1000000 | loss=2.3478 vocab=0.5661 token=1.1847 | 6.09 steps/s | VRAM=31.85 Go [2026-07-17 12:16:48] step 285750/1000000 | loss=2.2574 vocab=1.0743 token=1.2653 | 6.07 steps/s | VRAM=28.08 Go [2026-07-17 12:16:57] step 285800/1000000 | loss=2.3676 vocab=1.2278 token=1.4735 | 6.07 steps/s | VRAM=26.19 Go [2026-07-17 12:17:05] step 285850/1000000 | loss=2.2726 vocab=1.1189 token=1.4699 | 6.11 steps/s | VRAM=31.48 Go [2026-07-17 12:17:13] step 285900/1000000 | loss=2.4613 vocab=1.4434 token=1.3534 | 6.11 steps/s | VRAM=30.13 Go [2026-07-17 12:17:21] step 285950/1000000 | loss=2.2458 vocab=0.9112 token=1.4464 | 6.11 steps/s | VRAM=28.05 Go [2026-07-17 12:17:30] step 286000/1000000 | loss=2.3311 vocab=1.0272 token=1.6984 | 6.06 steps/s | VRAM=30.00 Go [2026-07-17 12:17:38] step 286050/1000000 | loss=2.3077 vocab=0.6986 token=1.1337 | 6.13 steps/s | VRAM=31.15 Go [2026-07-17 12:17:46] step 286100/1000000 | loss=2.3598 vocab=1.2868 token=1.4304 | 6.09 steps/s | VRAM=32.42 Go [2026-07-17 12:17:54] step 286150/1000000 | loss=2.3167 vocab=0.8438 token=0.9705 | 6.07 steps/s | VRAM=30.17 Go [2026-07-17 12:18:02] step 286200/1000000 | loss=2.3068 vocab=0.7705 token=1.3078 | 6.11 steps/s | VRAM=27.47 Go [2026-07-17 12:18:10] step 286250/1000000 | loss=2.2857 vocab=0.4959 token=0.7351 | 6.12 steps/s | VRAM=31.79 Go [2026-07-17 12:18:19] step 286300/1000000 | loss=2.3357 vocab=1.0928 token=1.6071 | 6.06 steps/s | VRAM=32.65 Go [2026-07-17 12:18:27] step 286350/1000000 | loss=2.3526 vocab=1.2846 token=1.4494 | 6.10 steps/s | VRAM=32.05 Go [2026-07-17 12:18:35] step 286400/1000000 | loss=2.3045 vocab=1.0923 token=1.6249 | 6.09 steps/s | VRAM=32.53 Go [2026-07-17 12:18:43] step 286450/1000000 | loss=2.3211 vocab=1.3185 token=1.2222 | 6.11 steps/s | VRAM=31.40 Go [2026-07-17 12:18:51] step 286500/1000000 | loss=2.3265 vocab=1.0874 token=1.7763 | 6.12 steps/s | VRAM=31.78 Go [2026-07-17 12:19:00] step 286550/1000000 | loss=2.2747 vocab=0.6881 token=1.6077 | 6.11 steps/s | VRAM=31.81 Go [2026-07-17 12:19:08] step 286600/1000000 | loss=2.2214 vocab=1.2729 token=1.4792 | 6.10 steps/s | VRAM=26.49 Go [2026-07-17 12:19:16] step 286650/1000000 | loss=2.3479 vocab=1.0042 token=1.5918 | 6.10 steps/s | VRAM=28.92 Go [2026-07-17 12:19:24] step 286700/1000000 | loss=2.4534 vocab=1.1352 token=1.7164 | 6.15 steps/s | VRAM=28.57 Go [2026-07-17 12:19:32] step 286750/1000000 | loss=2.3920 vocab=0.8330 token=1.2187 | 6.08 steps/s | VRAM=31.38 Go [2026-07-17 12:19:41] step 286800/1000000 | loss=2.3847 vocab=1.2380 token=1.2953 | 6.12 steps/s | VRAM=31.59 Go [2026-07-17 12:19:49] step 286850/1000000 | loss=2.2981 vocab=1.2143 token=1.2195 | 6.12 steps/s | VRAM=32.49 Go [2026-07-17 12:19:57] step 286900/1000000 | loss=2.3476 vocab=0.7734 token=1.6351 | 6.12 steps/s | VRAM=29.10 Go [2026-07-17 12:20:05] step 286950/1000000 | loss=2.2967 vocab=0.7804 token=1.4040 | 6.12 steps/s | VRAM=28.12 Go [2026-07-17 12:20:13] step 287000/1000000 | loss=2.4113 vocab=0.6485 token=1.4398 | 6.12 steps/s | VRAM=28.79 Go [2026-07-17 12:20:21] step 287050/1000000 | loss=2.3669 vocab=0.4055 token=1.1557 | 6.13 steps/s | VRAM=29.44 Go [2026-07-17 12:20:30] step 287100/1000000 | loss=2.2429 vocab=0.8910 token=1.1646 | 6.11 steps/s | VRAM=29.31 Go [2026-07-17 12:20:38] step 287150/1000000 | loss=2.3217 vocab=0.9546 token=1.2412 | 6.14 steps/s | VRAM=25.74 Go [2026-07-17 12:20:46] step 287200/1000000 | loss=2.2593 vocab=0.7474 token=1.3877 | 6.12 steps/s | VRAM=26.45 Go [2026-07-17 12:20:54] step 287250/1000000 | loss=2.3381 vocab=1.3306 token=1.5627 | 6.09 steps/s | VRAM=25.29 Go [2026-07-17 12:21:02] step 287300/1000000 | loss=2.3765 vocab=0.7583 token=1.6353 | 6.14 steps/s | VRAM=25.48 Go [2026-07-17 12:21:10] step 287350/1000000 | loss=2.3702 vocab=0.7382 token=1.0319 | 6.12 steps/s | VRAM=32.66 Go [2026-07-17 12:21:19] step 287400/1000000 | loss=2.3424 vocab=0.6612 token=1.1728 | 6.11 steps/s | VRAM=25.22 Go [2026-07-17 12:21:27] step 287450/1000000 | loss=2.2230 vocab=0.9409 token=1.5560 | 6.13 steps/s | VRAM=26.00 Go [2026-07-17 12:21:35] step 287500/1000000 | loss=2.2545 vocab=0.9103 token=1.6381 | 6.12 steps/s | VRAM=28.74 Go [2026-07-17 12:21:43] step 287550/1000000 | loss=2.3422 vocab=0.6948 token=1.4511 | 6.14 steps/s | VRAM=24.78 Go [2026-07-17 12:21:51] step 287600/1000000 | loss=2.3025 vocab=1.0994 token=1.2263 | 6.10 steps/s | VRAM=26.08 Go [2026-07-17 12:21:59] step 287650/1000000 | loss=2.2868 vocab=0.4311 token=1.1904 | 6.13 steps/s | VRAM=27.66 Go [2026-07-17 12:22:08] step 287700/1000000 | loss=2.3733 vocab=1.2219 token=1.6112 | 6.11 steps/s | VRAM=29.07 Go [2026-07-17 12:22:16] step 287750/1000000 | loss=2.3378 vocab=0.4405 token=0.7847 | 6.11 steps/s | VRAM=27.76 Go [2026-07-17 12:22:24] step 287800/1000000 | loss=2.4228 vocab=0.6410 token=1.5995 | 6.11 steps/s | VRAM=32.67 Go [2026-07-17 12:22:32] step 287850/1000000 | loss=2.2766 vocab=0.8772 token=1.0683 | 6.10 steps/s | VRAM=28.97 Go [2026-07-17 12:22:40] step 287900/1000000 | loss=2.3312 vocab=0.8262 token=1.3130 | 6.16 steps/s | VRAM=31.28 Go [2026-07-17 12:22:49] step 287950/1000000 | loss=2.3799 vocab=1.2465 token=1.6496 | 6.11 steps/s | VRAM=29.76 Go [2026-07-17 12:22:57] step 288000/1000000 | loss=2.3000 vocab=1.3931 token=1.7222 | 6.11 steps/s | VRAM=26.23 Go [2026-07-17 12:23:05] step 288050/1000000 | loss=2.2973 vocab=0.7637 token=1.2006 | 6.11 steps/s | VRAM=32.16 Go [2026-07-17 12:23:13] step 288100/1000000 | loss=2.2820 vocab=0.5869 token=1.1953 | 6.15 steps/s | VRAM=32.58 Go [2026-07-17 12:23:21] step 288150/1000000 | loss=2.3050 vocab=0.6897 token=1.2185 | 6.14 steps/s | VRAM=25.63 Go [2026-07-17 12:23:29] step 288200/1000000 | loss=2.2911 vocab=0.6238 token=1.3739 | 6.14 steps/s | VRAM=27.71 Go [2026-07-17 12:23:37] step 288250/1000000 | loss=2.3972 vocab=0.8662 token=1.4434 | 6.14 steps/s | VRAM=32.70 Go [2026-07-17 12:23:46] step 288300/1000000 | loss=2.3096 vocab=1.3107 token=1.4726 | 6.15 steps/s | VRAM=32.08 Go [2026-07-17 12:23:54] step 288350/1000000 | loss=2.3361 vocab=0.8209 token=1.2236 | 6.13 steps/s | VRAM=24.74 Go [2026-07-17 12:24:02] step 288400/1000000 | loss=2.2757 vocab=0.8025 token=1.1026 | 6.13 steps/s | VRAM=28.91 Go [2026-07-17 12:24:10] step 288450/1000000 | loss=2.4657 vocab=0.9027 token=1.0288 | 6.13 steps/s | VRAM=32.75 Go [2026-07-17 12:24:18] step 288500/1000000 | loss=2.3010 vocab=0.6799 token=1.4242 | 6.14 steps/s | VRAM=26.39 Go [2026-07-17 12:24:26] step 288550/1000000 | loss=2.3427 vocab=0.6041 token=1.1896 | 6.12 steps/s | VRAM=24.89 Go [2026-07-17 12:24:35] step 288600/1000000 | loss=2.2495 vocab=0.7009 token=1.3755 | 6.10 steps/s | VRAM=28.25 Go [2026-07-17 12:24:43] step 288650/1000000 | loss=2.3356 vocab=0.8325 token=1.2015 | 6.12 steps/s | VRAM=28.87 Go [2026-07-17 12:24:51] step 288700/1000000 | loss=2.3545 vocab=1.0291 token=1.3672 | 6.15 steps/s | VRAM=28.59 Go [2026-07-17 12:24:59] step 288750/1000000 | loss=2.3087 vocab=0.7292 token=1.2881 | 6.14 steps/s | VRAM=28.18 Go [2026-07-17 12:25:07] step 288800/1000000 | loss=2.2603 vocab=0.5145 token=1.7728 | 6.16 steps/s | VRAM=31.82 Go [2026-07-17 12:25:15] step 288850/1000000 | loss=2.2420 vocab=0.9197 token=1.7537 | 6.12 steps/s | VRAM=25.64 Go [2026-07-17 12:25:23] step 288900/1000000 | loss=2.3244 vocab=0.9860 token=1.4519 | 6.12 steps/s | VRAM=29.62 Go [2026-07-17 12:25:32] step 288950/1000000 | loss=2.2519 vocab=0.8981 token=1.4055 | 6.14 steps/s | VRAM=32.23 Go [2026-07-17 12:25:40] step 289000/1000000 | loss=2.3306 vocab=0.7574 token=0.8969 | 6.14 steps/s | VRAM=29.29 Go [2026-07-17 12:25:48] step 289050/1000000 | loss=2.3699 vocab=1.1014 token=1.5840 | 6.12 steps/s | VRAM=25.02 Go [2026-07-17 12:25:56] step 289100/1000000 | loss=2.4045 vocab=0.7478 token=1.6177 | 6.13 steps/s | VRAM=32.97 Go [2026-07-17 12:26:04] step 289150/1000000 | loss=2.2767 vocab=1.3009 token=1.8134 | 6.13 steps/s | VRAM=29.16 Go [2026-07-17 12:26:12] step 289200/1000000 | loss=2.3758 vocab=1.5402 token=1.7535 | 6.12 steps/s | VRAM=28.88 Go [2026-07-17 12:26:21] step 289250/1000000 | loss=2.2975 vocab=1.0851 token=1.4421 | 6.11 steps/s | VRAM=29.65 Go [2026-07-17 12:26:29] step 289300/1000000 | loss=2.3868 vocab=0.6786 token=1.5162 | 6.08 steps/s | VRAM=28.43 Go [2026-07-17 12:26:37] step 289350/1000000 | loss=2.3504 vocab=0.9217 token=1.8372 | 6.12 steps/s | VRAM=27.47 Go [2026-07-17 12:26:45] step 289400/1000000 | loss=2.3135 vocab=0.7271 token=1.1025 | 6.15 steps/s | VRAM=28.89 Go [2026-07-17 12:26:53] step 289450/1000000 | loss=2.2258 vocab=0.6026 token=0.8875 | 6.16 steps/s | VRAM=31.45 Go [2026-07-17 12:27:01] step 289500/1000000 | loss=2.4218 vocab=0.8803 token=1.1673 | 6.14 steps/s | VRAM=28.75 Go [2026-07-17 12:27:10] step 289550/1000000 | loss=2.2828 vocab=0.5423 token=0.8265 | 6.11 steps/s | VRAM=26.37 Go [2026-07-17 12:27:18] step 289600/1000000 | loss=2.3677 vocab=1.0375 token=0.9401 | 6.10 steps/s | VRAM=29.20 Go [2026-07-17 12:27:26] step 289650/1000000 | loss=2.3694 vocab=0.9492 token=1.0497 | 6.12 steps/s | VRAM=25.71 Go [2026-07-17 12:27:34] step 289700/1000000 | loss=2.2889 vocab=0.6028 token=1.4170 | 6.13 steps/s | VRAM=32.21 Go [2026-07-17 12:27:42] step 289750/1000000 | loss=2.2463 vocab=0.8982 token=1.2355 | 6.12 steps/s | VRAM=26.12 Go [2026-07-17 12:27:50] step 289800/1000000 | loss=2.1531 vocab=0.5930 token=1.0171 | 6.15 steps/s | VRAM=32.68 Go [2026-07-17 12:27:59] step 289850/1000000 | loss=2.2320 vocab=0.6222 token=0.8320 | 6.11 steps/s | VRAM=32.51 Go [2026-07-17 12:28:07] step 289900/1000000 | loss=2.2622 vocab=0.5723 token=0.7128 | 6.13 steps/s | VRAM=28.51 Go [2026-07-17 12:28:15] step 289950/1000000 | loss=2.2927 vocab=0.7259 token=1.4441 | 6.10 steps/s | VRAM=26.81 Go [2026-07-17 12:28:23] step 290000/1000000 | loss=2.2567 vocab=1.0916 token=1.5082 | 6.12 steps/s | VRAM=31.14 Go [step 290000] loss_token=1.5082 > min recent (1.3013) -> push saute [2026-07-17 12:28:31] step 290050/1000000 | loss=2.2802 vocab=0.8961 token=1.2964 | 6.12 steps/s | VRAM=32.81 Go [2026-07-17 12:28:39] step 290100/1000000 | loss=2.3341 vocab=0.7635 token=1.6670 | 6.12 steps/s | VRAM=33.02 Go [2026-07-17 12:28:48] step 290150/1000000 | loss=2.2815 vocab=1.5344 token=1.6468 | 6.10 steps/s | VRAM=23.82 Go [2026-07-17 12:28:56] step 290200/1000000 | loss=2.3469 vocab=0.9238 token=1.1435 | 6.14 steps/s | VRAM=25.25 Go [2026-07-17 12:29:04] step 290250/1000000 | loss=2.2062 vocab=0.3486 token=1.3665 | 6.17 steps/s | VRAM=27.70 Go [2026-07-17 12:29:12] step 290300/1000000 | loss=2.3046 vocab=0.8035 token=1.0233 | 6.13 steps/s | VRAM=24.92 Go [2026-07-17 12:29:20] step 290350/1000000 | loss=2.2731 vocab=0.5589 token=1.3531 | 6.14 steps/s | VRAM=30.50 Go [2026-07-17 12:29:28] step 290400/1000000 | loss=2.2919 vocab=0.5396 token=1.2150 | 6.17 steps/s | VRAM=29.79 Go [2026-07-17 12:29:36] step 290450/1000000 | loss=2.3347 vocab=1.1286 token=1.5892 | 6.16 steps/s | VRAM=25.25 Go [2026-07-17 12:29:45] step 290500/1000000 | loss=2.3232 vocab=0.6230 token=1.0387 | 6.15 steps/s | VRAM=32.04 Go [2026-07-17 12:29:53] step 290550/1000000 | loss=2.3182 vocab=0.7450 token=1.6150 | 6.13 steps/s | VRAM=28.06 Go [2026-07-17 12:30:01] step 290600/1000000 | loss=2.3497 vocab=0.9702 token=1.3070 | 6.14 steps/s | VRAM=26.54 Go [2026-07-17 12:30:09] step 290650/1000000 | loss=2.2640 vocab=1.0807 token=1.1653 | 6.13 steps/s | VRAM=32.43 Go [2026-07-17 12:30:17] step 290700/1000000 | loss=2.3497 vocab=1.2433 token=1.3532 | 6.14 steps/s | VRAM=28.57 Go [2026-07-17 12:30:25] step 290750/1000000 | loss=2.3576 vocab=1.4125 token=1.6539 | 6.06 steps/s | VRAM=26.42 Go [2026-07-17 12:30:34] step 290800/1000000 | loss=2.3228 vocab=0.9862 token=1.4628 | 6.14 steps/s | VRAM=32.31 Go [2026-07-17 12:30:42] step 290850/1000000 | loss=2.3520 vocab=0.7635 token=1.0296 | 6.12 steps/s | VRAM=26.21 Go [2026-07-17 12:30:50] step 290900/1000000 | loss=2.2952 vocab=1.0289 token=1.6299 | 6.14 steps/s | VRAM=24.00 Go [2026-07-17 12:30:58] step 290950/1000000 | loss=2.3133 vocab=0.9121 token=1.5270 | 6.15 steps/s | VRAM=32.69 Go [2026-07-17 12:31:06] step 291000/1000000 | loss=2.3703 vocab=0.6656 token=1.2153 | 6.13 steps/s | VRAM=32.42 Go [2026-07-17 12:31:14] step 291050/1000000 | loss=2.3380 vocab=0.7552 token=1.2545 | 6.15 steps/s | VRAM=31.80 Go [2026-07-17 12:31:22] step 291100/1000000 | loss=2.2595 vocab=0.7017 token=1.3108 | 6.13 steps/s | VRAM=32.69 Go [2026-07-17 12:31:31] step 291150/1000000 | loss=2.4215 vocab=0.9506 token=1.3543 | 6.12 steps/s | VRAM=32.54 Go [2026-07-17 12:31:39] step 291200/1000000 | loss=2.2976 vocab=1.0121 token=1.2804 | 6.13 steps/s | VRAM=28.70 Go [2026-07-17 12:31:47] step 291250/1000000 | loss=2.3195 vocab=0.4703 token=1.2599 | 6.13 steps/s | VRAM=32.07 Go [2026-07-17 12:31:55] step 291300/1000000 | loss=2.2418 vocab=0.7234 token=1.0769 | 6.13 steps/s | VRAM=28.86 Go [2026-07-17 12:32:03] step 291350/1000000 | loss=2.2960 vocab=0.7987 token=1.3340 | 6.10 steps/s | VRAM=27.06 Go [2026-07-17 12:32:11] step 291400/1000000 | loss=2.3105 vocab=0.5598 token=1.2087 | 6.15 steps/s | VRAM=32.54 Go [2026-07-17 12:32:20] step 291450/1000000 | loss=2.2907 vocab=1.1793 token=1.7424 | 6.16 steps/s | VRAM=27.61 Go [2026-07-17 12:32:28] step 291500/1000000 | loss=2.3658 vocab=1.1924 token=1.7476 | 6.11 steps/s | VRAM=24.87 Go [2026-07-17 12:32:36] step 291550/1000000 | loss=2.3563 vocab=0.4599 token=0.9921 | 6.13 steps/s | VRAM=32.46 Go [2026-07-17 12:32:44] step 291600/1000000 | loss=2.4292 vocab=1.0503 token=1.8055 | 6.10 steps/s | VRAM=28.72 Go [2026-07-17 12:32:52] step 291650/1000000 | loss=2.3465 vocab=0.9195 token=1.0470 | 6.13 steps/s | VRAM=29.02 Go [2026-07-17 12:33:00] step 291700/1000000 | loss=2.2071 vocab=0.6542 token=1.3004 | 6.12 steps/s | VRAM=25.37 Go [2026-07-17 12:33:09] step 291750/1000000 | loss=2.2578 vocab=0.7680 token=1.0774 | 6.10 steps/s | VRAM=29.25 Go [2026-07-17 12:33:17] step 291800/1000000 | loss=2.2527 vocab=1.0286 token=1.2918 | 6.13 steps/s | VRAM=26.00 Go [2026-07-17 12:33:25] step 291850/1000000 | loss=2.2264 vocab=0.8292 token=1.3901 | 6.13 steps/s | VRAM=32.53 Go [2026-07-17 12:33:33] step 291900/1000000 | loss=2.4297 vocab=1.5691 token=1.4606 | 6.12 steps/s | VRAM=29.05 Go [2026-07-17 12:33:41] step 291950/1000000 | loss=2.2329 vocab=1.4365 token=1.6223 | 6.11 steps/s | VRAM=28.28 Go [2026-07-17 12:33:49] step 292000/1000000 | loss=2.3388 vocab=0.9599 token=1.6957 | 6.13 steps/s | VRAM=32.09 Go [2026-07-17 12:33:58] step 292050/1000000 | loss=2.3315 vocab=1.3719 token=1.7360 | 6.14 steps/s | VRAM=32.92 Go [2026-07-17 12:34:06] step 292100/1000000 | loss=2.3919 vocab=0.9986 token=1.7425 | 6.11 steps/s | VRAM=28.44 Go [2026-07-17 12:34:14] step 292150/1000000 | loss=2.2830 vocab=0.9818 token=1.6035 | 6.15 steps/s | VRAM=28.90 Go [2026-07-17 12:34:22] step 292200/1000000 | loss=2.1935 vocab=1.0053 token=1.4047 | 6.10 steps/s | VRAM=29.87 Go [2026-07-17 12:34:30] step 292250/1000000 | loss=2.2253 vocab=1.4196 token=1.5570 | 6.12 steps/s | VRAM=25.26 Go [2026-07-17 12:34:38] step 292300/1000000 | loss=2.3171 vocab=1.5829 token=1.5734 | 6.12 steps/s | VRAM=29.45 Go [2026-07-17 12:34:47] step 292350/1000000 | loss=2.3167 vocab=1.4609 token=1.4723 | 6.13 steps/s | VRAM=26.14 Go [2026-07-17 12:34:55] step 292400/1000000 | loss=2.2870 vocab=1.0867 token=1.8376 | 6.13 steps/s | VRAM=28.85 Go [2026-07-17 12:35:03] step 292450/1000000 | loss=2.3624 vocab=0.7112 token=1.5675 | 6.11 steps/s | VRAM=28.93 Go [2026-07-17 12:35:11] step 292500/1000000 | loss=2.3248 vocab=0.6136 token=1.3230 | 6.11 steps/s | VRAM=28.96 Go [2026-07-17 12:35:19] step 292550/1000000 | loss=2.2685 vocab=0.6079 token=1.3929 | 6.12 steps/s | VRAM=26.01 Go [2026-07-17 12:35:27] step 292600/1000000 | loss=2.1670 vocab=0.7723 token=1.1085 | 6.12 steps/s | VRAM=32.67 Go [2026-07-17 12:35:36] step 292650/1000000 | loss=2.2920 vocab=1.0245 token=1.2368 | 6.11 steps/s | VRAM=32.22 Go [2026-07-17 12:35:44] step 292700/1000000 | loss=2.2668 vocab=0.9666 token=1.2464 | 6.14 steps/s | VRAM=31.12 Go [2026-07-17 12:35:52] step 292750/1000000 | loss=2.3464 vocab=0.5790 token=1.2635 | 6.11 steps/s | VRAM=28.64 Go [2026-07-17 12:36:00] step 292800/1000000 | loss=2.3669 vocab=0.9239 token=1.5223 | 6.13 steps/s | VRAM=29.40 Go [2026-07-17 12:36:08] step 292850/1000000 | loss=2.3578 vocab=0.4698 token=1.1206 | 6.14 steps/s | VRAM=30.21 Go [2026-07-17 12:36:16] step 292900/1000000 | loss=2.2977 vocab=1.0775 token=1.5248 | 6.09 steps/s | VRAM=32.99 Go [2026-07-17 12:36:25] step 292950/1000000 | loss=2.3296 vocab=0.5906 token=0.9362 | 6.10 steps/s | VRAM=28.59 Go [2026-07-17 12:36:33] step 293000/1000000 | loss=2.2687 vocab=0.8176 token=1.7291 | 6.13 steps/s | VRAM=28.86 Go [2026-07-17 12:36:41] step 293050/1000000 | loss=2.3461 vocab=0.7105 token=1.4572 | 6.11 steps/s | VRAM=25.78 Go [2026-07-17 12:36:49] step 293100/1000000 | loss=2.3254 vocab=0.7530 token=1.3803 | 6.11 steps/s | VRAM=32.41 Go [2026-07-17 12:36:57] step 293150/1000000 | loss=2.3155 vocab=1.1152 token=1.6714 | 6.12 steps/s | VRAM=28.85 Go [2026-07-17 12:37:06] step 293200/1000000 | loss=2.2757 vocab=0.9595 token=1.4984 | 6.11 steps/s | VRAM=28.06 Go [2026-07-17 12:37:14] step 293250/1000000 | loss=2.3848 vocab=0.9124 token=1.4047 | 6.13 steps/s | VRAM=32.26 Go [2026-07-17 12:37:22] step 293300/1000000 | loss=2.3785 vocab=0.9689 token=1.3932 | 6.11 steps/s | VRAM=29.46 Go [2026-07-17 12:37:30] step 293350/1000000 | loss=2.3623 vocab=0.5378 token=1.5103 | 6.12 steps/s | VRAM=26.02 Go [2026-07-17 12:37:38] step 293400/1000000 | loss=2.3419 vocab=0.9554 token=1.2108 | 6.12 steps/s | VRAM=32.33 Go [2026-07-17 12:37:46] step 293450/1000000 | loss=2.2519 vocab=1.0664 token=1.5934 | 6.11 steps/s | VRAM=32.17 Go [2026-07-17 12:37:55] step 293500/1000000 | loss=2.3818 vocab=0.8994 token=1.5289 | 6.09 steps/s | VRAM=29.39 Go [2026-07-17 12:38:03] step 293550/1000000 | loss=2.4105 vocab=0.7713 token=1.2731 | 6.10 steps/s | VRAM=30.85 Go [2026-07-17 12:38:11] step 293600/1000000 | loss=2.3754 vocab=0.7349 token=0.9261 | 6.09 steps/s | VRAM=27.94 Go [2026-07-17 12:38:19] step 293650/1000000 | loss=2.2695 vocab=1.4492 token=1.5765 | 6.15 steps/s | VRAM=31.87 Go [2026-07-17 12:38:27] step 293700/1000000 | loss=2.2000 vocab=0.6227 token=1.0462 | 6.09 steps/s | VRAM=26.32 Go [2026-07-17 12:38:36] step 293750/1000000 | loss=2.3515 vocab=1.0491 token=1.2717 | 6.13 steps/s | VRAM=31.61 Go [2026-07-17 12:38:44] step 293800/1000000 | loss=2.2626 vocab=0.9775 token=1.3968 | 6.12 steps/s | VRAM=30.03 Go [2026-07-17 12:38:52] step 293850/1000000 | loss=2.3109 vocab=1.2312 token=1.1834 | 6.11 steps/s | VRAM=29.05 Go [2026-07-17 12:39:00] step 293900/1000000 | loss=2.3136 vocab=0.5697 token=0.8626 | 6.13 steps/s | VRAM=25.56 Go [2026-07-17 12:39:08] step 293950/1000000 | loss=2.3987 vocab=1.1201 token=1.9047 | 6.15 steps/s | VRAM=28.55 Go [2026-07-17 12:39:16] step 294000/1000000 | loss=2.1503 vocab=0.9299 token=1.4780 | 6.15 steps/s | VRAM=26.43 Go [2026-07-17 12:39:24] step 294050/1000000 | loss=2.2524 vocab=1.1811 token=1.4576 | 6.15 steps/s | VRAM=31.46 Go [2026-07-17 12:39:33] step 294100/1000000 | loss=2.1747 vocab=0.7131 token=1.3782 | 6.14 steps/s | VRAM=31.86 Go [2026-07-17 12:39:41] step 294150/1000000 | loss=2.3070 vocab=1.0696 token=1.6271 | 6.15 steps/s | VRAM=31.45 Go [2026-07-17 12:39:49] step 294200/1000000 | loss=2.3855 vocab=1.0810 token=1.5415 | 6.13 steps/s | VRAM=29.13 Go [2026-07-17 12:39:57] step 294250/1000000 | loss=2.2740 vocab=0.9816 token=1.1661 | 6.15 steps/s | VRAM=32.39 Go [2026-07-17 12:40:05] step 294300/1000000 | loss=2.3057 vocab=0.6295 token=1.3823 | 6.15 steps/s | VRAM=32.08 Go [2026-07-17 12:40:13] step 294350/1000000 | loss=2.3117 vocab=1.0987 token=1.3073 | 6.13 steps/s | VRAM=25.07 Go [2026-07-17 12:40:21] step 294400/1000000 | loss=2.2857 vocab=0.9521 token=1.6349 | 6.17 steps/s | VRAM=32.38 Go [2026-07-17 12:40:30] step 294450/1000000 | loss=2.2897 vocab=1.0187 token=1.2662 | 6.15 steps/s | VRAM=28.92 Go [2026-07-17 12:40:38] step 294500/1000000 | loss=2.3924 vocab=0.9638 token=1.1866 | 6.11 steps/s | VRAM=29.51 Go [2026-07-17 12:40:46] step 294550/1000000 | loss=2.3790 vocab=1.0123 token=1.6533 | 6.13 steps/s | VRAM=28.71 Go [2026-07-17 12:40:54] step 294600/1000000 | loss=2.3462 vocab=0.9254 token=1.4386 | 6.09 steps/s | VRAM=28.58 Go [2026-07-17 12:41:02] step 294650/1000000 | loss=2.3777 vocab=1.0080 token=1.5732 | 6.12 steps/s | VRAM=26.24 Go [2026-07-17 12:41:10] step 294700/1000000 | loss=2.3548 vocab=0.6865 token=0.8439 | 6.16 steps/s | VRAM=31.87 Go [2026-07-17 12:41:19] step 294750/1000000 | loss=2.3814 vocab=0.7238 token=1.1545 | 6.13 steps/s | VRAM=28.54 Go [2026-07-17 12:41:27] step 294800/1000000 | loss=2.4658 vocab=1.1362 token=0.8550 | 6.11 steps/s | VRAM=33.01 Go [2026-07-17 12:41:35] step 294850/1000000 | loss=2.2968 vocab=1.0354 token=1.5852 | 6.11 steps/s | VRAM=29.76 Go [2026-07-17 12:41:43] step 294900/1000000 | loss=2.3316 vocab=0.8056 token=1.0327 | 6.11 steps/s | VRAM=28.58 Go [2026-07-17 12:41:51] step 294950/1000000 | loss=2.4813 vocab=1.0477 token=1.3062 | 6.15 steps/s | VRAM=27.37 Go [2026-07-17 12:41:59] step 295000/1000000 | loss=2.3660 vocab=0.4738 token=1.4455 | 6.11 steps/s | VRAM=28.57 Go [push] step 295000 OK (best-of-2, loss_token=1.4455) [2026-07-17 12:43:26] step 295050/1000000 | loss=2.3189 vocab=0.8829 token=1.5749 | 0.58 steps/s | VRAM=29.67 Go [2026-07-17 12:43:34] step 295100/1000000 | loss=2.2844 vocab=1.4872 token=1.2424 | 6.17 steps/s | VRAM=30.11 Go [2026-07-17 12:43:42] step 295150/1000000 | loss=2.3727 vocab=0.8965 token=1.3213 | 6.19 steps/s | VRAM=27.88 Go [2026-07-17 12:43:50] step 295200/1000000 | loss=2.3762 vocab=0.4374 token=1.1607 | 6.20 steps/s | VRAM=31.84 Go [2026-07-17 12:43:58] step 295250/1000000 | loss=2.3399 vocab=0.8907 token=1.2281 | 6.20 steps/s | VRAM=31.95 Go [2026-07-17 12:44:06] step 295300/1000000 | loss=2.3590 vocab=1.0437 token=0.9993 | 6.15 steps/s | VRAM=28.81 Go [2026-07-17 12:44:14] step 295350/1000000 | loss=2.3498 vocab=0.6731 token=1.2777 | 6.17 steps/s | VRAM=28.19 Go [2026-07-17 12:44:22] step 295400/1000000 | loss=2.2600 vocab=0.7875 token=1.4653 | 6.12 steps/s | VRAM=29.10 Go [2026-07-17 12:44:30] step 295450/1000000 | loss=2.4304 vocab=1.1485 token=1.7641 | 6.15 steps/s | VRAM=32.95 Go [2026-07-17 12:44:38] step 295500/1000000 | loss=2.3636 vocab=0.6173 token=1.3740 | 6.14 steps/s | VRAM=32.49 Go [2026-07-17 12:44:47] step 295550/1000000 | loss=2.2665 vocab=1.0410 token=1.7436 | 6.16 steps/s | VRAM=28.59 Go [2026-07-17 12:44:55] step 295600/1000000 | loss=2.3185 vocab=1.0390 token=1.6776 | 6.16 steps/s | VRAM=29.22 Go [2026-07-17 12:45:03] step 295650/1000000 | loss=2.2934 vocab=1.0060 token=1.5193 | 6.17 steps/s | VRAM=29.33 Go [2026-07-17 12:45:11] step 295700/1000000 | loss=2.3463 vocab=0.9283 token=1.5113 | 6.14 steps/s | VRAM=31.66 Go [2026-07-17 12:45:19] step 295750/1000000 | loss=2.4063 vocab=0.8825 token=1.5766 | 6.13 steps/s | VRAM=25.36 Go [2026-07-17 12:45:27] step 295800/1000000 | loss=2.3160 vocab=0.9513 token=0.9746 | 6.12 steps/s | VRAM=29.49 Go [2026-07-17 12:45:35] step 295850/1000000 | loss=2.3685 vocab=0.9271 token=1.1140 | 6.16 steps/s | VRAM=31.82 Go [2026-07-17 12:45:44] step 295900/1000000 | loss=2.3162 vocab=0.6399 token=1.2877 | 6.12 steps/s | VRAM=28.17 Go [2026-07-17 12:45:52] step 295950/1000000 | loss=2.2930 vocab=0.9817 token=0.8661 | 6.13 steps/s | VRAM=29.26 Go [2026-07-17 12:46:00] step 296000/1000000 | loss=2.3428 vocab=1.0846 token=1.2491 | 6.14 steps/s | VRAM=29.09 Go [2026-07-17 12:46:08] step 296050/1000000 | loss=2.3867 vocab=0.9989 token=1.7059 | 6.15 steps/s | VRAM=25.04 Go [2026-07-17 12:46:16] step 296100/1000000 | loss=2.2310 vocab=0.6932 token=1.2584 | 6.12 steps/s | VRAM=27.84 Go [2026-07-17 12:46:24] step 296150/1000000 | loss=2.2848 vocab=1.2743 token=1.8962 | 6.13 steps/s | VRAM=32.41 Go [2026-07-17 12:46:33] step 296200/1000000 | loss=2.2994 vocab=1.0259 token=1.0512 | 6.14 steps/s | VRAM=31.16 Go [2026-07-17 12:46:41] step 296250/1000000 | loss=2.2570 vocab=0.8050 token=1.0223 | 6.13 steps/s | VRAM=28.50 Go [2026-07-17 12:46:49] step 296300/1000000 | loss=2.3200 vocab=1.5752 token=1.5769 | 6.12 steps/s | VRAM=28.51 Go [2026-07-17 12:46:57] step 296350/1000000 | loss=2.3690 vocab=2.1535 token=1.5054 | 6.10 steps/s | VRAM=28.49 Go [2026-07-17 12:47:05] step 296400/1000000 | loss=2.2705 vocab=0.7588 token=1.5923 | 6.10 steps/s | VRAM=28.09 Go [2026-07-17 12:47:13] step 296450/1000000 | loss=2.2578 vocab=0.8996 token=1.8454 | 6.14 steps/s | VRAM=31.26 Go [2026-07-17 12:47:22] step 296500/1000000 | loss=2.2847 vocab=1.2095 token=1.6593 | 6.13 steps/s | VRAM=32.07 Go [2026-07-17 12:47:30] step 296550/1000000 | loss=2.3080 vocab=0.6743 token=1.3201 | 6.13 steps/s | VRAM=28.99 Go [2026-07-17 12:47:38] step 296600/1000000 | loss=2.2965 vocab=1.2711 token=2.0450 | 6.16 steps/s | VRAM=31.47 Go [2026-07-17 12:47:46] step 296650/1000000 | loss=2.4058 vocab=0.7847 token=1.2253 | 6.13 steps/s | VRAM=31.95 Go [2026-07-17 12:47:54] step 296700/1000000 | loss=2.3184 vocab=1.1835 token=1.5137 | 6.13 steps/s | VRAM=28.72 Go [2026-07-17 12:48:02] step 296750/1000000 | loss=2.3567 vocab=0.9450 token=1.7741 | 6.15 steps/s | VRAM=28.61 Go [2026-07-17 12:48:10] step 296800/1000000 | loss=2.3240 vocab=0.8782 token=1.5264 | 6.14 steps/s | VRAM=32.56 Go [2026-07-17 12:48:19] step 296850/1000000 | loss=2.3592 vocab=1.2765 token=1.2270 | 6.12 steps/s | VRAM=28.85 Go [2026-07-17 12:48:27] step 296900/1000000 | loss=2.3471 vocab=1.0446 token=1.3988 | 6.15 steps/s | VRAM=31.66 Go [2026-07-17 12:48:35] step 296950/1000000 | loss=2.2501 vocab=1.1918 token=1.4137 | 6.13 steps/s | VRAM=29.68 Go [2026-07-17 12:48:43] step 297000/1000000 | loss=2.2365 vocab=1.3555 token=1.7259 | 6.09 steps/s | VRAM=28.16 Go [2026-07-17 12:48:51] step 297050/1000000 | loss=2.2775 vocab=0.8085 token=0.9548 | 6.15 steps/s | VRAM=31.72 Go [2026-07-17 13:01:23] step 297100/1000000 | loss=0.6336 vocab=1.0049 token=1.4457 | 16.57 steps/s | VRAM=32.71 Go [2026-07-17 13:01:31] step 297150/1000000 | loss=2.3521 vocab=0.6709 token=1.0140 | 6.26 steps/s | VRAM=29.20 Go [2026-07-17 13:01:39] step 297200/1000000 | loss=2.3108 vocab=0.5964 token=1.4069 | 6.20 steps/s | VRAM=29.01 Go [2026-07-17 13:01:47] step 297250/1000000 | loss=2.3156 vocab=0.6211 token=1.2778 | 6.19 steps/s | VRAM=26.74 Go [2026-07-17 13:01:55] step 297300/1000000 | loss=2.3050 vocab=1.1282 token=1.5552 | 6.21 steps/s | VRAM=29.95 Go [2026-07-17 13:02:03] step 297350/1000000 | loss=2.4188 vocab=1.1228 token=1.5072 | 6.18 steps/s | VRAM=25.20 Go [2026-07-17 13:02:11] step 297400/1000000 | loss=2.2935 vocab=1.0618 token=1.2246 | 6.16 steps/s | VRAM=25.58 Go [2026-07-17 13:02:19] step 297450/1000000 | loss=2.3627 vocab=0.7451 token=1.2866 | 6.24 steps/s | VRAM=31.04 Go [2026-07-17 13:02:27] step 297500/1000000 | loss=2.3070 vocab=0.8588 token=1.2806 | 6.16 steps/s | VRAM=28.01 Go [2026-07-17 13:02:35] step 297550/1000000 | loss=2.2360 vocab=1.2998 token=1.3875 | 6.17 steps/s | VRAM=28.03 Go [2026-07-17 13:02:44] step 297600/1000000 | loss=2.3588 vocab=0.5142 token=1.3422 | 6.14 steps/s | VRAM=29.14 Go [2026-07-17 13:02:52] step 297650/1000000 | loss=2.2860 vocab=0.6900 token=1.5389 | 6.18 steps/s | VRAM=28.27 Go [2026-07-17 13:03:00] step 297700/1000000 | loss=2.3275 vocab=1.0475 token=1.9468 | 6.16 steps/s | VRAM=28.52 Go [2026-07-17 13:03:08] step 297750/1000000 | loss=2.3237 vocab=0.7443 token=1.2821 | 6.16 steps/s | VRAM=25.43 Go [2026-07-17 13:03:16] step 297800/1000000 | loss=2.2462 vocab=1.0723 token=1.5902 | 6.15 steps/s | VRAM=29.56 Go [2026-07-17 13:03:24] step 297850/1000000 | loss=2.3417 vocab=1.0892 token=1.3310 | 6.14 steps/s | VRAM=25.92 Go [2026-07-17 13:03:32] step 297900/1000000 | loss=2.3265 vocab=0.9693 token=1.6660 | 6.14 steps/s | VRAM=31.74 Go [2026-07-17 13:03:41] step 297950/1000000 | loss=2.3192 vocab=1.2199 token=1.2753 | 6.12 steps/s | VRAM=29.15 Go [2026-07-17 13:03:49] step 298000/1000000 | loss=2.2093 vocab=0.5145 token=1.2202 | 6.14 steps/s | VRAM=32.02 Go [2026-07-17 13:03:57] step 298050/1000000 | loss=2.2212 vocab=1.2839 token=1.1134 | 6.16 steps/s | VRAM=28.79 Go [2026-07-17 13:04:05] step 298100/1000000 | loss=2.2722 vocab=1.2285 token=1.7187 | 6.11 steps/s | VRAM=28.43 Go [2026-07-17 13:04:13] step 298150/1000000 | loss=2.2452 vocab=0.9248 token=1.2930 | 6.11 steps/s | VRAM=29.59 Go [2026-07-17 13:04:21] step 298200/1000000 | loss=2.3815 vocab=0.5866 token=1.2033 | 6.11 steps/s | VRAM=24.63 Go [2026-07-17 13:04:30] step 298250/1000000 | loss=2.2008 vocab=0.7473 token=1.5440 | 6.12 steps/s | VRAM=29.28 Go [2026-07-17 13:04:38] step 298300/1000000 | loss=2.3481 vocab=1.2101 token=1.5002 | 6.12 steps/s | VRAM=27.94 Go [2026-07-17 13:04:46] step 298350/1000000 | loss=2.3014 vocab=1.4562 token=1.2549 | 6.12 steps/s | VRAM=28.89 Go [2026-07-17 13:04:54] step 298400/1000000 | loss=2.3175 vocab=0.8096 token=1.3024 | 6.10 steps/s | VRAM=28.80 Go [2026-07-17 13:05:02] step 298450/1000000 | loss=2.3053 vocab=0.8103 token=1.0503 | 6.13 steps/s | VRAM=29.18 Go [2026-07-17 13:05:10] step 298500/1000000 | loss=2.3663 vocab=0.7148 token=1.2834 | 6.12 steps/s | VRAM=26.36 Go [2026-07-17 13:05:19] step 298550/1000000 | loss=2.3826 vocab=1.1214 token=1.2557 | 6.12 steps/s | VRAM=28.70 Go [2026-07-17 13:05:27] step 298600/1000000 | loss=2.2496 vocab=0.6491 token=0.9090 | 6.11 steps/s | VRAM=32.45 Go [2026-07-17 13:05:35] step 298650/1000000 | loss=2.3959 vocab=1.0677 token=1.2019 | 6.10 steps/s | VRAM=28.21 Go [2026-07-17 13:05:43] step 298700/1000000 | loss=2.3788 vocab=0.8228 token=1.0951 | 6.10 steps/s | VRAM=32.71 Go [2026-07-17 13:05:51] step 298750/1000000 | loss=2.2597 vocab=0.8427 token=1.4945 | 6.12 steps/s | VRAM=31.79 Go [2026-07-17 13:05:59] step 298800/1000000 | loss=2.3080 vocab=0.7060 token=1.1516 | 6.09 steps/s | VRAM=29.39 Go [2026-07-17 13:06:08] step 298850/1000000 | loss=2.3550 vocab=0.9117 token=1.3725 | 6.10 steps/s | VRAM=32.51 Go [2026-07-17 13:06:16] step 298900/1000000 | loss=2.2683 vocab=0.5772 token=0.8041 | 6.08 steps/s | VRAM=28.54 Go [2026-07-17 13:06:24] step 298950/1000000 | loss=2.2174 vocab=0.9738 token=1.4630 | 6.11 steps/s | VRAM=29.17 Go [2026-07-17 13:06:32] step 299000/1000000 | loss=2.3046 vocab=1.1248 token=1.1236 | 6.14 steps/s | VRAM=31.87 Go [2026-07-17 13:06:40] step 299050/1000000 | loss=2.2311 vocab=0.9325 token=1.0759 | 6.10 steps/s | VRAM=28.21 Go [2026-07-17 13:06:49] step 299100/1000000 | loss=2.3456 vocab=1.4663 token=1.6071 | 6.10 steps/s | VRAM=30.52 Go [2026-07-17 13:06:57] step 299150/1000000 | loss=2.3733 vocab=1.1617 token=1.4487 | 6.13 steps/s | VRAM=27.60 Go [2026-07-17 13:07:05] step 299200/1000000 | loss=2.3297 vocab=0.6051 token=1.2032 | 6.14 steps/s | VRAM=28.56 Go [2026-07-17 13:07:13] step 299250/1000000 | loss=2.3712 vocab=0.9650 token=1.8795 | 6.13 steps/s | VRAM=32.25 Go [2026-07-17 13:07:21] step 299300/1000000 | loss=2.3296 vocab=0.7989 token=1.5370 | 6.10 steps/s | VRAM=24.90 Go [2026-07-17 13:07:30] step 299350/1000000 | loss=2.4093 vocab=1.0117 token=1.3296 | 6.10 steps/s | VRAM=25.61 Go [2026-07-17 13:07:38] step 299400/1000000 | loss=2.3531 vocab=0.7656 token=1.3752 | 6.13 steps/s | VRAM=25.55 Go [2026-07-17 13:07:46] step 299450/1000000 | loss=2.3546 vocab=0.8483 token=1.0854 | 6.16 steps/s | VRAM=28.30 Go [2026-07-17 13:07:54] step 299500/1000000 | loss=2.2795 vocab=0.9053 token=1.3601 | 6.12 steps/s | VRAM=31.93 Go [2026-07-17 13:08:02] step 299550/1000000 | loss=2.3666 vocab=0.5849 token=1.0573 | 6.12 steps/s | VRAM=26.94 Go [2026-07-17 13:08:10] step 299600/1000000 | loss=2.4240 vocab=0.9855 token=1.9836 | 6.14 steps/s | VRAM=32.46 Go [2026-07-17 13:08:18] step 299650/1000000 | loss=2.3322 vocab=0.7004 token=1.5018 | 6.12 steps/s | VRAM=26.17 Go [2026-07-17 13:08:27] step 299700/1000000 | loss=2.2485 vocab=0.7523 token=1.2232 | 6.15 steps/s | VRAM=31.92 Go [2026-07-17 13:08:35] step 299750/1000000 | loss=2.3140 vocab=0.9446 token=1.4742 | 6.08 steps/s | VRAM=29.22 Go [2026-07-17 13:08:43] step 299800/1000000 | loss=2.3213 vocab=1.1731 token=1.8333 | 6.12 steps/s | VRAM=32.94 Go [2026-07-17 13:08:51] step 299850/1000000 | loss=2.3854 vocab=1.1816 token=1.3246 | 6.14 steps/s | VRAM=28.08 Go [2026-07-17 13:08:59] step 299900/1000000 | loss=2.2944 vocab=1.0246 token=1.4611 | 6.11 steps/s | VRAM=28.07 Go [2026-07-17 13:09:07] step 299950/1000000 | loss=2.2099 vocab=0.8832 token=1.2921 | 6.14 steps/s | VRAM=25.13 Go [2026-07-17 13:09:16] step 300000/1000000 | loss=2.3388 vocab=1.0953 token=1.3646 | 6.14 steps/s | VRAM=30.88 Go [push] step 300000 OK (best-of-2, loss_token=1.3646) [2026-07-17 13:11:17] step 300050/1000000 | loss=2.3342 vocab=0.2763 token=1.1116 | 0.41 steps/s | VRAM=32.04 Go [2026-07-17 13:11:25] step 300100/1000000 | loss=2.3268 vocab=0.9119 token=1.4969 | 6.17 steps/s | VRAM=28.56 Go [2026-07-17 13:11:33] step 300150/1000000 | loss=2.3340 vocab=0.7960 token=1.3724 | 6.18 steps/s | VRAM=28.75 Go [2026-07-17 13:11:41] step 300200/1000000 | loss=2.3268 vocab=0.6783 token=1.1172 | 6.20 steps/s | VRAM=31.66 Go [2026-07-17 13:11:50] step 300250/1000000 | loss=2.3863 vocab=0.8732 token=1.6753 | 6.15 steps/s | VRAM=26.56 Go [2026-07-17 13:11:58] step 300300/1000000 | loss=2.2975 vocab=0.9385 token=1.4812 | 6.17 steps/s | VRAM=25.56 Go [2026-07-17 13:12:06] step 300350/1000000 | loss=2.2870 vocab=0.7981 token=1.2543 | 6.21 steps/s | VRAM=31.35 Go [2026-07-17 13:12:14] step 300400/1000000 | loss=2.2585 vocab=0.6124 token=1.4256 | 6.12 steps/s | VRAM=26.59 Go [2026-07-17 13:12:22] step 300450/1000000 | loss=2.2274 vocab=0.6345 token=1.6405 | 6.18 steps/s | VRAM=29.31 Go [2026-07-17 13:12:30] step 300500/1000000 | loss=2.3573 vocab=0.7768 token=1.1345 | 6.18 steps/s | VRAM=25.40 Go [2026-07-17 13:12:38] step 300550/1000000 | loss=2.3292 vocab=1.2772 token=1.2652 | 6.14 steps/s | VRAM=25.09 Go [2026-07-17 13:12:46] step 300600/1000000 | loss=2.2815 vocab=0.3877 token=0.9971 | 6.17 steps/s | VRAM=25.06 Go [2026-07-17 13:12:54] step 300650/1000000 | loss=2.3593 vocab=0.9800 token=1.0898 | 6.15 steps/s | VRAM=28.37 Go [2026-07-17 13:13:03] step 300700/1000000 | loss=2.3983 vocab=0.4938 token=1.5806 | 6.11 steps/s | VRAM=28.77 Go [2026-07-17 13:13:11] step 300750/1000000 | loss=2.2366 vocab=0.6996 token=1.1921 | 6.14 steps/s | VRAM=32.07 Go [2026-07-17 13:13:19] step 300800/1000000 | loss=2.4408 vocab=1.1660 token=1.7891 | 6.15 steps/s | VRAM=32.51 Go [2026-07-17 13:13:27] step 300850/1000000 | loss=2.3022 vocab=0.6569 token=1.2868 | 6.14 steps/s | VRAM=28.41 Go [2026-07-17 13:13:35] step 300900/1000000 | loss=2.3479 vocab=0.5546 token=1.3038 | 6.11 steps/s | VRAM=25.51 Go [2026-07-17 13:13:43] step 300950/1000000 | loss=2.2749 vocab=1.0616 token=1.3611 | 6.13 steps/s | VRAM=28.59 Go [2026-07-17 13:13:52] step 301000/1000000 | loss=2.2791 vocab=0.7278 token=1.3611 | 6.13 steps/s | VRAM=31.84 Go [2026-07-17 13:14:00] step 301050/1000000 | loss=2.3454 vocab=1.0066 token=1.4746 | 6.12 steps/s | VRAM=28.50 Go [2026-07-17 13:14:08] step 301100/1000000 | loss=2.2238 vocab=0.5710 token=1.3464 | 6.14 steps/s | VRAM=31.94 Go [2026-07-17 13:14:16] step 301150/1000000 | loss=2.2971 vocab=1.1534 token=1.5202 | 6.14 steps/s | VRAM=27.76 Go [2026-07-17 13:14:24] step 301200/1000000 | loss=2.2497 vocab=0.8140 token=1.2763 | 6.11 steps/s | VRAM=26.54 Go [2026-07-17 13:14:32] step 301250/1000000 | loss=2.2817 vocab=0.9801 token=1.2836 | 6.14 steps/s | VRAM=28.33 Go [2026-07-17 13:14:41] step 301300/1000000 | loss=2.2592 vocab=1.1570 token=1.5408 | 6.12 steps/s | VRAM=26.54 Go [2026-07-17 13:14:49] step 301350/1000000 | loss=2.3211 vocab=1.1110 token=1.5494 | 6.14 steps/s | VRAM=28.22 Go [2026-07-17 13:14:57] step 301400/1000000 | loss=2.2851 vocab=0.8622 token=1.2801 | 6.14 steps/s | VRAM=25.46 Go [2026-07-17 13:15:05] step 301450/1000000 | loss=2.3704 vocab=0.7920 token=1.4202 | 6.13 steps/s | VRAM=26.55 Go [2026-07-17 13:15:13] step 301500/1000000 | loss=2.2151 vocab=1.0601 token=1.3080 | 6.12 steps/s | VRAM=28.28 Go [2026-07-17 13:15:21] step 301550/1000000 | loss=2.3515 vocab=1.1775 token=1.4901 | 6.14 steps/s | VRAM=32.78 Go [2026-07-17 13:15:29] step 301600/1000000 | loss=2.2459 vocab=0.8778 token=1.1020 | 6.13 steps/s | VRAM=25.64 Go [2026-07-17 13:15:38] step 301650/1000000 | loss=2.2694 vocab=1.5520 token=1.4839 | 6.12 steps/s | VRAM=32.90 Go [2026-07-17 13:15:46] step 301700/1000000 | loss=2.3812 vocab=0.5232 token=1.4258 | 6.14 steps/s | VRAM=31.85 Go [2026-07-17 13:15:54] step 301750/1000000 | loss=2.3678 vocab=0.7324 token=1.0150 | 6.10 steps/s | VRAM=26.35 Go [2026-07-17 13:16:02] step 301800/1000000 | loss=2.3179 vocab=1.1839 token=1.6052 | 6.14 steps/s | VRAM=28.93 Go [2026-07-17 13:16:10] step 301850/1000000 | loss=2.2574 vocab=1.0447 token=1.3798 | 6.13 steps/s | VRAM=31.27 Go [2026-07-17 13:16:18] step 301900/1000000 | loss=2.3069 vocab=1.1242 token=1.2701 | 6.12 steps/s | VRAM=25.99 Go [2026-07-17 13:16:27] step 301950/1000000 | loss=2.2970 vocab=1.1380 token=1.4511 | 6.13 steps/s | VRAM=29.06 Go [2026-07-17 13:16:35] step 302000/1000000 | loss=2.3710 vocab=0.8861 token=1.3040 | 6.11 steps/s | VRAM=28.43 Go [2026-07-17 13:16:43] step 302050/1000000 | loss=2.2417 vocab=0.7301 token=1.0170 | 6.10 steps/s | VRAM=23.32 Go [2026-07-17 13:16:51] step 302100/1000000 | loss=2.3081 vocab=0.9383 token=1.4284 | 6.14 steps/s | VRAM=28.36 Go [2026-07-17 13:16:59] step 302150/1000000 | loss=2.3197 vocab=0.8753 token=1.1862 | 6.14 steps/s | VRAM=32.68 Go [2026-07-17 13:17:07] step 302200/1000000 | loss=2.3433 vocab=1.2902 token=1.4016 | 6.14 steps/s | VRAM=28.60 Go [2026-07-17 13:17:16] step 302250/1000000 | loss=2.3578 vocab=0.5416 token=1.1620 | 6.10 steps/s | VRAM=27.07 Go [2026-07-17 13:17:24] step 302300/1000000 | loss=2.2934 vocab=0.6273 token=1.5203 | 6.17 steps/s | VRAM=24.76 Go [2026-07-17 13:17:32] step 302350/1000000 | loss=2.2707 vocab=0.8957 token=1.1899 | 6.14 steps/s | VRAM=26.48 Go [2026-07-17 13:17:40] step 302400/1000000 | loss=2.2576 vocab=0.7493 token=1.5223 | 6.11 steps/s | VRAM=29.83 Go [2026-07-17 13:17:48] step 302450/1000000 | loss=2.3306 vocab=0.8136 token=1.4422 | 6.10 steps/s | VRAM=28.50 Go [2026-07-17 13:17:56] step 302500/1000000 | loss=2.2026 vocab=0.4821 token=1.0518 | 6.09 steps/s | VRAM=26.67 Go [2026-07-17 13:18:05] step 302550/1000000 | loss=2.3602 vocab=0.5403 token=1.1601 | 6.14 steps/s | VRAM=32.67 Go [2026-07-17 13:18:13] step 302600/1000000 | loss=2.3113 vocab=1.2932 token=1.0625 | 6.10 steps/s | VRAM=29.27 Go [2026-07-17 13:18:21] step 302650/1000000 | loss=2.3274 vocab=0.9905 token=1.4785 | 6.10 steps/s | VRAM=29.55 Go [2026-07-17 13:18:29] step 302700/1000000 | loss=2.2269 vocab=1.1045 token=1.4951 | 6.11 steps/s | VRAM=30.70 Go [2026-07-17 13:18:37] step 302750/1000000 | loss=2.3843 vocab=0.6137 token=1.1131 | 6.10 steps/s | VRAM=32.66 Go [2026-07-17 13:18:45] step 302800/1000000 | loss=2.2739 vocab=0.8279 token=1.4821 | 6.14 steps/s | VRAM=31.08 Go [2026-07-17 13:18:54] step 302850/1000000 | loss=2.3633 vocab=1.0540 token=1.5139 | 6.11 steps/s | VRAM=29.09 Go [2026-07-17 13:19:02] step 302900/1000000 | loss=2.3743 vocab=0.8046 token=1.1409 | 6.12 steps/s | VRAM=28.46 Go [2026-07-17 13:19:10] step 302950/1000000 | loss=2.2813 vocab=0.9063 token=1.9235 | 6.11 steps/s | VRAM=30.11 Go [2026-07-17 13:19:18] step 303000/1000000 | loss=2.3491 vocab=0.9440 token=0.8412 | 6.15 steps/s | VRAM=29.81 Go [2026-07-17 13:19:26] step 303050/1000000 | loss=2.2629 vocab=0.8894 token=1.3337 | 6.11 steps/s | VRAM=28.81 Go [2026-07-17 13:19:34] step 303100/1000000 | loss=2.4339 vocab=1.0529 token=1.4082 | 6.14 steps/s | VRAM=32.59 Go [2026-07-17 13:19:43] step 303150/1000000 | loss=2.2280 vocab=0.6659 token=1.3051 | 6.13 steps/s | VRAM=25.99 Go [2026-07-17 13:19:51] step 303200/1000000 | loss=2.3411 vocab=0.8061 token=1.5371 | 6.15 steps/s | VRAM=24.53 Go [2026-07-17 13:19:59] step 303250/1000000 | loss=2.2708 vocab=0.8204 token=1.5158 | 6.14 steps/s | VRAM=25.36 Go [2026-07-17 13:20:07] step 303300/1000000 | loss=2.2654 vocab=0.9303 token=1.2744 | 6.11 steps/s | VRAM=28.71 Go [2026-07-17 13:20:15] step 303350/1000000 | loss=2.2787 vocab=0.7908 token=1.3131 | 6.13 steps/s | VRAM=27.89 Go [2026-07-17 13:20:23] step 303400/1000000 | loss=2.2744 vocab=0.8131 token=1.4101 | 6.13 steps/s | VRAM=31.05 Go [2026-07-17 13:20:32] step 303450/1000000 | loss=2.3107 vocab=1.3863 token=1.5298 | 6.12 steps/s | VRAM=29.35 Go [2026-07-17 13:20:40] step 303500/1000000 | loss=2.3538 vocab=0.9894 token=1.5499 | 6.12 steps/s | VRAM=29.09 Go [2026-07-17 13:20:48] step 303550/1000000 | loss=2.3782 vocab=1.2093 token=1.7109 | 6.14 steps/s | VRAM=27.76 Go [2026-07-17 13:20:56] step 303600/1000000 | loss=2.2252 vocab=0.5453 token=1.2574 | 6.13 steps/s | VRAM=28.51 Go [2026-07-17 13:21:04] step 303650/1000000 | loss=2.2675 vocab=0.8901 token=1.6078 | 6.13 steps/s | VRAM=32.04 Go [2026-07-17 13:21:12] step 303700/1000000 | loss=2.1998 vocab=0.6646 token=1.2377 | 6.13 steps/s | VRAM=32.15 Go [2026-07-17 13:21:21] step 303750/1000000 | loss=2.1278 vocab=0.7129 token=1.2144 | 6.12 steps/s | VRAM=28.43 Go [2026-07-17 13:21:29] step 303800/1000000 | loss=2.3996 vocab=0.8668 token=0.9132 | 6.12 steps/s | VRAM=28.02 Go [2026-07-17 13:21:37] step 303850/1000000 | loss=2.3387 vocab=0.6143 token=1.5822 | 6.10 steps/s | VRAM=26.78 Go [2026-07-17 13:21:45] step 303900/1000000 | loss=2.3047 vocab=1.2346 token=0.9489 | 6.13 steps/s | VRAM=29.14 Go [2026-07-17 13:21:53] step 303950/1000000 | loss=2.2911 vocab=0.9101 token=1.1046 | 6.10 steps/s | VRAM=28.96 Go [2026-07-17 13:22:01] step 304000/1000000 | loss=2.1739 vocab=0.5479 token=1.4878 | 6.14 steps/s | VRAM=28.65 Go [2026-07-17 13:22:09] step 304050/1000000 | loss=2.3167 vocab=0.6644 token=1.3085 | 6.14 steps/s | VRAM=25.59 Go [2026-07-17 13:22:18] step 304100/1000000 | loss=2.1752 vocab=0.9572 token=1.4791 | 6.17 steps/s | VRAM=28.44 Go [2026-07-17 13:22:26] step 304150/1000000 | loss=2.3211 vocab=1.0739 token=1.5845 | 6.13 steps/s | VRAM=28.91 Go [2026-07-17 13:22:34] step 304200/1000000 | loss=2.3032 vocab=0.8196 token=1.4075 | 6.15 steps/s | VRAM=29.08 Go [2026-07-17 13:22:42] step 304250/1000000 | loss=2.3373 vocab=1.0055 token=1.6311 | 6.12 steps/s | VRAM=28.15 Go [2026-07-17 13:22:50] step 304300/1000000 | loss=2.2551 vocab=0.8374 token=1.4494 | 6.13 steps/s | VRAM=32.00 Go [2026-07-17 13:22:58] step 304350/1000000 | loss=2.2774 vocab=0.8114 token=1.3681 | 6.15 steps/s | VRAM=28.44 Go [2026-07-17 13:23:06] step 304400/1000000 | loss=2.3453 vocab=0.9805 token=1.5175 | 6.13 steps/s | VRAM=25.99 Go [2026-07-17 13:23:15] step 304450/1000000 | loss=2.3478 vocab=0.7259 token=1.4528 | 6.13 steps/s | VRAM=32.12 Go [2026-07-17 13:23:23] step 304500/1000000 | loss=2.3502 vocab=0.8207 token=1.6608 | 6.14 steps/s | VRAM=29.03 Go [2026-07-17 13:23:31] step 304550/1000000 | loss=2.3011 vocab=0.9357 token=1.3922 | 6.14 steps/s | VRAM=28.56 Go [2026-07-17 13:23:39] step 304600/1000000 | loss=2.2098 vocab=1.2922 token=1.6247 | 6.15 steps/s | VRAM=31.88 Go [2026-07-17 13:23:47] step 304650/1000000 | loss=2.3432 vocab=0.7112 token=1.5880 | 6.15 steps/s | VRAM=32.25 Go [2026-07-17 13:23:55] step 304700/1000000 | loss=2.1730 vocab=1.0747 token=1.4988 | 6.15 steps/s | VRAM=31.87 Go [2026-07-17 13:24:04] step 304750/1000000 | loss=2.3625 vocab=1.0585 token=1.2034 | 6.12 steps/s | VRAM=26.34 Go [2026-07-17 13:24:12] step 304800/1000000 | loss=2.3088 vocab=0.9945 token=1.2515 | 6.14 steps/s | VRAM=28.52 Go [2026-07-17 13:24:20] step 304850/1000000 | loss=2.2968 vocab=1.1032 token=1.4763 | 6.15 steps/s | VRAM=31.74 Go [2026-07-17 13:24:28] step 304900/1000000 | loss=2.3063 vocab=1.1264 token=1.5487 | 6.18 steps/s | VRAM=31.03 Go [2026-07-17 13:24:36] step 304950/1000000 | loss=2.2343 vocab=0.8677 token=1.4027 | 6.14 steps/s | VRAM=29.74 Go [2026-07-17 13:24:44] step 305000/1000000 | loss=2.3074 vocab=1.1569 token=1.3272 | 6.09 steps/s | VRAM=26.43 Go [2026-07-17 13:24:52] step 305050/1000000 | loss=2.2337 vocab=0.4623 token=1.0998 | 6.14 steps/s | VRAM=31.92 Go [2026-07-17 13:25:01] step 305100/1000000 | loss=2.1966 vocab=0.4850 token=1.1321 | 6.14 steps/s | VRAM=29.03 Go [2026-07-17 13:25:09] step 305150/1000000 | loss=2.3139 vocab=0.8778 token=1.5992 | 6.11 steps/s | VRAM=27.34 Go [2026-07-17 13:25:17] step 305200/1000000 | loss=2.3839 vocab=0.7402 token=1.3542 | 6.15 steps/s | VRAM=29.46 Go [2026-07-17 13:25:25] step 305250/1000000 | loss=2.3285 vocab=0.5687 token=1.0791 | 6.16 steps/s | VRAM=28.76 Go [2026-07-17 13:25:33] step 305300/1000000 | loss=2.2681 vocab=1.0036 token=1.5868 | 6.15 steps/s | VRAM=28.66 Go [2026-07-17 13:25:41] step 305350/1000000 | loss=2.2920 vocab=0.5791 token=1.4511 | 6.13 steps/s | VRAM=25.98 Go [2026-07-17 13:25:49] step 305400/1000000 | loss=2.2361 vocab=0.9348 token=1.3240 | 6.13 steps/s | VRAM=28.16 Go [2026-07-17 13:25:58] step 305450/1000000 | loss=2.3583 vocab=1.3751 token=1.6433 | 6.12 steps/s | VRAM=27.01 Go [2026-07-17 13:26:06] step 305500/1000000 | loss=2.2737 vocab=1.7197 token=1.5036 | 6.12 steps/s | VRAM=32.42 Go [2026-07-17 13:26:14] step 305550/1000000 | loss=2.3020 vocab=0.7177 token=1.0636 | 6.13 steps/s | VRAM=28.71 Go [2026-07-17 13:26:22] step 305600/1000000 | loss=2.3257 vocab=1.1176 token=1.2453 | 6.16 steps/s | VRAM=29.20 Go [2026-07-17 13:26:30] step 305650/1000000 | loss=2.3452 vocab=1.4798 token=1.6434 | 6.14 steps/s | VRAM=27.70 Go [2026-07-17 13:26:38] step 305700/1000000 | loss=2.2761 vocab=0.9441 token=1.6866 | 6.12 steps/s | VRAM=29.10 Go [2026-07-17 13:26:46] step 305750/1000000 | loss=2.2453 vocab=0.9893 token=0.9755 | 6.15 steps/s | VRAM=28.45 Go [2026-07-17 13:26:55] step 305800/1000000 | loss=2.2634 vocab=0.9051 token=1.5292 | 6.09 steps/s | VRAM=26.57 Go [2026-07-17 13:27:03] step 305850/1000000 | loss=2.2795 vocab=0.8976 token=1.3149 | 6.13 steps/s | VRAM=29.07 Go [2026-07-17 13:27:11] step 305900/1000000 | loss=2.3915 vocab=0.9470 token=1.1149 | 6.12 steps/s | VRAM=29.20 Go [2026-07-17 13:27:19] step 305950/1000000 | loss=2.3278 vocab=0.8114 token=1.0702 | 6.12 steps/s | VRAM=32.92 Go [2026-07-17 13:27:27] step 306000/1000000 | loss=2.3525 vocab=1.1661 token=1.6432 | 6.16 steps/s | VRAM=24.96 Go [2026-07-17 13:27:35] step 306050/1000000 | loss=2.2696 vocab=0.8167 token=1.1073 | 6.17 steps/s | VRAM=31.42 Go [2026-07-17 13:27:43] step 306100/1000000 | loss=2.2189 vocab=1.2204 token=1.3477 | 6.15 steps/s | VRAM=29.41 Go [2026-07-17 13:27:52] step 306150/1000000 | loss=2.2440 vocab=0.4718 token=0.9043 | 6.13 steps/s | VRAM=29.40 Go [2026-07-17 13:28:00] step 306200/1000000 | loss=2.3970 vocab=1.1690 token=1.4022 | 6.09 steps/s | VRAM=26.97 Go [2026-07-17 13:28:08] step 306250/1000000 | loss=2.2268 vocab=0.6726 token=1.1349 | 6.15 steps/s | VRAM=25.34 Go [2026-07-17 13:28:16] step 306300/1000000 | loss=2.2698 vocab=0.6007 token=1.1389 | 6.14 steps/s | VRAM=27.81 Go [2026-07-17 13:28:24] step 306350/1000000 | loss=2.2830 vocab=0.9412 token=1.6336 | 6.16 steps/s | VRAM=28.90 Go [2026-07-17 13:28:32] step 306400/1000000 | loss=2.3328 vocab=0.8427 token=1.3896 | 6.14 steps/s | VRAM=29.22 Go [2026-07-17 13:28:41] step 306450/1000000 | loss=2.3682 vocab=1.0574 token=1.0761 | 6.15 steps/s | VRAM=31.34 Go [2026-07-17 13:28:49] step 306500/1000000 | loss=2.3535 vocab=1.1413 token=1.4359 | 6.12 steps/s | VRAM=32.50 Go [2026-07-17 13:28:57] step 306550/1000000 | loss=2.3052 vocab=0.8463 token=1.5953 | 6.12 steps/s | VRAM=32.06 Go [2026-07-17 13:29:05] step 306600/1000000 | loss=2.4157 vocab=0.6914 token=1.3533 | 6.15 steps/s | VRAM=25.51 Go [2026-07-17 13:29:13] step 306650/1000000 | loss=2.3525 vocab=0.8220 token=1.7089 | 6.11 steps/s | VRAM=25.29 Go [2026-07-17 13:29:21] step 306700/1000000 | loss=2.2332 vocab=0.5497 token=1.2077 | 6.13 steps/s | VRAM=28.00 Go [2026-07-17 13:29:29] step 306750/1000000 | loss=2.3292 vocab=0.8012 token=1.1646 | 6.13 steps/s | VRAM=26.12 Go [2026-07-17 13:29:38] step 306800/1000000 | loss=2.3439 vocab=1.0886 token=1.3178 | 6.15 steps/s | VRAM=24.92 Go [2026-07-17 13:29:46] step 306850/1000000 | loss=2.4247 vocab=2.6598 token=1.4667 | 6.12 steps/s | VRAM=26.77 Go [2026-07-17 13:29:54] step 306900/1000000 | loss=2.3594 vocab=0.8876 token=1.1729 | 6.16 steps/s | VRAM=28.45 Go [2026-07-17 13:30:02] step 306950/1000000 | loss=2.2710 vocab=0.5811 token=1.2442 | 6.16 steps/s | VRAM=30.94 Go [2026-07-17 13:30:10] step 307000/1000000 | loss=2.3196 vocab=0.8962 token=1.5092 | 6.14 steps/s | VRAM=29.92 Go [2026-07-17 13:30:18] step 307050/1000000 | loss=2.2462 vocab=1.4176 token=1.7068 | 6.14 steps/s | VRAM=32.24 Go [2026-07-17 13:30:26] step 307100/1000000 | loss=2.2403 vocab=0.7137 token=1.3093 | 6.12 steps/s | VRAM=31.22 Go [2026-07-17 13:30:35] step 307150/1000000 | loss=2.3529 vocab=1.5567 token=1.7056 | 6.13 steps/s | VRAM=27.99 Go [2026-07-17 13:30:43] step 307200/1000000 | loss=2.1839 vocab=1.3139 token=1.5922 | 6.13 steps/s | VRAM=32.00 Go [2026-07-17 13:30:51] step 307250/1000000 | loss=2.4174 vocab=0.6911 token=1.5805 | 6.13 steps/s | VRAM=24.70 Go [2026-07-17 13:30:59] step 307300/1000000 | loss=2.2412 vocab=1.0028 token=1.3330 | 6.17 steps/s | VRAM=31.29 Go [2026-07-17 13:31:07] step 307350/1000000 | loss=2.2627 vocab=0.7576 token=1.1333 | 6.10 steps/s | VRAM=32.87 Go [2026-07-17 13:31:15] step 307400/1000000 | loss=2.2526 vocab=0.4773 token=2.0986 | 6.15 steps/s | VRAM=32.22 Go [2026-07-17 13:31:24] step 307450/1000000 | loss=2.2273 vocab=0.8211 token=1.2293 | 6.12 steps/s | VRAM=32.71 Go [2026-07-17 13:31:32] step 307500/1000000 | loss=2.2414 vocab=1.0048 token=1.2866 | 6.10 steps/s | VRAM=28.06 Go [2026-07-17 13:31:40] step 307550/1000000 | loss=2.2602 vocab=0.5385 token=1.0259 | 6.14 steps/s | VRAM=28.61 Go [2026-07-17 13:31:48] step 307600/1000000 | loss=2.3037 vocab=1.9275 token=1.1397 | 6.12 steps/s | VRAM=28.09 Go [2026-07-17 13:31:56] step 307650/1000000 | loss=2.3060 vocab=0.7333 token=1.3822 | 6.14 steps/s | VRAM=25.52 Go [2026-07-17 13:32:04] step 307700/1000000 | loss=2.2972 vocab=1.0704 token=1.1534 | 6.11 steps/s | VRAM=32.42 Go [2026-07-17 13:32:13] step 307750/1000000 | loss=2.2536 vocab=0.3573 token=1.2465 | 6.14 steps/s | VRAM=32.26 Go [2026-07-17 13:32:21] step 307800/1000000 | loss=2.3031 vocab=0.9378 token=1.6652 | 6.10 steps/s | VRAM=26.53 Go [2026-07-17 13:32:29] step 307850/1000000 | loss=2.3063 vocab=0.9558 token=1.5841 | 6.15 steps/s | VRAM=29.76 Go [2026-07-17 13:32:37] step 307900/1000000 | loss=2.2539 vocab=1.5856 token=1.4193 | 6.12 steps/s | VRAM=28.84 Go [2026-07-17 13:32:45] step 307950/1000000 | loss=2.4113 vocab=0.8198 token=1.2571 | 6.11 steps/s | VRAM=26.45 Go [2026-07-17 13:32:53] step 308000/1000000 | loss=2.3098 vocab=1.1487 token=1.9075 | 6.12 steps/s | VRAM=24.84 Go [2026-07-17 13:33:02] step 308050/1000000 | loss=2.3066 vocab=1.0700 token=1.4951 | 6.12 steps/s | VRAM=26.48 Go [2026-07-17 13:33:10] step 308100/1000000 | loss=2.2788 vocab=0.8947 token=0.9544 | 6.14 steps/s | VRAM=32.73 Go [2026-07-17 13:33:18] step 308150/1000000 | loss=2.2172 vocab=0.6600 token=1.2929 | 6.09 steps/s | VRAM=29.98 Go [2026-07-17 13:33:26] step 308200/1000000 | loss=2.3176 vocab=0.8995 token=1.5662 | 6.12 steps/s | VRAM=26.59 Go [2026-07-17 13:33:34] step 308250/1000000 | loss=2.2262 vocab=0.6181 token=0.8044 | 6.13 steps/s | VRAM=28.57 Go [2026-07-17 13:33:42] step 308300/1000000 | loss=2.3277 vocab=1.2375 token=1.6441 | 6.11 steps/s | VRAM=32.41 Go [2026-07-17 13:33:51] step 308350/1000000 | loss=2.3027 vocab=1.1491 token=1.5333 | 6.11 steps/s | VRAM=28.15 Go [2026-07-17 13:33:59] step 308400/1000000 | loss=2.3419 vocab=0.9607 token=1.6620 | 6.15 steps/s | VRAM=30.88 Go [2026-07-17 13:34:07] step 308450/1000000 | loss=2.3377 vocab=0.4386 token=1.0754 | 6.12 steps/s | VRAM=24.76 Go [2026-07-17 13:34:15] step 308500/1000000 | loss=2.4326 vocab=1.3078 token=1.5343 | 6.13 steps/s | VRAM=26.23 Go [2026-07-17 13:34:23] step 308550/1000000 | loss=2.3366 vocab=0.6334 token=1.3747 | 6.16 steps/s | VRAM=31.62 Go [2026-07-17 13:34:31] step 308600/1000000 | loss=2.3053 vocab=0.5431 token=1.0865 | 6.15 steps/s | VRAM=28.44 Go [2026-07-17 13:34:39] step 308650/1000000 | loss=2.3256 vocab=1.1920 token=1.6423 | 6.12 steps/s | VRAM=26.29 Go [2026-07-17 13:34:48] step 308700/1000000 | loss=2.1661 vocab=0.9919 token=1.3492 | 6.12 steps/s | VRAM=27.96 Go [2026-07-17 13:34:56] step 308750/1000000 | loss=2.4358 vocab=0.8077 token=1.4471 | 6.10 steps/s | VRAM=26.23 Go [2026-07-17 13:35:04] step 308800/1000000 | loss=2.2507 vocab=1.1187 token=1.2712 | 6.13 steps/s | VRAM=31.78 Go [2026-07-17 13:35:12] step 308850/1000000 | loss=2.2526 vocab=0.5983 token=1.3063 | 6.12 steps/s | VRAM=32.56 Go [2026-07-17 13:35:20] step 308900/1000000 | loss=2.3452 vocab=0.8883 token=1.2140 | 6.14 steps/s | VRAM=28.36 Go [2026-07-17 13:35:29] step 308950/1000000 | loss=2.2310 vocab=1.2168 token=1.3606 | 6.10 steps/s | VRAM=27.15 Go [2026-07-17 13:35:37] step 309000/1000000 | loss=2.2560 vocab=0.4933 token=1.1945 | 6.12 steps/s | VRAM=29.05 Go [2026-07-17 13:35:45] step 309050/1000000 | loss=2.3266 vocab=1.7402 token=1.4636 | 6.15 steps/s | VRAM=29.28 Go [2026-07-17 13:35:53] step 309100/1000000 | loss=2.2376 vocab=0.8008 token=1.0865 | 6.13 steps/s | VRAM=32.40 Go [2026-07-17 13:36:01] step 309150/1000000 | loss=2.2587 vocab=2.0998 token=1.0695 | 6.15 steps/s | VRAM=28.12 Go [2026-07-17 13:36:09] step 309200/1000000 | loss=2.4292 vocab=1.4974 token=1.7171 | 6.14 steps/s | VRAM=31.94 Go [2026-07-17 13:36:17] step 309250/1000000 | loss=2.4050 vocab=0.5336 token=1.1819 | 6.08 steps/s | VRAM=29.30 Go [2026-07-17 13:36:26] step 309300/1000000 | loss=2.2079 vocab=0.5955 token=1.3385 | 6.12 steps/s | VRAM=29.11 Go [2026-07-17 13:36:34] step 309350/1000000 | loss=2.3150 vocab=1.0707 token=1.5742 | 6.10 steps/s | VRAM=27.47 Go [2026-07-17 13:36:42] step 309400/1000000 | loss=2.2589 vocab=0.6769 token=1.4242 | 6.16 steps/s | VRAM=31.56 Go [2026-07-17 13:36:50] step 309450/1000000 | loss=2.3086 vocab=1.1893 token=1.5517 | 6.17 steps/s | VRAM=31.91 Go [2026-07-17 13:36:58] step 309500/1000000 | loss=2.4071 vocab=0.7971 token=1.4156 | 6.11 steps/s | VRAM=25.91 Go [2026-07-17 13:37:06] step 309550/1000000 | loss=2.2509 vocab=1.2121 token=1.2574 | 6.13 steps/s | VRAM=32.17 Go [2026-07-17 13:37:15] step 309600/1000000 | loss=2.4311 vocab=0.9304 token=1.0014 | 6.13 steps/s | VRAM=28.61 Go [2026-07-17 13:37:23] step 309650/1000000 | loss=2.2778 vocab=0.7642 token=1.5910 | 6.10 steps/s | VRAM=26.81 Go [2026-07-17 13:37:31] step 309700/1000000 | loss=2.3121 vocab=1.5385 token=1.9530 | 6.13 steps/s | VRAM=28.31 Go [2026-07-17 13:37:39] step 309750/1000000 | loss=2.3215 vocab=0.9378 token=1.1855 | 6.14 steps/s | VRAM=31.69 Go [2026-07-17 13:37:47] step 309800/1000000 | loss=2.3562 vocab=1.5634 token=1.9170 | 6.15 steps/s | VRAM=32.64 Go [2026-07-17 13:37:55] step 309850/1000000 | loss=2.3303 vocab=1.1359 token=1.3863 | 6.15 steps/s | VRAM=28.80 Go [2026-07-17 13:38:03] step 309900/1000000 | loss=2.2232 vocab=1.3122 token=1.6844 | 6.13 steps/s | VRAM=29.95 Go [2026-07-17 13:38:12] step 309950/1000000 | loss=2.3255 vocab=0.6301 token=1.3253 | 6.11 steps/s | VRAM=23.52 Go [2026-07-17 13:38:20] step 310000/1000000 | loss=2.4559 vocab=0.7640 token=1.7312 | 6.13 steps/s | VRAM=27.74 Go [2026-07-17 13:38:28] step 310050/1000000 | loss=2.2329 vocab=0.8443 token=1.2409 | 6.15 steps/s | VRAM=28.34 Go [2026-07-17 13:38:36] step 310100/1000000 | loss=2.2427 vocab=0.7236 token=1.4615 | 6.13 steps/s | VRAM=32.60 Go [2026-07-17 13:38:44] step 310150/1000000 | loss=2.3273 vocab=1.5406 token=1.3689 | 6.15 steps/s | VRAM=32.39 Go [2026-07-17 13:38:52] step 310200/1000000 | loss=2.3148 vocab=0.8833 token=1.5188 | 6.14 steps/s | VRAM=32.51 Go [2026-07-17 13:39:00] step 310250/1000000 | loss=2.3045 vocab=0.9865 token=1.6156 | 6.16 steps/s | VRAM=28.98 Go [2026-07-17 13:39:09] step 310300/1000000 | loss=2.2672 vocab=0.8374 token=1.4842 | 6.12 steps/s | VRAM=25.39 Go [2026-07-17 13:39:17] step 310350/1000000 | loss=2.3472 vocab=1.2328 token=1.3633 | 6.15 steps/s | VRAM=29.07 Go [2026-07-17 13:39:25] step 310400/1000000 | loss=2.3155 vocab=1.5761 token=2.0630 | 6.11 steps/s | VRAM=26.17 Go [2026-07-17 13:39:33] step 310450/1000000 | loss=2.2772 vocab=0.8055 token=1.2789 | 6.15 steps/s | VRAM=28.79 Go [2026-07-17 13:39:41] step 310500/1000000 | loss=2.3340 vocab=1.0122 token=1.4414 | 6.18 steps/s | VRAM=28.42 Go [2026-07-17 13:39:49] step 310550/1000000 | loss=2.3935 vocab=1.5851 token=1.9092 | 6.18 steps/s | VRAM=32.09 Go [2026-07-17 13:39:57] step 310600/1000000 | loss=2.2789 vocab=0.8640 token=1.3752 | 6.17 steps/s | VRAM=28.70 Go [2026-07-17 13:40:05] step 310650/1000000 | loss=2.3137 vocab=1.3400 token=1.6393 | 6.15 steps/s | VRAM=32.48 Go [2026-07-17 13:40:14] step 310700/1000000 | loss=2.3280 vocab=1.2065 token=1.3256 | 6.10 steps/s | VRAM=32.53 Go [2026-07-17 13:40:22] step 310750/1000000 | loss=2.3394 vocab=0.6907 token=1.1888 | 6.15 steps/s | VRAM=28.10 Go [2026-07-17 13:40:30] step 310800/1000000 | loss=2.3041 vocab=1.0860 token=1.6141 | 6.16 steps/s | VRAM=28.10 Go [2026-07-17 13:40:38] step 310850/1000000 | loss=2.3041 vocab=0.8982 token=1.2381 | 6.13 steps/s | VRAM=26.22 Go [2026-07-17 13:40:46] step 310900/1000000 | loss=2.2803 vocab=0.8576 token=1.3775 | 6.17 steps/s | VRAM=24.48 Go [2026-07-17 13:40:54] step 310950/1000000 | loss=2.2915 vocab=0.6066 token=0.9992 | 6.14 steps/s | VRAM=27.82 Go [2026-07-17 13:41:03] step 311000/1000000 | loss=2.2809 vocab=0.9888 token=1.5246 | 6.12 steps/s | VRAM=26.11 Go [2026-07-17 13:41:11] step 311050/1000000 | loss=2.2498 vocab=1.4487 token=1.6539 | 6.15 steps/s | VRAM=28.44 Go [2026-07-17 13:41:19] step 311100/1000000 | loss=2.3323 vocab=1.1774 token=1.3969 | 6.13 steps/s | VRAM=27.44 Go [2026-07-17 13:41:27] step 311150/1000000 | loss=2.1843 vocab=0.7707 token=1.2922 | 6.18 steps/s | VRAM=32.04 Go [2026-07-17 13:41:35] step 311200/1000000 | loss=2.3054 vocab=1.1468 token=1.8284 | 6.18 steps/s | VRAM=27.89 Go [2026-07-17 13:41:43] step 311250/1000000 | loss=2.3170 vocab=1.3739 token=1.7408 | 6.12 steps/s | VRAM=25.92 Go [2026-07-17 13:41:51] step 311300/1000000 | loss=2.3204 vocab=1.1398 token=1.7224 | 6.13 steps/s | VRAM=26.64 Go [2026-07-17 13:41:59] step 311350/1000000 | loss=2.2766 vocab=0.9707 token=1.5762 | 6.13 steps/s | VRAM=28.81 Go [2026-07-17 13:42:08] step 311400/1000000 | loss=2.2484 vocab=0.6921 token=1.4055 | 6.15 steps/s | VRAM=28.81 Go [2026-07-17 13:42:16] step 311450/1000000 | loss=2.2465 vocab=0.7157 token=1.2462 | 6.15 steps/s | VRAM=28.76 Go [2026-07-17 13:42:24] step 311500/1000000 | loss=2.3278 vocab=0.6238 token=1.2234 | 6.10 steps/s | VRAM=31.66 Go [2026-07-17 13:42:32] step 311550/1000000 | loss=2.3436 vocab=0.4509 token=1.0353 | 6.13 steps/s | VRAM=28.14 Go [2026-07-17 13:42:40] step 311600/1000000 | loss=2.2766 vocab=0.9004 token=1.5225 | 6.15 steps/s | VRAM=25.56 Go [2026-07-17 13:42:48] step 311650/1000000 | loss=2.3015 vocab=1.0791 token=1.5889 | 6.16 steps/s | VRAM=25.13 Go [2026-07-17 13:42:56] step 311700/1000000 | loss=2.2021 vocab=0.7011 token=1.7145 | 6.14 steps/s | VRAM=28.51 Go [2026-07-17 13:43:05] step 311750/1000000 | loss=2.3845 vocab=1.3339 token=1.6591 | 6.14 steps/s | VRAM=26.11 Go [2026-07-17 13:43:13] step 311800/1000000 | loss=2.2189 vocab=0.9907 token=1.6887 | 6.15 steps/s | VRAM=31.31 Go [2026-07-17 13:43:21] step 311850/1000000 | loss=2.3495 vocab=0.9692 token=1.2731 | 6.15 steps/s | VRAM=32.81 Go [2026-07-17 13:43:29] step 311900/1000000 | loss=2.3321 vocab=1.3195 token=1.5358 | 6.14 steps/s | VRAM=28.16 Go [2026-07-17 13:43:37] step 311950/1000000 | loss=2.2840 vocab=0.6331 token=1.5534 | 6.09 steps/s | VRAM=26.34 Go [2026-07-17 13:43:45] step 312000/1000000 | loss=2.3140 vocab=1.1159 token=1.7862 | 6.15 steps/s | VRAM=28.48 Go [2026-07-17 13:43:54] step 312050/1000000 | loss=2.3243 vocab=0.6741 token=1.2160 | 6.15 steps/s | VRAM=28.71 Go [2026-07-17 13:44:02] step 312100/1000000 | loss=2.2201 vocab=0.7023 token=1.2536 | 6.12 steps/s | VRAM=29.52 Go [2026-07-17 13:44:10] step 312150/1000000 | loss=2.2762 vocab=0.8278 token=1.3978 | 6.13 steps/s | VRAM=32.90 Go [2026-07-17 13:44:18] step 312200/1000000 | loss=2.2907 vocab=0.7089 token=1.4769 | 6.12 steps/s | VRAM=25.02 Go [2026-07-17 13:44:26] step 312250/1000000 | loss=2.2546 vocab=0.9456 token=1.4295 | 6.13 steps/s | VRAM=32.39 Go [2026-07-17 13:44:34] step 312300/1000000 | loss=2.2122 vocab=0.7123 token=1.2099 | 6.13 steps/s | VRAM=32.57 Go [2026-07-17 13:44:42] step 312350/1000000 | loss=2.2752 vocab=0.6343 token=1.2087 | 6.14 steps/s | VRAM=28.76 Go [2026-07-17 13:44:51] step 312400/1000000 | loss=2.2903 vocab=0.3999 token=0.9879 | 6.13 steps/s | VRAM=26.16 Go [2026-07-17 13:44:59] step 312450/1000000 | loss=2.4613 vocab=0.7047 token=1.1677 | 6.11 steps/s | VRAM=28.61 Go [2026-07-17 13:45:07] step 312500/1000000 | loss=2.3216 vocab=1.0954 token=1.7185 | 6.09 steps/s | VRAM=26.35 Go [2026-07-17 13:45:15] step 312550/1000000 | loss=2.3761 vocab=1.0600 token=1.5837 | 6.15 steps/s | VRAM=28.47 Go [2026-07-17 13:45:23] step 312600/1000000 | loss=2.3586 vocab=1.3528 token=1.7409 | 6.13 steps/s | VRAM=31.80 Go [2026-07-17 13:45:31] step 312650/1000000 | loss=2.3904 vocab=1.0039 token=1.5436 | 6.15 steps/s | VRAM=24.29 Go [2026-07-17 13:45:40] step 312700/1000000 | loss=2.4110 vocab=1.3190 token=1.2524 | 6.14 steps/s | VRAM=30.17 Go [2026-07-17 13:45:48] step 312750/1000000 | loss=2.2822 vocab=1.1552 token=1.3588 | 6.09 steps/s | VRAM=26.02 Go [2026-07-17 13:45:56] step 312800/1000000 | loss=2.3020 vocab=0.7221 token=1.0550 | 6.15 steps/s | VRAM=32.39 Go [2026-07-17 13:46:04] step 312850/1000000 | loss=2.3128 vocab=0.8068 token=1.6565 | 6.15 steps/s | VRAM=28.77 Go [2026-07-17 13:46:12] step 312900/1000000 | loss=2.3509 vocab=1.3608 token=1.6155 | 6.13 steps/s | VRAM=28.26 Go [2026-07-17 13:46:20] step 312950/1000000 | loss=2.2794 vocab=1.1918 token=1.0668 | 6.14 steps/s | VRAM=26.32 Go [2026-07-17 13:46:29] step 313000/1000000 | loss=2.2624 vocab=0.6461 token=1.4878 | 6.12 steps/s | VRAM=28.75 Go [2026-07-17 13:46:37] step 313050/1000000 | loss=2.3368 vocab=0.8595 token=1.2998 | 6.13 steps/s | VRAM=28.66 Go [2026-07-17 13:46:45] step 313100/1000000 | loss=2.2546 vocab=1.2700 token=1.3708 | 6.14 steps/s | VRAM=25.60 Go [2026-07-17 13:46:53] step 313150/1000000 | loss=2.3828 vocab=0.7004 token=1.1807 | 6.14 steps/s | VRAM=32.41 Go [2026-07-17 13:47:01] step 313200/1000000 | loss=2.3223 vocab=0.6061 token=1.4545 | 6.13 steps/s | VRAM=28.55 Go [2026-07-17 13:47:09] step 313250/1000000 | loss=2.2899 vocab=0.8707 token=1.0886 | 6.14 steps/s | VRAM=32.18 Go [2026-07-17 13:47:17] step 313300/1000000 | loss=2.3732 vocab=0.9705 token=1.0720 | 6.12 steps/s | VRAM=23.31 Go [2026-07-17 13:47:26] step 313350/1000000 | loss=2.2879 vocab=0.6628 token=1.1777 | 6.14 steps/s | VRAM=27.98 Go [2026-07-17 13:47:34] step 313400/1000000 | loss=2.3506 vocab=1.6058 token=2.4199 | 6.14 steps/s | VRAM=28.26 Go [2026-07-17 13:47:42] step 313450/1000000 | loss=2.3441 vocab=1.3835 token=0.9763 | 6.16 steps/s | VRAM=31.14 Go [2026-07-17 13:47:50] step 313500/1000000 | loss=2.3069 vocab=0.6944 token=1.5158 | 6.12 steps/s | VRAM=32.82 Go [2026-07-17 13:47:58] step 313550/1000000 | loss=2.3616 vocab=0.5656 token=1.3046 | 6.13 steps/s | VRAM=28.49 Go [2026-07-17 13:48:06] step 313600/1000000 | loss=2.4405 vocab=1.5484 token=1.8788 | 6.11 steps/s | VRAM=32.70 Go [2026-07-17 13:48:15] step 313650/1000000 | loss=2.3157 vocab=0.7144 token=0.8579 | 6.11 steps/s | VRAM=25.45 Go [2026-07-17 13:48:23] step 313700/1000000 | loss=2.2618 vocab=0.9667 token=1.5550 | 6.15 steps/s | VRAM=30.05 Go [2026-07-17 13:48:31] step 313750/1000000 | loss=2.3363 vocab=1.6051 token=1.8098 | 6.12 steps/s | VRAM=28.06 Go [2026-07-17 13:48:39] step 313800/1000000 | loss=2.3032 vocab=1.1125 token=1.1873 | 6.16 steps/s | VRAM=32.09 Go [2026-07-17 13:48:47] step 313850/1000000 | loss=2.1816 vocab=1.2171 token=1.4990 | 6.11 steps/s | VRAM=29.76 Go [2026-07-17 13:48:55] step 313900/1000000 | loss=2.3408 vocab=1.3595 token=1.4271 | 6.15 steps/s | VRAM=29.18 Go [2026-07-17 13:49:03] step 313950/1000000 | loss=2.3121 vocab=0.8834 token=1.5390 | 6.13 steps/s | VRAM=28.97 Go [2026-07-17 13:49:12] step 314000/1000000 | loss=2.3462 vocab=0.4650 token=1.1959 | 6.14 steps/s | VRAM=28.35 Go [2026-07-17 13:49:20] step 314050/1000000 | loss=2.3321 vocab=0.5573 token=1.3542 | 6.13 steps/s | VRAM=29.43 Go [2026-07-17 13:49:28] step 314100/1000000 | loss=2.3378 vocab=1.0821 token=1.5495 | 6.13 steps/s | VRAM=25.96 Go [2026-07-17 13:49:36] step 314150/1000000 | loss=2.3544 vocab=0.8098 token=1.5035 | 6.14 steps/s | VRAM=26.50 Go [2026-07-17 13:49:44] step 314200/1000000 | loss=2.2525 vocab=1.0275 token=2.0173 | 6.12 steps/s | VRAM=26.14 Go [2026-07-17 13:49:52] step 314250/1000000 | loss=2.2872 vocab=1.0321 token=1.4530 | 6.18 steps/s | VRAM=30.80 Go [2026-07-17 13:50:00] step 314300/1000000 | loss=2.1930 vocab=1.0393 token=1.7044 | 6.13 steps/s | VRAM=29.03 Go [2026-07-17 13:50:09] step 314350/1000000 | loss=2.3741 vocab=1.0172 token=1.4122 | 6.13 steps/s | VRAM=28.21 Go [2026-07-17 13:50:17] step 314400/1000000 | loss=2.3045 vocab=0.6280 token=1.1237 | 6.11 steps/s | VRAM=26.22 Go [2026-07-17 13:50:25] step 314450/1000000 | loss=2.2929 vocab=1.7550 token=1.5045 | 6.10 steps/s | VRAM=26.37 Go [2026-07-17 13:50:33] step 314500/1000000 | loss=2.2508 vocab=1.0939 token=1.2412 | 6.12 steps/s | VRAM=29.31 Go [2026-07-17 13:50:41] step 314550/1000000 | loss=2.4247 vocab=0.6870 token=1.1321 | 6.13 steps/s | VRAM=27.82 Go [2026-07-17 13:50:49] step 314600/1000000 | loss=2.2556 vocab=1.3436 token=1.9102 | 6.15 steps/s | VRAM=25.36 Go [2026-07-17 13:50:58] step 314650/1000000 | loss=2.3180 vocab=0.5988 token=1.2684 | 6.12 steps/s | VRAM=29.10 Go [2026-07-17 13:51:06] step 314700/1000000 | loss=2.2719 vocab=1.2602 token=1.3691 | 6.13 steps/s | VRAM=28.93 Go [2026-07-17 13:51:14] step 314750/1000000 | loss=2.3515 vocab=0.7829 token=1.3460 | 6.15 steps/s | VRAM=31.17 Go [2026-07-17 13:51:22] step 314800/1000000 | loss=2.3143 vocab=0.5890 token=1.0354 | 6.11 steps/s | VRAM=26.52 Go [2026-07-17 13:51:30] step 314850/1000000 | loss=2.3983 vocab=1.1506 token=1.4342 | 6.16 steps/s | VRAM=28.15 Go [2026-07-17 13:51:38] step 314900/1000000 | loss=2.2577 vocab=0.6793 token=1.3111 | 6.12 steps/s | VRAM=32.17 Go [2026-07-17 13:51:47] step 314950/1000000 | loss=2.2598 vocab=0.8980 token=1.3033 | 6.14 steps/s | VRAM=24.04 Go [2026-07-17 13:51:55] step 315000/1000000 | loss=2.3098 vocab=1.0498 token=1.1396 | 6.13 steps/s | VRAM=26.51 Go [2026-07-17 13:52:03] step 315050/1000000 | loss=2.3066 vocab=0.9385 token=1.5960 | 6.15 steps/s | VRAM=27.70 Go [2026-07-17 13:52:11] step 315100/1000000 | loss=2.2660 vocab=0.7250 token=1.2540 | 6.12 steps/s | VRAM=28.62 Go [2026-07-17 13:52:19] step 315150/1000000 | loss=2.3380 vocab=1.1626 token=1.8924 | 6.13 steps/s | VRAM=29.44 Go [2026-07-17 13:52:27] step 315200/1000000 | loss=2.3337 vocab=0.7394 token=1.3664 | 6.16 steps/s | VRAM=27.45 Go [2026-07-17 13:52:35] step 315250/1000000 | loss=2.3102 vocab=1.1236 token=1.7093 | 6.11 steps/s | VRAM=28.13 Go [2026-07-17 13:52:44] step 315300/1000000 | loss=2.3668 vocab=0.9541 token=1.4256 | 6.11 steps/s | VRAM=29.56 Go [2026-07-17 13:52:52] step 315350/1000000 | loss=2.3308 vocab=0.5848 token=1.1492 | 6.12 steps/s | VRAM=29.51 Go [2026-07-17 13:53:00] step 315400/1000000 | loss=2.2989 vocab=0.9164 token=1.6861 | 6.13 steps/s | VRAM=28.32 Go [2026-07-17 13:53:08] step 315450/1000000 | loss=2.2711 vocab=0.6768 token=1.3824 | 6.13 steps/s | VRAM=25.35 Go [2026-07-17 13:53:16] step 315500/1000000 | loss=2.2836 vocab=0.4703 token=1.4650 | 6.12 steps/s | VRAM=28.50 Go [2026-07-17 13:53:24] step 315550/1000000 | loss=2.3424 vocab=0.6496 token=1.0216 | 6.12 steps/s | VRAM=26.73 Go [2026-07-17 13:53:33] step 315600/1000000 | loss=2.3286 vocab=0.8138 token=1.2972 | 6.13 steps/s | VRAM=28.06 Go [2026-07-17 13:53:41] step 315650/1000000 | loss=2.3626 vocab=0.6277 token=1.3034 | 6.15 steps/s | VRAM=31.32 Go [2026-07-17 13:53:49] step 315700/1000000 | loss=2.2908 vocab=1.2343 token=1.2411 | 6.12 steps/s | VRAM=25.00 Go [2026-07-17 13:53:57] step 315750/1000000 | loss=2.2343 vocab=0.4696 token=0.8924 | 6.16 steps/s | VRAM=31.33 Go [2026-07-17 13:54:05] step 315800/1000000 | loss=2.2564 vocab=0.6711 token=0.9147 | 6.12 steps/s | VRAM=28.87 Go [2026-07-17 13:54:13] step 315850/1000000 | loss=2.2594 vocab=0.6178 token=1.2329 | 6.10 steps/s | VRAM=28.37 Go [2026-07-17 13:54:22] step 315900/1000000 | loss=2.3808 vocab=0.6683 token=1.8798 | 6.09 steps/s | VRAM=29.34 Go [2026-07-17 13:54:30] step 315950/1000000 | loss=2.3368 vocab=0.8337 token=1.4564 | 6.10 steps/s | VRAM=29.49 Go [2026-07-17 13:54:38] step 316000/1000000 | loss=2.2297 vocab=0.6779 token=1.5791 | 6.13 steps/s | VRAM=25.23 Go [2026-07-17 13:54:46] step 316050/1000000 | loss=2.2310 vocab=1.0641 token=1.7392 | 6.15 steps/s | VRAM=25.07 Go [2026-07-17 13:54:54] step 316100/1000000 | loss=2.3803 vocab=1.2021 token=1.6344 | 6.12 steps/s | VRAM=33.00 Go [2026-07-17 13:55:02] step 316150/1000000 | loss=2.2666 vocab=1.0081 token=1.2240 | 6.12 steps/s | VRAM=24.62 Go [2026-07-17 13:55:11] step 316200/1000000 | loss=2.3329 vocab=0.9498 token=1.1900 | 6.13 steps/s | VRAM=29.20 Go [2026-07-17 13:55:19] step 316250/1000000 | loss=2.3912 vocab=0.5922 token=1.1368 | 6.11 steps/s | VRAM=26.48 Go [2026-07-17 13:55:27] step 316300/1000000 | loss=2.3164 vocab=0.5518 token=1.0502 | 6.14 steps/s | VRAM=29.53 Go [2026-07-17 13:55:35] step 316350/1000000 | loss=2.2641 vocab=1.2711 token=1.7359 | 6.12 steps/s | VRAM=30.23 Go [2026-07-17 13:55:43] step 316400/1000000 | loss=2.2709 vocab=0.9993 token=1.5597 | 6.15 steps/s | VRAM=31.66 Go [2026-07-17 13:55:51] step 316450/1000000 | loss=2.3338 vocab=0.6122 token=1.3983 | 6.11 steps/s | VRAM=29.16 Go [2026-07-17 13:56:00] step 316500/1000000 | loss=2.2477 vocab=0.7263 token=1.3748 | 6.13 steps/s | VRAM=32.73 Go [2026-07-17 13:56:08] step 316550/1000000 | loss=2.1968 vocab=0.7268 token=1.0595 | 6.13 steps/s | VRAM=28.66 Go [2026-07-17 13:56:16] step 316600/1000000 | loss=2.3117 vocab=1.2301 token=1.6089 | 6.14 steps/s | VRAM=32.03 Go [2026-07-17 13:56:24] step 316650/1000000 | loss=2.4106 vocab=1.0791 token=1.2886 | 6.11 steps/s | VRAM=26.25 Go [2026-07-17 13:56:32] step 316700/1000000 | loss=2.3928 vocab=0.9473 token=1.3839 | 6.11 steps/s | VRAM=29.15 Go [2026-07-17 13:56:40] step 316750/1000000 | loss=2.3237 vocab=0.6506 token=1.0023 | 6.12 steps/s | VRAM=32.94 Go [2026-07-17 13:56:49] step 316800/1000000 | loss=2.3229 vocab=1.2905 token=1.3140 | 6.11 steps/s | VRAM=31.92 Go [2026-07-17 13:56:57] step 316850/1000000 | loss=2.2280 vocab=0.9645 token=1.4250 | 6.11 steps/s | VRAM=26.20 Go [2026-07-17 13:57:05] step 316900/1000000 | loss=2.3633 vocab=0.7741 token=1.7655 | 6.12 steps/s | VRAM=26.50 Go [2026-07-17 13:57:13] step 316950/1000000 | loss=2.3199 vocab=0.9950 token=1.5354 | 6.15 steps/s | VRAM=32.15 Go [2026-07-17 13:57:21] step 317000/1000000 | loss=2.2514 vocab=1.0309 token=1.3206 | 6.15 steps/s | VRAM=31.43 Go [2026-07-17 13:57:29] step 317050/1000000 | loss=2.3155 vocab=0.8446 token=1.4110 | 6.14 steps/s | VRAM=31.24 Go [2026-07-17 13:57:37] step 317100/1000000 | loss=2.2616 vocab=0.9588 token=1.5457 | 6.14 steps/s | VRAM=28.30 Go [2026-07-17 13:57:46] step 317150/1000000 | loss=2.2860 vocab=1.1643 token=1.3293 | 6.13 steps/s | VRAM=29.18 Go [2026-07-17 13:57:54] step 317200/1000000 | loss=2.2354 vocab=1.3563 token=1.2392 | 6.12 steps/s | VRAM=31.91 Go [2026-07-17 13:58:02] step 317250/1000000 | loss=2.3203 vocab=0.9342 token=1.0421 | 6.11 steps/s | VRAM=28.58 Go [2026-07-17 13:58:10] step 317300/1000000 | loss=2.1451 vocab=0.6875 token=1.5641 | 6.11 steps/s | VRAM=30.25 Go [2026-07-17 13:58:18] step 317350/1000000 | loss=2.2438 vocab=0.6739 token=1.2150 | 6.14 steps/s | VRAM=26.62 Go [2026-07-17 13:58:26] step 317400/1000000 | loss=2.3217 vocab=1.0628 token=1.0873 | 6.12 steps/s | VRAM=28.77 Go [2026-07-17 13:58:35] step 317450/1000000 | loss=2.2685 vocab=1.1636 token=1.3873 | 6.12 steps/s | VRAM=32.38 Go [2026-07-17 13:58:43] step 317500/1000000 | loss=2.3502 vocab=0.9152 token=1.2033 | 6.13 steps/s | VRAM=26.33 Go [2026-07-17 13:58:51] step 317550/1000000 | loss=2.2814 vocab=0.5403 token=1.3286 | 6.15 steps/s | VRAM=28.48 Go [2026-07-17 13:58:59] step 317600/1000000 | loss=2.2727 vocab=1.2648 token=1.6411 | 6.14 steps/s | VRAM=28.59 Go [2026-07-17 13:59:07] step 317650/1000000 | loss=2.2698 vocab=0.4825 token=1.2181 | 6.11 steps/s | VRAM=29.51 Go [2026-07-17 13:59:15] step 317700/1000000 | loss=2.2828 vocab=1.0618 token=1.2557 | 6.13 steps/s | VRAM=31.89 Go [2026-07-17 13:59:24] step 317750/1000000 | loss=2.2056 vocab=0.9195 token=1.3016 | 6.14 steps/s | VRAM=28.63 Go [2026-07-17 13:59:32] step 317800/1000000 | loss=2.3126 vocab=0.8737 token=1.2898 | 6.14 steps/s | VRAM=27.90 Go [2026-07-17 13:59:40] step 317850/1000000 | loss=2.4102 vocab=0.8164 token=1.4995 | 6.13 steps/s | VRAM=24.98 Go [2026-07-17 13:59:48] step 317900/1000000 | loss=2.2840 vocab=0.7193 token=1.1196 | 6.13 steps/s | VRAM=28.42 Go [2026-07-17 13:59:56] step 317950/1000000 | loss=2.2240 vocab=0.8107 token=1.2797 | 6.12 steps/s | VRAM=26.30 Go [2026-07-17 14:00:04] step 318000/1000000 | loss=2.3274 vocab=0.8735 token=1.2116 | 6.14 steps/s | VRAM=28.77 Go [2026-07-17 14:00:12] step 318050/1000000 | loss=2.2672 vocab=0.4942 token=1.6350 | 6.14 steps/s | VRAM=28.40 Go [2026-07-17 14:00:21] step 318100/1000000 | loss=2.3321 vocab=1.0322 token=1.3624 | 6.14 steps/s | VRAM=26.02 Go [2026-07-17 14:00:29] step 318150/1000000 | loss=2.2375 vocab=0.7159 token=1.3435 | 6.15 steps/s | VRAM=28.83 Go [2026-07-17 14:00:37] step 318200/1000000 | loss=2.2571 vocab=0.5428 token=1.6603 | 6.16 steps/s | VRAM=32.47 Go [2026-07-17 14:00:45] step 318250/1000000 | loss=2.1724 vocab=0.7038 token=1.1804 | 6.14 steps/s | VRAM=28.67 Go [2026-07-17 14:00:53] step 318300/1000000 | loss=2.3321 vocab=0.8921 token=1.3078 | 6.11 steps/s | VRAM=23.95 Go [2026-07-17 14:01:01] step 318350/1000000 | loss=2.2102 vocab=0.7746 token=1.0944 | 6.18 steps/s | VRAM=28.63 Go [2026-07-17 14:01:09] step 318400/1000000 | loss=2.2615 vocab=1.1642 token=1.3258 | 6.18 steps/s | VRAM=32.54 Go [2026-07-17 14:01:18] step 318450/1000000 | loss=2.3740 vocab=0.6237 token=0.8279 | 6.13 steps/s | VRAM=32.84 Go [2026-07-17 14:01:26] step 318500/1000000 | loss=2.3388 vocab=1.1620 token=1.4810 | 6.14 steps/s | VRAM=29.61 Go [2026-07-17 14:01:34] step 318550/1000000 | loss=2.2624 vocab=0.6925 token=1.3993 | 6.12 steps/s | VRAM=28.30 Go [2026-07-17 14:01:42] step 318600/1000000 | loss=2.2399 vocab=1.0799 token=1.3995 | 6.11 steps/s | VRAM=27.49 Go [2026-07-17 14:01:50] step 318650/1000000 | loss=2.2773 vocab=0.8193 token=1.2049 | 6.15 steps/s | VRAM=28.59 Go [2026-07-17 14:01:58] step 318700/1000000 | loss=2.3850 vocab=1.4546 token=1.2874 | 6.14 steps/s | VRAM=30.80 Go [2026-07-17 14:02:06] step 318750/1000000 | loss=2.3664 vocab=1.0438 token=1.3534 | 6.16 steps/s | VRAM=31.39 Go [2026-07-17 14:02:15] step 318800/1000000 | loss=2.2218 vocab=0.9667 token=0.9713 | 6.12 steps/s | VRAM=28.62 Go [2026-07-17 14:02:23] step 318850/1000000 | loss=2.3202 vocab=0.4710 token=0.8302 | 6.13 steps/s | VRAM=25.08 Go [2026-07-17 14:02:31] step 318900/1000000 | loss=2.2303 vocab=0.5075 token=1.0238 | 6.12 steps/s | VRAM=29.60 Go [2026-07-17 14:02:39] step 318950/1000000 | loss=2.2721 vocab=0.9249 token=0.9181 | 6.14 steps/s | VRAM=26.11 Go [2026-07-17 14:02:47] step 319000/1000000 | loss=2.3327 vocab=1.0755 token=1.6010 | 6.13 steps/s | VRAM=29.15 Go [2026-07-17 14:02:55] step 319050/1000000 | loss=2.3690 vocab=1.2861 token=1.5708 | 6.11 steps/s | VRAM=28.77 Go [2026-07-17 14:03:04] step 319100/1000000 | loss=2.2673 vocab=1.0238 token=1.5598 | 6.15 steps/s | VRAM=29.16 Go [2026-07-17 14:03:12] step 319150/1000000 | loss=2.3141 vocab=1.1343 token=1.6597 | 6.13 steps/s | VRAM=28.40 Go [2026-07-17 14:03:20] step 319200/1000000 | loss=2.4121 vocab=1.0392 token=1.4206 | 6.11 steps/s | VRAM=27.25 Go [2026-07-17 14:03:28] step 319250/1000000 | loss=2.2944 vocab=0.8704 token=1.4696 | 6.15 steps/s | VRAM=28.62 Go [2026-07-17 14:03:36] step 319300/1000000 | loss=2.2785 vocab=0.6382 token=1.4138 | 6.13 steps/s | VRAM=31.14 Go [2026-07-17 14:03:44] step 319350/1000000 | loss=2.3545 vocab=0.6799 token=1.4438 | 6.14 steps/s | VRAM=31.66 Go [2026-07-17 14:03:52] step 319400/1000000 | loss=2.3783 vocab=0.8431 token=1.5201 | 6.12 steps/s | VRAM=28.80 Go [2026-07-17 14:04:01] step 319450/1000000 | loss=2.2942 vocab=0.5181 token=1.0673 | 6.13 steps/s | VRAM=28.02 Go [2026-07-17 14:04:09] step 319500/1000000 | loss=2.3412 vocab=1.0857 token=1.3579 | 6.13 steps/s | VRAM=27.94 Go [2026-07-17 14:04:17] step 319550/1000000 | loss=2.3955 vocab=1.2325 token=1.3218 | 6.12 steps/s | VRAM=29.15 Go [2026-07-17 14:04:25] step 319600/1000000 | loss=2.2433 vocab=1.1019 token=1.3334 | 6.14 steps/s | VRAM=28.14 Go [2026-07-17 14:04:33] step 319650/1000000 | loss=2.4072 vocab=0.6965 token=1.3025 | 6.14 steps/s | VRAM=32.57 Go [2026-07-17 14:04:41] step 319700/1000000 | loss=2.3000 vocab=0.8293 token=1.2623 | 6.11 steps/s | VRAM=32.81 Go [2026-07-17 14:04:50] step 319750/1000000 | loss=2.3125 vocab=1.2730 token=1.6106 | 6.10 steps/s | VRAM=25.03 Go [2026-07-17 14:04:58] step 319800/1000000 | loss=2.2320 vocab=0.6616 token=1.3109 | 6.14 steps/s | VRAM=26.45 Go [2026-07-17 14:05:06] step 319850/1000000 | loss=2.3644 vocab=1.2127 token=1.7589 | 6.18 steps/s | VRAM=32.48 Go [2026-07-17 14:05:14] step 319900/1000000 | loss=2.2837 vocab=0.5725 token=1.4125 | 6.14 steps/s | VRAM=31.59 Go [2026-07-17 14:05:22] step 319950/1000000 | loss=2.4406 vocab=0.9526 token=1.4425 | 6.10 steps/s | VRAM=26.03 Go [2026-07-17 14:05:30] step 320000/1000000 | loss=2.2658 vocab=0.6626 token=0.9585 | 6.15 steps/s | VRAM=31.93 Go [2026-07-17 14:05:38] step 320050/1000000 | loss=2.2511 vocab=0.8478 token=1.3750 | 6.14 steps/s | VRAM=31.31 Go [2026-07-17 14:05:47] step 320100/1000000 | loss=2.2357 vocab=0.8880 token=1.6242 | 6.13 steps/s | VRAM=28.92 Go [2026-07-17 14:05:55] step 320150/1000000 | loss=2.3218 vocab=0.7411 token=1.6303 | 6.14 steps/s | VRAM=28.98 Go [2026-07-17 14:06:03] step 320200/1000000 | loss=2.2216 vocab=0.8856 token=1.2968 | 6.14 steps/s | VRAM=29.34 Go [2026-07-17 14:06:11] step 320250/1000000 | loss=2.3101 vocab=0.7835 token=1.1619 | 6.16 steps/s | VRAM=31.64 Go [2026-07-17 14:06:19] step 320300/1000000 | loss=2.3702 vocab=0.7165 token=1.3889 | 6.12 steps/s | VRAM=28.24 Go [2026-07-17 14:06:27] step 320350/1000000 | loss=2.2952 vocab=1.4127 token=1.7496 | 6.16 steps/s | VRAM=31.30 Go [2026-07-17 14:06:35] step 320400/1000000 | loss=2.3401 vocab=0.9206 token=1.5161 | 6.14 steps/s | VRAM=31.26 Go [2026-07-17 14:06:44] step 320450/1000000 | loss=2.2434 vocab=1.2008 token=1.5520 | 6.12 steps/s | VRAM=28.33 Go [2026-07-17 14:06:52] step 320500/1000000 | loss=2.2720 vocab=1.0008 token=0.9544 | 6.12 steps/s | VRAM=29.30 Go [2026-07-17 14:07:00] step 320550/1000000 | loss=2.2893 vocab=0.4776 token=1.3567 | 6.14 steps/s | VRAM=28.68 Go [2026-07-17 14:07:08] step 320600/1000000 | loss=2.1908 vocab=0.8527 token=1.0943 | 6.14 steps/s | VRAM=28.48 Go [2026-07-17 14:07:16] step 320650/1000000 | loss=2.1464 vocab=1.1407 token=1.8115 | 6.09 steps/s | VRAM=29.18 Go [2026-07-17 14:07:24] step 320700/1000000 | loss=2.3195 vocab=1.0614 token=1.8051 | 6.14 steps/s | VRAM=31.99 Go [2026-07-17 14:07:33] step 320750/1000000 | loss=2.3817 vocab=0.6637 token=1.1873 | 6.13 steps/s | VRAM=27.82 Go [2026-07-17 14:07:41] step 320800/1000000 | loss=2.2849 vocab=1.1677 token=1.5540 | 6.12 steps/s | VRAM=28.73 Go [2026-07-17 14:07:49] step 320850/1000000 | loss=2.2048 vocab=0.7878 token=1.3430 | 6.16 steps/s | VRAM=25.46 Go [2026-07-17 14:07:57] step 320900/1000000 | loss=2.3217 vocab=0.8781 token=1.7304 | 6.10 steps/s | VRAM=26.47 Go [2026-07-17 14:08:05] step 320950/1000000 | loss=2.2379 vocab=0.7335 token=1.0428 | 6.12 steps/s | VRAM=31.97 Go [2026-07-17 14:08:13] step 321000/1000000 | loss=2.3413 vocab=1.6663 token=1.5175 | 6.14 steps/s | VRAM=29.12 Go [2026-07-17 14:08:22] step 321050/1000000 | loss=2.2919 vocab=0.7878 token=1.3421 | 6.12 steps/s | VRAM=29.33 Go [2026-07-17 14:08:30] step 321100/1000000 | loss=2.2606 vocab=1.0026 token=1.1239 | 6.15 steps/s | VRAM=32.53 Go [2026-07-17 14:08:38] step 321150/1000000 | loss=2.2689 vocab=0.7182 token=1.7796 | 6.12 steps/s | VRAM=29.24 Go [2026-07-17 14:08:46] step 321200/1000000 | loss=2.3541 vocab=0.6496 token=1.5749 | 6.16 steps/s | VRAM=31.24 Go [2026-07-17 14:08:54] step 321250/1000000 | loss=2.2301 vocab=1.1320 token=1.4828 | 6.15 steps/s | VRAM=28.25 Go [2026-07-17 14:09:02] step 321300/1000000 | loss=2.2680 vocab=0.9665 token=1.4448 | 6.13 steps/s | VRAM=32.45 Go [2026-07-17 14:09:10] step 321350/1000000 | loss=2.2145 vocab=0.9743 token=1.3342 | 6.13 steps/s | VRAM=29.06 Go [2026-07-17 14:09:19] step 321400/1000000 | loss=2.2872 vocab=0.8036 token=1.3755 | 6.13 steps/s | VRAM=29.81 Go [2026-07-17 14:09:27] step 321450/1000000 | loss=2.2706 vocab=0.7880 token=1.5024 | 6.13 steps/s | VRAM=32.46 Go [2026-07-17 14:09:35] step 321500/1000000 | loss=2.2256 vocab=1.0032 token=1.3044 | 6.14 steps/s | VRAM=28.22 Go [2026-07-17 14:09:43] step 321550/1000000 | loss=2.3585 vocab=0.7591 token=1.4300 | 6.12 steps/s | VRAM=29.42 Go [2026-07-17 14:09:51] step 321600/1000000 | loss=2.3055 vocab=1.0314 token=1.5538 | 6.13 steps/s | VRAM=26.25 Go [2026-07-17 14:09:59] step 321650/1000000 | loss=2.2693 vocab=0.7067 token=1.1211 | 6.14 steps/s | VRAM=24.95 Go [2026-07-17 14:10:08] step 321700/1000000 | loss=2.2960 vocab=1.3989 token=1.6296 | 6.12 steps/s | VRAM=25.93 Go [2026-07-17 14:10:16] step 321750/1000000 | loss=2.3383 vocab=0.6678 token=0.9672 | 6.15 steps/s | VRAM=31.01 Go [2026-07-17 14:10:24] step 321800/1000000 | loss=2.2925 vocab=0.5487 token=1.3065 | 6.13 steps/s | VRAM=28.46 Go [2026-07-17 14:10:32] step 321850/1000000 | loss=2.3373 vocab=0.7564 token=1.3135 | 6.15 steps/s | VRAM=31.37 Go [2026-07-17 14:10:40] step 321900/1000000 | loss=2.2766 vocab=0.9346 token=1.6566 | 6.14 steps/s | VRAM=28.57 Go [2026-07-17 14:10:48] step 321950/1000000 | loss=2.2288 vocab=1.0361 token=1.4950 | 6.13 steps/s | VRAM=28.01 Go [2026-07-17 14:10:56] step 322000/1000000 | loss=2.3072 vocab=1.0016 token=1.5516 | 6.15 steps/s | VRAM=29.13 Go [2026-07-17 14:11:05] step 322050/1000000 | loss=2.2215 vocab=1.1696 token=1.4298 | 6.11 steps/s | VRAM=28.21 Go [2026-07-17 14:11:13] step 322100/1000000 | loss=2.3166 vocab=0.8775 token=1.3823 | 6.14 steps/s | VRAM=29.00 Go [2026-07-17 14:11:21] step 322150/1000000 | loss=2.2833 vocab=1.1263 token=1.2260 | 6.14 steps/s | VRAM=22.51 Go [2026-07-17 14:11:29] step 322200/1000000 | loss=2.3676 vocab=1.0897 token=1.3790 | 6.15 steps/s | VRAM=27.74 Go [2026-07-17 14:11:37] step 322250/1000000 | loss=2.2587 vocab=1.0986 token=1.4570 | 6.13 steps/s | VRAM=32.26 Go [2026-07-17 14:11:45] step 322300/1000000 | loss=2.3617 vocab=0.6503 token=0.9427 | 6.12 steps/s | VRAM=29.75 Go [2026-07-17 14:11:53] step 322350/1000000 | loss=2.3472 vocab=0.8681 token=1.6007 | 6.12 steps/s | VRAM=28.79 Go [2026-07-17 14:12:02] step 322400/1000000 | loss=2.3608 vocab=1.0410 token=1.3412 | 6.13 steps/s | VRAM=32.80 Go [2026-07-17 14:12:10] step 322450/1000000 | loss=2.2158 vocab=0.5749 token=1.4696 | 6.14 steps/s | VRAM=29.04 Go [2026-07-17 14:12:18] step 322500/1000000 | loss=2.3358 vocab=0.9271 token=1.4593 | 6.13 steps/s | VRAM=28.98 Go [2026-07-17 14:12:26] step 322550/1000000 | loss=2.4335 vocab=0.7961 token=1.4656 | 6.14 steps/s | VRAM=32.60 Go [2026-07-17 14:12:34] step 322600/1000000 | loss=2.2984 vocab=1.0344 token=1.6585 | 6.14 steps/s | VRAM=28.54 Go [2026-07-17 14:12:42] step 322650/1000000 | loss=2.3844 vocab=1.2644 token=1.4219 | 6.14 steps/s | VRAM=26.48 Go [2026-07-17 14:12:50] step 322700/1000000 | loss=2.2715 vocab=0.7428 token=1.3216 | 6.16 steps/s | VRAM=26.36 Go [2026-07-17 14:12:59] step 322750/1000000 | loss=2.3074 vocab=0.6086 token=1.1125 | 6.13 steps/s | VRAM=28.62 Go [2026-07-17 14:13:07] step 322800/1000000 | loss=2.3732 vocab=1.5641 token=1.5272 | 6.14 steps/s | VRAM=26.24 Go [2026-07-17 14:13:15] step 322850/1000000 | loss=2.3040 vocab=0.5564 token=1.4867 | 6.14 steps/s | VRAM=29.05 Go [2026-07-17 14:13:23] step 322900/1000000 | loss=2.3036 vocab=0.5706 token=1.3112 | 6.13 steps/s | VRAM=28.95 Go [2026-07-17 14:13:31] step 322950/1000000 | loss=2.2059 vocab=0.8274 token=1.4653 | 6.16 steps/s | VRAM=31.98 Go [2026-07-17 14:13:39] step 323000/1000000 | loss=2.3462 vocab=1.0761 token=1.6925 | 6.13 steps/s | VRAM=32.64 Go [2026-07-17 14:13:48] step 323050/1000000 | loss=2.3385 vocab=1.2321 token=1.2990 | 6.12 steps/s | VRAM=25.72 Go [2026-07-17 14:13:56] step 323100/1000000 | loss=2.3739 vocab=0.9009 token=1.7830 | 6.16 steps/s | VRAM=32.12 Go [2026-07-17 14:14:04] step 323150/1000000 | loss=2.2068 vocab=1.1805 token=1.7736 | 6.12 steps/s | VRAM=30.33 Go [2026-07-17 14:14:12] step 323200/1000000 | loss=2.2966 vocab=0.8058 token=1.4719 | 6.14 steps/s | VRAM=28.21 Go [2026-07-17 14:14:20] step 323250/1000000 | loss=2.3191 vocab=1.0732 token=1.1189 | 6.16 steps/s | VRAM=25.70 Go [2026-07-17 14:14:28] step 323300/1000000 | loss=2.3672 vocab=1.0388 token=1.8529 | 6.16 steps/s | VRAM=26.10 Go [2026-07-17 14:14:36] step 323350/1000000 | loss=2.4319 vocab=0.7022 token=1.6727 | 6.13 steps/s | VRAM=27.99 Go [2026-07-17 14:14:44] step 323400/1000000 | loss=2.3151 vocab=0.6222 token=1.2848 | 6.15 steps/s | VRAM=27.42 Go [2026-07-17 14:14:53] step 323450/1000000 | loss=2.3398 vocab=0.6196 token=1.2870 | 6.15 steps/s | VRAM=28.58 Go [2026-07-17 14:15:01] step 323500/1000000 | loss=2.3440 vocab=0.7811 token=1.3285 | 6.10 steps/s | VRAM=30.50 Go [2026-07-17 14:15:09] step 323550/1000000 | loss=2.2751 vocab=1.4967 token=1.4713 | 6.17 steps/s | VRAM=26.17 Go [2026-07-17 14:15:17] step 323600/1000000 | loss=2.3426 vocab=0.9528 token=1.5605 | 6.16 steps/s | VRAM=30.18 Go [2026-07-17 14:15:25] step 323650/1000000 | loss=2.2520 vocab=0.9830 token=1.6783 | 6.14 steps/s | VRAM=31.66 Go [2026-07-17 14:15:33] step 323700/1000000 | loss=2.1989 vocab=0.8408 token=1.5811 | 6.14 steps/s | VRAM=28.43 Go [2026-07-17 14:15:41] step 323750/1000000 | loss=2.2853 vocab=0.7094 token=1.5850 | 6.12 steps/s | VRAM=29.25 Go [2026-07-17 14:15:50] step 323800/1000000 | loss=2.2804 vocab=0.7207 token=1.3429 | 6.12 steps/s | VRAM=29.49 Go [2026-07-17 14:15:58] step 323850/1000000 | loss=2.2326 vocab=1.2172 token=1.4705 | 6.12 steps/s | VRAM=28.93 Go [2026-07-17 14:16:06] step 323900/1000000 | loss=2.3542 vocab=1.0124 token=1.5202 | 6.16 steps/s | VRAM=31.85 Go [2026-07-17 14:16:14] step 323950/1000000 | loss=2.3300 vocab=1.5034 token=1.4906 | 6.15 steps/s | VRAM=32.13 Go [2026-07-17 14:16:22] step 324000/1000000 | loss=2.3538 vocab=1.6011 token=1.9174 | 6.11 steps/s | VRAM=30.20 Go [2026-07-17 14:16:30] step 324050/1000000 | loss=2.2717 vocab=0.9886 token=1.7943 | 6.17 steps/s | VRAM=32.32 Go [2026-07-17 14:16:39] step 324100/1000000 | loss=2.3379 vocab=0.7360 token=1.2926 | 6.12 steps/s | VRAM=29.13 Go [2026-07-17 14:16:47] step 324150/1000000 | loss=2.3243 vocab=0.9846 token=1.1316 | 6.14 steps/s | VRAM=29.05 Go [2026-07-17 14:16:55] step 324200/1000000 | loss=2.3019 vocab=0.8270 token=1.2181 | 6.15 steps/s | VRAM=26.78 Go [2026-07-17 14:17:03] step 324250/1000000 | loss=2.3305 vocab=0.7150 token=1.2802 | 6.14 steps/s | VRAM=32.10 Go [2026-07-17 14:17:11] step 324300/1000000 | loss=2.2860 vocab=0.6208 token=1.3065 | 6.13 steps/s | VRAM=25.00 Go [2026-07-17 14:17:19] step 324350/1000000 | loss=2.4289 vocab=1.2682 token=1.4745 | 6.14 steps/s | VRAM=29.16 Go [2026-07-17 14:17:27] step 324400/1000000 | loss=2.3818 vocab=0.8325 token=1.0549 | 6.13 steps/s | VRAM=28.67 Go [2026-07-17 14:17:36] step 324450/1000000 | loss=2.2535 vocab=0.9497 token=1.3949 | 6.10 steps/s | VRAM=26.45 Go [2026-07-17 14:17:44] step 324500/1000000 | loss=2.3249 vocab=1.2545 token=1.3622 | 6.14 steps/s | VRAM=28.94 Go [2026-07-17 14:17:52] step 324550/1000000 | loss=2.2771 vocab=0.9212 token=1.2872 | 6.15 steps/s | VRAM=28.76 Go [2026-07-17 14:18:00] step 324600/1000000 | loss=2.2082 vocab=1.0391 token=1.3919 | 6.16 steps/s | VRAM=31.74 Go [2026-07-17 14:18:08] step 324650/1000000 | loss=2.2628 vocab=0.9153 token=1.4385 | 6.14 steps/s | VRAM=32.59 Go [2026-07-17 14:18:16] step 324700/1000000 | loss=2.2770 vocab=1.0610 token=1.0263 | 6.16 steps/s | VRAM=28.14 Go [2026-07-17 14:18:24] step 324750/1000000 | loss=2.3425 vocab=0.9495 token=1.1010 | 6.15 steps/s | VRAM=31.92 Go [2026-07-17 14:18:33] step 324800/1000000 | loss=2.3466 vocab=0.7107 token=1.2062 | 6.12 steps/s | VRAM=28.46 Go [2026-07-17 14:18:41] step 324850/1000000 | loss=2.3158 vocab=0.8541 token=1.3174 | 6.12 steps/s | VRAM=27.08 Go [2026-07-17 14:18:49] step 324900/1000000 | loss=2.3110 vocab=1.1345 token=1.6311 | 6.13 steps/s | VRAM=28.22 Go [2026-07-17 14:18:57] step 324950/1000000 | loss=2.2337 vocab=0.6626 token=1.5042 | 6.13 steps/s | VRAM=25.40 Go [2026-07-17 14:19:05] step 325000/1000000 | loss=2.2840 vocab=1.0245 token=1.4442 | 6.16 steps/s | VRAM=29.42 Go [2026-07-17 14:19:13] step 325050/1000000 | loss=2.1957 vocab=0.8193 token=1.5546 | 6.14 steps/s | VRAM=29.00 Go [2026-07-17 14:19:21] step 325100/1000000 | loss=2.4165 vocab=0.9625 token=0.9129 | 6.13 steps/s | VRAM=28.36 Go [2026-07-17 14:19:30] step 325150/1000000 | loss=2.2995 vocab=0.7230 token=1.5545 | 6.11 steps/s | VRAM=26.07 Go [2026-07-17 14:19:38] step 325200/1000000 | loss=2.2538 vocab=0.6030 token=1.2640 | 6.13 steps/s | VRAM=26.12 Go [2026-07-17 14:19:46] step 325250/1000000 | loss=2.2803 vocab=1.0574 token=1.1115 | 6.10 steps/s | VRAM=26.96 Go [2026-07-17 14:19:54] step 325300/1000000 | loss=2.3478 vocab=0.5400 token=1.4957 | 6.14 steps/s | VRAM=29.10 Go [2026-07-17 14:20:02] step 325350/1000000 | loss=2.3201 vocab=1.2903 token=1.4341 | 6.13 steps/s | VRAM=25.53 Go [2026-07-17 14:20:10] step 325400/1000000 | loss=2.3291 vocab=0.7509 token=1.3204 | 6.13 steps/s | VRAM=29.08 Go [2026-07-17 14:20:19] step 325450/1000000 | loss=2.3435 vocab=0.8903 token=1.1345 | 6.11 steps/s | VRAM=28.26 Go [2026-07-17 14:20:27] step 325500/1000000 | loss=2.3321 vocab=1.3625 token=1.6656 | 6.11 steps/s | VRAM=26.17 Go [2026-07-17 14:20:35] step 325550/1000000 | loss=2.3347 vocab=0.9602 token=1.5174 | 6.13 steps/s | VRAM=26.58 Go [2026-07-17 14:20:43] step 325600/1000000 | loss=2.3732 vocab=0.6613 token=1.2553 | 6.11 steps/s | VRAM=25.90 Go [2026-07-17 14:20:51] step 325650/1000000 | loss=2.2527 vocab=0.9604 token=1.5327 | 6.13 steps/s | VRAM=32.22 Go [2026-07-17 14:20:59] step 325700/1000000 | loss=2.2427 vocab=0.6086 token=1.4540 | 6.15 steps/s | VRAM=32.11 Go [2026-07-17 14:21:08] step 325750/1000000 | loss=2.3131 vocab=0.8070 token=1.4300 | 6.14 steps/s | VRAM=31.40 Go [2026-07-17 14:21:16] step 325800/1000000 | loss=2.3782 vocab=1.2200 token=1.1928 | 6.14 steps/s | VRAM=28.68 Go [2026-07-17 14:21:24] step 325850/1000000 | loss=2.3288 vocab=0.7086 token=1.0651 | 6.11 steps/s | VRAM=27.04 Go [2026-07-17 14:21:32] step 325900/1000000 | loss=2.3141 vocab=1.1651 token=1.0704 | 6.16 steps/s | VRAM=25.03 Go [2026-07-17 14:21:40] step 325950/1000000 | loss=2.1943 vocab=1.0784 token=1.5132 | 6.15 steps/s | VRAM=32.39 Go [2026-07-17 14:21:48] step 326000/1000000 | loss=2.2485 vocab=0.7797 token=1.3796 | 6.12 steps/s | VRAM=26.98 Go [2026-07-17 14:21:56] step 326050/1000000 | loss=2.3137 vocab=1.4972 token=2.0912 | 6.15 steps/s | VRAM=32.05 Go [2026-07-17 14:22:05] step 326100/1000000 | loss=2.3163 vocab=0.7160 token=1.3687 | 6.16 steps/s | VRAM=32.14 Go [2026-07-17 14:22:13] step 326150/1000000 | loss=2.1595 vocab=0.7308 token=1.6725 | 6.16 steps/s | VRAM=28.32 Go [2026-07-17 14:22:21] step 326200/1000000 | loss=2.3766 vocab=1.3023 token=1.4425 | 6.13 steps/s | VRAM=30.31 Go [2026-07-17 14:22:29] step 326250/1000000 | loss=2.2218 vocab=0.6845 token=1.1378 | 6.11 steps/s | VRAM=28.65 Go [2026-07-17 14:22:37] step 326300/1000000 | loss=2.2719 vocab=1.2336 token=1.5634 | 6.15 steps/s | VRAM=29.03 Go [2026-07-17 14:22:45] step 326350/1000000 | loss=2.2392 vocab=1.3772 token=1.5727 | 6.16 steps/s | VRAM=32.38 Go [2026-07-17 14:22:53] step 326400/1000000 | loss=2.3168 vocab=1.1031 token=1.7350 | 6.13 steps/s | VRAM=31.83 Go [2026-07-17 14:23:02] step 326450/1000000 | loss=2.3643 vocab=1.0752 token=1.2868 | 6.14 steps/s | VRAM=26.59 Go [2026-07-17 14:23:10] step 326500/1000000 | loss=2.3533 vocab=0.9011 token=1.2904 | 6.15 steps/s | VRAM=28.98 Go [2026-07-17 14:23:18] step 326550/1000000 | loss=2.2730 vocab=0.7340 token=1.1397 | 6.16 steps/s | VRAM=28.18 Go [2026-07-17 14:23:26] step 326600/1000000 | loss=2.2537 vocab=1.0579 token=1.2299 | 6.16 steps/s | VRAM=28.17 Go [2026-07-17 14:23:34] step 326650/1000000 | loss=2.3694 vocab=0.5984 token=1.0798 | 6.12 steps/s | VRAM=28.95 Go [2026-07-17 14:23:42] step 326700/1000000 | loss=2.3026 vocab=1.0649 token=1.3208 | 6.15 steps/s | VRAM=28.58 Go [2026-07-17 14:23:50] step 326750/1000000 | loss=2.2722 vocab=1.0970 token=1.5236 | 6.16 steps/s | VRAM=31.20 Go [2026-07-17 14:23:59] step 326800/1000000 | loss=2.3158 vocab=1.7365 token=1.7772 | 6.13 steps/s | VRAM=29.10 Go [2026-07-17 14:24:07] step 326850/1000000 | loss=2.3724 vocab=0.8625 token=1.5693 | 6.13 steps/s | VRAM=29.34 Go [2026-07-17 14:24:15] step 326900/1000000 | loss=2.3454 vocab=0.6957 token=1.4120 | 6.11 steps/s | VRAM=29.50 Go [2026-07-17 14:24:23] step 326950/1000000 | loss=2.2765 vocab=0.7652 token=1.6585 | 6.13 steps/s | VRAM=27.96 Go [2026-07-17 14:24:31] step 327000/1000000 | loss=2.2727 vocab=1.2613 token=1.1137 | 6.15 steps/s | VRAM=25.17 Go [2026-07-17 14:24:39] step 327050/1000000 | loss=2.2392 vocab=0.6539 token=1.3215 | 6.14 steps/s | VRAM=25.70 Go [2026-07-17 14:24:47] step 327100/1000000 | loss=2.2901 vocab=0.6782 token=1.1055 | 6.12 steps/s | VRAM=25.85 Go [2026-07-17 14:24:56] step 327150/1000000 | loss=2.3185 vocab=0.9348 token=1.4820 | 6.11 steps/s | VRAM=29.77 Go [2026-07-17 14:25:04] step 327200/1000000 | loss=2.3895 vocab=1.1329 token=1.5373 | 6.14 steps/s | VRAM=28.15 Go [2026-07-17 14:25:12] step 327250/1000000 | loss=2.3008 vocab=0.8307 token=1.5022 | 6.13 steps/s | VRAM=28.48 Go [2026-07-17 14:25:20] step 327300/1000000 | loss=2.3218 vocab=1.2262 token=1.9604 | 6.10 steps/s | VRAM=25.87 Go [2026-07-17 14:25:28] step 327350/1000000 | loss=2.1871 vocab=0.5275 token=1.3599 | 6.15 steps/s | VRAM=28.20 Go [2026-07-17 14:25:36] step 327400/1000000 | loss=2.3151 vocab=0.7328 token=1.2484 | 6.12 steps/s | VRAM=29.49 Go [2026-07-17 14:25:45] step 327450/1000000 | loss=2.3997 vocab=0.9504 token=1.2730 | 6.12 steps/s | VRAM=29.34 Go [2026-07-17 14:25:53] step 327500/1000000 | loss=2.2692 vocab=1.0052 token=1.2693 | 6.14 steps/s | VRAM=28.60 Go [2026-07-17 14:26:01] step 327550/1000000 | loss=2.3678 vocab=0.6592 token=1.4574 | 6.09 steps/s | VRAM=32.91 Go [2026-07-17 14:26:09] step 327600/1000000 | loss=2.3225 vocab=0.7567 token=1.1282 | 6.12 steps/s | VRAM=28.40 Go [2026-07-17 14:26:17] step 327650/1000000 | loss=2.2365 vocab=0.6179 token=1.0530 | 6.13 steps/s | VRAM=32.64 Go [2026-07-17 14:26:25] step 327700/1000000 | loss=2.1654 vocab=0.6815 token=0.8717 | 6.14 steps/s | VRAM=25.27 Go [2026-07-17 14:26:34] step 327750/1000000 | loss=2.2993 vocab=1.1378 token=1.6716 | 6.14 steps/s | VRAM=27.74 Go [2026-07-17 14:26:42] step 327800/1000000 | loss=2.3169 vocab=0.8952 token=1.4052 | 6.13 steps/s | VRAM=28.52 Go [2026-07-17 14:26:50] step 327850/1000000 | loss=2.3216 vocab=0.6194 token=1.3384 | 6.14 steps/s | VRAM=28.51 Go [2026-07-17 14:26:58] step 327900/1000000 | loss=2.4199 vocab=0.4988 token=1.0186 | 6.13 steps/s | VRAM=28.29 Go [2026-07-17 14:27:06] step 327950/1000000 | loss=2.3620 vocab=0.6735 token=1.2038 | 6.13 steps/s | VRAM=28.93 Go [2026-07-17 14:27:14] step 328000/1000000 | loss=2.3157 vocab=0.5830 token=0.9135 | 6.12 steps/s | VRAM=28.75 Go [2026-07-17 14:27:22] step 328050/1000000 | loss=2.1897 vocab=0.8236 token=1.0743 | 6.16 steps/s | VRAM=28.98 Go [2026-07-17 14:27:31] step 328100/1000000 | loss=2.3424 vocab=0.5308 token=1.1502 | 6.12 steps/s | VRAM=28.82 Go [2026-07-17 14:27:39] step 328150/1000000 | loss=2.2504 vocab=0.9711 token=1.4961 | 6.15 steps/s | VRAM=31.35 Go [2026-07-17 14:27:47] step 328200/1000000 | loss=2.2891 vocab=1.1803 token=1.3207 | 6.12 steps/s | VRAM=29.59 Go [2026-07-17 14:27:55] step 328250/1000000 | loss=2.2705 vocab=1.4635 token=1.0336 | 6.13 steps/s | VRAM=29.44 Go [2026-07-17 14:28:03] step 328300/1000000 | loss=2.3629 vocab=1.0974 token=1.2179 | 6.12 steps/s | VRAM=28.50 Go [2026-07-17 14:28:11] step 328350/1000000 | loss=2.3315 vocab=0.9694 token=1.4032 | 6.12 steps/s | VRAM=29.10 Go [2026-07-17 14:28:20] step 328400/1000000 | loss=2.3022 vocab=0.8765 token=1.3087 | 6.11 steps/s | VRAM=28.92 Go [2026-07-17 14:28:28] step 328450/1000000 | loss=2.3694 vocab=1.5647 token=1.1629 | 6.12 steps/s | VRAM=28.76 Go [2026-07-17 14:28:36] step 328500/1000000 | loss=2.2593 vocab=1.2716 token=1.5181 | 6.13 steps/s | VRAM=27.67 Go [2026-07-17 14:28:44] step 328550/1000000 | loss=2.2510 vocab=0.9819 token=1.1613 | 6.16 steps/s | VRAM=32.10 Go [2026-07-17 14:28:52] step 328600/1000000 | loss=2.3316 vocab=1.6551 token=1.7600 | 6.13 steps/s | VRAM=31.91 Go [2026-07-17 14:29:00] step 328650/1000000 | loss=2.4301 vocab=1.1096 token=1.4242 | 6.16 steps/s | VRAM=32.53 Go [2026-07-17 14:29:08] step 328700/1000000 | loss=2.2609 vocab=0.8921 token=1.7613 | 6.16 steps/s | VRAM=27.40 Go [2026-07-17 14:29:17] step 328750/1000000 | loss=2.2538 vocab=0.6438 token=1.4083 | 6.14 steps/s | VRAM=29.25 Go [2026-07-17 14:29:25] step 328800/1000000 | loss=2.2366 vocab=1.1185 token=1.1410 | 6.12 steps/s | VRAM=29.03 Go [2026-07-17 14:29:33] step 328850/1000000 | loss=2.3122 vocab=0.6838 token=1.1838 | 6.11 steps/s | VRAM=28.48 Go [2026-07-17 14:29:41] step 328900/1000000 | loss=2.3282 vocab=0.5612 token=1.3600 | 6.11 steps/s | VRAM=27.93 Go [2026-07-17 14:29:49] step 328950/1000000 | loss=2.3706 vocab=1.3986 token=1.8492 | 6.14 steps/s | VRAM=29.51 Go [2026-07-17 14:29:57] step 329000/1000000 | loss=2.1746 vocab=0.9918 token=1.5575 | 6.14 steps/s | VRAM=27.91 Go [2026-07-17 14:30:06] step 329050/1000000 | loss=2.3631 vocab=1.0568 token=1.5932 | 6.12 steps/s | VRAM=29.25 Go [2026-07-17 14:30:14] step 329100/1000000 | loss=2.2974 vocab=1.1796 token=1.2124 | 6.12 steps/s | VRAM=27.99 Go [2026-07-17 14:30:22] step 329150/1000000 | loss=2.1755 vocab=1.1328 token=1.2098 | 6.15 steps/s | VRAM=32.29 Go [2026-07-17 14:30:30] step 329200/1000000 | loss=2.2809 vocab=0.6326 token=1.4790 | 6.12 steps/s | VRAM=28.69 Go [2026-07-17 14:30:38] step 329250/1000000 | loss=2.2648 vocab=0.7933 token=1.3839 | 6.11 steps/s | VRAM=29.37 Go [2026-07-17 14:30:46] step 329300/1000000 | loss=2.3565 vocab=0.9417 token=1.6826 | 6.14 steps/s | VRAM=29.07 Go [2026-07-17 14:30:55] step 329350/1000000 | loss=2.2672 vocab=1.0104 token=1.4826 | 6.17 steps/s | VRAM=27.37 Go [2026-07-17 14:31:03] step 329400/1000000 | loss=2.3142 vocab=0.7287 token=1.2095 | 6.11 steps/s | VRAM=24.51 Go [2026-07-17 14:31:11] step 329450/1000000 | loss=2.2433 vocab=0.7220 token=0.8204 | 6.12 steps/s | VRAM=28.31 Go [2026-07-17 14:31:19] step 329500/1000000 | loss=2.3398 vocab=0.7266 token=1.2041 | 6.15 steps/s | VRAM=31.50 Go [2026-07-17 14:31:27] step 329550/1000000 | loss=2.3240 vocab=1.3554 token=1.3970 | 6.10 steps/s | VRAM=26.67 Go [2026-07-17 14:31:35] step 329600/1000000 | loss=2.3707 vocab=0.9503 token=1.4056 | 6.13 steps/s | VRAM=28.09 Go [2026-07-17 14:31:43] step 329650/1000000 | loss=2.3831 vocab=1.2204 token=1.4828 | 6.15 steps/s | VRAM=32.55 Go [2026-07-17 14:31:52] step 329700/1000000 | loss=2.3103 vocab=0.6821 token=1.4799 | 6.14 steps/s | VRAM=31.97 Go [2026-07-17 14:32:00] step 329750/1000000 | loss=2.2973 vocab=0.6675 token=1.0623 | 6.14 steps/s | VRAM=29.07 Go [2026-07-17 14:32:08] step 329800/1000000 | loss=2.3002 vocab=1.0704 token=1.8042 | 6.17 steps/s | VRAM=31.94 Go [2026-07-17 14:32:16] step 329850/1000000 | loss=2.2692 vocab=0.9200 token=1.2577 | 6.10 steps/s | VRAM=28.49 Go [2026-07-17 14:32:24] step 329900/1000000 | loss=2.3203 vocab=0.9936 token=1.7418 | 6.15 steps/s | VRAM=26.43 Go [2026-07-17 14:32:32] step 329950/1000000 | loss=2.2257 vocab=0.8550 token=0.9031 | 6.13 steps/s | VRAM=29.01 Go [2026-07-17 14:32:41] step 330000/1000000 | loss=2.3834 vocab=0.3805 token=1.1534 | 6.10 steps/s | VRAM=28.73 Go [2026-07-17 14:32:49] step 330050/1000000 | loss=2.2443 vocab=1.4797 token=1.3323 | 6.11 steps/s | VRAM=27.36 Go [2026-07-17 14:32:57] step 330100/1000000 | loss=2.3050 vocab=1.1434 token=1.4892 | 6.17 steps/s | VRAM=27.75 Go [2026-07-17 14:33:05] step 330150/1000000 | loss=2.2856 vocab=0.8866 token=1.3616 | 6.14 steps/s | VRAM=33.00 Go [2026-07-17 14:33:13] step 330200/1000000 | loss=2.3310 vocab=1.4699 token=1.6019 | 6.11 steps/s | VRAM=23.29 Go [2026-07-17 14:33:21] step 330250/1000000 | loss=2.3829 vocab=0.8189 token=1.4493 | 6.17 steps/s | VRAM=27.73 Go [2026-07-17 14:33:29] step 330300/1000000 | loss=2.2772 vocab=0.9893 token=1.6197 | 6.12 steps/s | VRAM=25.76 Go [2026-07-17 14:33:38] step 330350/1000000 | loss=2.3050 vocab=0.9770 token=1.3553 | 6.14 steps/s | VRAM=27.97 Go [2026-07-17 14:33:46] step 330400/1000000 | loss=2.4159 vocab=1.4028 token=1.7459 | 6.14 steps/s | VRAM=32.71 Go [2026-07-17 14:33:54] step 330450/1000000 | loss=2.2914 vocab=1.0766 token=1.6152 | 6.12 steps/s | VRAM=32.10 Go [2026-07-17 14:34:02] step 330500/1000000 | loss=2.3061 vocab=0.9274 token=1.2910 | 6.14 steps/s | VRAM=28.52 Go [2026-07-17 14:34:10] step 330550/1000000 | loss=2.2771 vocab=1.0974 token=1.5618 | 6.13 steps/s | VRAM=31.74 Go [2026-07-17 14:34:18] step 330600/1000000 | loss=2.3317 vocab=0.5845 token=1.5264 | 6.13 steps/s | VRAM=28.03 Go [2026-07-17 14:34:26] step 330650/1000000 | loss=2.3364 vocab=0.9647 token=1.1635 | 6.14 steps/s | VRAM=32.51 Go [2026-07-17 14:34:35] step 330700/1000000 | loss=2.2578 vocab=1.3926 token=1.4954 | 6.13 steps/s | VRAM=29.50 Go [2026-07-17 14:34:43] step 330750/1000000 | loss=2.3673 vocab=0.7790 token=1.4008 | 6.11 steps/s | VRAM=28.57 Go [2026-07-17 14:34:51] step 330800/1000000 | loss=2.3462 vocab=0.6412 token=1.4238 | 6.11 steps/s | VRAM=28.43 Go [2026-07-17 14:34:59] step 330850/1000000 | loss=2.2463 vocab=0.7571 token=1.4665 | 6.12 steps/s | VRAM=29.10 Go [2026-07-17 14:35:07] step 330900/1000000 | loss=2.1860 vocab=0.6003 token=1.3671 | 6.12 steps/s | VRAM=28.36 Go [2026-07-17 14:35:16] step 330950/1000000 | loss=2.2973 vocab=0.6758 token=1.3176 | 6.11 steps/s | VRAM=30.03 Go [2026-07-17 14:35:24] step 331000/1000000 | loss=2.2285 vocab=0.9612 token=1.2215 | 6.11 steps/s | VRAM=28.08 Go [2026-07-17 14:35:32] step 331050/1000000 | loss=2.2654 vocab=1.2558 token=1.3473 | 6.13 steps/s | VRAM=32.12 Go [2026-07-17 14:35:40] step 331100/1000000 | loss=2.2721 vocab=0.6849 token=1.4563 | 6.15 steps/s | VRAM=31.48 Go [2026-07-17 14:35:48] step 331150/1000000 | loss=2.3160 vocab=0.9338 token=1.3737 | 6.14 steps/s | VRAM=25.22 Go [2026-07-17 14:35:56] step 331200/1000000 | loss=2.3661 vocab=1.1386 token=1.2717 | 6.10 steps/s | VRAM=32.15 Go [2026-07-17 14:36:05] step 331250/1000000 | loss=2.2559 vocab=0.6662 token=1.3948 | 6.10 steps/s | VRAM=29.01 Go [2026-07-17 14:36:13] step 331300/1000000 | loss=2.3373 vocab=0.8944 token=1.3225 | 6.13 steps/s | VRAM=28.15 Go [2026-07-17 14:36:21] step 331350/1000000 | loss=2.2869 vocab=0.8016 token=1.2651 | 6.12 steps/s | VRAM=27.60 Go [2026-07-17 14:36:29] step 331400/1000000 | loss=2.2120 vocab=0.8567 token=1.0913 | 6.10 steps/s | VRAM=31.13 Go [2026-07-17 14:36:37] step 331450/1000000 | loss=2.2362 vocab=0.8972 token=1.3853 | 6.18 steps/s | VRAM=26.56 Go [2026-07-17 14:36:45] step 331500/1000000 | loss=2.3301 vocab=0.6601 token=1.6477 | 6.12 steps/s | VRAM=32.79 Go [2026-07-17 14:36:54] step 331550/1000000 | loss=2.3897 vocab=1.0289 token=1.6508 | 6.09 steps/s | VRAM=25.41 Go [2026-07-17 14:37:02] step 331600/1000000 | loss=2.3150 vocab=0.9284 token=1.3120 | 6.07 steps/s | VRAM=28.75 Go [2026-07-17 14:37:10] step 331650/1000000 | loss=2.2770 vocab=0.8627 token=1.3685 | 6.16 steps/s | VRAM=32.07 Go [2026-07-17 14:37:18] step 331700/1000000 | loss=2.3578 vocab=1.0194 token=1.4273 | 6.11 steps/s | VRAM=31.59 Go [2026-07-17 14:37:26] step 331750/1000000 | loss=2.2333 vocab=0.8624 token=1.1933 | 6.12 steps/s | VRAM=28.88 Go [2026-07-17 14:37:34] step 331800/1000000 | loss=2.2649 vocab=0.8852 token=1.4251 | 6.11 steps/s | VRAM=30.05 Go [2026-07-17 14:37:43] step 331850/1000000 | loss=2.1731 vocab=0.5183 token=1.1459 | 6.12 steps/s | VRAM=31.79 Go [2026-07-17 14:37:51] step 331900/1000000 | loss=2.2154 vocab=1.0385 token=1.6700 | 6.13 steps/s | VRAM=26.06 Go [2026-07-17 14:37:59] step 331950/1000000 | loss=2.2529 vocab=1.0343 token=1.3250 | 6.15 steps/s | VRAM=32.35 Go [2026-07-17 14:38:07] step 332000/1000000 | loss=2.3029 vocab=0.7096 token=1.4527 | 6.15 steps/s | VRAM=30.88 Go [2026-07-17 14:38:15] step 332050/1000000 | loss=2.3507 vocab=0.8114 token=1.5128 | 6.12 steps/s | VRAM=25.40 Go [2026-07-17 14:38:23] step 332100/1000000 | loss=2.2801 vocab=1.1405 token=1.5277 | 6.13 steps/s | VRAM=26.02 Go [2026-07-17 14:38:31] step 332150/1000000 | loss=2.3607 vocab=1.3142 token=1.5949 | 6.14 steps/s | VRAM=29.08 Go [2026-07-17 14:38:40] step 332200/1000000 | loss=2.3029 vocab=1.0962 token=1.6119 | 6.11 steps/s | VRAM=32.76 Go [2026-07-17 14:38:48] step 332250/1000000 | loss=2.2996 vocab=0.7498 token=1.4269 | 6.12 steps/s | VRAM=25.93 Go [2026-07-17 14:38:56] step 332300/1000000 | loss=2.2879 vocab=0.6751 token=1.5317 | 6.12 steps/s | VRAM=28.54 Go [2026-07-17 14:39:04] step 332350/1000000 | loss=2.3097 vocab=1.2861 token=1.0820 | 6.12 steps/s | VRAM=28.35 Go [2026-07-17 14:39:12] step 332400/1000000 | loss=2.3090 vocab=0.9972 token=1.3604 | 6.09 steps/s | VRAM=26.11 Go [2026-07-17 14:39:20] step 332450/1000000 | loss=2.2383 vocab=0.8788 token=1.5373 | 6.14 steps/s | VRAM=32.26 Go [2026-07-17 14:39:29] step 332500/1000000 | loss=2.3403 vocab=1.0808 token=1.2766 | 6.12 steps/s | VRAM=28.76 Go [2026-07-17 14:39:37] step 332550/1000000 | loss=2.2635 vocab=1.2125 token=1.3490 | 6.11 steps/s | VRAM=29.85 Go [2026-07-17 14:39:45] step 332600/1000000 | loss=2.3797 vocab=1.2409 token=1.1880 | 6.09 steps/s | VRAM=28.51 Go [2026-07-17 14:39:53] step 332650/1000000 | loss=2.2623 vocab=1.4346 token=1.3493 | 6.11 steps/s | VRAM=25.69 Go [2026-07-17 14:40:01] step 332700/1000000 | loss=2.3260 vocab=1.0117 token=1.2818 | 6.11 steps/s | VRAM=25.47 Go [2026-07-17 14:40:10] step 332750/1000000 | loss=2.2508 vocab=0.8982 token=1.3757 | 6.12 steps/s | VRAM=29.09 Go [2026-07-17 14:40:18] step 332800/1000000 | loss=2.2589 vocab=0.4776 token=1.2456 | 6.11 steps/s | VRAM=28.48 Go [2026-07-17 14:40:26] step 332850/1000000 | loss=2.4526 vocab=0.7293 token=1.4191 | 6.14 steps/s | VRAM=28.62 Go [2026-07-17 14:40:34] step 332900/1000000 | loss=2.2574 vocab=0.5309 token=0.7625 | 6.10 steps/s | VRAM=25.96 Go [2026-07-17 14:40:42] step 332950/1000000 | loss=2.2741 vocab=0.9707 token=1.1849 | 6.15 steps/s | VRAM=32.54 Go [2026-07-17 14:40:50] step 333000/1000000 | loss=2.3589 vocab=0.7355 token=1.3627 | 6.13 steps/s | VRAM=32.49 Go [2026-07-17 14:40:59] step 333050/1000000 | loss=2.3559 vocab=0.9357 token=1.3420 | 6.09 steps/s | VRAM=29.49 Go [2026-07-17 14:41:07] step 333100/1000000 | loss=2.1654 vocab=0.7428 token=1.3833 | 6.10 steps/s | VRAM=26.81 Go [2026-07-17 14:41:15] step 333150/1000000 | loss=2.3249 vocab=1.4022 token=1.6164 | 6.10 steps/s | VRAM=28.00 Go [2026-07-17 14:41:23] step 333200/1000000 | loss=2.2606 vocab=1.0243 token=1.0915 | 6.12 steps/s | VRAM=28.87 Go [2026-07-17 14:41:31] step 333250/1000000 | loss=2.3572 vocab=0.4684 token=1.1909 | 6.12 steps/s | VRAM=29.11 Go [2026-07-17 14:41:40] step 333300/1000000 | loss=2.2711 vocab=0.9869 token=1.2857 | 6.10 steps/s | VRAM=29.64 Go [2026-07-17 14:41:48] step 333350/1000000 | loss=2.3193 vocab=1.1626 token=1.5843 | 6.11 steps/s | VRAM=29.06 Go [2026-07-17 14:41:56] step 333400/1000000 | loss=2.2420 vocab=1.0266 token=1.2104 | 6.12 steps/s | VRAM=32.48 Go [2026-07-17 14:42:04] step 333450/1000000 | loss=2.2358 vocab=0.8557 token=1.5405 | 6.13 steps/s | VRAM=28.23 Go [2026-07-17 14:42:12] step 333500/1000000 | loss=2.2841 vocab=0.8318 token=1.1942 | 6.10 steps/s | VRAM=32.63 Go [2026-07-17 14:42:20] step 333550/1000000 | loss=2.2874 vocab=0.9196 token=1.5728 | 6.09 steps/s | VRAM=29.61 Go [2026-07-17 14:42:29] step 333600/1000000 | loss=2.2716 vocab=1.0098 token=1.5571 | 6.13 steps/s | VRAM=32.69 Go [2026-07-17 14:42:37] step 333650/1000000 | loss=2.2670 vocab=0.5169 token=1.4345 | 6.12 steps/s | VRAM=29.17 Go [2026-07-17 14:42:45] step 333700/1000000 | loss=2.2631 vocab=0.9436 token=1.3331 | 6.12 steps/s | VRAM=29.82 Go [2026-07-17 14:42:53] step 333750/1000000 | loss=2.3359 vocab=0.9920 token=1.1792 | 6.12 steps/s | VRAM=28.62 Go [2026-07-17 14:43:01] step 333800/1000000 | loss=2.1570 vocab=1.0121 token=0.9807 | 6.12 steps/s | VRAM=26.49 Go [2026-07-17 14:43:09] step 333850/1000000 | loss=2.1793 vocab=0.8710 token=1.1656 | 6.15 steps/s | VRAM=31.50 Go [2026-07-17 14:43:18] step 333900/1000000 | loss=2.3194 vocab=1.3069 token=1.5631 | 6.08 steps/s | VRAM=29.28 Go [2026-07-17 14:43:26] step 333950/1000000 | loss=2.4070 vocab=1.1670 token=1.3799 | 6.12 steps/s | VRAM=29.13 Go [2026-07-17 14:43:34] step 334000/1000000 | loss=2.3445 vocab=1.0438 token=1.3497 | 6.12 steps/s | VRAM=28.02 Go [2026-07-17 14:43:42] step 334050/1000000 | loss=2.1744 vocab=1.3529 token=1.7267 | 6.10 steps/s | VRAM=32.52 Go [2026-07-17 14:43:50] step 334100/1000000 | loss=2.3272 vocab=1.1062 token=1.3034 | 6.11 steps/s | VRAM=26.29 Go [2026-07-17 14:43:59] step 334150/1000000 | loss=2.4012 vocab=0.9307 token=1.5598 | 6.10 steps/s | VRAM=25.45 Go [2026-07-17 14:44:07] step 334200/1000000 | loss=2.4112 vocab=1.3657 token=1.2501 | 6.16 steps/s | VRAM=30.97 Go [2026-07-17 14:44:15] step 334250/1000000 | loss=2.3154 vocab=0.9740 token=1.7401 | 6.12 steps/s | VRAM=31.60 Go [2026-07-17 14:44:23] step 334300/1000000 | loss=2.3725 vocab=0.7235 token=1.2779 | 6.08 steps/s | VRAM=28.84 Go [2026-07-17 14:44:31] step 334350/1000000 | loss=2.3062 vocab=0.5213 token=1.4407 | 6.14 steps/s | VRAM=27.89 Go [2026-07-17 14:44:39] step 334400/1000000 | loss=2.1734 vocab=1.1257 token=1.2486 | 6.11 steps/s | VRAM=25.71 Go [2026-07-17 14:44:48] step 334450/1000000 | loss=2.3464 vocab=0.6116 token=1.3760 | 6.11 steps/s | VRAM=28.54 Go [2026-07-17 14:44:56] step 334500/1000000 | loss=2.2749 vocab=1.1039 token=1.7110 | 6.11 steps/s | VRAM=32.63 Go [2026-07-17 14:45:04] step 334550/1000000 | loss=2.2460 vocab=1.0079 token=1.2909 | 6.12 steps/s | VRAM=31.86 Go [2026-07-17 14:45:12] step 334600/1000000 | loss=2.4249 vocab=1.0788 token=1.3228 | 6.08 steps/s | VRAM=26.01 Go [2026-07-17 14:45:20] step 334650/1000000 | loss=2.3000 vocab=0.8575 token=1.2000 | 6.10 steps/s | VRAM=29.58 Go [2026-07-17 14:45:28] step 334700/1000000 | loss=2.2668 vocab=0.9552 token=1.2090 | 6.15 steps/s | VRAM=33.01 Go [2026-07-17 14:45:37] step 334750/1000000 | loss=2.2966 vocab=1.8465 token=1.2744 | 6.12 steps/s | VRAM=27.59 Go [2026-07-17 14:45:45] step 334800/1000000 | loss=2.2518 vocab=0.6186 token=1.1583 | 6.17 steps/s | VRAM=31.60 Go [2026-07-17 14:45:53] step 334850/1000000 | loss=2.2408 vocab=1.0582 token=1.2385 | 6.14 steps/s | VRAM=31.01 Go [2026-07-17 14:46:01] step 334900/1000000 | loss=2.3601 vocab=0.4681 token=1.2186 | 6.14 steps/s | VRAM=28.14 Go [2026-07-17 14:46:09] step 334950/1000000 | loss=2.4006 vocab=0.7273 token=1.1741 | 6.12 steps/s | VRAM=31.48 Go [2026-07-17 14:46:17] step 335000/1000000 | loss=2.2427 vocab=0.8600 token=1.4010 | 6.15 steps/s | VRAM=32.27 Go [2026-07-17 14:46:25] step 335050/1000000 | loss=2.1784 vocab=1.1830 token=1.1687 | 6.13 steps/s | VRAM=28.47 Go [2026-07-17 14:46:34] step 335100/1000000 | loss=2.4562 vocab=0.8664 token=1.3199 | 6.09 steps/s | VRAM=28.19 Go [2026-07-17 14:46:42] step 335150/1000000 | loss=2.3183 vocab=0.8245 token=1.6412 | 6.10 steps/s | VRAM=32.48 Go [2026-07-17 14:46:50] step 335200/1000000 | loss=2.4204 vocab=0.7704 token=1.1612 | 6.11 steps/s | VRAM=32.45 Go [2026-07-17 14:46:58] step 335250/1000000 | loss=2.2223 vocab=0.5405 token=1.1444 | 6.12 steps/s | VRAM=26.13 Go [2026-07-17 14:47:06] step 335300/1000000 | loss=2.2081 vocab=0.8990 token=1.2215 | 6.13 steps/s | VRAM=29.80 Go [2026-07-17 14:47:14] step 335350/1000000 | loss=2.2736 vocab=0.9766 token=1.0645 | 6.15 steps/s | VRAM=28.73 Go [2026-07-17 14:47:23] step 335400/1000000 | loss=2.2725 vocab=0.5128 token=1.1260 | 6.12 steps/s | VRAM=28.41 Go [2026-07-17 14:47:31] step 335450/1000000 | loss=2.4049 vocab=1.3547 token=1.4117 | 6.13 steps/s | VRAM=28.87 Go [2026-07-17 14:47:39] step 335500/1000000 | loss=2.3089 vocab=1.1902 token=1.4676 | 6.11 steps/s | VRAM=28.26 Go [2026-07-17 14:47:47] step 335550/1000000 | loss=2.2856 vocab=1.2567 token=1.7569 | 6.14 steps/s | VRAM=31.54 Go [2026-07-17 14:47:55] step 335600/1000000 | loss=2.2818 vocab=1.0970 token=1.7282 | 6.11 steps/s | VRAM=28.61 Go [2026-07-17 14:48:04] step 335650/1000000 | loss=2.1989 vocab=1.4519 token=1.6409 | 6.11 steps/s | VRAM=28.87 Go [2026-07-17 14:48:12] step 335700/1000000 | loss=2.2878 vocab=1.0299 token=1.2712 | 6.07 steps/s | VRAM=28.20 Go [2026-07-17 14:48:20] step 335750/1000000 | loss=2.3751 vocab=0.8080 token=1.1904 | 6.15 steps/s | VRAM=31.97 Go [2026-07-17 14:48:28] step 335800/1000000 | loss=2.3935 vocab=1.1065 token=1.5808 | 6.15 steps/s | VRAM=31.36 Go [2026-07-17 14:48:36] step 335850/1000000 | loss=2.2209 vocab=1.2214 token=1.5001 | 6.14 steps/s | VRAM=26.22 Go [2026-07-17 14:48:44] step 335900/1000000 | loss=2.2593 vocab=0.6545 token=1.3810 | 6.12 steps/s | VRAM=32.59 Go [2026-07-17 14:48:53] step 335950/1000000 | loss=2.3835 vocab=1.0352 token=1.5205 | 6.09 steps/s | VRAM=26.35 Go [2026-07-17 14:49:01] step 336000/1000000 | loss=2.3551 vocab=0.9108 token=0.9766 | 6.13 steps/s | VRAM=32.81 Go [2026-07-17 14:49:09] step 336050/1000000 | loss=2.3416 vocab=0.7122 token=1.2567 | 6.13 steps/s | VRAM=31.97 Go [2026-07-17 14:49:17] step 336100/1000000 | loss=2.2861 vocab=0.8530 token=1.2110 | 6.10 steps/s | VRAM=29.38 Go [2026-07-17 14:49:25] step 336150/1000000 | loss=2.3152 vocab=0.9537 token=1.4502 | 6.11 steps/s | VRAM=32.50 Go [2026-07-17 14:49:33] step 336200/1000000 | loss=2.2684 vocab=0.9780 token=1.2662 | 6.12 steps/s | VRAM=28.28 Go [2026-07-17 14:49:42] step 336250/1000000 | loss=2.2490 vocab=0.4823 token=1.2055 | 6.11 steps/s | VRAM=28.21 Go [2026-07-17 14:49:50] step 336300/1000000 | loss=2.3235 vocab=0.6622 token=1.3013 | 6.12 steps/s | VRAM=26.14 Go [2026-07-17 14:49:58] step 336350/1000000 | loss=2.3273 vocab=1.1885 token=1.1155 | 6.11 steps/s | VRAM=29.01 Go [2026-07-17 14:50:06] step 336400/1000000 | loss=2.3359 vocab=0.6925 token=1.2568 | 6.10 steps/s | VRAM=28.82 Go [2026-07-17 14:50:14] step 336450/1000000 | loss=2.3343 vocab=0.9126 token=1.3693 | 6.12 steps/s | VRAM=28.72 Go [2026-07-17 14:50:23] step 336500/1000000 | loss=2.3523 vocab=1.2752 token=1.6802 | 6.11 steps/s | VRAM=32.65 Go [2026-07-17 14:50:31] step 336550/1000000 | loss=2.2638 vocab=0.8108 token=1.5354 | 6.11 steps/s | VRAM=32.56 Go [2026-07-17 14:50:39] step 336600/1000000 | loss=2.2974 vocab=0.7258 token=1.2934 | 6.10 steps/s | VRAM=26.58 Go [2026-07-17 14:50:47] step 336650/1000000 | loss=2.2899 vocab=1.0258 token=1.2178 | 6.13 steps/s | VRAM=32.03 Go [2026-07-17 14:50:55] step 336700/1000000 | loss=2.2923 vocab=1.1596 token=1.3209 | 6.14 steps/s | VRAM=28.12 Go [2026-07-17 14:51:03] step 336750/1000000 | loss=2.2382 vocab=0.9676 token=1.4881 | 6.15 steps/s | VRAM=26.96 Go [2026-07-17 14:51:11] step 336800/1000000 | loss=2.3641 vocab=0.7667 token=1.5045 | 6.13 steps/s | VRAM=27.83 Go [2026-07-17 14:51:20] step 336850/1000000 | loss=2.2980 vocab=1.1773 token=1.7791 | 6.12 steps/s | VRAM=28.75 Go [2026-07-17 14:51:28] step 336900/1000000 | loss=2.1724 vocab=0.8697 token=1.3782 | 6.13 steps/s | VRAM=29.70 Go [2026-07-17 14:51:36] step 336950/1000000 | loss=2.2644 vocab=1.2000 token=1.3496 | 6.12 steps/s | VRAM=32.73 Go [2026-07-17 14:51:44] step 337000/1000000 | loss=2.1936 vocab=0.7570 token=1.1378 | 6.10 steps/s | VRAM=29.08 Go [2026-07-17 14:51:52] step 337050/1000000 | loss=2.2988 vocab=0.3958 token=1.2687 | 6.14 steps/s | VRAM=28.20 Go [2026-07-17 14:52:01] step 337100/1000000 | loss=2.2831 vocab=0.9394 token=1.3389 | 6.09 steps/s | VRAM=26.54 Go [2026-07-17 14:52:09] step 337150/1000000 | loss=2.3240 vocab=1.1256 token=1.2981 | 6.12 steps/s | VRAM=28.67 Go [2026-07-17 14:52:17] step 337200/1000000 | loss=2.2647 vocab=0.8719 token=1.4228 | 6.12 steps/s | VRAM=32.27 Go [2026-07-17 14:52:25] step 337250/1000000 | loss=2.4152 vocab=1.3221 token=1.6724 | 6.13 steps/s | VRAM=26.65 Go [2026-07-17 14:52:33] step 337300/1000000 | loss=2.3179 vocab=1.0325 token=1.3907 | 6.13 steps/s | VRAM=27.92 Go [2026-07-17 14:52:41] step 337350/1000000 | loss=2.3137 vocab=0.7657 token=1.3271 | 6.09 steps/s | VRAM=26.26 Go [2026-07-17 14:52:50] step 337400/1000000 | loss=2.3321 vocab=0.7216 token=1.3657 | 6.14 steps/s | VRAM=31.72 Go [2026-07-17 14:52:58] step 337450/1000000 | loss=2.2897 vocab=0.9606 token=1.5435 | 6.09 steps/s | VRAM=26.42 Go [2026-07-17 14:53:06] step 337500/1000000 | loss=2.3778 vocab=0.7626 token=1.1273 | 6.13 steps/s | VRAM=31.69 Go [2026-07-17 14:53:14] step 337550/1000000 | loss=2.2408 vocab=0.9202 token=1.4701 | 6.14 steps/s | VRAM=27.73 Go [2026-07-17 14:53:22] step 337600/1000000 | loss=2.3452 vocab=0.6921 token=1.5105 | 6.10 steps/s | VRAM=25.53 Go [2026-07-17 14:53:30] step 337650/1000000 | loss=2.3202 vocab=1.0604 token=1.4727 | 6.12 steps/s | VRAM=29.08 Go [2026-07-17 14:53:39] step 337700/1000000 | loss=2.3279 vocab=1.1997 token=1.4818 | 6.10 steps/s | VRAM=28.51 Go [2026-07-17 14:53:47] step 337750/1000000 | loss=2.3163 vocab=1.8089 token=1.4384 | 6.11 steps/s | VRAM=28.00 Go [2026-07-17 14:53:55] step 337800/1000000 | loss=2.2617 vocab=0.9676 token=1.4515 | 6.10 steps/s | VRAM=29.32 Go [2026-07-17 14:54:03] step 337850/1000000 | loss=2.2668 vocab=0.9695 token=1.3365 | 6.10 steps/s | VRAM=27.04 Go [2026-07-17 14:54:11] step 337900/1000000 | loss=2.3902 vocab=0.8593 token=1.2452 | 6.13 steps/s | VRAM=32.54 Go [2026-07-17 14:54:20] step 337950/1000000 | loss=2.2299 vocab=1.7235 token=1.3832 | 6.11 steps/s | VRAM=28.84 Go [2026-07-17 14:54:28] step 338000/1000000 | loss=2.4272 vocab=0.7685 token=1.1275 | 6.12 steps/s | VRAM=31.68 Go [2026-07-17 14:54:36] step 338050/1000000 | loss=2.3798 vocab=0.6992 token=0.9943 | 6.11 steps/s | VRAM=27.88 Go [2026-07-17 14:54:44] step 338100/1000000 | loss=2.3145 vocab=0.7415 token=1.3305 | 6.12 steps/s | VRAM=29.10 Go [2026-07-17 14:54:52] step 338150/1000000 | loss=2.3084 vocab=0.9570 token=1.3921 | 6.08 steps/s | VRAM=28.58 Go [2026-07-17 14:55:00] step 338200/1000000 | loss=2.3420 vocab=0.8596 token=1.6356 | 6.12 steps/s | VRAM=29.39 Go [2026-07-17 14:55:09] step 338250/1000000 | loss=2.2788 vocab=0.7537 token=1.3968 | 6.12 steps/s | VRAM=24.36 Go [2026-07-17 14:55:17] step 338300/1000000 | loss=2.2743 vocab=0.9377 token=1.1600 | 6.12 steps/s | VRAM=25.85 Go [2026-07-17 14:55:25] step 338350/1000000 | loss=2.3562 vocab=0.9863 token=1.3823 | 6.10 steps/s | VRAM=27.24 Go [2026-07-17 14:55:33] step 338400/1000000 | loss=2.2071 vocab=0.8217 token=1.2590 | 6.15 steps/s | VRAM=31.41 Go [2026-07-17 14:55:41] step 338450/1000000 | loss=2.3440 vocab=0.6412 token=1.2341 | 6.08 steps/s | VRAM=26.72 Go [2026-07-17 14:55:50] step 338500/1000000 | loss=2.3950 vocab=1.1156 token=1.4666 | 6.12 steps/s | VRAM=28.84 Go [2026-07-17 14:55:58] step 338550/1000000 | loss=2.2040 vocab=0.8882 token=1.5814 | 6.14 steps/s | VRAM=26.22 Go [2026-07-17 14:56:06] step 338600/1000000 | loss=2.2686 vocab=0.9900 token=1.4597 | 6.14 steps/s | VRAM=29.77 Go [2026-07-17 14:56:14] step 338650/1000000 | loss=2.2612 vocab=0.9023 token=1.2251 | 6.10 steps/s | VRAM=29.11 Go [2026-07-17 14:56:22] step 338700/1000000 | loss=2.2925 vocab=0.7417 token=1.5631 | 6.13 steps/s | VRAM=27.77 Go [2026-07-17 14:56:30] step 338750/1000000 | loss=2.2825 vocab=1.5841 token=1.8455 | 6.11 steps/s | VRAM=28.58 Go [2026-07-17 14:56:39] step 338800/1000000 | loss=2.2501 vocab=1.1598 token=1.4491 | 6.14 steps/s | VRAM=29.32 Go [2026-07-17 14:56:47] step 338850/1000000 | loss=2.2505 vocab=1.3849 token=1.4795 | 6.15 steps/s | VRAM=31.76 Go [2026-07-17 14:56:55] step 338900/1000000 | loss=2.2959 vocab=0.8266 token=1.2569 | 6.09 steps/s | VRAM=25.18 Go [2026-07-17 14:57:03] step 338950/1000000 | loss=2.3097 vocab=1.3147 token=1.5400 | 6.12 steps/s | VRAM=32.62 Go [2026-07-17 14:57:11] step 339000/1000000 | loss=2.2963 vocab=0.8739 token=1.2507 | 6.10 steps/s | VRAM=26.32 Go [2026-07-17 14:57:19] step 339050/1000000 | loss=2.4374 vocab=0.8846 token=1.1112 | 6.12 steps/s | VRAM=32.16 Go [2026-07-17 14:57:28] step 339100/1000000 | loss=2.2521 vocab=0.7469 token=1.1886 | 6.15 steps/s | VRAM=31.58 Go [2026-07-17 14:57:36] step 339150/1000000 | loss=2.2683 vocab=1.4676 token=1.3489 | 6.11 steps/s | VRAM=29.67 Go [2026-07-17 14:57:44] step 339200/1000000 | loss=2.3122 vocab=1.3995 token=1.5799 | 6.15 steps/s | VRAM=32.35 Go [2026-07-17 14:57:52] step 339250/1000000 | loss=2.2586 vocab=0.6855 token=1.1245 | 6.14 steps/s | VRAM=31.45 Go [2026-07-17 14:58:00] step 339300/1000000 | loss=2.1788 vocab=1.2036 token=1.5542 | 6.14 steps/s | VRAM=32.30 Go [2026-07-17 14:58:08] step 339350/1000000 | loss=2.3838 vocab=1.1784 token=1.5298 | 6.11 steps/s | VRAM=25.97 Go [2026-07-17 14:58:16] step 339400/1000000 | loss=2.3160 vocab=0.7162 token=1.3156 | 6.13 steps/s | VRAM=26.17 Go [2026-07-17 14:58:25] step 339450/1000000 | loss=2.2840 vocab=0.7452 token=1.2230 | 6.12 steps/s | VRAM=24.54 Go [2026-07-17 14:58:33] step 339500/1000000 | loss=2.2789 vocab=0.9822 token=1.2965 | 6.11 steps/s | VRAM=30.79 Go [2026-07-17 14:58:41] step 339550/1000000 | loss=2.3870 vocab=1.3015 token=1.5540 | 6.13 steps/s | VRAM=28.01 Go [2026-07-17 14:58:49] step 339600/1000000 | loss=2.3112 vocab=0.9133 token=1.3589 | 6.11 steps/s | VRAM=26.60 Go [2026-07-17 14:58:57] step 339650/1000000 | loss=2.3536 vocab=0.8301 token=1.3676 | 6.14 steps/s | VRAM=31.27 Go [2026-07-17 14:59:05] step 339700/1000000 | loss=2.2637 vocab=1.2133 token=1.4771 | 6.14 steps/s | VRAM=32.64 Go [2026-07-17 14:59:14] step 339750/1000000 | loss=2.2076 vocab=0.8608 token=1.3226 | 6.14 steps/s | VRAM=27.83 Go [2026-07-17 14:59:22] step 339800/1000000 | loss=2.2976 vocab=1.2234 token=1.2077 | 6.13 steps/s | VRAM=25.85 Go [2026-07-17 14:59:30] step 339850/1000000 | loss=2.2800 vocab=0.9970 token=1.6725 | 6.10 steps/s | VRAM=27.91 Go [2026-07-17 14:59:38] step 339900/1000000 | loss=2.3053 vocab=0.3894 token=1.2586 | 6.11 steps/s | VRAM=28.37 Go [2026-07-17 14:59:46] step 339950/1000000 | loss=2.2972 vocab=1.2225 token=1.4830 | 6.15 steps/s | VRAM=28.76 Go [2026-07-17 14:59:54] step 340000/1000000 | loss=2.3415 vocab=0.5205 token=1.4116 | 6.15 steps/s | VRAM=31.10 Go [2026-07-17 15:00:03] step 340050/1000000 | loss=2.2949 vocab=1.2008 token=1.6131 | 6.05 steps/s | VRAM=29.72 Go [2026-07-17 15:00:11] step 340100/1000000 | loss=2.3349 vocab=0.4075 token=1.1357 | 6.11 steps/s | VRAM=29.93 Go [2026-07-17 15:00:19] step 340150/1000000 | loss=2.3995 vocab=0.7026 token=1.4389 | 6.10 steps/s | VRAM=23.23 Go [2026-07-17 15:00:27] step 340200/1000000 | loss=2.2670 vocab=1.0904 token=1.7917 | 6.13 steps/s | VRAM=29.29 Go [2026-07-17 15:00:35] step 340250/1000000 | loss=2.4289 vocab=1.2155 token=1.5455 | 6.10 steps/s | VRAM=29.21 Go [2026-07-17 15:00:44] step 340300/1000000 | loss=2.2595 vocab=1.0221 token=1.0707 | 6.14 steps/s | VRAM=32.21 Go [2026-07-17 15:00:52] step 340350/1000000 | loss=2.2549 vocab=1.2320 token=1.4693 | 6.12 steps/s | VRAM=29.95 Go [2026-07-17 15:01:00] step 340400/1000000 | loss=2.3080 vocab=0.6424 token=0.9029 | 6.14 steps/s | VRAM=28.27 Go [2026-07-17 15:01:08] step 340450/1000000 | loss=2.4167 vocab=1.0360 token=1.6129 | 6.13 steps/s | VRAM=25.64 Go [2026-07-17 15:01:16] step 340500/1000000 | loss=2.2628 vocab=1.1689 token=1.3684 | 6.13 steps/s | VRAM=24.06 Go [2026-07-17 15:01:24] step 340550/1000000 | loss=2.1889 vocab=1.1289 token=1.1433 | 6.12 steps/s | VRAM=25.47 Go [2026-07-17 15:01:32] step 340600/1000000 | loss=2.2511 vocab=0.6039 token=1.1055 | 6.14 steps/s | VRAM=28.59 Go [2026-07-17 15:01:41] step 340650/1000000 | loss=2.3214 vocab=1.0342 token=1.5809 | 6.12 steps/s | VRAM=31.48 Go [2026-07-17 15:01:49] step 340700/1000000 | loss=2.3530 vocab=0.8528 token=1.6852 | 6.11 steps/s | VRAM=29.85 Go [2026-07-17 15:01:57] step 340750/1000000 | loss=2.2738 vocab=0.9318 token=1.7210 | 6.14 steps/s | VRAM=28.07 Go [2026-07-17 15:02:05] step 340800/1000000 | loss=2.3825 vocab=1.0899 token=2.0672 | 6.16 steps/s | VRAM=25.11 Go [2026-07-17 15:02:13] step 340850/1000000 | loss=2.3560 vocab=1.1133 token=1.6129 | 6.10 steps/s | VRAM=28.39 Go [2026-07-17 15:02:21] step 340900/1000000 | loss=2.2141 vocab=0.6932 token=1.3801 | 6.11 steps/s | VRAM=26.87 Go [2026-07-17 15:02:30] step 340950/1000000 | loss=2.3411 vocab=0.7539 token=1.3794 | 6.13 steps/s | VRAM=28.10 Go [2026-07-17 15:02:38] step 341000/1000000 | loss=2.2084 vocab=0.9064 token=1.2203 | 6.12 steps/s | VRAM=31.91 Go [2026-07-17 15:02:46] step 341050/1000000 | loss=2.3685 vocab=1.5608 token=1.8865 | 6.12 steps/s | VRAM=29.28 Go [2026-07-17 15:02:54] step 341100/1000000 | loss=2.2737 vocab=1.1151 token=1.1916 | 6.12 steps/s | VRAM=31.66 Go [2026-07-17 15:03:02] step 341150/1000000 | loss=2.2286 vocab=0.9014 token=1.0666 | 6.11 steps/s | VRAM=25.96 Go [2026-07-17 15:03:10] step 341200/1000000 | loss=2.2946 vocab=1.2200 token=1.3896 | 6.15 steps/s | VRAM=31.55 Go [2026-07-17 15:03:19] step 341250/1000000 | loss=2.4198 vocab=0.7918 token=1.2966 | 6.13 steps/s | VRAM=24.81 Go [2026-07-17 15:03:27] step 341300/1000000 | loss=2.2879 vocab=1.7708 token=1.7776 | 6.10 steps/s | VRAM=28.67 Go [2026-07-17 15:03:35] step 341350/1000000 | loss=2.4405 vocab=1.0323 token=1.5680 | 6.08 steps/s | VRAM=29.55 Go [2026-07-17 15:03:43] step 341400/1000000 | loss=2.3105 vocab=0.5561 token=1.3781 | 6.14 steps/s | VRAM=24.76 Go [2026-07-17 15:03:51] step 341450/1000000 | loss=2.2053 vocab=0.4106 token=0.9204 | 6.10 steps/s | VRAM=28.31 Go [2026-07-17 15:04:00] step 341500/1000000 | loss=2.3706 vocab=0.8805 token=1.5896 | 6.08 steps/s | VRAM=26.71 Go [2026-07-17 15:04:08] step 341550/1000000 | loss=2.3631 vocab=0.9436 token=1.3956 | 6.11 steps/s | VRAM=28.74 Go [2026-07-17 15:04:16] step 341600/1000000 | loss=2.2796 vocab=1.2520 token=1.3787 | 6.11 steps/s | VRAM=32.06 Go [2026-07-17 15:04:24] step 341650/1000000 | loss=2.3125 vocab=0.7954 token=1.4821 | 6.08 steps/s | VRAM=25.75 Go [2026-07-17 15:04:32] step 341700/1000000 | loss=2.2248 vocab=1.1272 token=1.3627 | 6.10 steps/s | VRAM=27.80 Go [2026-07-17 15:04:41] step 341750/1000000 | loss=2.3128 vocab=0.8615 token=1.4185 | 6.12 steps/s | VRAM=29.40 Go [2026-07-17 15:04:49] step 341800/1000000 | loss=2.2840 vocab=0.5487 token=1.4752 | 6.12 steps/s | VRAM=28.12 Go [2026-07-17 15:04:57] step 341850/1000000 | loss=2.3195 vocab=0.7376 token=1.6723 | 6.12 steps/s | VRAM=27.37 Go [2026-07-17 15:05:05] step 341900/1000000 | loss=2.3823 vocab=1.1652 token=1.6240 | 6.15 steps/s | VRAM=26.25 Go [2026-07-17 15:05:13] step 341950/1000000 | loss=2.2280 vocab=1.2546 token=1.3504 | 6.09 steps/s | VRAM=28.33 Go [2026-07-17 15:05:21] step 342000/1000000 | loss=2.3832 vocab=1.3532 token=1.2169 | 6.11 steps/s | VRAM=32.51 Go [2026-07-17 15:05:30] step 342050/1000000 | loss=2.2257 vocab=0.8324 token=1.4822 | 6.14 steps/s | VRAM=27.41 Go [2026-07-17 15:05:38] step 342100/1000000 | loss=2.2603 vocab=0.9521 token=1.1741 | 6.13 steps/s | VRAM=32.65 Go [2026-07-17 15:05:46] step 342150/1000000 | loss=2.3807 vocab=1.1176 token=1.5033 | 6.08 steps/s | VRAM=29.15 Go [2026-07-17 15:05:54] step 342200/1000000 | loss=2.0811 vocab=1.1088 token=1.1667 | 6.10 steps/s | VRAM=29.11 Go [2026-07-17 15:06:02] step 342250/1000000 | loss=2.3646 vocab=1.0373 token=1.3305 | 6.10 steps/s | VRAM=32.08 Go [2026-07-17 15:06:11] step 342300/1000000 | loss=2.2425 vocab=1.2936 token=1.7014 | 6.10 steps/s | VRAM=31.95 Go [2026-07-17 15:06:19] step 342350/1000000 | loss=2.3004 vocab=1.1974 token=1.6507 | 6.09 steps/s | VRAM=30.00 Go [2026-07-17 15:06:27] step 342400/1000000 | loss=2.3027 vocab=1.4175 token=1.3401 | 6.16 steps/s | VRAM=30.32 Go [2026-07-17 15:06:35] step 342450/1000000 | loss=2.2598 vocab=0.7422 token=1.2936 | 6.12 steps/s | VRAM=31.59 Go [2026-07-17 15:06:43] step 342500/1000000 | loss=2.2196 vocab=0.7767 token=1.3460 | 6.11 steps/s | VRAM=29.19 Go [2026-07-17 15:06:51] step 342550/1000000 | loss=2.3798 vocab=0.9741 token=1.5479 | 6.09 steps/s | VRAM=31.85 Go [2026-07-17 15:07:00] step 342600/1000000 | loss=2.2851 vocab=0.5524 token=1.1741 | 6.10 steps/s | VRAM=31.85 Go [2026-07-17 15:07:08] step 342650/1000000 | loss=2.2320 vocab=0.8147 token=0.6957 | 6.14 steps/s | VRAM=32.18 Go [2026-07-17 15:07:16] step 342700/1000000 | loss=2.2569 vocab=1.0153 token=1.6560 | 6.08 steps/s | VRAM=29.06 Go [2026-07-17 15:07:24] step 342750/1000000 | loss=2.1654 vocab=0.4575 token=1.1381 | 6.10 steps/s | VRAM=29.25 Go [2026-07-17 15:07:32] step 342800/1000000 | loss=2.2445 vocab=0.5865 token=1.0694 | 6.09 steps/s | VRAM=32.88 Go [2026-07-17 15:07:41] step 342850/1000000 | loss=2.2871 vocab=0.6703 token=1.2370 | 6.14 steps/s | VRAM=28.20 Go [2026-07-17 15:07:49] step 342900/1000000 | loss=2.3030 vocab=1.5093 token=1.5567 | 6.11 steps/s | VRAM=24.90 Go [2026-07-17 15:07:57] step 342950/1000000 | loss=2.2673 vocab=1.0488 token=1.4162 | 6.10 steps/s | VRAM=29.66 Go [2026-07-17 15:08:05] step 343000/1000000 | loss=2.2777 vocab=1.2246 token=1.2274 | 6.11 steps/s | VRAM=29.06 Go [2026-07-17 15:08:13] step 343050/1000000 | loss=2.2923 vocab=0.8699 token=1.3342 | 6.13 steps/s | VRAM=31.95 Go [2026-07-17 15:08:21] step 343100/1000000 | loss=2.3741 vocab=0.4827 token=1.1058 | 6.09 steps/s | VRAM=32.35 Go [2026-07-17 15:08:30] step 343150/1000000 | loss=2.3443 vocab=1.1010 token=1.6841 | 6.10 steps/s | VRAM=32.62 Go [2026-07-17 15:08:38] step 343200/1000000 | loss=2.3009 vocab=1.0479 token=1.1299 | 6.06 steps/s | VRAM=28.74 Go [2026-07-17 15:08:46] step 343250/1000000 | loss=2.3293 vocab=0.8905 token=1.1467 | 6.09 steps/s | VRAM=26.06 Go [2026-07-17 15:08:54] step 343300/1000000 | loss=2.2804 vocab=1.5060 token=1.3080 | 6.08 steps/s | VRAM=28.64 Go [2026-07-17 15:09:03] step 343350/1000000 | loss=2.3701 vocab=1.2331 token=1.4672 | 6.10 steps/s | VRAM=28.14 Go [2026-07-17 15:09:11] step 343400/1000000 | loss=2.4385 vocab=0.8936 token=1.4222 | 6.11 steps/s | VRAM=28.27 Go [2026-07-17 15:09:19] step 343450/1000000 | loss=2.2627 vocab=0.8278 token=1.4796 | 6.08 steps/s | VRAM=26.22 Go [2026-07-17 15:09:27] step 343500/1000000 | loss=2.3318 vocab=1.0079 token=1.7083 | 6.10 steps/s | VRAM=27.99 Go [2026-07-17 15:09:35] step 343550/1000000 | loss=2.2832 vocab=1.1158 token=1.0160 | 6.10 steps/s | VRAM=28.39 Go [2026-07-17 15:09:43] step 343600/1000000 | loss=2.3447 vocab=0.6036 token=1.2081 | 6.14 steps/s | VRAM=28.08 Go [2026-07-17 15:09:52] step 343650/1000000 | loss=2.3466 vocab=1.0636 token=1.2796 | 6.12 steps/s | VRAM=29.09 Go [2026-07-17 15:10:00] step 343700/1000000 | loss=2.2687 vocab=1.2487 token=1.5983 | 6.15 steps/s | VRAM=26.14 Go [2026-07-17 15:10:08] step 343750/1000000 | loss=2.3691 vocab=0.9652 token=1.5275 | 6.11 steps/s | VRAM=31.59 Go [2026-07-17 15:10:16] step 343800/1000000 | loss=2.3111 vocab=0.3308 token=0.9642 | 6.08 steps/s | VRAM=29.89 Go [2026-07-17 15:10:24] step 343850/1000000 | loss=2.2894 vocab=0.8450 token=1.1939 | 6.13 steps/s | VRAM=27.90 Go [2026-07-17 15:10:33] step 343900/1000000 | loss=2.2889 vocab=0.9894 token=1.1474 | 6.12 steps/s | VRAM=32.19 Go [2026-07-17 15:10:41] step 343950/1000000 | loss=2.3078 vocab=0.7279 token=1.4713 | 6.10 steps/s | VRAM=26.18 Go [2026-07-17 15:10:49] step 344000/1000000 | loss=2.3109 vocab=0.9922 token=1.3302 | 6.11 steps/s | VRAM=28.49 Go [2026-07-17 15:10:57] step 344050/1000000 | loss=2.3389 vocab=1.0845 token=1.4950 | 6.11 steps/s | VRAM=28.83 Go [2026-07-17 15:11:05] step 344100/1000000 | loss=2.2177 vocab=0.9233 token=1.3658 | 6.14 steps/s | VRAM=27.52 Go [2026-07-17 15:11:13] step 344150/1000000 | loss=2.2027 vocab=1.0023 token=1.3000 | 6.15 steps/s | VRAM=32.85 Go [2026-07-17 15:11:22] step 344200/1000000 | loss=2.3181 vocab=1.1725 token=1.3193 | 6.11 steps/s | VRAM=26.35 Go [2026-07-17 15:11:30] step 344250/1000000 | loss=2.2919 vocab=0.9028 token=1.4951 | 6.12 steps/s | VRAM=31.74 Go [2026-07-17 15:11:38] step 344300/1000000 | loss=2.2366 vocab=1.4578 token=1.6407 | 6.14 steps/s | VRAM=28.42 Go [2026-07-17 15:11:46] step 344350/1000000 | loss=2.2515 vocab=0.7860 token=1.0707 | 6.11 steps/s | VRAM=29.54 Go [2026-07-17 15:11:54] step 344400/1000000 | loss=2.1750 vocab=0.7145 token=1.1051 | 6.11 steps/s | VRAM=29.88 Go [2026-07-17 15:12:02] step 344450/1000000 | loss=2.4395 vocab=1.1977 token=1.4900 | 6.15 steps/s | VRAM=31.73 Go [2026-07-17 15:12:10] step 344500/1000000 | loss=2.3198 vocab=1.5842 token=1.6185 | 6.16 steps/s | VRAM=32.12 Go [2026-07-17 15:12:19] step 344550/1000000 | loss=2.3273 vocab=0.8449 token=1.4365 | 6.12 steps/s | VRAM=29.03 Go [2026-07-17 15:12:27] step 344600/1000000 | loss=2.2450 vocab=0.3667 token=0.8957 | 6.12 steps/s | VRAM=29.43 Go [2026-07-17 15:12:35] step 344650/1000000 | loss=2.3236 vocab=1.3055 token=1.2636 | 6.16 steps/s | VRAM=31.36 Go [2026-07-17 15:12:43] step 344700/1000000 | loss=2.3607 vocab=1.0546 token=1.3400 | 6.12 steps/s | VRAM=30.00 Go [2026-07-17 15:12:51] step 344750/1000000 | loss=2.2786 vocab=0.7869 token=1.4970 | 6.12 steps/s | VRAM=28.55 Go [2026-07-17 15:12:59] step 344800/1000000 | loss=2.3122 vocab=1.1670 token=1.3814 | 6.14 steps/s | VRAM=25.27 Go [2026-07-17 15:13:08] step 344850/1000000 | loss=2.3604 vocab=0.7632 token=1.6687 | 6.15 steps/s | VRAM=25.11 Go [2026-07-17 15:13:16] step 344900/1000000 | loss=2.2346 vocab=0.5595 token=1.0817 | 6.13 steps/s | VRAM=32.22 Go [2026-07-17 15:13:24] step 344950/1000000 | loss=2.2526 vocab=0.8017 token=1.2447 | 6.17 steps/s | VRAM=32.39 Go [2026-07-17 15:13:32] step 345000/1000000 | loss=2.2146 vocab=0.9584 token=1.3247 | 6.13 steps/s | VRAM=29.50 Go [2026-07-17 15:13:40] step 345050/1000000 | loss=2.2733 vocab=1.0622 token=1.2102 | 6.13 steps/s | VRAM=28.98 Go [2026-07-17 15:13:48] step 345100/1000000 | loss=2.2916 vocab=0.6618 token=1.2792 | 6.14 steps/s | VRAM=26.59 Go [2026-07-17 15:13:56] step 345150/1000000 | loss=2.3196 vocab=1.0261 token=1.5445 | 6.16 steps/s | VRAM=27.73 Go [2026-07-17 15:14:05] step 345200/1000000 | loss=2.4048 vocab=1.2954 token=1.8001 | 6.13 steps/s | VRAM=28.68 Go [2026-07-17 15:14:13] step 345250/1000000 | loss=2.1993 vocab=0.6954 token=1.4631 | 6.13 steps/s | VRAM=26.43 Go [2026-07-17 15:14:21] step 345300/1000000 | loss=2.2291 vocab=0.6615 token=1.0311 | 6.14 steps/s | VRAM=32.67 Go [2026-07-17 15:14:29] step 345350/1000000 | loss=2.3173 vocab=0.5040 token=1.1849 | 6.13 steps/s | VRAM=29.40 Go [2026-07-17 15:14:37] step 345400/1000000 | loss=2.3350 vocab=1.3495 token=1.5640 | 6.16 steps/s | VRAM=32.16 Go [2026-07-17 15:14:45] step 345450/1000000 | loss=2.4003 vocab=1.1504 token=1.2031 | 6.13 steps/s | VRAM=32.79 Go [2026-07-17 15:14:53] step 345500/1000000 | loss=2.3495 vocab=0.8138 token=1.3031 | 6.13 steps/s | VRAM=29.36 Go [2026-07-17 15:15:02] step 345550/1000000 | loss=2.2784 vocab=0.7773 token=1.2673 | 6.15 steps/s | VRAM=32.97 Go [2026-07-17 15:15:10] step 345600/1000000 | loss=2.2749 vocab=0.7038 token=1.2639 | 6.13 steps/s | VRAM=31.43 Go [2026-07-17 15:15:18] step 345650/1000000 | loss=2.2526 vocab=1.1625 token=1.2862 | 6.11 steps/s | VRAM=26.16 Go [2026-07-17 15:15:26] step 345700/1000000 | loss=2.2914 vocab=1.3706 token=1.7125 | 6.12 steps/s | VRAM=28.88 Go [2026-07-17 15:15:34] step 345750/1000000 | loss=2.1614 vocab=0.6949 token=1.6661 | 6.14 steps/s | VRAM=29.93 Go [2026-07-17 15:15:42] step 345800/1000000 | loss=2.3067 vocab=0.6087 token=1.0460 | 6.15 steps/s | VRAM=31.48 Go [2026-07-17 15:15:50] step 345850/1000000 | loss=2.2528 vocab=0.9681 token=1.2477 | 6.13 steps/s | VRAM=28.41 Go [2026-07-17 15:15:59] step 345900/1000000 | loss=2.2349 vocab=0.8430 token=1.2862 | 6.12 steps/s | VRAM=28.62 Go [2026-07-17 15:16:07] step 345950/1000000 | loss=2.2272 vocab=0.6902 token=1.2989 | 6.14 steps/s | VRAM=26.07 Go [2026-07-17 15:16:15] step 346000/1000000 | loss=2.2621 vocab=0.9881 token=1.3141 | 6.12 steps/s | VRAM=28.87 Go [2026-07-17 15:16:23] step 346050/1000000 | loss=2.3173 vocab=0.5205 token=1.4438 | 6.10 steps/s | VRAM=29.05 Go [2026-07-17 15:16:31] step 346100/1000000 | loss=2.2481 vocab=1.0493 token=1.5686 | 6.14 steps/s | VRAM=32.44 Go [2026-07-17 15:16:39] step 346150/1000000 | loss=2.3179 vocab=0.8147 token=1.3072 | 6.14 steps/s | VRAM=28.76 Go [2026-07-17 15:16:48] step 346200/1000000 | loss=2.2278 vocab=0.8561 token=1.1343 | 6.11 steps/s | VRAM=32.88 Go [2026-07-17 15:16:56] step 346250/1000000 | loss=2.3559 vocab=1.4299 token=1.5790 | 6.13 steps/s | VRAM=25.51 Go [2026-07-17 15:17:04] step 346300/1000000 | loss=2.2203 vocab=1.2230 token=1.6820 | 6.14 steps/s | VRAM=27.96 Go [2026-07-17 15:17:12] step 346350/1000000 | loss=2.3636 vocab=0.6399 token=1.4487 | 6.15 steps/s | VRAM=29.12 Go [2026-07-17 15:17:20] step 346400/1000000 | loss=2.3436 vocab=1.7796 token=1.2777 | 6.09 steps/s | VRAM=26.34 Go [2026-07-17 15:17:28] step 346450/1000000 | loss=2.2592 vocab=1.4368 token=1.9420 | 6.15 steps/s | VRAM=31.25 Go [2026-07-17 15:17:37] step 346500/1000000 | loss=2.3539 vocab=1.2283 token=1.5565 | 6.11 steps/s | VRAM=28.55 Go [2026-07-17 15:17:45] step 346550/1000000 | loss=2.2712 vocab=0.5139 token=1.4196 | 6.15 steps/s | VRAM=28.35 Go [2026-07-17 15:17:53] step 346600/1000000 | loss=2.2984 vocab=0.5456 token=1.3783 | 6.14 steps/s | VRAM=29.43 Go [2026-07-17 15:18:01] step 346650/1000000 | loss=2.2214 vocab=0.8659 token=1.3514 | 6.14 steps/s | VRAM=29.02 Go [2026-07-17 15:18:09] step 346700/1000000 | loss=2.3144 vocab=1.2119 token=1.2359 | 6.15 steps/s | VRAM=32.59 Go [2026-07-17 15:18:17] step 346750/1000000 | loss=2.3565 vocab=0.7421 token=1.3041 | 6.08 steps/s | VRAM=30.19 Go [2026-07-17 15:18:26] step 346800/1000000 | loss=2.3674 vocab=1.2116 token=1.3265 | 6.11 steps/s | VRAM=29.15 Go [2026-07-17 15:18:34] step 346850/1000000 | loss=2.3960 vocab=1.1403 token=1.6584 | 6.12 steps/s | VRAM=28.49 Go [2026-07-17 15:18:42] step 346900/1000000 | loss=2.3279 vocab=0.8543 token=1.3071 | 6.12 steps/s | VRAM=28.69 Go [2026-07-17 15:18:50] step 346950/1000000 | loss=2.2631 vocab=1.0268 token=1.2641 | 6.11 steps/s | VRAM=29.34 Go [2026-07-17 15:18:58] step 347000/1000000 | loss=2.2817 vocab=0.6806 token=1.4064 | 6.10 steps/s | VRAM=28.98 Go [2026-07-17 15:19:06] step 347050/1000000 | loss=2.3591 vocab=0.8809 token=1.1981 | 6.14 steps/s | VRAM=27.29 Go [2026-07-17 15:19:15] step 347100/1000000 | loss=2.3641 vocab=1.3640 token=1.6922 | 6.12 steps/s | VRAM=28.36 Go [2026-07-17 15:19:23] step 347150/1000000 | loss=2.4028 vocab=1.0397 token=1.4695 | 6.11 steps/s | VRAM=29.71 Go [2026-07-17 15:19:31] step 347200/1000000 | loss=2.3073 vocab=0.8920 token=1.5109 | 6.15 steps/s | VRAM=23.99 Go [2026-07-17 15:19:39] step 347250/1000000 | loss=2.3209 vocab=1.3188 token=1.7000 | 6.14 steps/s | VRAM=31.59 Go [2026-07-17 15:19:47] step 347300/1000000 | loss=2.3046 vocab=1.3119 token=1.4383 | 6.09 steps/s | VRAM=23.67 Go [2026-07-17 15:19:55] step 347350/1000000 | loss=2.3127 vocab=0.7996 token=1.4432 | 6.15 steps/s | VRAM=31.56 Go [2026-07-17 15:20:04] step 347400/1000000 | loss=2.2374 vocab=0.6972 token=1.5023 | 6.12 steps/s | VRAM=25.60 Go [2026-07-17 15:20:12] step 347450/1000000 | loss=2.2682 vocab=1.3246 token=1.1771 | 6.11 steps/s | VRAM=32.43 Go [2026-07-17 15:20:20] step 347500/1000000 | loss=2.3481 vocab=0.9261 token=1.4358 | 6.12 steps/s | VRAM=28.76 Go [2026-07-17 15:20:28] step 347550/1000000 | loss=2.2652 vocab=1.4365 token=1.7060 | 6.15 steps/s | VRAM=28.79 Go [2026-07-17 15:20:36] step 347600/1000000 | loss=2.3037 vocab=1.2503 token=1.4661 | 6.13 steps/s | VRAM=28.55 Go [2026-07-17 15:20:44] step 347650/1000000 | loss=2.3722 vocab=0.7814 token=1.5652 | 6.12 steps/s | VRAM=25.30 Go [2026-07-17 15:20:53] step 347700/1000000 | loss=2.2580 vocab=1.2040 token=1.3046 | 6.12 steps/s | VRAM=32.59 Go [2026-07-17 15:21:01] step 347750/1000000 | loss=2.2726 vocab=0.9155 token=1.3652 | 6.09 steps/s | VRAM=26.30 Go [2026-07-17 15:21:09] step 347800/1000000 | loss=2.2750 vocab=0.9952 token=1.3570 | 6.13 steps/s | VRAM=28.97 Go [2026-07-17 15:21:17] step 347850/1000000 | loss=2.3683 vocab=1.2162 token=0.8825 | 6.08 steps/s | VRAM=24.00 Go [2026-07-17 15:21:25] step 347900/1000000 | loss=2.3083 vocab=0.8341 token=1.2426 | 6.14 steps/s | VRAM=27.69 Go [2026-07-17 15:21:33] step 347950/1000000 | loss=2.3846 vocab=1.3235 token=1.5181 | 6.13 steps/s | VRAM=29.24 Go [2026-07-17 15:21:42] step 348000/1000000 | loss=2.2894 vocab=0.9543 token=1.2856 | 6.07 steps/s | VRAM=26.56 Go [2026-07-17 15:21:50] step 348050/1000000 | loss=2.2724 vocab=1.2410 token=1.5478 | 6.16 steps/s | VRAM=31.67 Go [2026-07-17 15:21:58] step 348100/1000000 | loss=2.3758 vocab=1.0906 token=1.3876 | 6.14 steps/s | VRAM=31.12 Go [2026-07-17 15:22:06] step 348150/1000000 | loss=2.2399 vocab=1.0640 token=1.3639 | 6.11 steps/s | VRAM=26.15 Go [2026-07-17 15:22:14] step 348200/1000000 | loss=2.3645 vocab=1.0882 token=1.6026 | 6.15 steps/s | VRAM=31.74 Go [2026-07-17 15:22:22] step 348250/1000000 | loss=2.2496 vocab=1.0341 token=1.3851 | 6.11 steps/s | VRAM=28.03 Go [2026-07-17 15:22:31] step 348300/1000000 | loss=2.2384 vocab=0.6960 token=1.0277 | 6.14 steps/s | VRAM=32.11 Go [2026-07-17 15:22:39] step 348350/1000000 | loss=2.2314 vocab=1.1343 token=1.5287 | 6.13 steps/s | VRAM=32.19 Go [2026-07-17 15:22:47] step 348400/1000000 | loss=2.3362 vocab=0.7857 token=1.3252 | 6.11 steps/s | VRAM=32.44 Go [2026-07-17 15:22:55] step 348450/1000000 | loss=2.2894 vocab=1.2776 token=1.5562 | 6.11 steps/s | VRAM=28.94 Go [2026-07-17 15:23:03] step 348500/1000000 | loss=2.3618 vocab=1.4676 token=1.7601 | 6.12 steps/s | VRAM=25.35 Go [2026-07-17 15:23:11] step 348550/1000000 | loss=2.2475 vocab=0.9004 token=1.3432 | 6.15 steps/s | VRAM=31.62 Go [2026-07-17 15:23:20] step 348600/1000000 | loss=2.2904 vocab=0.4445 token=1.1069 | 6.12 steps/s | VRAM=32.42 Go [2026-07-17 15:23:28] step 348650/1000000 | loss=2.2296 vocab=1.0221 token=1.5876 | 6.13 steps/s | VRAM=28.14 Go [2026-07-17 15:23:36] step 348700/1000000 | loss=2.3559 vocab=0.7237 token=1.3424 | 6.11 steps/s | VRAM=23.35 Go [2026-07-17 15:23:44] step 348750/1000000 | loss=2.3574 vocab=1.1082 token=1.5823 | 6.11 steps/s | VRAM=26.60 Go [2026-07-17 15:23:52] step 348800/1000000 | loss=2.1967 vocab=0.6883 token=1.1410 | 6.15 steps/s | VRAM=32.16 Go [2026-07-17 15:24:00] step 348850/1000000 | loss=2.2245 vocab=1.0470 token=1.3978 | 6.11 steps/s | VRAM=25.42 Go [2026-07-17 15:24:09] step 348900/1000000 | loss=2.3491 vocab=1.3663 token=1.4685 | 6.13 steps/s | VRAM=27.98 Go [2026-07-17 15:24:17] step 348950/1000000 | loss=2.3019 vocab=1.2901 token=1.4268 | 6.12 steps/s | VRAM=28.89 Go [2026-07-17 15:24:25] step 349000/1000000 | loss=2.3986 vocab=0.8191 token=1.1845 | 6.11 steps/s | VRAM=23.45 Go [2026-07-17 15:24:33] step 349050/1000000 | loss=2.3255 vocab=1.2996 token=1.4711 | 6.16 steps/s | VRAM=31.48 Go [2026-07-17 15:24:41] step 349100/1000000 | loss=2.3175 vocab=1.1087 token=1.3436 | 6.13 steps/s | VRAM=26.07 Go [2026-07-17 15:24:49] step 349150/1000000 | loss=2.3109 vocab=0.8462 token=1.0919 | 6.11 steps/s | VRAM=26.86 Go [2026-07-17 15:24:57] step 349200/1000000 | loss=2.4137 vocab=0.8955 token=1.2375 | 6.15 steps/s | VRAM=29.88 Go [2026-07-17 15:25:06] step 349250/1000000 | loss=2.2789 vocab=0.9423 token=1.2477 | 6.16 steps/s | VRAM=24.99 Go [2026-07-17 15:25:14] step 349300/1000000 | loss=2.2749 vocab=0.6122 token=1.2454 | 6.16 steps/s | VRAM=24.99 Go [2026-07-17 15:25:22] step 349350/1000000 | loss=2.3517 vocab=0.7701 token=1.3348 | 6.13 steps/s | VRAM=28.91 Go [2026-07-17 15:25:30] step 349400/1000000 | loss=2.2752 vocab=0.7063 token=1.4136 | 6.10 steps/s | VRAM=28.28 Go [2026-07-17 15:25:38] step 349450/1000000 | loss=2.3052 vocab=0.7742 token=1.4770 | 6.14 steps/s | VRAM=32.38 Go [2026-07-17 15:25:46] step 349500/1000000 | loss=2.1999 vocab=1.0897 token=1.3329 | 6.13 steps/s | VRAM=32.46 Go [2026-07-17 15:25:55] step 349550/1000000 | loss=2.1490 vocab=1.4478 token=1.9078 | 6.12 steps/s | VRAM=28.93 Go [2026-07-17 15:26:03] step 349600/1000000 | loss=2.2414 vocab=0.8697 token=1.4533 | 6.14 steps/s | VRAM=28.79 Go [2026-07-17 15:26:11] step 349650/1000000 | loss=2.2610 vocab=1.1104 token=1.5465 | 6.10 steps/s | VRAM=26.86 Go [2026-07-17 15:26:19] step 349700/1000000 | loss=2.2830 vocab=0.9678 token=1.2941 | 6.14 steps/s | VRAM=31.50 Go [2026-07-17 15:26:27] step 349750/1000000 | loss=2.3722 vocab=1.0742 token=1.3105 | 6.16 steps/s | VRAM=31.88 Go [2026-07-17 15:26:35] step 349800/1000000 | loss=2.2813 vocab=1.1427 token=1.4198 | 6.15 steps/s | VRAM=29.48 Go [2026-07-17 15:26:43] step 349850/1000000 | loss=2.3106 vocab=1.3813 token=1.5577 | 6.16 steps/s | VRAM=30.67 Go [2026-07-17 15:26:52] step 349900/1000000 | loss=2.2933 vocab=0.9085 token=1.1454 | 6.15 steps/s | VRAM=26.08 Go [2026-07-17 15:27:00] step 349950/1000000 | loss=2.3231 vocab=0.9220 token=1.3286 | 6.15 steps/s | VRAM=28.25 Go [2026-07-17 15:27:08] step 350000/1000000 | loss=2.1535 vocab=0.7811 token=1.0593 | 6.12 steps/s | VRAM=29.33 Go [push] step 350000 OK (best-of-2, loss_token=1.0593) [2026-07-17 15:28:56] step 350050/1000000 | loss=2.2679 vocab=1.0570 token=1.5811 | 0.46 steps/s | VRAM=28.91 Go [2026-07-17 15:29:04] step 350100/1000000 | loss=2.2099 vocab=1.1242 token=1.4475 | 6.19 steps/s | VRAM=28.53 Go [2026-07-17 15:29:12] step 350150/1000000 | loss=2.3082 vocab=1.2887 token=1.4833 | 6.18 steps/s | VRAM=28.53 Go [2026-07-17 15:29:20] step 350200/1000000 | loss=2.2672 vocab=1.4169 token=1.8687 | 6.19 steps/s | VRAM=31.72 Go [2026-07-17 15:29:28] step 350250/1000000 | loss=2.2288 vocab=0.8335 token=1.2055 | 6.18 steps/s | VRAM=28.67 Go [2026-07-17 15:29:36] step 350300/1000000 | loss=2.3070 vocab=0.8269 token=1.0814 | 6.17 steps/s | VRAM=32.81 Go [2026-07-17 15:29:44] step 350350/1000000 | loss=2.2935 vocab=0.8407 token=1.4372 | 6.15 steps/s | VRAM=30.31 Go [2026-07-17 15:29:52] step 350400/1000000 | loss=2.2077 vocab=0.7093 token=1.2547 | 6.18 steps/s | VRAM=32.16 Go [2026-07-17 15:30:00] step 350450/1000000 | loss=2.2612 vocab=0.7539 token=1.0281 | 6.17 steps/s | VRAM=24.58 Go [2026-07-17 15:30:09] step 350500/1000000 | loss=2.3156 vocab=1.1731 token=1.5903 | 6.13 steps/s | VRAM=28.99 Go [2026-07-17 15:30:17] step 350550/1000000 | loss=2.3774 vocab=1.5072 token=1.3997 | 6.14 steps/s | VRAM=28.55 Go [2026-07-17 15:30:25] step 350600/1000000 | loss=2.3235 vocab=0.5944 token=1.3037 | 6.11 steps/s | VRAM=26.72 Go [2026-07-17 15:30:33] step 350650/1000000 | loss=2.2840 vocab=0.5695 token=1.2318 | 6.15 steps/s | VRAM=29.24 Go [2026-07-17 15:30:41] step 350700/1000000 | loss=2.2684 vocab=0.9693 token=1.4325 | 6.16 steps/s | VRAM=31.29 Go [2026-07-17 15:30:49] step 350750/1000000 | loss=2.2743 vocab=1.5002 token=1.2213 | 6.11 steps/s | VRAM=28.83 Go [2026-07-17 15:30:57] step 350800/1000000 | loss=2.3551 vocab=1.2788 token=1.2310 | 6.08 steps/s | VRAM=27.16 Go [2026-07-17 15:31:06] step 350850/1000000 | loss=2.2112 vocab=1.0747 token=1.8969 | 6.13 steps/s | VRAM=31.32 Go [2026-07-17 15:31:14] step 350900/1000000 | loss=2.2561 vocab=0.6126 token=1.2207 | 6.13 steps/s | VRAM=25.12 Go [2026-07-17 15:31:22] step 350950/1000000 | loss=2.2935 vocab=0.9665 token=1.4336 | 6.11 steps/s | VRAM=28.35 Go [2026-07-17 15:31:30] step 351000/1000000 | loss=2.1861 vocab=0.7849 token=1.4892 | 6.15 steps/s | VRAM=27.35 Go [2026-07-17 15:31:38] step 351050/1000000 | loss=2.1826 vocab=0.4945 token=1.1561 | 6.11 steps/s | VRAM=28.47 Go [2026-07-17 15:31:46] step 351100/1000000 | loss=2.2988 vocab=0.7834 token=1.1536 | 6.14 steps/s | VRAM=29.27 Go [2026-07-17 15:31:55] step 351150/1000000 | loss=2.2151 vocab=0.5911 token=1.2397 | 6.09 steps/s | VRAM=28.92 Go [2026-07-17 15:32:03] step 351200/1000000 | loss=2.2839 vocab=0.6536 token=1.5701 | 6.12 steps/s | VRAM=27.98 Go [2026-07-17 15:32:11] step 351250/1000000 | loss=2.3132 vocab=0.8402 token=1.1153 | 6.10 steps/s | VRAM=26.33 Go [2026-07-17 15:32:19] step 351300/1000000 | loss=2.3362 vocab=0.7167 token=1.3510 | 6.11 steps/s | VRAM=24.71 Go [2026-07-17 15:32:27] step 351350/1000000 | loss=2.2657 vocab=0.9583 token=1.5375 | 6.13 steps/s | VRAM=28.79 Go [2026-07-17 15:32:36] step 351400/1000000 | loss=2.2567 vocab=1.3556 token=1.5522 | 6.10 steps/s | VRAM=29.04 Go [2026-07-17 15:32:44] step 351450/1000000 | loss=2.3000 vocab=0.9637 token=1.2308 | 6.16 steps/s | VRAM=31.73 Go [2026-07-17 15:32:52] step 351500/1000000 | loss=2.2915 vocab=0.9383 token=1.4398 | 6.10 steps/s | VRAM=28.31 Go [2026-07-17 15:33:00] step 351550/1000000 | loss=2.2778 vocab=0.4500 token=0.9453 | 6.13 steps/s | VRAM=32.03 Go [2026-07-17 15:33:08] step 351600/1000000 | loss=2.2445 vocab=1.3854 token=1.4887 | 6.15 steps/s | VRAM=28.87 Go [2026-07-17 15:33:16] step 351650/1000000 | loss=2.2571 vocab=0.8081 token=1.5369 | 6.11 steps/s | VRAM=28.98 Go [2026-07-17 15:33:25] step 351700/1000000 | loss=2.2633 vocab=0.8715 token=1.3474 | 6.11 steps/s | VRAM=25.39 Go [2026-07-17 15:33:33] step 351750/1000000 | loss=2.3411 vocab=1.9082 token=1.6843 | 6.14 steps/s | VRAM=25.59 Go [2026-07-17 15:33:41] step 351800/1000000 | loss=2.3312 vocab=0.8483 token=1.1716 | 6.14 steps/s | VRAM=29.03 Go [2026-07-17 15:33:49] step 351850/1000000 | loss=2.2607 vocab=1.0122 token=1.3743 | 6.12 steps/s | VRAM=28.59 Go [2026-07-17 15:33:57] step 351900/1000000 | loss=2.2358 vocab=0.4795 token=0.8667 | 6.11 steps/s | VRAM=31.72 Go [2026-07-17 15:34:05] step 351950/1000000 | loss=2.2382 vocab=0.7607 token=1.5327 | 6.11 steps/s | VRAM=24.15 Go [2026-07-17 15:34:14] step 352000/1000000 | loss=2.4066 vocab=1.2092 token=1.2977 | 6.11 steps/s | VRAM=25.90 Go [2026-07-17 15:34:22] step 352050/1000000 | loss=2.2286 vocab=1.1775 token=1.5602 | 6.14 steps/s | VRAM=31.95 Go [2026-07-17 15:34:30] step 352100/1000000 | loss=2.2671 vocab=1.0232 token=1.1838 | 6.13 steps/s | VRAM=31.48 Go [2026-07-17 15:34:38] step 352150/1000000 | loss=2.3343 vocab=0.6095 token=1.3621 | 6.13 steps/s | VRAM=32.13 Go [2026-07-17 15:34:46] step 352200/1000000 | loss=2.3392 vocab=1.3918 token=1.1299 | 6.11 steps/s | VRAM=25.73 Go [2026-07-17 15:34:54] step 352250/1000000 | loss=2.1938 vocab=1.2214 token=1.7940 | 6.11 steps/s | VRAM=26.16 Go [2026-07-17 15:35:03] step 352300/1000000 | loss=2.2549 vocab=1.1749 token=1.2316 | 6.14 steps/s | VRAM=28.51 Go [2026-07-17 15:35:11] step 352350/1000000 | loss=2.2430 vocab=0.7463 token=1.2735 | 6.13 steps/s | VRAM=27.19 Go [2026-07-17 15:35:19] step 352400/1000000 | loss=2.2139 vocab=1.2283 token=1.8245 | 6.15 steps/s | VRAM=25.17 Go [2026-07-17 15:35:27] step 352450/1000000 | loss=2.3273 vocab=0.7467 token=1.5970 | 6.14 steps/s | VRAM=28.49 Go [2026-07-17 15:35:35] step 352500/1000000 | loss=2.2789 vocab=0.5993 token=1.2689 | 6.14 steps/s | VRAM=29.15 Go [2026-07-17 15:35:43] step 352550/1000000 | loss=2.3312 vocab=1.4137 token=1.3548 | 6.16 steps/s | VRAM=32.13 Go [2026-07-17 15:35:51] step 352600/1000000 | loss=2.3054 vocab=0.6958 token=1.2272 | 6.13 steps/s | VRAM=29.30 Go [2026-07-17 15:36:00] step 352650/1000000 | loss=2.3498 vocab=1.1817 token=1.3143 | 6.13 steps/s | VRAM=28.39 Go [2026-07-17 15:36:08] step 352700/1000000 | loss=2.3066 vocab=0.7610 token=0.9700 | 6.15 steps/s | VRAM=28.26 Go [2026-07-17 15:36:16] step 352750/1000000 | loss=2.2235 vocab=0.7597 token=1.1221 | 6.17 steps/s | VRAM=32.43 Go [2026-07-17 15:36:24] step 352800/1000000 | loss=2.2686 vocab=1.1425 token=1.4473 | 6.15 steps/s | VRAM=28.68 Go [2026-07-17 15:36:32] step 352850/1000000 | loss=2.2922 vocab=0.7774 token=1.3346 | 6.14 steps/s | VRAM=28.44 Go [2026-07-17 15:36:40] step 352900/1000000 | loss=2.3728 vocab=0.7835 token=1.4982 | 6.16 steps/s | VRAM=31.92 Go [2026-07-17 15:36:48] step 352950/1000000 | loss=2.3379 vocab=0.9754 token=1.0458 | 6.13 steps/s | VRAM=29.27 Go [2026-07-17 15:36:56] step 353000/1000000 | loss=2.2541 vocab=0.7949 token=1.4593 | 6.14 steps/s | VRAM=28.90 Go [2026-07-17 15:37:05] step 353050/1000000 | loss=2.3331 vocab=0.7426 token=1.1432 | 6.12 steps/s | VRAM=32.55 Go [2026-07-17 15:37:13] step 353100/1000000 | loss=2.2916 vocab=0.9631 token=1.5537 | 6.09 steps/s | VRAM=25.17 Go [2026-07-17 15:37:21] step 353150/1000000 | loss=2.3345 vocab=0.6434 token=1.3904 | 6.10 steps/s | VRAM=27.29 Go [2026-07-17 15:37:29] step 353200/1000000 | loss=2.3144 vocab=0.8420 token=1.3566 | 6.15 steps/s | VRAM=25.24 Go [2026-07-17 15:37:37] step 353250/1000000 | loss=2.3434 vocab=0.5594 token=1.1943 | 6.13 steps/s | VRAM=32.62 Go [2026-07-17 15:37:45] step 353300/1000000 | loss=2.2863 vocab=1.0936 token=1.2808 | 6.14 steps/s | VRAM=31.87 Go [2026-07-17 15:37:54] step 353350/1000000 | loss=2.2370 vocab=0.6749 token=1.2056 | 6.12 steps/s | VRAM=31.39 Go [2026-07-17 15:38:02] step 353400/1000000 | loss=2.2530 vocab=0.6500 token=1.1190 | 6.11 steps/s | VRAM=28.90 Go [2026-07-17 15:38:10] step 353450/1000000 | loss=2.2538 vocab=0.8381 token=1.3835 | 6.14 steps/s | VRAM=28.66 Go [2026-07-17 15:38:18] step 353500/1000000 | loss=2.2652 vocab=1.1218 token=1.5763 | 6.10 steps/s | VRAM=26.09 Go [2026-07-17 15:38:26] step 353550/1000000 | loss=2.3739 vocab=0.9861 token=1.5139 | 6.13 steps/s | VRAM=32.39 Go [2026-07-17 15:38:34] step 353600/1000000 | loss=2.2921 vocab=0.8884 token=1.3727 | 6.15 steps/s | VRAM=27.68 Go [2026-07-17 15:38:43] step 353650/1000000 | loss=2.3360 vocab=0.8859 token=1.7991 | 6.11 steps/s | VRAM=31.74 Go [2026-07-17 15:38:51] step 353700/1000000 | loss=2.3558 vocab=1.1636 token=1.5202 | 6.12 steps/s | VRAM=26.85 Go [2026-07-17 15:38:59] step 353750/1000000 | loss=2.2092 vocab=1.0171 token=1.8962 | 6.12 steps/s | VRAM=27.95 Go [2026-07-17 15:39:07] step 353800/1000000 | loss=2.3921 vocab=0.9461 token=1.2323 | 6.10 steps/s | VRAM=30.15 Go [2026-07-17 15:39:15] step 353850/1000000 | loss=2.2804 vocab=0.7385 token=1.2536 | 6.10 steps/s | VRAM=26.90 Go [2026-07-17 15:39:23] step 353900/1000000 | loss=2.2757 vocab=0.9468 token=1.1878 | 6.16 steps/s | VRAM=31.34 Go [2026-07-17 15:39:32] step 353950/1000000 | loss=2.4006 vocab=0.6956 token=1.4692 | 6.15 steps/s | VRAM=28.03 Go [2026-07-17 15:39:40] step 354000/1000000 | loss=2.3776 vocab=1.0793 token=1.2811 | 6.13 steps/s | VRAM=31.90 Go [2026-07-17 15:39:48] step 354050/1000000 | loss=2.2992 vocab=0.8035 token=1.5112 | 6.10 steps/s | VRAM=29.15 Go [2026-07-17 15:39:56] step 354100/1000000 | loss=2.2748 vocab=0.7850 token=1.5665 | 6.13 steps/s | VRAM=31.20 Go [2026-07-17 15:40:04] step 354150/1000000 | loss=2.2334 vocab=0.9118 token=1.3580 | 6.11 steps/s | VRAM=29.67 Go [2026-07-17 15:40:12] step 354200/1000000 | loss=2.2744 vocab=0.7375 token=1.3464 | 6.12 steps/s | VRAM=28.03 Go [2026-07-17 15:40:21] step 354250/1000000 | loss=2.3156 vocab=1.3069 token=1.2879 | 6.11 steps/s | VRAM=32.58 Go [2026-07-17 15:40:29] step 354300/1000000 | loss=2.3628 vocab=1.4909 token=1.3419 | 6.09 steps/s | VRAM=26.18 Go [2026-07-17 15:40:37] step 354350/1000000 | loss=2.2453 vocab=0.9706 token=1.4514 | 6.11 steps/s | VRAM=28.93 Go [2026-07-17 15:40:45] step 354400/1000000 | loss=2.2757 vocab=0.9237 token=1.2625 | 6.14 steps/s | VRAM=28.98 Go [2026-07-17 15:40:53] step 354450/1000000 | loss=2.2142 vocab=1.5648 token=1.3141 | 6.11 steps/s | VRAM=31.80 Go [2026-07-17 15:41:01] step 354500/1000000 | loss=2.3772 vocab=0.9187 token=1.3845 | 6.15 steps/s | VRAM=31.46 Go [2026-07-17 15:41:10] step 354550/1000000 | loss=2.2933 vocab=0.7501 token=1.5574 | 6.14 steps/s | VRAM=30.95 Go [2026-07-17 15:41:18] step 354600/1000000 | loss=2.4543 vocab=0.8470 token=1.5163 | 6.12 steps/s | VRAM=28.10 Go [2026-07-17 15:41:26] step 354650/1000000 | loss=2.2188 vocab=0.6150 token=1.4012 | 6.18 steps/s | VRAM=29.89 Go [2026-07-17 15:41:34] step 354700/1000000 | loss=2.3321 vocab=0.7712 token=1.0764 | 6.14 steps/s | VRAM=28.84 Go [2026-07-17 15:41:42] step 354750/1000000 | loss=2.2741 vocab=1.1007 token=1.2219 | 6.13 steps/s | VRAM=29.21 Go [2026-07-17 15:41:50] step 354800/1000000 | loss=2.2142 vocab=0.8809 token=1.1861 | 6.14 steps/s | VRAM=29.16 Go [2026-07-17 15:41:58] step 354850/1000000 | loss=2.4172 vocab=0.9082 token=1.1880 | 6.13 steps/s | VRAM=26.64 Go [2026-07-17 15:42:07] step 354900/1000000 | loss=2.2513 vocab=1.0635 token=1.3048 | 6.13 steps/s | VRAM=31.92 Go [2026-07-17 15:42:15] step 354950/1000000 | loss=2.3496 vocab=1.1803 token=1.5385 | 6.09 steps/s | VRAM=26.33 Go [2026-07-17 15:42:23] step 355000/1000000 | loss=2.4029 vocab=1.3852 token=1.4135 | 6.14 steps/s | VRAM=26.31 Go [2026-07-17 15:42:31] step 355050/1000000 | loss=2.2285 vocab=0.8085 token=1.1307 | 6.15 steps/s | VRAM=31.97 Go [2026-07-17 15:42:39] step 355100/1000000 | loss=2.3170 vocab=1.5505 token=1.3134 | 6.13 steps/s | VRAM=29.49 Go [2026-07-17 15:42:47] step 355150/1000000 | loss=2.2558 vocab=0.6090 token=1.1389 | 6.13 steps/s | VRAM=32.31 Go [2026-07-17 15:42:56] step 355200/1000000 | loss=2.2635 vocab=1.0276 token=1.5396 | 6.12 steps/s | VRAM=28.56 Go [2026-07-17 15:43:04] step 355250/1000000 | loss=2.3153 vocab=1.4036 token=1.3439 | 6.16 steps/s | VRAM=29.19 Go [2026-07-17 15:43:12] step 355300/1000000 | loss=2.2528 vocab=0.9848 token=1.3899 | 6.10 steps/s | VRAM=26.84 Go [2026-07-17 15:43:20] step 355350/1000000 | loss=2.2997 vocab=1.0608 token=1.4858 | 6.15 steps/s | VRAM=29.45 Go [2026-07-17 15:43:28] step 355400/1000000 | loss=2.4475 vocab=1.4115 token=1.5429 | 6.12 steps/s | VRAM=32.98 Go [2026-07-17 15:43:36] step 355450/1000000 | loss=2.3014 vocab=1.2041 token=1.6393 | 6.09 steps/s | VRAM=32.19 Go [2026-07-17 15:43:45] step 355500/1000000 | loss=2.1880 vocab=0.4462 token=1.3519 | 6.15 steps/s | VRAM=32.78 Go [2026-07-17 15:43:53] step 355550/1000000 | loss=2.2986 vocab=0.5941 token=1.3934 | 6.17 steps/s | VRAM=32.16 Go [2026-07-17 15:44:01] step 355600/1000000 | loss=2.3788 vocab=0.7487 token=1.4542 | 6.11 steps/s | VRAM=27.95 Go [2026-07-17 15:44:09] step 355650/1000000 | loss=2.3335 vocab=0.9741 token=1.6872 | 6.13 steps/s | VRAM=29.41 Go [2026-07-17 15:44:17] step 355700/1000000 | loss=2.2896 vocab=1.1848 token=1.1924 | 6.12 steps/s | VRAM=32.67 Go [2026-07-17 15:44:25] step 355750/1000000 | loss=2.3200 vocab=1.9792 token=1.7215 | 6.14 steps/s | VRAM=32.79 Go [2026-07-17 15:44:34] step 355800/1000000 | loss=2.2602 vocab=1.0452 token=1.5402 | 6.13 steps/s | VRAM=31.93 Go [2026-07-17 15:44:42] step 355850/1000000 | loss=2.2399 vocab=1.2867 token=1.2911 | 6.10 steps/s | VRAM=28.35 Go [2026-07-17 15:44:50] step 355900/1000000 | loss=2.2356 vocab=0.9040 token=1.3287 | 6.13 steps/s | VRAM=28.22 Go [2026-07-17 15:44:58] step 355950/1000000 | loss=2.4433 vocab=1.5129 token=1.6212 | 6.16 steps/s | VRAM=27.25 Go [2026-07-17 15:45:06] step 356000/1000000 | loss=2.3269 vocab=0.7710 token=1.2289 | 6.14 steps/s | VRAM=31.10 Go [2026-07-17 15:45:14] step 356050/1000000 | loss=2.2620 vocab=0.7558 token=0.9419 | 6.15 steps/s | VRAM=30.21 Go [2026-07-17 15:45:22] step 356100/1000000 | loss=2.4624 vocab=0.9860 token=1.3015 | 6.12 steps/s | VRAM=32.46 Go [2026-07-17 15:45:31] step 356150/1000000 | loss=2.2812 vocab=0.8648 token=1.3518 | 6.14 steps/s | VRAM=28.16 Go [2026-07-17 15:45:39] step 356200/1000000 | loss=2.2433 vocab=0.7295 token=1.1704 | 6.15 steps/s | VRAM=27.88 Go [2026-07-17 15:45:47] step 356250/1000000 | loss=2.3358 vocab=0.9697 token=1.6040 | 6.16 steps/s | VRAM=27.24 Go [2026-07-17 15:45:55] step 356300/1000000 | loss=2.2501 vocab=0.6976 token=1.1800 | 6.14 steps/s | VRAM=32.97 Go [2026-07-17 15:46:03] step 356350/1000000 | loss=2.3458 vocab=0.7599 token=1.2198 | 6.11 steps/s | VRAM=26.45 Go [2026-07-17 15:46:11] step 356400/1000000 | loss=2.2755 vocab=1.0105 token=1.1674 | 6.12 steps/s | VRAM=28.57 Go [2026-07-17 15:46:19] step 356450/1000000 | loss=2.2688 vocab=1.3846 token=1.7275 | 6.12 steps/s | VRAM=28.63 Go [2026-07-17 15:46:28] step 356500/1000000 | loss=2.2574 vocab=0.7205 token=0.9628 | 6.16 steps/s | VRAM=25.75 Go [2026-07-17 15:46:36] step 356550/1000000 | loss=2.2369 vocab=0.7250 token=1.4378 | 6.15 steps/s | VRAM=29.05 Go [2026-07-17 15:46:44] step 356600/1000000 | loss=2.2735 vocab=1.3874 token=1.6771 | 6.10 steps/s | VRAM=26.32 Go [2026-07-17 15:46:52] step 356650/1000000 | loss=2.1616 vocab=0.5843 token=0.8425 | 6.14 steps/s | VRAM=27.85 Go [2026-07-17 15:47:00] step 356700/1000000 | loss=2.2492 vocab=1.1218 token=1.5861 | 6.10 steps/s | VRAM=26.93 Go [2026-07-17 15:47:08] step 356750/1000000 | loss=2.2070 vocab=0.8748 token=1.4760 | 6.14 steps/s | VRAM=28.59 Go [2026-07-17 15:47:17] step 356800/1000000 | loss=2.3788 vocab=1.3926 token=1.4553 | 6.11 steps/s | VRAM=32.65 Go [2026-07-17 15:47:25] step 356850/1000000 | loss=2.2462 vocab=0.7479 token=1.1399 | 6.15 steps/s | VRAM=30.78 Go [2026-07-17 15:47:33] step 356900/1000000 | loss=2.2679 vocab=1.2570 token=1.3540 | 6.13 steps/s | VRAM=29.41 Go [2026-07-17 15:47:41] step 356950/1000000 | loss=2.3323 vocab=0.8790 token=1.5587 | 6.08 steps/s | VRAM=24.18 Go [2026-07-17 15:47:49] step 357000/1000000 | loss=2.2715 vocab=1.0796 token=1.6293 | 6.11 steps/s | VRAM=29.41 Go [2026-07-17 15:47:57] step 357050/1000000 | loss=2.2496 vocab=0.2945 token=1.2014 | 6.13 steps/s | VRAM=25.35 Go [2026-07-17 15:48:06] step 357100/1000000 | loss=2.3420 vocab=0.6414 token=1.2512 | 6.15 steps/s | VRAM=28.36 Go [2026-07-17 15:48:14] step 357150/1000000 | loss=2.2946 vocab=1.3297 token=1.5332 | 6.17 steps/s | VRAM=31.14 Go [2026-07-17 15:48:22] step 357200/1000000 | loss=2.2704 vocab=0.4087 token=1.1241 | 6.15 steps/s | VRAM=32.69 Go [2026-07-17 15:48:30] step 357250/1000000 | loss=2.2607 vocab=0.4680 token=0.8576 | 6.10 steps/s | VRAM=28.00 Go [2026-07-17 15:48:38] step 357300/1000000 | loss=2.2738 vocab=0.6876 token=1.2069 | 6.13 steps/s | VRAM=32.75 Go [2026-07-17 15:48:46] step 357350/1000000 | loss=2.4007 vocab=0.6958 token=0.8949 | 6.13 steps/s | VRAM=32.40 Go [2026-07-17 15:48:54] step 357400/1000000 | loss=2.2630 vocab=1.1524 token=1.8220 | 6.11 steps/s | VRAM=32.69 Go [2026-07-17 15:49:03] step 357450/1000000 | loss=2.4706 vocab=0.7218 token=1.6197 | 6.14 steps/s | VRAM=31.66 Go [2026-07-17 15:49:11] step 357500/1000000 | loss=2.2700 vocab=1.0107 token=1.4325 | 6.09 steps/s | VRAM=25.99 Go [2026-07-17 15:49:19] step 357550/1000000 | loss=2.2122 vocab=0.8250 token=1.0154 | 6.11 steps/s | VRAM=26.90 Go [2026-07-17 15:49:27] step 357600/1000000 | loss=2.2792 vocab=1.0417 token=1.4832 | 6.15 steps/s | VRAM=28.90 Go [2026-07-17 15:49:35] step 357650/1000000 | loss=2.3169 vocab=0.8676 token=1.4215 | 6.14 steps/s | VRAM=31.35 Go [2026-07-17 15:49:43] step 357700/1000000 | loss=2.4092 vocab=0.9379 token=1.7720 | 6.13 steps/s | VRAM=28.51 Go [2026-07-17 15:49:52] step 357750/1000000 | loss=2.3671 vocab=0.4582 token=0.7037 | 6.11 steps/s | VRAM=29.10 Go [2026-07-17 15:50:00] step 357800/1000000 | loss=2.2841 vocab=0.6365 token=1.7257 | 6.09 steps/s | VRAM=26.28 Go [2026-07-17 15:50:08] step 357850/1000000 | loss=2.3087 vocab=0.8016 token=1.4081 | 6.12 steps/s | VRAM=28.11 Go [2026-07-17 15:50:16] step 357900/1000000 | loss=2.3562 vocab=1.1061 token=1.5255 | 6.14 steps/s | VRAM=28.00 Go [2026-07-17 15:50:24] step 357950/1000000 | loss=2.3821 vocab=0.8867 token=1.8133 | 6.12 steps/s | VRAM=33.01 Go [2026-07-17 15:50:32] step 358000/1000000 | loss=2.2958 vocab=2.2705 token=1.5642 | 6.15 steps/s | VRAM=31.61 Go [2026-07-17 15:50:41] step 358050/1000000 | loss=2.2050 vocab=0.6204 token=0.8914 | 6.12 steps/s | VRAM=32.33 Go [2026-07-17 15:50:49] step 358100/1000000 | loss=2.3295 vocab=0.9363 token=1.6129 | 6.13 steps/s | VRAM=29.17 Go [2026-07-17 15:50:57] step 358150/1000000 | loss=2.3428 vocab=0.6851 token=1.2512 | 6.16 steps/s | VRAM=28.88 Go [2026-07-17 15:51:05] step 358200/1000000 | loss=2.2972 vocab=1.1754 token=1.1253 | 6.13 steps/s | VRAM=29.17 Go [2026-07-17 15:51:13] step 358250/1000000 | loss=2.3515 vocab=0.4148 token=1.1476 | 6.12 steps/s | VRAM=29.40 Go [2026-07-17 15:51:21] step 358300/1000000 | loss=2.3298 vocab=0.6002 token=1.3586 | 6.13 steps/s | VRAM=29.49 Go [2026-07-17 15:51:30] step 358350/1000000 | loss=2.1865 vocab=0.6507 token=0.8067 | 6.14 steps/s | VRAM=32.76 Go [2026-07-17 15:51:38] step 358400/1000000 | loss=2.1693 vocab=0.5833 token=1.2540 | 6.12 steps/s | VRAM=24.83 Go [2026-07-17 15:51:46] step 358450/1000000 | loss=2.2709 vocab=0.8502 token=1.3889 | 6.16 steps/s | VRAM=28.89 Go [2026-07-17 15:51:54] step 358500/1000000 | loss=2.2321 vocab=0.8691 token=1.2924 | 6.14 steps/s | VRAM=28.13 Go [2026-07-17 15:52:02] step 358550/1000000 | loss=2.2821 vocab=1.1439 token=1.2135 | 6.13 steps/s | VRAM=27.83 Go [2026-07-17 15:52:10] step 358600/1000000 | loss=2.2926 vocab=0.9631 token=1.9225 | 6.14 steps/s | VRAM=32.49 Go [2026-07-17 15:52:18] step 358650/1000000 | loss=2.2829 vocab=0.7509 token=1.6684 | 6.16 steps/s | VRAM=29.09 Go [2026-07-17 15:52:27] step 358700/1000000 | loss=2.3035 vocab=1.1667 token=1.3712 | 6.11 steps/s | VRAM=27.69 Go [2026-07-17 15:52:35] step 358750/1000000 | loss=2.3263 vocab=0.8745 token=1.4708 | 6.17 steps/s | VRAM=27.17 Go [2026-07-17 15:52:43] step 358800/1000000 | loss=2.3544 vocab=1.2446 token=1.4138 | 6.12 steps/s | VRAM=28.24 Go [2026-07-17 15:52:51] step 358850/1000000 | loss=2.3551 vocab=1.3183 token=1.3011 | 6.14 steps/s | VRAM=27.20 Go [2026-07-17 15:52:59] step 358900/1000000 | loss=2.2858 vocab=0.7731 token=1.4429 | 6.14 steps/s | VRAM=31.20 Go [2026-07-17 15:53:07] step 358950/1000000 | loss=2.3319 vocab=1.9825 token=1.3945 | 6.14 steps/s | VRAM=28.60 Go [2026-07-17 15:53:15] step 359000/1000000 | loss=2.3412 vocab=1.0252 token=1.0415 | 6.09 steps/s | VRAM=26.75 Go [2026-07-17 15:53:24] step 359050/1000000 | loss=2.3917 vocab=0.8194 token=1.6054 | 6.14 steps/s | VRAM=28.62 Go [2026-07-17 15:53:32] step 359100/1000000 | loss=2.2185 vocab=1.0261 token=1.5194 | 6.10 steps/s | VRAM=29.87 Go [2026-07-17 15:53:40] step 359150/1000000 | loss=2.2640 vocab=1.0047 token=1.5361 | 6.11 steps/s | VRAM=29.01 Go [2026-07-17 15:53:48] step 359200/1000000 | loss=2.3179 vocab=1.0039 token=1.3720 | 6.12 steps/s | VRAM=31.86 Go [2026-07-17 15:53:56] step 359250/1000000 | loss=2.3280 vocab=1.1719 token=2.0281 | 6.12 steps/s | VRAM=26.25 Go [2026-07-17 15:54:05] step 359300/1000000 | loss=2.3607 vocab=0.9325 token=1.2327 | 6.12 steps/s | VRAM=28.75 Go [2026-07-17 15:54:13] step 359350/1000000 | loss=2.2373 vocab=1.0024 token=1.3974 | 6.13 steps/s | VRAM=28.93 Go [2026-07-17 15:54:21] step 359400/1000000 | loss=2.3543 vocab=1.2716 token=1.5181 | 6.10 steps/s | VRAM=26.15 Go [2026-07-17 15:54:29] step 359450/1000000 | loss=2.3614 vocab=0.8811 token=1.7223 | 6.16 steps/s | VRAM=31.56 Go [2026-07-17 15:54:37] step 359500/1000000 | loss=2.3109 vocab=1.5014 token=1.4353 | 6.14 steps/s | VRAM=28.83 Go [2026-07-17 15:54:45] step 359550/1000000 | loss=2.2156 vocab=0.9093 token=1.5221 | 6.14 steps/s | VRAM=29.46 Go [2026-07-17 15:54:53] step 359600/1000000 | loss=2.2764 vocab=1.7431 token=1.5547 | 6.14 steps/s | VRAM=31.39 Go [2026-07-17 15:55:02] step 359650/1000000 | loss=2.3429 vocab=0.8466 token=1.0483 | 6.09 steps/s | VRAM=32.66 Go [2026-07-17 15:55:10] step 359700/1000000 | loss=2.2790 vocab=1.0818 token=1.5040 | 6.12 steps/s | VRAM=26.03 Go [2026-07-17 15:55:18] step 359750/1000000 | loss=2.1638 vocab=0.4851 token=1.2171 | 6.14 steps/s | VRAM=26.44 Go [2026-07-17 15:55:26] step 359800/1000000 | loss=2.2911 vocab=0.2609 token=0.8960 | 6.15 steps/s | VRAM=32.31 Go [2026-07-17 15:55:34] step 359850/1000000 | loss=2.3490 vocab=0.6211 token=0.9647 | 6.14 steps/s | VRAM=24.96 Go [2026-07-17 15:55:42] step 359900/1000000 | loss=2.3328 vocab=0.8051 token=1.1941 | 6.11 steps/s | VRAM=30.04 Go [2026-07-17 15:55:51] step 359950/1000000 | loss=2.3283 vocab=0.7159 token=1.1305 | 6.13 steps/s | VRAM=28.64 Go [2026-07-17 15:55:59] step 360000/1000000 | loss=2.3032 vocab=0.3721 token=1.0882 | 6.12 steps/s | VRAM=28.19 Go [2026-07-17 15:56:07] step 360050/1000000 | loss=2.4188 vocab=1.2637 token=1.4507 | 6.08 steps/s | VRAM=27.23 Go [2026-07-17 15:56:15] step 360100/1000000 | loss=2.2346 vocab=0.5068 token=1.0942 | 6.13 steps/s | VRAM=27.72 Go [2026-07-17 15:56:23] step 360150/1000000 | loss=2.3635 vocab=0.6805 token=1.6156 | 6.12 steps/s | VRAM=26.12 Go [2026-07-17 15:56:31] step 360200/1000000 | loss=2.3394 vocab=0.7818 token=1.1590 | 6.11 steps/s | VRAM=32.16 Go [2026-07-17 15:56:40] step 360250/1000000 | loss=2.3578 vocab=0.9807 token=1.5607 | 6.11 steps/s | VRAM=28.80 Go [2026-07-17 15:56:48] step 360300/1000000 | loss=2.3494 vocab=0.7019 token=1.1229 | 6.11 steps/s | VRAM=28.25 Go [2026-07-17 15:56:56] step 360350/1000000 | loss=2.2973 vocab=0.6321 token=1.6325 | 6.12 steps/s | VRAM=26.87 Go [2026-07-17 15:57:04] step 360400/1000000 | loss=2.2672 vocab=0.7264 token=1.3013 | 6.15 steps/s | VRAM=31.31 Go [2026-07-17 15:57:12] step 360450/1000000 | loss=2.2574 vocab=0.5023 token=1.0556 | 6.10 steps/s | VRAM=32.38 Go [2026-07-17 15:57:21] step 360500/1000000 | loss=2.3324 vocab=1.0208 token=1.4042 | 6.11 steps/s | VRAM=28.60 Go [2026-07-17 15:57:29] step 360550/1000000 | loss=2.3316 vocab=0.5614 token=1.7550 | 6.12 steps/s | VRAM=25.00 Go [2026-07-17 15:57:37] step 360600/1000000 | loss=2.1888 vocab=1.0740 token=1.5694 | 6.10 steps/s | VRAM=29.10 Go [2026-07-17 15:57:45] step 360650/1000000 | loss=2.2250 vocab=1.1199 token=1.1258 | 6.08 steps/s | VRAM=29.98 Go [2026-07-17 15:57:53] step 360700/1000000 | loss=2.2756 vocab=0.4695 token=1.6604 | 6.14 steps/s | VRAM=32.05 Go [2026-07-17 15:58:01] step 360750/1000000 | loss=2.3941 vocab=0.4428 token=0.8342 | 6.11 steps/s | VRAM=31.70 Go [2026-07-17 15:58:10] step 360800/1000000 | loss=2.2552 vocab=0.6405 token=1.3132 | 6.13 steps/s | VRAM=25.51 Go [2026-07-17 15:58:18] step 360850/1000000 | loss=2.2272 vocab=0.6590 token=1.0978 | 6.11 steps/s | VRAM=28.83 Go [2026-07-17 15:58:26] step 360900/1000000 | loss=2.3319 vocab=0.4749 token=1.3708 | 6.13 steps/s | VRAM=27.91 Go [2026-07-17 15:58:34] step 360950/1000000 | loss=2.3414 vocab=0.8666 token=1.2279 | 6.16 steps/s | VRAM=32.49 Go [2026-07-17 15:58:42] step 361000/1000000 | loss=2.2640 vocab=1.1286 token=1.4525 | 6.16 steps/s | VRAM=24.89 Go [2026-07-17 15:58:50] step 361050/1000000 | loss=2.2376 vocab=1.1204 token=1.5800 | 6.14 steps/s | VRAM=30.86 Go [2026-07-17 15:58:58] step 361100/1000000 | loss=2.2495 vocab=0.6253 token=1.0794 | 6.11 steps/s | VRAM=25.51 Go [2026-07-17 15:59:07] step 361150/1000000 | loss=2.3259 vocab=0.9226 token=1.3236 | 6.14 steps/s | VRAM=30.49 Go [2026-07-17 15:59:15] step 361200/1000000 | loss=2.3127 vocab=0.9994 token=1.1428 | 6.11 steps/s | VRAM=29.01 Go [2026-07-17 15:59:23] step 361250/1000000 | loss=2.3422 vocab=0.8207 token=1.4938 | 6.11 steps/s | VRAM=29.19 Go [2026-07-17 15:59:31] step 361300/1000000 | loss=2.2212 vocab=0.9132 token=1.3432 | 6.14 steps/s | VRAM=25.23 Go [2026-07-17 15:59:39] step 361350/1000000 | loss=2.3366 vocab=0.6404 token=1.2137 | 6.12 steps/s | VRAM=28.24 Go [2026-07-17 15:59:47] step 361400/1000000 | loss=2.2026 vocab=0.7077 token=1.6051 | 6.11 steps/s | VRAM=29.00 Go [2026-07-17 15:59:56] step 361450/1000000 | loss=2.3538 vocab=0.7373 token=1.3200 | 6.10 steps/s | VRAM=31.56 Go [2026-07-17 16:00:04] step 361500/1000000 | loss=2.3886 vocab=0.5986 token=1.2657 | 6.11 steps/s | VRAM=31.29 Go [2026-07-17 16:00:12] step 361550/1000000 | loss=2.3572 vocab=0.8844 token=1.1833 | 6.07 steps/s | VRAM=28.97 Go [2026-07-17 16:00:20] step 361600/1000000 | loss=2.3202 vocab=1.1101 token=1.9175 | 6.10 steps/s | VRAM=28.77 Go [2026-07-17 16:00:28] step 361650/1000000 | loss=2.1704 vocab=0.6181 token=1.4266 | 6.13 steps/s | VRAM=32.80 Go [2026-07-17 16:00:37] step 361700/1000000 | loss=2.2877 vocab=1.3996 token=1.5954 | 6.12 steps/s | VRAM=32.34 Go [2026-07-17 16:00:45] step 361750/1000000 | loss=2.3592 vocab=1.1395 token=1.2502 | 6.11 steps/s | VRAM=30.19 Go [2026-07-17 16:00:53] step 361800/1000000 | loss=2.3000 vocab=0.7498 token=1.4366 | 6.17 steps/s | VRAM=31.33 Go [2026-07-17 16:01:01] step 361850/1000000 | loss=2.3234 vocab=1.2451 token=1.4240 | 6.15 steps/s | VRAM=31.58 Go [2026-07-17 16:01:09] step 361900/1000000 | loss=2.3232 vocab=0.7124 token=0.9604 | 6.07 steps/s | VRAM=28.90 Go [2026-07-17 16:01:17] step 361950/1000000 | loss=2.2848 vocab=0.9652 token=1.7385 | 6.12 steps/s | VRAM=28.52 Go [2026-07-17 16:01:26] step 362000/1000000 | loss=2.2885 vocab=0.9292 token=1.3545 | 6.12 steps/s | VRAM=30.64 Go [2026-07-17 16:01:34] step 362050/1000000 | loss=2.1678 vocab=0.8845 token=1.2382 | 6.12 steps/s | VRAM=28.00 Go [2026-07-17 16:01:42] step 362100/1000000 | loss=2.3162 vocab=1.1291 token=1.0751 | 6.09 steps/s | VRAM=28.03 Go [2026-07-17 16:01:50] step 362150/1000000 | loss=2.1430 vocab=0.7412 token=1.2242 | 6.14 steps/s | VRAM=30.78 Go [2026-07-17 16:01:58] step 362200/1000000 | loss=2.2343 vocab=0.7674 token=1.3356 | 6.16 steps/s | VRAM=32.05 Go [2026-07-17 16:02:06] step 362250/1000000 | loss=2.2969 vocab=1.0708 token=1.2957 | 6.13 steps/s | VRAM=25.07 Go [2026-07-17 16:02:15] step 362300/1000000 | loss=2.3669 vocab=0.8554 token=1.3021 | 6.15 steps/s | VRAM=31.94 Go [2026-07-17 16:02:23] step 362350/1000000 | loss=2.2375 vocab=1.0444 token=1.4738 | 6.13 steps/s | VRAM=31.00 Go [2026-07-17 16:02:31] step 362400/1000000 | loss=2.3605 vocab=1.4586 token=1.2333 | 6.12 steps/s | VRAM=25.10 Go [2026-07-17 16:02:39] step 362450/1000000 | loss=2.2978 vocab=0.9181 token=1.3827 | 6.12 steps/s | VRAM=31.60 Go [2026-07-17 16:02:47] step 362500/1000000 | loss=2.3598 vocab=1.2350 token=1.2672 | 6.11 steps/s | VRAM=26.65 Go [2026-07-17 16:02:55] step 362550/1000000 | loss=2.2353 vocab=0.8421 token=1.5064 | 6.16 steps/s | VRAM=25.75 Go [2026-07-17 16:03:03] step 362600/1000000 | loss=2.3083 vocab=0.4134 token=1.2113 | 6.15 steps/s | VRAM=29.00 Go [2026-07-17 16:03:12] step 362650/1000000 | loss=2.2937 vocab=0.9412 token=0.9970 | 6.12 steps/s | VRAM=29.10 Go [2026-07-17 16:03:20] step 362700/1000000 | loss=2.2970 vocab=0.7445 token=1.4969 | 6.12 steps/s | VRAM=26.08 Go [2026-07-17 16:03:28] step 362750/1000000 | loss=2.3160 vocab=0.5098 token=1.3962 | 6.15 steps/s | VRAM=27.98 Go [2026-07-17 16:03:36] step 362800/1000000 | loss=2.2011 vocab=0.6187 token=1.1164 | 6.13 steps/s | VRAM=28.13 Go [2026-07-17 16:03:44] step 362850/1000000 | loss=2.2416 vocab=0.7683 token=1.3270 | 6.12 steps/s | VRAM=27.35 Go [2026-07-17 16:03:52] step 362900/1000000 | loss=2.3960 vocab=1.3749 token=1.8413 | 6.13 steps/s | VRAM=30.07 Go [2026-07-17 16:04:01] step 362950/1000000 | loss=2.2385 vocab=1.0253 token=1.0685 | 6.12 steps/s | VRAM=28.78 Go [2026-07-17 16:04:09] step 363000/1000000 | loss=2.4170 vocab=1.3954 token=1.3927 | 6.11 steps/s | VRAM=26.31 Go [2026-07-17 16:04:17] step 363050/1000000 | loss=2.2978 vocab=0.7992 token=1.4456 | 6.11 steps/s | VRAM=25.82 Go [2026-07-17 16:04:25] step 363100/1000000 | loss=2.3571 vocab=1.1971 token=1.3568 | 6.13 steps/s | VRAM=29.46 Go [2026-07-17 16:04:33] step 363150/1000000 | loss=2.2465 vocab=0.7329 token=1.1758 | 6.12 steps/s | VRAM=25.94 Go [2026-07-17 16:04:41] step 363200/1000000 | loss=2.2766 vocab=0.7241 token=1.1833 | 6.12 steps/s | VRAM=28.14 Go [2026-07-17 16:04:50] step 363250/1000000 | loss=2.2654 vocab=0.6681 token=1.1535 | 6.12 steps/s | VRAM=32.13 Go [2026-07-17 16:04:58] step 363300/1000000 | loss=2.3512 vocab=1.3782 token=1.9331 | 6.11 steps/s | VRAM=25.76 Go [2026-07-17 16:05:06] step 363350/1000000 | loss=2.4064 vocab=1.0726 token=1.1622 | 6.13 steps/s | VRAM=28.57 Go [2026-07-17 16:05:14] step 363400/1000000 | loss=2.3593 vocab=0.8823 token=1.6226 | 6.11 steps/s | VRAM=31.51 Go [2026-07-17 16:05:22] step 363450/1000000 | loss=2.3727 vocab=0.8516 token=1.4651 | 6.13 steps/s | VRAM=29.44 Go [2026-07-17 16:05:30] step 363500/1000000 | loss=2.2828 vocab=0.6304 token=1.2947 | 6.12 steps/s | VRAM=28.11 Go [2026-07-17 16:05:39] step 363550/1000000 | loss=2.2815 vocab=0.7852 token=1.5301 | 6.12 steps/s | VRAM=25.08 Go [2026-07-17 16:05:47] step 363600/1000000 | loss=2.3790 vocab=1.1107 token=1.3981 | 6.11 steps/s | VRAM=31.96 Go [2026-07-17 16:05:55] step 363650/1000000 | loss=2.2475 vocab=0.9298 token=1.0188 | 6.11 steps/s | VRAM=32.12 Go [2026-07-17 16:06:03] step 363700/1000000 | loss=2.3704 vocab=0.9685 token=1.6760 | 6.12 steps/s | VRAM=31.83 Go [2026-07-17 16:06:11] step 363750/1000000 | loss=2.2964 vocab=0.9687 token=1.1524 | 6.11 steps/s | VRAM=32.12 Go [2026-07-17 16:06:20] step 363800/1000000 | loss=2.3658 vocab=0.7559 token=1.2722 | 6.10 steps/s | VRAM=29.11 Go [2026-07-17 16:06:28] step 363850/1000000 | loss=2.2493 vocab=0.9228 token=1.1749 | 6.09 steps/s | VRAM=29.29 Go [2026-07-17 16:06:36] step 363900/1000000 | loss=2.4086 vocab=0.9177 token=1.2160 | 6.09 steps/s | VRAM=28.29 Go [2026-07-17 16:06:44] step 363950/1000000 | loss=2.3790 vocab=0.5884 token=1.1460 | 6.13 steps/s | VRAM=28.36 Go [2026-07-17 16:06:52] step 364000/1000000 | loss=2.3496 vocab=0.8335 token=1.2648 | 6.10 steps/s | VRAM=29.62 Go [2026-07-17 16:07:01] step 364050/1000000 | loss=2.3525 vocab=0.8969 token=1.5541 | 6.08 steps/s | VRAM=28.98 Go [2026-07-17 16:07:09] step 364100/1000000 | loss=2.2977 vocab=1.1777 token=1.6710 | 6.11 steps/s | VRAM=26.94 Go [2026-07-17 16:07:17] step 364150/1000000 | loss=2.3118 vocab=0.5033 token=1.3256 | 6.15 steps/s | VRAM=30.60 Go [2026-07-17 16:07:25] step 364200/1000000 | loss=2.3057 vocab=1.3271 token=1.3888 | 6.11 steps/s | VRAM=32.81 Go [2026-07-17 16:07:33] step 364250/1000000 | loss=2.3123 vocab=1.0338 token=1.6545 | 6.11 steps/s | VRAM=28.95 Go [2026-07-17 16:07:41] step 364300/1000000 | loss=2.2178 vocab=0.7099 token=0.8903 | 6.14 steps/s | VRAM=24.76 Go [2026-07-17 16:07:50] step 364350/1000000 | loss=2.4176 vocab=0.9507 token=1.4649 | 6.12 steps/s | VRAM=28.36 Go [2026-07-17 16:07:58] step 364400/1000000 | loss=2.2752 vocab=0.6888 token=1.5369 | 6.07 steps/s | VRAM=26.70 Go [2026-07-17 16:08:06] step 364450/1000000 | loss=2.3622 vocab=0.9635 token=1.6410 | 6.11 steps/s | VRAM=25.39 Go [2026-07-17 16:08:14] step 364500/1000000 | loss=2.4377 vocab=1.2804 token=1.3616 | 6.12 steps/s | VRAM=25.25 Go [2026-07-17 16:08:22] step 364550/1000000 | loss=2.1585 vocab=0.9156 token=1.3103 | 6.10 steps/s | VRAM=25.85 Go [2026-07-17 16:08:31] step 364600/1000000 | loss=2.3016 vocab=1.0435 token=1.3368 | 6.10 steps/s | VRAM=29.39 Go [2026-07-17 16:08:39] step 364650/1000000 | loss=2.3050 vocab=0.6132 token=1.3216 | 6.08 steps/s | VRAM=29.67 Go [2026-07-17 16:08:47] step 364700/1000000 | loss=2.2423 vocab=0.4804 token=1.1708 | 6.13 steps/s | VRAM=24.85 Go [2026-07-17 16:08:55] step 364750/1000000 | loss=2.3691 vocab=0.8547 token=1.0115 | 6.13 steps/s | VRAM=32.26 Go [2026-07-17 16:09:03] step 364800/1000000 | loss=2.2584 vocab=1.1262 token=1.5901 | 6.08 steps/s | VRAM=26.62 Go [2026-07-17 16:09:11] step 364850/1000000 | loss=2.2346 vocab=0.9566 token=1.5410 | 6.14 steps/s | VRAM=26.11 Go [2026-07-17 16:09:20] step 364900/1000000 | loss=2.2660 vocab=0.8352 token=1.5228 | 6.14 steps/s | VRAM=24.49 Go [2026-07-17 16:09:28] step 364950/1000000 | loss=2.1966 vocab=1.0226 token=1.3162 | 6.12 steps/s | VRAM=25.52 Go [2026-07-17 16:09:36] step 365000/1000000 | loss=2.3490 vocab=0.5725 token=1.0230 | 6.13 steps/s | VRAM=28.11 Go [2026-07-17 16:09:44] step 365050/1000000 | loss=2.2946 vocab=0.7103 token=1.6267 | 6.11 steps/s | VRAM=28.87 Go [2026-07-17 16:09:52] step 365100/1000000 | loss=2.2722 vocab=1.2162 token=1.2709 | 6.13 steps/s | VRAM=32.28 Go [2026-07-17 16:10:00] step 365150/1000000 | loss=2.2669 vocab=0.4807 token=1.3138 | 6.12 steps/s | VRAM=32.85 Go [2026-07-17 16:10:09] step 365200/1000000 | loss=2.2130 vocab=1.0101 token=1.3710 | 6.11 steps/s | VRAM=28.39 Go [2026-07-17 16:10:17] step 365250/1000000 | loss=2.3695 vocab=0.7424 token=1.3419 | 6.11 steps/s | VRAM=29.53 Go [2026-07-17 16:10:25] step 365300/1000000 | loss=2.3017 vocab=0.8359 token=1.4752 | 6.13 steps/s | VRAM=29.00 Go [2026-07-17 16:10:33] step 365350/1000000 | loss=2.2524 vocab=1.0360 token=1.5647 | 6.13 steps/s | VRAM=28.16 Go [2026-07-17 16:10:41] step 365400/1000000 | loss=2.1232 vocab=0.8740 token=1.3568 | 6.14 steps/s | VRAM=26.02 Go [2026-07-17 16:10:49] step 365450/1000000 | loss=2.3216 vocab=0.9078 token=1.1510 | 6.09 steps/s | VRAM=26.20 Go [2026-07-17 16:10:58] step 365500/1000000 | loss=2.3116 vocab=0.9149 token=1.3509 | 6.10 steps/s | VRAM=28.01 Go [2026-07-17 16:11:06] step 365550/1000000 | loss=2.3603 vocab=1.0321 token=1.1878 | 6.10 steps/s | VRAM=26.88 Go [2026-07-17 16:11:14] step 365600/1000000 | loss=2.3638 vocab=1.3397 token=1.2496 | 6.12 steps/s | VRAM=32.25 Go [2026-07-17 16:11:22] step 365650/1000000 | loss=2.3018 vocab=0.7674 token=1.1990 | 6.11 steps/s | VRAM=25.03 Go [2026-07-17 16:11:30] step 365700/1000000 | loss=2.3259 vocab=0.7970 token=1.7015 | 6.12 steps/s | VRAM=28.95 Go [2026-07-17 16:11:39] step 365750/1000000 | loss=2.3705 vocab=0.9432 token=1.3015 | 6.10 steps/s | VRAM=32.67 Go [2026-07-17 16:11:47] step 365800/1000000 | loss=2.3388 vocab=0.8723 token=1.4490 | 6.12 steps/s | VRAM=29.15 Go [2026-07-17 16:11:55] step 365850/1000000 | loss=2.1986 vocab=1.0233 token=1.0435 | 6.08 steps/s | VRAM=26.72 Go [2026-07-17 16:12:03] step 365900/1000000 | loss=2.2528 vocab=1.0643 token=1.5746 | 6.12 steps/s | VRAM=29.00 Go [2026-07-17 16:12:11] step 365950/1000000 | loss=2.1757 vocab=1.0288 token=1.2520 | 6.11 steps/s | VRAM=26.29 Go [2026-07-17 16:12:19] step 366000/1000000 | loss=2.4105 vocab=1.1409 token=1.4219 | 6.11 steps/s | VRAM=28.91 Go [2026-07-17 16:12:28] step 366050/1000000 | loss=2.2960 vocab=1.1136 token=1.0469 | 6.11 steps/s | VRAM=32.82 Go [2026-07-17 16:12:36] step 366100/1000000 | loss=2.3141 vocab=0.8727 token=1.4248 | 6.08 steps/s | VRAM=26.62 Go [2026-07-17 16:12:44] step 366150/1000000 | loss=2.2215 vocab=0.9910 token=1.3206 | 6.10 steps/s | VRAM=29.76 Go [2026-07-17 16:12:52] step 366200/1000000 | loss=2.2670 vocab=1.1421 token=1.5435 | 6.09 steps/s | VRAM=28.98 Go [2026-07-17 16:13:00] step 366250/1000000 | loss=2.3457 vocab=0.4845 token=1.3383 | 6.11 steps/s | VRAM=28.44 Go [2026-07-17 16:13:09] step 366300/1000000 | loss=2.2888 vocab=0.6503 token=1.0151 | 6.10 steps/s | VRAM=32.61 Go [2026-07-17 16:13:17] step 366350/1000000 | loss=2.2690 vocab=0.7797 token=1.1924 | 6.09 steps/s | VRAM=32.73 Go [2026-07-17 16:13:25] step 366400/1000000 | loss=2.2955 vocab=0.9964 token=1.1597 | 6.09 steps/s | VRAM=28.51 Go [2026-07-17 16:13:33] step 366450/1000000 | loss=2.2900 vocab=1.1345 token=1.5662 | 6.10 steps/s | VRAM=28.69 Go [2026-07-17 16:13:41] step 366500/1000000 | loss=2.3345 vocab=0.7596 token=1.0700 | 6.09 steps/s | VRAM=27.21 Go [2026-07-17 16:13:50] step 366550/1000000 | loss=2.3950 vocab=0.7618 token=1.4545 | 6.12 steps/s | VRAM=32.62 Go [2026-07-17 16:13:58] step 366600/1000000 | loss=2.2824 vocab=0.5568 token=1.5859 | 6.11 steps/s | VRAM=28.80 Go [2026-07-17 16:14:06] step 366650/1000000 | loss=2.3123 vocab=1.3020 token=1.8138 | 6.12 steps/s | VRAM=32.61 Go [2026-07-17 16:14:14] step 366700/1000000 | loss=2.3416 vocab=0.8411 token=1.4926 | 6.09 steps/s | VRAM=26.16 Go [2026-07-17 16:14:22] step 366750/1000000 | loss=2.1834 vocab=1.4286 token=1.4803 | 6.11 steps/s | VRAM=30.90 Go [2026-07-17 16:14:31] step 366800/1000000 | loss=2.3376 vocab=1.0643 token=1.5127 | 6.09 steps/s | VRAM=29.41 Go [2026-07-17 16:14:39] step 366850/1000000 | loss=2.3043 vocab=0.6278 token=1.0485 | 6.13 steps/s | VRAM=28.68 Go [2026-07-17 16:14:47] step 366900/1000000 | loss=2.2919 vocab=0.8907 token=1.2050 | 6.12 steps/s | VRAM=32.61 Go [2026-07-17 16:14:55] step 366950/1000000 | loss=2.2772 vocab=0.8344 token=1.6035 | 6.11 steps/s | VRAM=25.90 Go [2026-07-17 16:15:03] step 367000/1000000 | loss=2.2573 vocab=0.8611 token=1.2806 | 6.12 steps/s | VRAM=31.84 Go [2026-07-17 16:15:11] step 367050/1000000 | loss=2.2832 vocab=1.2195 token=1.1936 | 6.13 steps/s | VRAM=31.35 Go [2026-07-17 16:15:20] step 367100/1000000 | loss=2.2401 vocab=0.9336 token=0.8840 | 6.09 steps/s | VRAM=29.19 Go [2026-07-17 16:15:28] step 367150/1000000 | loss=2.2017 vocab=0.7672 token=1.2584 | 6.10 steps/s | VRAM=25.55 Go [2026-07-17 16:15:36] step 367200/1000000 | loss=2.2945 vocab=0.7327 token=0.9950 | 6.12 steps/s | VRAM=28.36 Go [2026-07-17 16:15:44] step 367250/1000000 | loss=2.2588 vocab=0.7511 token=1.4592 | 6.10 steps/s | VRAM=31.77 Go [2026-07-17 16:15:52] step 367300/1000000 | loss=2.2432 vocab=1.0246 token=1.3411 | 6.10 steps/s | VRAM=32.73 Go [2026-07-17 16:16:01] step 367350/1000000 | loss=2.2305 vocab=0.9627 token=1.0232 | 6.08 steps/s | VRAM=29.72 Go [2026-07-17 16:16:09] step 367400/1000000 | loss=2.2785 vocab=1.0737 token=1.3195 | 6.13 steps/s | VRAM=31.76 Go [2026-07-17 16:16:17] step 367450/1000000 | loss=2.4512 vocab=0.8578 token=1.4924 | 6.11 steps/s | VRAM=29.50 Go [2026-07-17 16:16:25] step 367500/1000000 | loss=2.3268 vocab=0.8929 token=1.4292 | 6.08 steps/s | VRAM=26.03 Go [2026-07-17 16:16:33] step 367550/1000000 | loss=2.2642 vocab=0.4390 token=1.1935 | 6.12 steps/s | VRAM=29.83 Go [2026-07-17 16:16:42] step 367600/1000000 | loss=2.2712 vocab=0.5934 token=1.1669 | 6.10 steps/s | VRAM=26.17 Go [2026-07-17 16:16:50] step 367650/1000000 | loss=2.3073 vocab=0.8089 token=1.0901 | 6.12 steps/s | VRAM=28.53 Go [2026-07-17 16:16:58] step 367700/1000000 | loss=2.2618 vocab=0.5722 token=1.4540 | 6.15 steps/s | VRAM=31.25 Go [2026-07-17 16:17:06] step 367750/1000000 | loss=2.3164 vocab=0.7928 token=1.5727 | 6.09 steps/s | VRAM=26.57 Go [2026-07-17 16:17:14] step 367800/1000000 | loss=2.3021 vocab=1.0217 token=1.2622 | 6.11 steps/s | VRAM=31.17 Go [2026-07-17 16:17:22] step 367850/1000000 | loss=2.3473 vocab=0.9547 token=2.0101 | 6.11 steps/s | VRAM=32.30 Go [2026-07-17 16:17:31] step 367900/1000000 | loss=2.2998 vocab=0.9476 token=0.9568 | 6.11 steps/s | VRAM=28.93 Go [2026-07-17 16:17:39] step 367950/1000000 | loss=2.3057 vocab=1.0150 token=1.5774 | 6.08 steps/s | VRAM=27.07 Go [2026-07-17 16:17:47] step 368000/1000000 | loss=2.2521 vocab=1.1694 token=1.6118 | 6.12 steps/s | VRAM=25.77 Go [2026-07-17 16:17:55] step 368050/1000000 | loss=2.3095 vocab=1.1686 token=1.4685 | 6.13 steps/s | VRAM=29.42 Go [2026-07-17 16:18:03] step 368100/1000000 | loss=2.3115 vocab=0.8316 token=1.3709 | 6.14 steps/s | VRAM=31.77 Go [2026-07-17 16:18:12] step 368150/1000000 | loss=2.3133 vocab=1.4049 token=1.4244 | 6.07 steps/s | VRAM=29.38 Go [2026-07-17 16:18:20] step 368200/1000000 | loss=2.1915 vocab=0.7302 token=1.0423 | 6.08 steps/s | VRAM=29.86 Go [2026-07-17 16:18:28] step 368250/1000000 | loss=2.2597 vocab=0.9123 token=1.5037 | 6.13 steps/s | VRAM=24.95 Go [2026-07-17 16:18:36] step 368300/1000000 | loss=2.2549 vocab=0.7836 token=1.1511 | 6.13 steps/s | VRAM=32.39 Go [2026-07-17 16:18:44] step 368350/1000000 | loss=2.3486 vocab=0.7464 token=1.5360 | 6.12 steps/s | VRAM=32.22 Go [2026-07-17 16:18:53] step 368400/1000000 | loss=2.2912 vocab=0.9816 token=1.1255 | 6.07 steps/s | VRAM=26.82 Go [2026-07-17 16:19:01] step 368450/1000000 | loss=2.2112 vocab=0.4119 token=0.9197 | 6.15 steps/s | VRAM=25.19 Go [2026-07-17 16:19:09] step 368500/1000000 | loss=2.3031 vocab=1.0875 token=0.9117 | 6.13 steps/s | VRAM=28.13 Go [2026-07-17 16:19:17] step 368550/1000000 | loss=2.3095 vocab=0.8463 token=1.6076 | 6.11 steps/s | VRAM=28.45 Go [2026-07-17 16:19:25] step 368600/1000000 | loss=2.2990 vocab=1.1444 token=1.4316 | 6.10 steps/s | VRAM=25.91 Go [2026-07-17 16:19:33] step 368650/1000000 | loss=2.2507 vocab=0.5173 token=1.0220 | 6.13 steps/s | VRAM=32.02 Go [2026-07-17 16:19:42] step 368700/1000000 | loss=2.3117 vocab=0.6707 token=1.0245 | 6.10 steps/s | VRAM=29.22 Go [2026-07-17 16:19:50] step 368750/1000000 | loss=2.2681 vocab=0.9373 token=1.5586 | 6.13 steps/s | VRAM=32.43 Go [2026-07-17 16:19:58] step 368800/1000000 | loss=2.3323 vocab=0.9526 token=1.2171 | 6.11 steps/s | VRAM=31.39 Go [2026-07-17 16:20:06] step 368850/1000000 | loss=2.2294 vocab=1.2896 token=1.4734 | 6.12 steps/s | VRAM=32.47 Go [2026-07-17 16:20:14] step 368900/1000000 | loss=2.2370 vocab=0.7865 token=1.1265 | 6.09 steps/s | VRAM=26.52 Go [2026-07-17 16:20:22] step 368950/1000000 | loss=2.3042 vocab=0.7728 token=1.2580 | 6.12 steps/s | VRAM=32.17 Go [2026-07-17 16:20:31] step 369000/1000000 | loss=2.3799 vocab=0.8067 token=1.3185 | 6.12 steps/s | VRAM=24.69 Go [2026-07-17 16:20:39] step 369050/1000000 | loss=2.3475 vocab=0.8467 token=1.3090 | 6.08 steps/s | VRAM=28.95 Go [2026-07-17 16:20:47] step 369100/1000000 | loss=2.3413 vocab=0.7984 token=1.8267 | 6.14 steps/s | VRAM=32.17 Go [2026-07-17 16:20:55] step 369150/1000000 | loss=2.2500 vocab=0.8179 token=1.4247 | 6.12 steps/s | VRAM=32.76 Go [2026-07-17 16:21:03] step 369200/1000000 | loss=2.2032 vocab=0.8134 token=1.6753 | 6.09 steps/s | VRAM=28.13 Go [2026-07-17 16:21:12] step 369250/1000000 | loss=2.3176 vocab=1.3707 token=1.2204 | 6.09 steps/s | VRAM=28.75 Go [2026-07-17 16:21:20] step 369300/1000000 | loss=2.2295 vocab=0.7019 token=0.7876 | 6.11 steps/s | VRAM=32.23 Go [2026-07-17 16:21:28] step 369350/1000000 | loss=2.3411 vocab=1.0377 token=1.1689 | 6.12 steps/s | VRAM=27.98 Go [2026-07-17 16:21:36] step 369400/1000000 | loss=2.2806 vocab=0.8310 token=1.2188 | 6.09 steps/s | VRAM=26.53 Go [2026-07-17 16:21:44] step 369450/1000000 | loss=2.2920 vocab=0.5767 token=1.5940 | 6.14 steps/s | VRAM=31.57 Go [2026-07-17 16:21:52] step 369500/1000000 | loss=2.1786 vocab=0.7738 token=1.6800 | 6.13 steps/s | VRAM=26.06 Go [2026-07-17 16:22:01] step 369550/1000000 | loss=2.2577 vocab=0.8105 token=1.3506 | 6.15 steps/s | VRAM=32.38 Go [2026-07-17 16:22:09] step 369600/1000000 | loss=2.2995 vocab=0.5782 token=1.3468 | 6.09 steps/s | VRAM=28.72 Go [2026-07-17 16:22:17] step 369650/1000000 | loss=2.2533 vocab=1.2291 token=1.6523 | 6.11 steps/s | VRAM=26.50 Go [2026-07-17 16:22:25] step 369700/1000000 | loss=2.2885 vocab=0.6761 token=1.3701 | 6.12 steps/s | VRAM=32.04 Go [2026-07-17 16:22:33] step 369750/1000000 | loss=2.2965 vocab=1.3420 token=1.4400 | 6.12 steps/s | VRAM=26.27 Go [2026-07-17 16:22:41] step 369800/1000000 | loss=2.2030 vocab=0.6524 token=1.0811 | 6.10 steps/s | VRAM=26.27 Go [2026-07-17 16:22:50] step 369850/1000000 | loss=2.2419 vocab=0.6552 token=1.0752 | 6.12 steps/s | VRAM=31.95 Go [2026-07-17 16:22:58] step 369900/1000000 | loss=2.4251 vocab=1.2605 token=1.4235 | 6.12 steps/s | VRAM=25.90 Go [2026-07-17 16:23:06] step 369950/1000000 | loss=2.3877 vocab=0.9323 token=1.4824 | 6.10 steps/s | VRAM=29.24 Go [2026-07-17 16:23:14] step 370000/1000000 | loss=2.3455 vocab=0.7876 token=1.2019 | 6.11 steps/s | VRAM=28.29 Go [2026-07-17 16:23:22] step 370050/1000000 | loss=2.2116 vocab=0.8748 token=1.3882 | 6.13 steps/s | VRAM=31.81 Go [2026-07-17 16:23:30] step 370100/1000000 | loss=2.2128 vocab=0.7612 token=1.2287 | 6.13 steps/s | VRAM=31.48 Go [2026-07-17 16:23:39] step 370150/1000000 | loss=2.3619 vocab=1.2394 token=1.5659 | 6.11 steps/s | VRAM=29.03 Go [2026-07-17 16:23:47] step 370200/1000000 | loss=2.3453 vocab=0.8355 token=1.4025 | 6.09 steps/s | VRAM=26.18 Go [2026-07-17 16:23:55] step 370250/1000000 | loss=2.1213 vocab=0.6250 token=1.0437 | 6.13 steps/s | VRAM=31.17 Go [2026-07-17 16:24:03] step 370300/1000000 | loss=2.4800 vocab=1.8591 token=1.8923 | 6.14 steps/s | VRAM=28.37 Go [2026-07-17 16:24:11] step 370350/1000000 | loss=2.3295 vocab=1.3166 token=1.3091 | 6.13 steps/s | VRAM=25.06 Go [2026-07-17 16:24:20] step 370400/1000000 | loss=2.2703 vocab=0.9555 token=1.3824 | 6.11 steps/s | VRAM=28.00 Go [2026-07-17 16:24:28] step 370450/1000000 | loss=2.3942 vocab=0.6713 token=1.5895 | 6.09 steps/s | VRAM=29.35 Go [2026-07-17 16:24:36] step 370500/1000000 | loss=2.2612 vocab=1.1946 token=1.1809 | 6.12 steps/s | VRAM=32.71 Go [2026-07-17 16:24:44] step 370550/1000000 | loss=2.2136 vocab=0.7333 token=1.5668 | 6.09 steps/s | VRAM=26.59 Go [2026-07-17 16:24:52] step 370600/1000000 | loss=2.2612 vocab=0.7240 token=1.2431 | 6.13 steps/s | VRAM=28.86 Go [2026-07-17 16:25:00] step 370650/1000000 | loss=2.3455 vocab=1.1773 token=1.8199 | 6.10 steps/s | VRAM=32.47 Go [2026-07-17 16:25:09] step 370700/1000000 | loss=2.3103 vocab=0.7476 token=1.0205 | 6.09 steps/s | VRAM=25.77 Go [2026-07-17 16:25:17] step 370750/1000000 | loss=2.2487 vocab=0.5351 token=0.8291 | 6.11 steps/s | VRAM=23.24 Go [2026-07-17 16:25:25] step 370800/1000000 | loss=2.2146 vocab=0.4618 token=1.1862 | 6.11 steps/s | VRAM=26.61 Go [2026-07-17 16:25:33] step 370850/1000000 | loss=2.2543 vocab=0.9889 token=1.5066 | 6.12 steps/s | VRAM=28.40 Go [2026-07-17 16:25:41] step 370900/1000000 | loss=2.2969 vocab=0.6686 token=1.0708 | 6.14 steps/s | VRAM=28.63 Go [2026-07-17 16:25:50] step 370950/1000000 | loss=2.3497 vocab=1.3078 token=1.3422 | 6.12 steps/s | VRAM=28.60 Go [2026-07-17 16:25:58] step 371000/1000000 | loss=2.2674 vocab=0.7670 token=1.4684 | 6.11 steps/s | VRAM=25.86 Go [2026-07-17 16:26:06] step 371050/1000000 | loss=2.2099 vocab=0.6546 token=1.4366 | 6.15 steps/s | VRAM=30.50 Go [2026-07-17 16:26:14] step 371100/1000000 | loss=2.2452 vocab=1.0393 token=1.1374 | 6.13 steps/s | VRAM=29.32 Go [2026-07-17 16:26:22] step 371150/1000000 | loss=2.3431 vocab=0.8401 token=1.5517 | 6.12 steps/s | VRAM=26.36 Go [2026-07-17 16:26:30] step 371200/1000000 | loss=2.2845 vocab=0.9462 token=1.1538 | 6.11 steps/s | VRAM=26.21 Go [2026-07-17 16:26:39] step 371250/1000000 | loss=2.4166 vocab=0.7435 token=1.2438 | 6.11 steps/s | VRAM=32.26 Go [2026-07-17 16:26:47] step 371300/1000000 | loss=2.2481 vocab=1.1291 token=1.5908 | 6.15 steps/s | VRAM=32.26 Go [2026-07-17 16:26:55] step 371350/1000000 | loss=2.3634 vocab=0.8417 token=1.1884 | 6.13 steps/s | VRAM=31.25 Go [2026-07-17 16:27:03] step 371400/1000000 | loss=2.1722 vocab=1.0875 token=1.4282 | 6.13 steps/s | VRAM=31.61 Go [2026-07-17 16:27:11] step 371450/1000000 | loss=2.2856 vocab=0.9002 token=1.2003 | 6.11 steps/s | VRAM=26.33 Go [2026-07-17 16:27:19] step 371500/1000000 | loss=2.3586 vocab=1.0830 token=1.2483 | 6.10 steps/s | VRAM=25.58 Go [2026-07-17 16:27:28] step 371550/1000000 | loss=2.2833 vocab=0.6346 token=1.3386 | 6.11 steps/s | VRAM=28.48 Go [2026-07-17 16:27:36] step 371600/1000000 | loss=2.2289 vocab=0.8847 token=0.8525 | 6.11 steps/s | VRAM=28.46 Go [2026-07-17 16:27:44] step 371650/1000000 | loss=2.3316 vocab=2.1388 token=1.1923 | 6.08 steps/s | VRAM=27.25 Go [2026-07-17 16:27:52] step 371700/1000000 | loss=2.2951 vocab=0.7147 token=1.2388 | 6.13 steps/s | VRAM=25.72 Go [2026-07-17 16:28:00] step 371750/1000000 | loss=2.3329 vocab=1.0434 token=1.4630 | 6.13 steps/s | VRAM=29.66 Go [2026-07-17 16:28:08] step 371800/1000000 | loss=2.2888 vocab=0.8015 token=1.0874 | 6.15 steps/s | VRAM=32.82 Go [2026-07-17 16:28:17] step 371850/1000000 | loss=2.3032 vocab=1.2071 token=1.7197 | 6.11 steps/s | VRAM=28.33 Go [2026-07-17 16:28:25] step 371900/1000000 | loss=2.2698 vocab=1.1943 token=1.7243 | 6.09 steps/s | VRAM=26.87 Go [2026-07-17 16:28:33] step 371950/1000000 | loss=2.3791 vocab=1.3369 token=1.6753 | 6.13 steps/s | VRAM=27.15 Go [2026-07-17 16:28:41] step 372000/1000000 | loss=2.3032 vocab=0.8053 token=1.1640 | 6.08 steps/s | VRAM=26.00 Go [2026-07-17 16:28:49] step 372050/1000000 | loss=2.2411 vocab=1.0889 token=1.5270 | 6.13 steps/s | VRAM=27.38 Go [2026-07-17 16:28:58] step 372100/1000000 | loss=2.2455 vocab=0.6445 token=1.3124 | 6.10 steps/s | VRAM=26.31 Go [2026-07-17 16:29:06] step 372150/1000000 | loss=2.3140 vocab=0.9245 token=1.1107 | 6.12 steps/s | VRAM=32.41 Go [2026-07-17 16:29:14] step 372200/1000000 | loss=2.3068 vocab=0.8986 token=0.8983 | 6.13 steps/s | VRAM=24.99 Go [2026-07-17 16:29:22] step 372250/1000000 | loss=2.1904 vocab=1.3647 token=1.7891 | 6.12 steps/s | VRAM=28.91 Go [2026-07-17 16:29:30] step 372300/1000000 | loss=2.3096 vocab=0.7462 token=1.3064 | 6.13 steps/s | VRAM=31.46 Go [2026-07-17 16:29:38] step 372350/1000000 | loss=2.2426 vocab=0.9893 token=1.5966 | 6.09 steps/s | VRAM=23.00 Go [2026-07-17 16:29:47] step 372400/1000000 | loss=2.3330 vocab=1.4209 token=1.3189 | 6.12 steps/s | VRAM=28.65 Go [2026-07-17 16:29:55] step 372450/1000000 | loss=2.3998 vocab=1.0836 token=1.1129 | 6.09 steps/s | VRAM=28.33 Go [2026-07-17 16:30:03] step 372500/1000000 | loss=2.3011 vocab=1.2266 token=1.4402 | 6.13 steps/s | VRAM=25.61 Go [2026-07-17 16:30:11] step 372550/1000000 | loss=2.2036 vocab=0.6905 token=1.3841 | 6.13 steps/s | VRAM=31.46 Go [2026-07-17 16:30:19] step 372600/1000000 | loss=2.2537 vocab=0.8616 token=1.2967 | 6.13 steps/s | VRAM=31.63 Go [2026-07-17 16:30:27] step 372650/1000000 | loss=2.3008 vocab=0.9124 token=0.9747 | 6.12 steps/s | VRAM=29.39 Go [2026-07-17 16:30:36] step 372700/1000000 | loss=2.2543 vocab=0.5220 token=1.4517 | 6.14 steps/s | VRAM=32.39 Go [2026-07-17 16:30:44] step 372750/1000000 | loss=2.2518 vocab=1.3262 token=1.5230 | 6.12 steps/s | VRAM=31.77 Go [2026-07-17 16:30:52] step 372800/1000000 | loss=2.1827 vocab=0.9946 token=1.4603 | 6.08 steps/s | VRAM=26.97 Go [2026-07-17 16:31:00] step 372850/1000000 | loss=2.1805 vocab=1.3362 token=1.1531 | 6.13 steps/s | VRAM=32.98 Go [2026-07-17 16:31:08] step 372900/1000000 | loss=2.3350 vocab=0.7208 token=1.4411 | 6.12 steps/s | VRAM=29.32 Go [2026-07-17 16:31:16] step 372950/1000000 | loss=2.3985 vocab=0.8881 token=1.5782 | 6.09 steps/s | VRAM=23.33 Go [2026-07-17 16:31:25] step 373000/1000000 | loss=2.2706 vocab=0.7278 token=1.3595 | 6.13 steps/s | VRAM=27.99 Go [2026-07-17 16:31:33] step 373050/1000000 | loss=2.3262 vocab=0.5588 token=1.0947 | 6.09 steps/s | VRAM=26.45 Go [2026-07-17 16:31:41] step 373100/1000000 | loss=2.2772 vocab=0.8840 token=1.6719 | 6.13 steps/s | VRAM=31.42 Go [2026-07-17 16:31:49] step 373150/1000000 | loss=2.2999 vocab=0.6851 token=1.1935 | 6.09 steps/s | VRAM=25.47 Go [2026-07-17 16:31:57] step 373200/1000000 | loss=2.3071 vocab=0.7868 token=1.3703 | 6.12 steps/s | VRAM=31.56 Go [2026-07-17 16:32:06] step 373250/1000000 | loss=2.3484 vocab=0.4209 token=0.6878 | 6.11 steps/s | VRAM=29.81 Go [2026-07-17 16:32:14] step 373300/1000000 | loss=2.3025 vocab=1.5200 token=1.3681 | 6.13 steps/s | VRAM=28.90 Go [2026-07-17 16:32:22] step 373350/1000000 | loss=2.3159 vocab=1.1321 token=1.5537 | 6.10 steps/s | VRAM=32.93 Go [2026-07-17 16:32:30] step 373400/1000000 | loss=2.3477 vocab=0.7585 token=1.1483 | 6.12 steps/s | VRAM=32.54 Go [2026-07-17 16:32:38] step 373450/1000000 | loss=2.4270 vocab=1.1002 token=1.9549 | 6.08 steps/s | VRAM=29.75 Go [2026-07-17 16:32:46] step 373500/1000000 | loss=2.3137 vocab=0.8843 token=1.4371 | 6.13 steps/s | VRAM=32.16 Go [2026-07-17 16:32:55] step 373550/1000000 | loss=2.3139 vocab=0.9236 token=1.4167 | 6.12 steps/s | VRAM=26.26 Go [2026-07-17 16:33:03] step 373600/1000000 | loss=2.3020 vocab=0.8677 token=1.1849 | 6.11 steps/s | VRAM=28.16 Go [2026-07-17 16:33:11] step 373650/1000000 | loss=2.3901 vocab=0.8559 token=1.2478 | 6.15 steps/s | VRAM=30.95 Go [2026-07-17 16:33:19] step 373700/1000000 | loss=2.3114 vocab=0.8285 token=1.2796 | 6.15 steps/s | VRAM=31.99 Go [2026-07-17 16:33:27] step 373750/1000000 | loss=2.2984 vocab=0.8382 token=1.4559 | 6.11 steps/s | VRAM=31.55 Go [2026-07-17 16:33:35] step 373800/1000000 | loss=2.3272 vocab=0.3907 token=1.4053 | 6.10 steps/s | VRAM=26.49 Go [2026-07-17 16:33:44] step 373850/1000000 | loss=2.2156 vocab=0.4061 token=1.3788 | 6.11 steps/s | VRAM=28.48 Go [2026-07-17 16:33:52] step 373900/1000000 | loss=2.3278 vocab=1.5032 token=1.6403 | 6.11 steps/s | VRAM=23.43 Go [2026-07-17 16:34:00] step 373950/1000000 | loss=2.2482 vocab=0.7754 token=0.7555 | 6.13 steps/s | VRAM=32.66 Go [2026-07-17 16:34:08] step 374000/1000000 | loss=2.2237 vocab=0.9638 token=1.5521 | 6.12 steps/s | VRAM=28.44 Go [2026-07-17 16:34:16] step 374050/1000000 | loss=2.3520 vocab=1.3511 token=1.8574 | 6.12 steps/s | VRAM=32.41 Go [2026-07-17 16:34:24] step 374100/1000000 | loss=2.2484 vocab=1.1363 token=1.6052 | 6.10 steps/s | VRAM=28.96 Go [2026-07-17 16:34:33] step 374150/1000000 | loss=2.2832 vocab=0.8679 token=1.4435 | 6.13 steps/s | VRAM=27.89 Go [2026-07-17 16:34:41] step 374200/1000000 | loss=2.1884 vocab=1.1166 token=1.4725 | 6.10 steps/s | VRAM=29.10 Go [2026-07-17 16:34:49] step 374250/1000000 | loss=2.2561 vocab=0.9383 token=1.3028 | 6.11 steps/s | VRAM=28.42 Go [2026-07-17 16:34:57] step 374300/1000000 | loss=2.2732 vocab=0.6499 token=0.6434 | 6.10 steps/s | VRAM=28.86 Go [2026-07-17 16:35:05] step 374350/1000000 | loss=2.3982 vocab=1.4633 token=1.4533 | 6.10 steps/s | VRAM=28.94 Go [2026-07-17 16:35:14] step 374400/1000000 | loss=2.2521 vocab=1.0270 token=0.9907 | 6.10 steps/s | VRAM=26.92 Go [2026-07-17 16:35:22] step 374450/1000000 | loss=2.2860 vocab=0.7887 token=1.3756 | 6.14 steps/s | VRAM=31.82 Go [2026-07-17 16:35:30] step 374500/1000000 | loss=2.3473 vocab=1.5588 token=1.4982 | 6.13 steps/s | VRAM=29.11 Go [2026-07-17 16:35:38] step 374550/1000000 | loss=2.3872 vocab=1.1636 token=1.4398 | 6.12 steps/s | VRAM=28.06 Go [2026-07-17 16:35:46] step 374600/1000000 | loss=2.2432 vocab=0.6971 token=1.2220 | 6.13 steps/s | VRAM=31.74 Go [2026-07-17 16:35:54] step 374650/1000000 | loss=2.3230 vocab=0.7151 token=0.8265 | 6.14 steps/s | VRAM=26.60 Go [2026-07-17 16:36:03] step 374700/1000000 | loss=2.3271 vocab=0.8019 token=1.5575 | 6.15 steps/s | VRAM=27.95 Go [2026-07-17 16:36:11] step 374750/1000000 | loss=2.2662 vocab=0.6754 token=1.3267 | 6.08 steps/s | VRAM=26.38 Go [2026-07-17 16:36:19] step 374800/1000000 | loss=2.3319 vocab=0.8378 token=1.3239 | 6.11 steps/s | VRAM=31.50 Go [2026-07-17 16:36:27] step 374850/1000000 | loss=2.3218 vocab=0.9205 token=1.3407 | 6.09 steps/s | VRAM=26.48 Go [2026-07-17 16:36:35] step 374900/1000000 | loss=2.2508 vocab=0.8518 token=1.4213 | 6.13 steps/s | VRAM=24.85 Go [2026-07-17 16:36:43] step 374950/1000000 | loss=2.3080 vocab=1.1718 token=1.4623 | 6.12 steps/s | VRAM=26.40 Go [2026-07-17 16:36:52] step 375000/1000000 | loss=2.2641 vocab=1.2653 token=1.4561 | 6.09 steps/s | VRAM=28.24 Go [2026-07-17 16:37:00] step 375050/1000000 | loss=2.3228 vocab=1.0993 token=1.2804 | 6.13 steps/s | VRAM=28.37 Go [2026-07-17 16:37:08] step 375100/1000000 | loss=2.2981 vocab=0.3983 token=1.1173 | 6.13 steps/s | VRAM=28.69 Go [2026-07-17 16:37:16] step 375150/1000000 | loss=2.2075 vocab=0.9337 token=1.5675 | 6.13 steps/s | VRAM=27.83 Go [2026-07-17 16:37:24] step 375200/1000000 | loss=2.3575 vocab=0.6784 token=1.4136 | 6.08 steps/s | VRAM=25.86 Go [2026-07-17 16:37:33] step 375250/1000000 | loss=2.3492 vocab=0.9800 token=1.3995 | 6.12 steps/s | VRAM=27.59 Go [2026-07-17 16:37:41] step 375300/1000000 | loss=2.3079 vocab=1.1413 token=1.8386 | 6.11 steps/s | VRAM=26.44 Go [2026-07-17 16:37:49] step 375350/1000000 | loss=2.3370 vocab=0.7202 token=1.1237 | 6.10 steps/s | VRAM=29.46 Go [2026-07-17 16:37:57] step 375400/1000000 | loss=2.3129 vocab=0.8814 token=1.5394 | 6.12 steps/s | VRAM=28.24 Go [2026-07-17 16:38:05] step 375450/1000000 | loss=2.3362 vocab=0.7050 token=1.4602 | 6.11 steps/s | VRAM=26.24 Go [2026-07-17 16:38:13] step 375500/1000000 | loss=2.2297 vocab=0.8937 token=1.1604 | 6.12 steps/s | VRAM=27.90 Go [2026-07-17 16:38:22] step 375550/1000000 | loss=2.3084 vocab=0.6179 token=1.0963 | 6.13 steps/s | VRAM=28.03 Go [2026-07-17 16:38:30] step 375600/1000000 | loss=2.2942 vocab=0.5037 token=1.4411 | 6.13 steps/s | VRAM=32.12 Go [2026-07-17 16:38:38] step 375650/1000000 | loss=2.2393 vocab=1.9153 token=1.3610 | 6.10 steps/s | VRAM=28.76 Go [2026-07-17 16:38:46] step 375700/1000000 | loss=2.3496 vocab=0.8452 token=1.5831 | 6.08 steps/s | VRAM=27.76 Go [2026-07-17 16:38:54] step 375750/1000000 | loss=2.3342 vocab=1.3378 token=1.4413 | 6.11 steps/s | VRAM=31.28 Go [2026-07-17 16:39:03] step 375800/1000000 | loss=2.2363 vocab=0.9127 token=1.3872 | 6.10 steps/s | VRAM=32.09 Go [2026-07-17 16:39:11] step 375850/1000000 | loss=2.1984 vocab=0.5535 token=1.2599 | 6.11 steps/s | VRAM=32.29 Go [2026-07-17 16:39:19] step 375900/1000000 | loss=2.2115 vocab=1.0830 token=1.4917 | 6.09 steps/s | VRAM=23.45 Go [2026-07-17 16:39:27] step 375950/1000000 | loss=2.2122 vocab=1.4371 token=1.5982 | 6.16 steps/s | VRAM=24.22 Go [2026-07-17 16:39:35] step 376000/1000000 | loss=2.1909 vocab=0.5538 token=1.1272 | 6.10 steps/s | VRAM=27.46 Go [2026-07-17 16:39:43] step 376050/1000000 | loss=2.2618 vocab=1.1035 token=1.2835 | 6.14 steps/s | VRAM=32.64 Go [2026-07-17 16:39:52] step 376100/1000000 | loss=2.2651 vocab=1.2772 token=1.0653 | 6.08 steps/s | VRAM=29.74 Go [2026-07-17 16:40:00] step 376150/1000000 | loss=2.3956 vocab=0.7893 token=1.0297 | 6.09 steps/s | VRAM=29.32 Go [2026-07-17 16:40:08] step 376200/1000000 | loss=2.3129 vocab=0.8337 token=1.1279 | 6.11 steps/s | VRAM=28.33 Go [2026-07-17 16:40:16] step 376250/1000000 | loss=2.3310 vocab=0.6134 token=1.2536 | 6.10 steps/s | VRAM=33.01 Go [2026-07-17 16:40:24] step 376300/1000000 | loss=2.3645 vocab=1.2594 token=1.4107 | 6.12 steps/s | VRAM=28.64 Go [2026-07-17 16:40:33] step 376350/1000000 | loss=2.2865 vocab=0.5806 token=1.2763 | 6.11 steps/s | VRAM=31.69 Go [2026-07-17 16:40:41] step 376400/1000000 | loss=2.1943 vocab=0.5561 token=0.9291 | 6.14 steps/s | VRAM=32.38 Go [2026-07-17 16:40:49] step 376450/1000000 | loss=2.2189 vocab=1.0620 token=1.2269 | 6.12 steps/s | VRAM=28.94 Go [2026-07-17 16:40:57] step 376500/1000000 | loss=2.2285 vocab=1.1711 token=1.3351 | 6.10 steps/s | VRAM=28.66 Go [2026-07-17 16:41:05] step 376550/1000000 | loss=2.2922 vocab=0.9154 token=1.3420 | 6.11 steps/s | VRAM=26.22 Go [2026-07-17 16:41:13] step 376600/1000000 | loss=2.2206 vocab=0.6410 token=1.1473 | 6.14 steps/s | VRAM=29.19 Go [2026-07-17 16:41:22] step 376650/1000000 | loss=2.3394 vocab=1.0085 token=1.5206 | 6.14 steps/s | VRAM=31.35 Go [2026-07-17 16:41:30] step 376700/1000000 | loss=2.3720 vocab=0.8712 token=1.3746 | 6.12 steps/s | VRAM=23.12 Go [2026-07-17 16:41:38] step 376750/1000000 | loss=2.2938 vocab=0.9818 token=1.4051 | 6.11 steps/s | VRAM=29.88 Go [2026-07-17 16:41:46] step 376800/1000000 | loss=2.3251 vocab=0.9814 token=1.4131 | 6.13 steps/s | VRAM=28.23 Go [2026-07-17 16:41:54] step 376850/1000000 | loss=2.1363 vocab=0.7510 token=1.2410 | 6.12 steps/s | VRAM=28.37 Go [2026-07-17 16:42:02] step 376900/1000000 | loss=2.2901 vocab=0.9891 token=1.2164 | 6.14 steps/s | VRAM=31.94 Go [2026-07-17 16:42:11] step 376950/1000000 | loss=2.2673 vocab=0.9324 token=1.6851 | 6.11 steps/s | VRAM=28.10 Go [2026-07-17 16:42:19] step 377000/1000000 | loss=2.3052 vocab=0.7208 token=1.0980 | 6.13 steps/s | VRAM=29.12 Go [2026-07-17 16:42:27] step 377050/1000000 | loss=2.2503 vocab=0.7037 token=1.6478 | 6.14 steps/s | VRAM=32.48 Go [2026-07-17 16:42:35] step 377100/1000000 | loss=2.4266 vocab=0.7893 token=1.5885 | 6.13 steps/s | VRAM=28.96 Go [2026-07-17 16:42:43] step 377150/1000000 | loss=2.3394 vocab=0.9315 token=1.6452 | 6.08 steps/s | VRAM=31.16 Go [2026-07-17 16:42:51] step 377200/1000000 | loss=2.3990 vocab=0.8972 token=1.3282 | 6.10 steps/s | VRAM=28.53 Go [2026-07-17 16:43:00] step 377250/1000000 | loss=2.5035 vocab=1.8208 token=1.5924 | 6.12 steps/s | VRAM=25.41 Go [2026-07-17 16:43:08] step 377300/1000000 | loss=2.2887 vocab=0.9132 token=1.4527 | 6.16 steps/s | VRAM=31.30 Go [2026-07-17 16:43:16] step 377350/1000000 | loss=2.3366 vocab=1.2378 token=1.2377 | 6.10 steps/s | VRAM=32.62 Go [2026-07-17 16:43:24] step 377400/1000000 | loss=2.2536 vocab=1.4755 token=1.6607 | 6.13 steps/s | VRAM=30.19 Go [2026-07-17 16:43:32] step 377450/1000000 | loss=2.2099 vocab=1.2605 token=1.4167 | 6.11 steps/s | VRAM=28.34 Go [2026-07-17 16:43:40] step 377500/1000000 | loss=2.1680 vocab=0.5192 token=1.5415 | 6.11 steps/s | VRAM=28.49 Go [2026-07-17 16:43:49] step 377550/1000000 | loss=2.3400 vocab=0.8310 token=1.1762 | 6.13 steps/s | VRAM=31.42 Go [2026-07-17 16:43:57] step 377600/1000000 | loss=2.3036 vocab=0.9343 token=1.1066 | 6.12 steps/s | VRAM=27.84 Go [2026-07-17 16:44:05] step 377650/1000000 | loss=2.3950 vocab=0.8910 token=1.2784 | 6.11 steps/s | VRAM=28.53 Go [2026-07-17 16:44:13] step 377700/1000000 | loss=2.3155 vocab=0.8482 token=1.5694 | 6.08 steps/s | VRAM=25.30 Go [2026-07-17 16:44:21] step 377750/1000000 | loss=2.2045 vocab=0.4661 token=1.1324 | 6.14 steps/s | VRAM=28.48 Go [2026-07-17 16:44:30] step 377800/1000000 | loss=2.4297 vocab=0.9086 token=1.2648 | 6.12 steps/s | VRAM=26.42 Go [2026-07-17 16:44:38] step 377850/1000000 | loss=2.2883 vocab=0.9293 token=1.6629 | 6.13 steps/s | VRAM=28.75 Go [2026-07-17 16:44:46] step 377900/1000000 | loss=2.2919 vocab=1.2791 token=1.1890 | 6.11 steps/s | VRAM=29.31 Go [2026-07-17 16:44:54] step 377950/1000000 | loss=2.2038 vocab=0.9030 token=1.2793 | 6.15 steps/s | VRAM=30.74 Go [2026-07-17 16:45:02] step 378000/1000000 | loss=2.2417 vocab=0.5029 token=1.1241 | 6.11 steps/s | VRAM=26.97 Go [2026-07-17 16:45:10] step 378050/1000000 | loss=2.2530 vocab=0.6487 token=1.2308 | 6.12 steps/s | VRAM=27.61 Go [2026-07-17 16:45:18] step 378100/1000000 | loss=2.3825 vocab=0.9168 token=2.1177 | 6.14 steps/s | VRAM=25.87 Go [2026-07-17 16:45:27] step 378150/1000000 | loss=2.1869 vocab=0.8492 token=1.3337 | 6.13 steps/s | VRAM=32.04 Go [2026-07-17 16:45:35] step 378200/1000000 | loss=2.2971 vocab=0.7732 token=1.2866 | 6.13 steps/s | VRAM=29.53 Go [2026-07-17 16:45:43] step 378250/1000000 | loss=2.2873 vocab=1.2651 token=1.6165 | 6.12 steps/s | VRAM=25.36 Go [2026-07-17 16:45:51] step 378300/1000000 | loss=2.2607 vocab=1.2506 token=1.5154 | 6.15 steps/s | VRAM=31.66 Go [2026-07-17 16:45:59] step 378350/1000000 | loss=2.2783 vocab=0.9088 token=1.5487 | 6.13 steps/s | VRAM=25.87 Go [2026-07-17 16:46:07] step 378400/1000000 | loss=2.2511 vocab=0.7381 token=0.9989 | 6.10 steps/s | VRAM=29.41 Go [2026-07-17 16:46:16] step 378450/1000000 | loss=2.2837 vocab=0.5700 token=1.3008 | 6.11 steps/s | VRAM=26.55 Go [2026-07-17 16:46:24] step 378500/1000000 | loss=2.2480 vocab=0.9255 token=1.2646 | 6.10 steps/s | VRAM=29.05 Go [2026-07-17 16:46:32] step 378550/1000000 | loss=2.2110 vocab=0.9967 token=1.5858 | 6.15 steps/s | VRAM=31.10 Go [2026-07-17 16:46:40] step 378600/1000000 | loss=2.2859 vocab=1.0091 token=1.5315 | 6.13 steps/s | VRAM=27.73 Go [2026-07-17 16:46:48] step 378650/1000000 | loss=2.3847 vocab=1.1340 token=1.6281 | 6.10 steps/s | VRAM=31.48 Go [2026-07-17 16:46:56] step 378700/1000000 | loss=2.2304 vocab=0.6001 token=0.7856 | 6.13 steps/s | VRAM=28.93 Go [2026-07-17 16:47:05] step 378750/1000000 | loss=2.2366 vocab=0.7519 token=1.3801 | 6.11 steps/s | VRAM=29.78 Go [2026-07-17 16:47:13] step 378800/1000000 | loss=2.2683 vocab=1.3428 token=1.2830 | 6.13 steps/s | VRAM=31.73 Go [2026-07-17 16:47:21] step 378850/1000000 | loss=2.3479 vocab=0.8589 token=1.4201 | 6.13 steps/s | VRAM=32.99 Go [2026-07-17 16:47:29] step 378900/1000000 | loss=2.1643 vocab=1.0736 token=1.5223 | 6.07 steps/s | VRAM=29.68 Go [2026-07-17 16:47:37] step 378950/1000000 | loss=2.2704 vocab=0.8189 token=1.2925 | 6.11 steps/s | VRAM=26.10 Go [2026-07-17 16:47:46] step 379000/1000000 | loss=2.2585 vocab=0.7371 token=1.4293 | 6.13 steps/s | VRAM=24.92 Go [2026-07-17 16:47:54] step 379050/1000000 | loss=2.3174 vocab=1.0892 token=1.5768 | 6.11 steps/s | VRAM=25.05 Go [2026-07-17 16:48:02] step 379100/1000000 | loss=2.3100 vocab=0.9228 token=1.0986 | 6.14 steps/s | VRAM=32.05 Go [2026-07-17 16:48:10] step 379150/1000000 | loss=2.3560 vocab=1.2915 token=1.6904 | 6.09 steps/s | VRAM=28.76 Go [2026-07-17 16:48:18] step 379200/1000000 | loss=2.2309 vocab=0.8918 token=1.6589 | 6.12 steps/s | VRAM=26.31 Go [2026-07-17 16:48:26] step 379250/1000000 | loss=2.4280 vocab=0.7982 token=1.3121 | 6.12 steps/s | VRAM=28.57 Go [2026-07-17 16:48:35] step 379300/1000000 | loss=2.1656 vocab=0.7174 token=1.0585 | 6.13 steps/s | VRAM=32.42 Go [2026-07-17 16:48:43] step 379350/1000000 | loss=2.2194 vocab=1.0802 token=1.4109 | 6.09 steps/s | VRAM=25.82 Go [2026-07-17 16:48:51] step 379400/1000000 | loss=2.3358 vocab=1.0094 token=1.3843 | 6.14 steps/s | VRAM=32.29 Go [2026-07-17 16:48:59] step 379450/1000000 | loss=2.3119 vocab=1.0122 token=1.2076 | 6.12 steps/s | VRAM=31.50 Go [2026-07-17 16:49:07] step 379500/1000000 | loss=2.3336 vocab=1.4050 token=1.5147 | 6.11 steps/s | VRAM=26.71 Go [2026-07-17 16:49:15] step 379550/1000000 | loss=2.1910 vocab=1.0101 token=1.4676 | 6.15 steps/s | VRAM=30.24 Go [2026-07-17 16:49:24] step 379600/1000000 | loss=2.2362 vocab=0.9183 token=1.4710 | 6.14 steps/s | VRAM=31.77 Go [2026-07-17 16:49:32] step 379650/1000000 | loss=2.3882 vocab=0.8318 token=1.4204 | 6.11 steps/s | VRAM=28.26 Go [2026-07-17 16:49:40] step 379700/1000000 | loss=2.2579 vocab=1.0399 token=1.9692 | 6.09 steps/s | VRAM=29.20 Go [2026-07-17 16:49:48] step 379750/1000000 | loss=2.2713 vocab=0.8444 token=1.2197 | 6.14 steps/s | VRAM=32.14 Go [2026-07-17 16:49:56] step 379800/1000000 | loss=2.3302 vocab=0.9884 token=1.3114 | 6.12 steps/s | VRAM=29.67 Go [2026-07-17 16:50:04] step 379850/1000000 | loss=2.3237 vocab=0.8356 token=1.5949 | 6.11 steps/s | VRAM=30.22 Go [2026-07-17 16:50:13] step 379900/1000000 | loss=2.2810 vocab=0.8706 token=1.1905 | 6.13 steps/s | VRAM=27.94 Go [2026-07-17 16:50:21] step 379950/1000000 | loss=2.2149 vocab=0.5487 token=1.2768 | 6.13 steps/s | VRAM=32.89 Go [2026-07-17 16:50:29] step 380000/1000000 | loss=2.2401 vocab=0.9506 token=1.6850 | 6.14 steps/s | VRAM=25.87 Go [2026-07-17 16:50:37] step 380050/1000000 | loss=2.2365 vocab=1.3062 token=1.5536 | 6.09 steps/s | VRAM=29.01 Go [2026-07-17 16:50:45] step 380100/1000000 | loss=2.3043 vocab=1.3134 token=1.6553 | 6.12 steps/s | VRAM=28.44 Go [2026-07-17 16:50:53] step 380150/1000000 | loss=2.3866 vocab=0.9609 token=1.2455 | 6.11 steps/s | VRAM=26.04 Go [2026-07-17 16:51:02] step 380200/1000000 | loss=2.1665 vocab=1.1205 token=1.5057 | 6.15 steps/s | VRAM=32.29 Go [2026-07-17 16:51:10] step 380250/1000000 | loss=2.2553 vocab=1.1597 token=1.4786 | 6.15 steps/s | VRAM=25.28 Go [2026-07-17 16:51:18] step 380300/1000000 | loss=2.2423 vocab=1.0931 token=1.0771 | 6.13 steps/s | VRAM=31.36 Go [2026-07-17 16:51:26] step 380350/1000000 | loss=2.3404 vocab=1.1069 token=1.7779 | 6.12 steps/s | VRAM=29.02 Go [2026-07-17 16:51:34] step 380400/1000000 | loss=2.3509 vocab=0.9628 token=1.2394 | 6.11 steps/s | VRAM=23.52 Go [2026-07-17 16:51:42] step 380450/1000000 | loss=2.2795 vocab=0.9284 token=1.9380 | 6.12 steps/s | VRAM=25.51 Go [2026-07-17 16:51:51] step 380500/1000000 | loss=2.3238 vocab=0.9381 token=1.8194 | 6.11 steps/s | VRAM=27.69 Go [2026-07-17 16:51:59] step 380550/1000000 | loss=2.2899 vocab=0.5971 token=1.0240 | 6.10 steps/s | VRAM=29.25 Go [2026-07-17 16:52:07] step 380600/1000000 | loss=2.3589 vocab=0.9162 token=1.4583 | 6.12 steps/s | VRAM=28.34 Go [2026-07-17 16:52:15] step 380650/1000000 | loss=2.2922 vocab=0.5689 token=1.3748 | 6.13 steps/s | VRAM=29.18 Go [2026-07-17 16:52:23] step 380700/1000000 | loss=2.3575 vocab=1.2436 token=1.7892 | 6.10 steps/s | VRAM=28.92 Go [2026-07-17 16:52:31] step 380750/1000000 | loss=2.3358 vocab=1.1499 token=1.6357 | 6.14 steps/s | VRAM=28.59 Go [2026-07-17 16:52:40] step 380800/1000000 | loss=2.2716 vocab=1.1223 token=1.5663 | 6.10 steps/s | VRAM=29.62 Go [2026-07-17 16:52:48] step 380850/1000000 | loss=2.2747 vocab=0.7829 token=1.1975 | 6.10 steps/s | VRAM=28.91 Go [2026-07-17 16:52:56] step 380900/1000000 | loss=2.3332 vocab=0.6921 token=0.9676 | 6.15 steps/s | VRAM=31.85 Go [2026-07-17 16:53:04] step 380950/1000000 | loss=2.3943 vocab=0.9471 token=1.3881 | 6.12 steps/s | VRAM=29.12 Go [2026-07-17 16:53:12] step 381000/1000000 | loss=2.3133 vocab=0.8335 token=1.3714 | 6.13 steps/s | VRAM=32.20 Go [2026-07-17 16:53:21] step 381050/1000000 | loss=2.2310 vocab=0.3968 token=1.1327 | 6.03 steps/s | VRAM=28.44 Go [2026-07-17 16:53:29] step 381100/1000000 | loss=2.2273 vocab=0.6621 token=1.2825 | 6.07 steps/s | VRAM=33.00 Go [2026-07-17 16:53:37] step 381150/1000000 | loss=2.2911 vocab=1.2064 token=1.8722 | 6.10 steps/s | VRAM=25.01 Go [2026-07-17 16:53:45] step 381200/1000000 | loss=2.2848 vocab=1.0238 token=1.4997 | 6.05 steps/s | VRAM=30.17 Go [2026-07-17 16:53:53] step 381250/1000000 | loss=2.2843 vocab=1.3098 token=1.2815 | 6.09 steps/s | VRAM=28.07 Go [2026-07-17 16:54:02] step 381300/1000000 | loss=2.4088 vocab=0.8000 token=1.6381 | 6.07 steps/s | VRAM=26.32 Go [2026-07-17 16:54:10] step 381350/1000000 | loss=2.1942 vocab=1.2267 token=1.3537 | 6.17 steps/s | VRAM=31.95 Go [2026-07-17 16:54:18] step 381400/1000000 | loss=2.2973 vocab=0.9358 token=1.5156 | 6.14 steps/s | VRAM=31.35 Go [2026-07-17 16:54:26] step 381450/1000000 | loss=2.2042 vocab=0.9895 token=0.9677 | 6.12 steps/s | VRAM=31.46 Go [2026-07-17 16:54:34] step 381500/1000000 | loss=2.2194 vocab=1.0316 token=1.2409 | 6.09 steps/s | VRAM=29.80 Go [2026-07-17 16:54:42] step 381550/1000000 | loss=2.2538 vocab=0.7316 token=1.2789 | 6.13 steps/s | VRAM=32.19 Go [2026-07-17 16:54:51] step 381600/1000000 | loss=2.2705 vocab=1.2923 token=1.4731 | 6.13 steps/s | VRAM=32.77 Go [2026-07-17 16:54:59] step 381650/1000000 | loss=2.3197 vocab=1.1348 token=1.4659 | 6.14 steps/s | VRAM=31.86 Go [2026-07-17 16:55:07] step 381700/1000000 | loss=2.2474 vocab=0.9429 token=1.3805 | 6.10 steps/s | VRAM=29.24 Go [2026-07-17 16:55:15] step 381750/1000000 | loss=2.1646 vocab=0.6737 token=1.1326 | 6.14 steps/s | VRAM=29.28 Go [2026-07-17 16:55:23] step 381800/1000000 | loss=2.3250 vocab=0.4225 token=1.4065 | 6.13 steps/s | VRAM=32.14 Go [2026-07-17 16:55:31] step 381850/1000000 | loss=2.3859 vocab=0.8259 token=1.5697 | 6.12 steps/s | VRAM=28.26 Go [2026-07-17 16:55:40] step 381900/1000000 | loss=2.3179 vocab=1.0092 token=1.3129 | 6.15 steps/s | VRAM=32.49 Go [2026-07-17 16:55:48] step 381950/1000000 | loss=2.2996 vocab=0.8589 token=1.4499 | 6.15 steps/s | VRAM=25.48 Go [2026-07-17 16:55:56] step 382000/1000000 | loss=2.3004 vocab=1.3694 token=1.2810 | 6.15 steps/s | VRAM=28.37 Go [2026-07-17 16:56:04] step 382050/1000000 | loss=2.1920 vocab=1.0950 token=1.2538 | 6.11 steps/s | VRAM=26.22 Go [2026-07-17 16:56:12] step 382100/1000000 | loss=2.3442 vocab=0.8573 token=1.3210 | 6.11 steps/s | VRAM=28.50 Go [2026-07-17 16:56:20] step 382150/1000000 | loss=2.2194 vocab=1.4362 token=1.3818 | 6.14 steps/s | VRAM=32.74 Go [2026-07-17 16:56:28] step 382200/1000000 | loss=2.1849 vocab=0.8225 token=1.4294 | 6.15 steps/s | VRAM=28.45 Go [2026-07-17 16:56:37] step 382250/1000000 | loss=2.2003 vocab=0.9046 token=1.4663 | 6.10 steps/s | VRAM=29.08 Go [2026-07-17 16:56:45] step 382300/1000000 | loss=2.2783 vocab=0.9375 token=1.6123 | 6.15 steps/s | VRAM=29.14 Go [2026-07-17 16:56:53] step 382350/1000000 | loss=2.3616 vocab=0.9971 token=1.2910 | 6.14 steps/s | VRAM=28.37 Go [2026-07-17 16:57:01] step 382400/1000000 | loss=2.1728 vocab=0.8912 token=1.1267 | 6.13 steps/s | VRAM=28.92 Go [2026-07-17 16:57:09] step 382450/1000000 | loss=2.1860 vocab=1.0460 token=1.1394 | 6.15 steps/s | VRAM=22.71 Go [2026-07-17 16:57:17] step 382500/1000000 | loss=2.3067 vocab=0.4050 token=1.3471 | 6.11 steps/s | VRAM=28.38 Go [2026-07-17 16:57:26] step 382550/1000000 | loss=2.3011 vocab=0.9564 token=1.1420 | 6.10 steps/s | VRAM=28.19 Go [2026-07-17 16:57:34] step 382600/1000000 | loss=2.2644 vocab=1.1525 token=1.4912 | 6.11 steps/s | VRAM=28.07 Go [2026-07-17 16:57:42] step 382650/1000000 | loss=2.1044 vocab=0.5483 token=1.2281 | 6.13 steps/s | VRAM=32.45 Go [2026-07-17 16:57:50] step 382700/1000000 | loss=2.2569 vocab=0.8701 token=1.4981 | 6.13 steps/s | VRAM=28.85 Go [2026-07-17 16:57:58] step 382750/1000000 | loss=2.3086 vocab=0.9235 token=1.4545 | 6.11 steps/s | VRAM=28.69 Go [2026-07-17 16:58:06] step 382800/1000000 | loss=2.2435 vocab=0.5498 token=1.0760 | 6.13 steps/s | VRAM=28.82 Go [2026-07-17 16:58:15] step 382850/1000000 | loss=2.1523 vocab=0.3944 token=1.1663 | 6.14 steps/s | VRAM=32.20 Go [2026-07-17 16:58:23] step 382900/1000000 | loss=2.3202 vocab=1.2561 token=1.0960 | 6.09 steps/s | VRAM=28.28 Go [2026-07-17 16:58:31] step 382950/1000000 | loss=2.3031 vocab=1.1095 token=1.3528 | 6.11 steps/s | VRAM=28.72 Go [2026-07-17 16:58:39] step 383000/1000000 | loss=2.2823 vocab=0.5087 token=1.2889 | 6.11 steps/s | VRAM=30.25 Go [2026-07-17 16:58:47] step 383050/1000000 | loss=2.2160 vocab=0.7515 token=1.4336 | 6.15 steps/s | VRAM=32.41 Go [2026-07-17 16:58:56] step 383100/1000000 | loss=2.3677 vocab=0.7186 token=1.4048 | 6.11 steps/s | VRAM=26.57 Go [2026-07-17 16:59:04] step 383150/1000000 | loss=2.2207 vocab=1.0557 token=1.0769 | 6.12 steps/s | VRAM=31.82 Go [2026-07-17 16:59:12] step 383200/1000000 | loss=2.2806 vocab=0.9542 token=1.1401 | 6.13 steps/s | VRAM=32.38 Go [2026-07-17 16:59:20] step 383250/1000000 | loss=2.2481 vocab=0.8577 token=1.2724 | 6.16 steps/s | VRAM=29.23 Go [2026-07-17 16:59:28] step 383300/1000000 | loss=2.2387 vocab=1.0258 token=1.6069 | 6.16 steps/s | VRAM=24.93 Go [2026-07-17 16:59:36] step 383350/1000000 | loss=2.2646 vocab=0.8249 token=0.8963 | 6.12 steps/s | VRAM=26.13 Go [2026-07-17 16:59:44] step 383400/1000000 | loss=2.2992 vocab=0.8648 token=1.3777 | 6.13 steps/s | VRAM=26.65 Go [2026-07-17 16:59:53] step 383450/1000000 | loss=2.2953 vocab=0.8667 token=1.4089 | 6.11 steps/s | VRAM=24.46 Go [2026-07-17 17:00:01] step 383500/1000000 | loss=2.3427 vocab=0.9622 token=1.7560 | 6.15 steps/s | VRAM=25.47 Go [2026-07-17 17:00:09] step 383550/1000000 | loss=2.2479 vocab=0.6723 token=1.5036 | 6.11 steps/s | VRAM=32.41 Go [2026-07-17 17:00:17] step 383600/1000000 | loss=2.3100 vocab=0.5698 token=1.2934 | 6.09 steps/s | VRAM=29.00 Go [2026-07-17 17:00:25] step 383650/1000000 | loss=2.3989 vocab=1.0624 token=1.2266 | 6.10 steps/s | VRAM=26.10 Go [2026-07-17 17:00:34] step 383700/1000000 | loss=2.4321 vocab=0.9020 token=1.3878 | 6.08 steps/s | VRAM=28.11 Go [2026-07-17 17:00:42] step 383750/1000000 | loss=2.2704 vocab=0.8168 token=1.5927 | 6.10 steps/s | VRAM=29.18 Go [2026-07-17 17:00:50] step 383800/1000000 | loss=2.2501 vocab=0.7357 token=1.2094 | 6.10 steps/s | VRAM=28.95 Go [2026-07-17 17:00:58] step 383850/1000000 | loss=2.3080 vocab=0.9830 token=1.0305 | 6.13 steps/s | VRAM=28.18 Go [2026-07-17 17:01:06] step 383900/1000000 | loss=2.3408 vocab=0.5993 token=1.3272 | 6.16 steps/s | VRAM=28.29 Go [2026-07-17 17:01:14] step 383950/1000000 | loss=2.2460 vocab=0.6736 token=1.2585 | 6.11 steps/s | VRAM=27.05 Go [2026-07-17 17:01:23] step 384000/1000000 | loss=2.1629 vocab=0.6928 token=0.7918 | 6.13 steps/s | VRAM=32.50 Go [2026-07-17 17:01:31] step 384050/1000000 | loss=2.2919 vocab=0.8490 token=1.2766 | 6.14 steps/s | VRAM=31.63 Go [2026-07-17 17:01:39] step 384100/1000000 | loss=2.2458 vocab=1.0304 token=1.5904 | 6.13 steps/s | VRAM=26.15 Go [2026-07-17 17:01:47] step 384150/1000000 | loss=2.3052 vocab=0.7788 token=1.4023 | 6.13 steps/s | VRAM=31.71 Go [2026-07-17 17:01:55] step 384200/1000000 | loss=2.2273 vocab=0.8788 token=0.8565 | 6.11 steps/s | VRAM=27.34 Go [2026-07-17 17:02:03] step 384250/1000000 | loss=2.3563 vocab=0.5152 token=1.0849 | 6.09 steps/s | VRAM=29.23 Go [2026-07-17 17:02:12] step 384300/1000000 | loss=2.3282 vocab=0.8792 token=1.4285 | 6.11 steps/s | VRAM=26.04 Go [2026-07-17 17:02:20] step 384350/1000000 | loss=2.2615 vocab=1.0361 token=1.8706 | 6.07 steps/s | VRAM=26.51 Go [2026-07-17 17:02:28] step 384400/1000000 | loss=2.1931 vocab=0.6016 token=1.1515 | 6.12 steps/s | VRAM=29.39 Go [2026-07-17 17:02:36] step 384450/1000000 | loss=2.1759 vocab=1.0418 token=1.3651 | 6.10 steps/s | VRAM=28.27 Go [2026-07-17 17:02:44] step 384500/1000000 | loss=2.3544 vocab=1.1854 token=1.4966 | 6.09 steps/s | VRAM=28.42 Go [2026-07-17 17:02:53] step 384550/1000000 | loss=2.2865 vocab=1.1641 token=1.4852 | 6.10 steps/s | VRAM=29.79 Go [2026-07-17 17:03:01] step 384600/1000000 | loss=2.2947 vocab=0.7148 token=1.3780 | 6.08 steps/s | VRAM=26.17 Go [2026-07-17 17:03:09] step 384650/1000000 | loss=2.3178 vocab=1.0276 token=1.6038 | 6.10 steps/s | VRAM=28.71 Go [2026-07-17 17:03:17] step 384700/1000000 | loss=2.3302 vocab=0.8343 token=1.4953 | 6.11 steps/s | VRAM=30.17 Go [2026-07-17 17:03:25] step 384750/1000000 | loss=2.2085 vocab=1.4024 token=1.5791 | 6.09 steps/s | VRAM=28.67 Go [2026-07-17 17:03:34] step 384800/1000000 | loss=2.2986 vocab=0.7291 token=1.1372 | 6.11 steps/s | VRAM=27.93 Go [2026-07-17 17:03:42] step 384850/1000000 | loss=2.2219 vocab=1.4253 token=1.2007 | 6.11 steps/s | VRAM=28.20 Go [2026-07-17 17:03:50] step 384900/1000000 | loss=2.2654 vocab=0.8127 token=1.6476 | 6.12 steps/s | VRAM=30.88 Go [2026-07-17 17:03:58] step 384950/1000000 | loss=2.2602 vocab=0.7533 token=1.6122 | 6.09 steps/s | VRAM=25.20 Go [2026-07-17 17:04:06] step 385000/1000000 | loss=2.3321 vocab=0.8453 token=1.5579 | 6.12 steps/s | VRAM=29.33 Go [2026-07-17 17:04:15] step 385050/1000000 | loss=2.2790 vocab=1.4946 token=1.6009 | 6.11 steps/s | VRAM=26.01 Go [2026-07-17 17:04:23] step 385100/1000000 | loss=2.2448 vocab=0.9203 token=1.1206 | 6.16 steps/s | VRAM=31.13 Go [2026-07-17 17:04:31] step 385150/1000000 | loss=2.2341 vocab=0.9225 token=1.7208 | 6.09 steps/s | VRAM=29.90 Go [2026-07-17 17:04:39] step 385200/1000000 | loss=2.3317 vocab=1.0629 token=0.9881 | 6.11 steps/s | VRAM=28.34 Go [2026-07-17 17:04:47] step 385250/1000000 | loss=2.1967 vocab=0.6838 token=1.1970 | 6.14 steps/s | VRAM=28.65 Go [2026-07-17 17:04:55] step 385300/1000000 | loss=2.2081 vocab=0.8918 token=1.3348 | 6.11 steps/s | VRAM=32.91 Go [2026-07-17 17:05:04] step 385350/1000000 | loss=2.2773 vocab=1.2087 token=1.7937 | 6.14 steps/s | VRAM=25.62 Go [2026-07-17 17:05:12] step 385400/1000000 | loss=2.3279 vocab=0.7221 token=1.5071 | 6.13 steps/s | VRAM=32.56 Go [2026-07-17 17:05:20] step 385450/1000000 | loss=2.2856 vocab=1.2341 token=1.5131 | 6.09 steps/s | VRAM=27.76 Go [2026-07-17 17:05:28] step 385500/1000000 | loss=2.2799 vocab=1.1458 token=1.3631 | 6.12 steps/s | VRAM=32.17 Go [2026-07-17 17:05:36] step 385550/1000000 | loss=2.1589 vocab=0.7775 token=0.8210 | 6.13 steps/s | VRAM=26.24 Go [2026-07-17 17:05:44] step 385600/1000000 | loss=2.2468 vocab=0.9867 token=1.3678 | 6.13 steps/s | VRAM=32.32 Go [2026-07-17 17:05:53] step 385650/1000000 | loss=2.2191 vocab=0.4404 token=1.1489 | 6.10 steps/s | VRAM=28.99 Go [2026-07-17 17:06:01] step 385700/1000000 | loss=2.3260 vocab=0.9177 token=1.5653 | 6.10 steps/s | VRAM=28.42 Go [2026-07-17 17:06:09] step 385750/1000000 | loss=2.3867 vocab=1.4577 token=1.1904 | 6.09 steps/s | VRAM=29.31 Go [2026-07-17 17:06:17] step 385800/1000000 | loss=2.3591 vocab=0.8535 token=1.3903 | 6.12 steps/s | VRAM=28.44 Go [2026-07-17 17:06:25] step 385850/1000000 | loss=2.2696 vocab=1.0716 token=1.6665 | 6.14 steps/s | VRAM=32.17 Go [2026-07-17 17:06:33] step 385900/1000000 | loss=2.3450 vocab=0.6299 token=1.4049 | 6.13 steps/s | VRAM=26.53 Go [2026-07-17 17:06:42] step 385950/1000000 | loss=2.3336 vocab=1.2757 token=1.4887 | 6.13 steps/s | VRAM=29.64 Go [2026-07-17 17:06:50] step 386000/1000000 | loss=2.2909 vocab=0.9001 token=1.4677 | 6.12 steps/s | VRAM=29.77 Go [2026-07-17 17:06:58] step 386050/1000000 | loss=2.3787 vocab=1.0441 token=1.1991 | 6.14 steps/s | VRAM=31.38 Go [2026-07-17 17:07:06] step 386100/1000000 | loss=2.2784 vocab=0.7180 token=1.3032 | 6.15 steps/s | VRAM=29.35 Go [2026-07-17 17:07:14] step 386150/1000000 | loss=2.2942 vocab=0.7193 token=1.4234 | 6.14 steps/s | VRAM=32.44 Go [2026-07-17 17:07:22] step 386200/1000000 | loss=2.2387 vocab=0.7215 token=1.1391 | 6.11 steps/s | VRAM=26.34 Go [2026-07-17 17:07:31] step 386250/1000000 | loss=2.2649 vocab=0.8919 token=1.2233 | 6.14 steps/s | VRAM=32.26 Go [2026-07-17 17:07:39] step 386300/1000000 | loss=2.3833 vocab=0.9969 token=1.4891 | 6.13 steps/s | VRAM=31.65 Go [2026-07-17 17:07:47] step 386350/1000000 | loss=2.2056 vocab=0.5864 token=1.1613 | 6.14 steps/s | VRAM=31.25 Go [2026-07-17 17:07:55] step 386400/1000000 | loss=2.2198 vocab=1.0983 token=1.4928 | 6.13 steps/s | VRAM=32.54 Go [2026-07-17 17:08:03] step 386450/1000000 | loss=2.3170 vocab=1.1861 token=1.6564 | 6.13 steps/s | VRAM=32.49 Go [2026-07-17 17:08:11] step 386500/1000000 | loss=2.1883 vocab=0.4453 token=1.4286 | 6.13 steps/s | VRAM=31.79 Go [2026-07-17 17:08:19] step 386550/1000000 | loss=2.2545 vocab=1.2823 token=1.1070 | 6.11 steps/s | VRAM=28.48 Go [2026-07-17 17:08:28] step 386600/1000000 | loss=2.3127 vocab=1.3803 token=1.6182 | 6.12 steps/s | VRAM=26.24 Go [2026-07-17 17:08:36] step 386650/1000000 | loss=2.1603 vocab=0.7006 token=1.3869 | 6.14 steps/s | VRAM=28.94 Go [2026-07-17 17:08:44] step 386700/1000000 | loss=2.2917 vocab=0.8549 token=1.7314 | 6.14 steps/s | VRAM=24.49 Go [2026-07-17 17:08:52] step 386750/1000000 | loss=2.3304 vocab=0.8651 token=1.6914 | 6.15 steps/s | VRAM=32.03 Go [2026-07-17 17:09:00] step 386800/1000000 | loss=2.2478 vocab=1.1399 token=1.5029 | 6.14 steps/s | VRAM=31.87 Go [2026-07-17 17:09:08] step 386850/1000000 | loss=2.3409 vocab=0.9006 token=1.6033 | 6.15 steps/s | VRAM=28.80 Go [2026-07-17 17:09:17] step 386900/1000000 | loss=2.3284 vocab=0.7503 token=1.2343 | 6.11 steps/s | VRAM=29.08 Go [2026-07-17 17:09:25] step 386950/1000000 | loss=2.3057 vocab=0.7758 token=1.4413 | 6.11 steps/s | VRAM=28.72 Go [2026-07-17 17:09:33] step 387000/1000000 | loss=2.2532 vocab=0.9883 token=1.4116 | 6.14 steps/s | VRAM=29.35 Go [2026-07-17 17:09:41] step 387050/1000000 | loss=2.3226 vocab=1.0758 token=1.0118 | 6.14 steps/s | VRAM=28.79 Go [2026-07-17 17:09:49] step 387100/1000000 | loss=2.2476 vocab=1.5823 token=1.3780 | 6.12 steps/s | VRAM=27.93 Go [2026-07-17 17:09:57] step 387150/1000000 | loss=2.2459 vocab=0.6457 token=0.9209 | 6.13 steps/s | VRAM=31.86 Go [2026-07-17 17:10:06] step 387200/1000000 | loss=2.2738 vocab=0.7073 token=1.2143 | 6.10 steps/s | VRAM=25.19 Go [2026-07-17 17:10:14] step 387250/1000000 | loss=2.2404 vocab=0.7560 token=1.1886 | 6.14 steps/s | VRAM=32.08 Go [2026-07-17 17:10:22] step 387300/1000000 | loss=2.2933 vocab=0.9576 token=1.6995 | 6.13 steps/s | VRAM=32.52 Go [2026-07-17 17:10:30] step 387350/1000000 | loss=2.2600 vocab=0.8917 token=1.4750 | 6.10 steps/s | VRAM=29.68 Go [2026-07-17 17:10:38] step 387400/1000000 | loss=2.2884 vocab=0.5740 token=1.0505 | 6.12 steps/s | VRAM=26.34 Go [2026-07-17 17:10:46] step 387450/1000000 | loss=2.2616 vocab=0.9000 token=1.2494 | 6.09 steps/s | VRAM=26.39 Go [2026-07-17 17:10:55] step 387500/1000000 | loss=2.3167 vocab=1.1910 token=1.1920 | 6.15 steps/s | VRAM=32.38 Go [2026-07-17 17:11:03] step 387550/1000000 | loss=2.3254 vocab=1.3471 token=1.3065 | 6.14 steps/s | VRAM=28.70 Go [2026-07-17 17:11:11] step 387600/1000000 | loss=2.2905 vocab=0.8643 token=0.9885 | 6.16 steps/s | VRAM=27.75 Go [2026-07-17 17:11:19] step 387650/1000000 | loss=2.2415 vocab=1.1689 token=1.2951 | 6.13 steps/s | VRAM=29.74 Go [2026-07-17 17:11:27] step 387700/1000000 | loss=2.2595 vocab=1.1481 token=1.1455 | 6.11 steps/s | VRAM=28.62 Go [2026-07-17 17:11:35] step 387750/1000000 | loss=2.3945 vocab=0.8436 token=1.4554 | 6.13 steps/s | VRAM=32.37 Go [2026-07-17 17:11:43] step 387800/1000000 | loss=2.2348 vocab=1.1865 token=1.4370 | 6.11 steps/s | VRAM=31.77 Go [2026-07-17 17:11:52] step 387850/1000000 | loss=2.3696 vocab=1.2687 token=1.4763 | 6.11 steps/s | VRAM=26.79 Go [2026-07-17 17:12:00] step 387900/1000000 | loss=2.2723 vocab=0.9225 token=1.0149 | 6.13 steps/s | VRAM=28.57 Go [2026-07-17 17:12:08] step 387950/1000000 | loss=2.2565 vocab=0.8911 token=1.1210 | 6.12 steps/s | VRAM=25.88 Go [2026-07-17 17:12:16] step 388000/1000000 | loss=2.3690 vocab=0.7026 token=1.4553 | 6.13 steps/s | VRAM=32.55 Go [2026-07-17 17:12:24] step 388050/1000000 | loss=2.1684 vocab=0.5871 token=1.0243 | 6.14 steps/s | VRAM=31.82 Go [2026-07-17 17:12:32] step 388100/1000000 | loss=2.3276 vocab=0.6374 token=1.3126 | 6.13 steps/s | VRAM=32.95 Go [2026-07-17 17:12:41] step 388150/1000000 | loss=2.2404 vocab=0.8661 token=1.2018 | 6.12 steps/s | VRAM=29.18 Go [2026-07-17 17:12:49] step 388200/1000000 | loss=2.2900 vocab=0.9861 token=1.3294 | 6.11 steps/s | VRAM=27.29 Go [2026-07-17 17:12:57] step 388250/1000000 | loss=2.2623 vocab=0.8121 token=1.3291 | 6.13 steps/s | VRAM=31.59 Go [2026-07-17 17:13:05] step 388300/1000000 | loss=2.1855 vocab=0.9323 token=1.4621 | 6.14 steps/s | VRAM=26.00 Go [2026-07-17 17:13:13] step 388350/1000000 | loss=2.2242 vocab=1.0540 token=1.1182 | 6.14 steps/s | VRAM=25.34 Go [2026-07-17 17:13:21] step 388400/1000000 | loss=2.2165 vocab=0.9654 token=1.4147 | 6.13 steps/s | VRAM=32.42 Go [2026-07-17 17:13:30] step 388450/1000000 | loss=2.2798 vocab=1.1182 token=1.8382 | 6.13 steps/s | VRAM=26.27 Go [2026-07-17 17:13:38] step 388500/1000000 | loss=2.3048 vocab=0.8755 token=1.1677 | 6.14 steps/s | VRAM=31.24 Go [2026-07-17 17:13:46] step 388550/1000000 | loss=2.3281 vocab=0.6906 token=1.3033 | 6.09 steps/s | VRAM=23.85 Go [2026-07-17 17:13:54] step 388600/1000000 | loss=2.3058 vocab=0.8953 token=1.3275 | 6.16 steps/s | VRAM=32.03 Go [2026-07-17 17:14:02] step 388650/1000000 | loss=2.3594 vocab=1.1821 token=1.4713 | 6.13 steps/s | VRAM=26.80 Go [2026-07-17 17:14:10] step 388700/1000000 | loss=2.3031 vocab=1.2256 token=1.6393 | 6.15 steps/s | VRAM=29.46 Go [2026-07-17 17:14:18] step 388750/1000000 | loss=2.2510 vocab=0.6376 token=1.4094 | 6.19 steps/s | VRAM=28.27 Go [2026-07-17 17:14:27] step 388800/1000000 | loss=2.2617 vocab=0.9918 token=1.4561 | 6.13 steps/s | VRAM=29.38 Go [2026-07-17 17:14:35] step 388850/1000000 | loss=2.2865 vocab=0.7463 token=1.2700 | 6.12 steps/s | VRAM=31.76 Go [2026-07-17 17:14:43] step 388900/1000000 | loss=2.2632 vocab=1.0279 token=1.5626 | 6.09 steps/s | VRAM=25.98 Go [2026-07-17 17:14:51] step 388950/1000000 | loss=2.2230 vocab=0.6885 token=2.0117 | 6.12 steps/s | VRAM=26.32 Go [2026-07-17 17:14:59] step 389000/1000000 | loss=2.3435 vocab=1.0100 token=1.4144 | 6.12 steps/s | VRAM=31.20 Go [2026-07-17 17:15:07] step 389050/1000000 | loss=2.3086 vocab=0.9574 token=1.0612 | 6.12 steps/s | VRAM=28.95 Go [2026-07-17 17:15:16] step 389100/1000000 | loss=2.3282 vocab=0.6586 token=1.6737 | 6.15 steps/s | VRAM=28.42 Go [2026-07-17 17:15:24] step 389150/1000000 | loss=2.2538 vocab=1.0721 token=1.2561 | 6.12 steps/s | VRAM=25.78 Go [2026-07-17 17:15:32] step 389200/1000000 | loss=2.2638 vocab=0.8424 token=1.3604 | 6.13 steps/s | VRAM=28.52 Go [2026-07-17 17:15:40] step 389250/1000000 | loss=2.2289 vocab=0.9233 token=1.4326 | 6.13 steps/s | VRAM=25.59 Go [2026-07-17 17:15:48] step 389300/1000000 | loss=2.3183 vocab=1.0856 token=1.4396 | 6.13 steps/s | VRAM=28.00 Go [2026-07-17 17:15:56] step 389350/1000000 | loss=2.2767 vocab=0.8298 token=1.4866 | 6.13 steps/s | VRAM=26.18 Go [2026-07-17 17:16:05] step 389400/1000000 | loss=2.2319 vocab=0.8825 token=1.3736 | 6.13 steps/s | VRAM=25.02 Go [2026-07-17 17:16:13] step 389450/1000000 | loss=2.3372 vocab=1.2337 token=1.6815 | 6.11 steps/s | VRAM=29.71 Go [2026-07-17 17:16:21] step 389500/1000000 | loss=2.2785 vocab=0.6920 token=0.8659 | 6.12 steps/s | VRAM=28.82 Go [2026-07-17 17:16:29] step 389550/1000000 | loss=2.2374 vocab=1.4614 token=1.7398 | 6.14 steps/s | VRAM=25.70 Go [2026-07-17 17:16:37] step 389600/1000000 | loss=2.2916 vocab=0.6076 token=1.5103 | 6.16 steps/s | VRAM=24.64 Go [2026-07-17 17:16:45] step 389650/1000000 | loss=2.2850 vocab=0.9293 token=1.0904 | 6.11 steps/s | VRAM=32.65 Go [2026-07-17 17:16:53] step 389700/1000000 | loss=2.2583 vocab=1.1929 token=1.4645 | 6.13 steps/s | VRAM=27.95 Go [2026-07-17 17:17:02] step 389750/1000000 | loss=2.3073 vocab=1.1784 token=1.0817 | 6.10 steps/s | VRAM=25.57 Go [2026-07-17 17:17:10] step 389800/1000000 | loss=2.3737 vocab=0.6058 token=1.1968 | 6.15 steps/s | VRAM=31.95 Go [2026-07-17 17:17:18] step 389850/1000000 | loss=2.3030 vocab=0.5973 token=1.4941 | 6.17 steps/s | VRAM=27.68 Go [2026-07-17 17:17:26] step 389900/1000000 | loss=2.4038 vocab=0.6652 token=1.6672 | 6.15 steps/s | VRAM=28.83 Go [2026-07-17 17:17:34] step 389950/1000000 | loss=2.3091 vocab=0.7450 token=1.4575 | 6.15 steps/s | VRAM=29.82 Go [2026-07-17 17:17:42] step 390000/1000000 | loss=2.2041 vocab=0.6274 token=1.5851 | 6.17 steps/s | VRAM=31.68 Go [2026-07-17 17:17:50] step 390050/1000000 | loss=2.2975 vocab=0.5867 token=1.5113 | 6.12 steps/s | VRAM=26.07 Go [2026-07-17 17:17:59] step 390100/1000000 | loss=2.1223 vocab=0.7864 token=1.3813 | 6.14 steps/s | VRAM=28.47 Go [2026-07-17 17:18:07] step 390150/1000000 | loss=2.3297 vocab=1.3481 token=1.2563 | 6.13 steps/s | VRAM=28.72 Go [2026-07-17 17:18:15] step 390200/1000000 | loss=2.2992 vocab=0.9399 token=1.3710 | 6.16 steps/s | VRAM=27.90 Go [2026-07-17 17:18:23] step 390250/1000000 | loss=2.3082 vocab=1.2236 token=1.6390 | 6.12 steps/s | VRAM=28.34 Go [2026-07-17 17:18:31] step 390300/1000000 | loss=2.2670 vocab=1.2287 token=1.2547 | 6.17 steps/s | VRAM=32.35 Go [2026-07-17 17:18:39] step 390350/1000000 | loss=2.2993 vocab=1.3181 token=1.4939 | 6.14 steps/s | VRAM=27.76 Go [2026-07-17 17:18:47] step 390400/1000000 | loss=2.4241 vocab=1.2716 token=2.1552 | 6.14 steps/s | VRAM=28.87 Go [2026-07-17 17:18:56] step 390450/1000000 | loss=2.2290 vocab=0.7233 token=1.2926 | 6.17 steps/s | VRAM=32.41 Go [2026-07-17 17:19:04] step 390500/1000000 | loss=2.2861 vocab=0.5236 token=1.3168 | 6.16 steps/s | VRAM=25.67 Go [2026-07-17 17:19:12] step 390550/1000000 | loss=2.3409 vocab=0.8821 token=1.0985 | 6.12 steps/s | VRAM=32.80 Go [2026-07-17 17:19:20] step 390600/1000000 | loss=2.3088 vocab=0.4882 token=0.7698 | 6.16 steps/s | VRAM=24.50 Go [2026-07-17 17:19:28] step 390650/1000000 | loss=2.3169 vocab=0.7624 token=1.2137 | 6.15 steps/s | VRAM=28.96 Go [2026-07-17 17:19:36] step 390700/1000000 | loss=2.3056 vocab=0.8780 token=1.6807 | 6.14 steps/s | VRAM=26.06 Go [2026-07-17 17:19:44] step 390750/1000000 | loss=2.2920 vocab=1.0884 token=1.5587 | 6.12 steps/s | VRAM=29.96 Go [2026-07-17 17:19:53] step 390800/1000000 | loss=2.2963 vocab=0.7813 token=1.0973 | 6.14 steps/s | VRAM=26.35 Go [2026-07-17 17:20:01] step 390850/1000000 | loss=2.2007 vocab=0.5706 token=1.3867 | 6.10 steps/s | VRAM=25.20 Go [2026-07-17 17:20:09] step 390900/1000000 | loss=2.3026 vocab=0.7541 token=1.1671 | 6.16 steps/s | VRAM=27.98 Go [2026-07-17 17:20:17] step 390950/1000000 | loss=2.2620 vocab=0.7748 token=1.3385 | 6.14 steps/s | VRAM=29.08 Go [2026-07-17 17:20:25] step 391000/1000000 | loss=2.3371 vocab=0.9702 token=1.2968 | 6.14 steps/s | VRAM=25.98 Go [2026-07-17 17:20:33] step 391050/1000000 | loss=2.4996 vocab=0.6630 token=1.3614 | 6.14 steps/s | VRAM=29.03 Go [2026-07-17 17:20:41] step 391100/1000000 | loss=2.2161 vocab=1.1665 token=1.5007 | 6.14 steps/s | VRAM=31.92 Go [2026-07-17 17:20:50] step 391150/1000000 | loss=2.3425 vocab=1.0245 token=1.4201 | 6.14 steps/s | VRAM=28.33 Go [2026-07-17 17:20:58] step 391200/1000000 | loss=2.3548 vocab=1.2344 token=1.6588 | 6.13 steps/s | VRAM=26.07 Go [2026-07-17 17:21:06] step 391250/1000000 | loss=2.2816 vocab=1.1976 token=1.5915 | 6.18 steps/s | VRAM=28.01 Go [2026-07-17 17:21:14] step 391300/1000000 | loss=2.3197 vocab=1.1322 token=1.3423 | 6.12 steps/s | VRAM=27.86 Go [2026-07-17 17:21:22] step 391350/1000000 | loss=2.3550 vocab=0.9320 token=1.4916 | 6.17 steps/s | VRAM=32.98 Go [2026-07-17 17:21:30] step 391400/1000000 | loss=2.3057 vocab=0.6410 token=1.2466 | 6.13 steps/s | VRAM=32.45 Go [2026-07-17 17:21:38] step 391450/1000000 | loss=2.2828 vocab=0.8892 token=1.1190 | 6.09 steps/s | VRAM=26.89 Go [2026-07-17 17:21:47] step 391500/1000000 | loss=2.2783 vocab=0.6409 token=0.9453 | 6.15 steps/s | VRAM=28.75 Go [2026-07-17 17:21:55] step 391550/1000000 | loss=2.2469 vocab=1.1029 token=1.3435 | 6.12 steps/s | VRAM=28.90 Go [2026-07-17 17:22:03] step 391600/1000000 | loss=2.2569 vocab=0.7679 token=1.2104 | 6.13 steps/s | VRAM=29.71 Go [2026-07-17 17:22:11] step 391650/1000000 | loss=2.2985 vocab=1.2025 token=1.4413 | 6.12 steps/s | VRAM=28.06 Go [2026-07-17 17:22:19] step 391700/1000000 | loss=2.3904 vocab=0.5790 token=1.4869 | 6.15 steps/s | VRAM=26.32 Go [2026-07-17 17:22:27] step 391750/1000000 | loss=2.2691 vocab=0.5124 token=0.9590 | 6.13 steps/s | VRAM=31.85 Go [2026-07-17 17:22:36] step 391800/1000000 | loss=2.3227 vocab=0.9281 token=1.1051 | 6.12 steps/s | VRAM=26.52 Go [2026-07-17 17:22:44] step 391850/1000000 | loss=2.2555 vocab=1.0217 token=1.7373 | 6.14 steps/s | VRAM=28.93 Go [2026-07-17 17:22:52] step 391900/1000000 | loss=2.3013 vocab=0.9672 token=1.4316 | 6.13 steps/s | VRAM=28.12 Go [2026-07-17 17:23:00] step 391950/1000000 | loss=2.3489 vocab=1.0838 token=1.5899 | 6.13 steps/s | VRAM=29.37 Go [2026-07-17 17:23:08] step 392000/1000000 | loss=2.3415 vocab=0.8410 token=1.1843 | 6.13 steps/s | VRAM=32.14 Go [2026-07-17 17:23:16] step 392050/1000000 | loss=2.4487 vocab=0.4982 token=0.9599 | 6.12 steps/s | VRAM=32.34 Go [2026-07-17 17:23:25] step 392100/1000000 | loss=2.3406 vocab=1.2397 token=1.5050 | 6.13 steps/s | VRAM=31.56 Go [2026-07-17 17:23:33] step 392150/1000000 | loss=2.3205 vocab=0.7427 token=1.4181 | 6.12 steps/s | VRAM=32.47 Go [2026-07-17 17:23:41] step 392200/1000000 | loss=2.3686 vocab=0.6560 token=1.2377 | 6.11 steps/s | VRAM=28.12 Go [2026-07-17 17:23:49] step 392250/1000000 | loss=2.2726 vocab=0.4516 token=1.5041 | 6.14 steps/s | VRAM=31.28 Go [2026-07-17 17:23:57] step 392300/1000000 | loss=2.3120 vocab=0.9083 token=1.1512 | 6.07 steps/s | VRAM=30.31 Go [2026-07-17 17:24:05] step 392350/1000000 | loss=2.2600 vocab=0.9483 token=1.4474 | 6.16 steps/s | VRAM=31.53 Go [2026-07-17 17:24:14] step 392400/1000000 | loss=2.3108 vocab=1.0550 token=1.3070 | 6.10 steps/s | VRAM=28.40 Go [2026-07-17 17:24:22] step 392450/1000000 | loss=2.2380 vocab=0.9665 token=1.5491 | 6.16 steps/s | VRAM=30.57 Go [2026-07-17 17:24:30] step 392500/1000000 | loss=2.3357 vocab=0.9949 token=1.4354 | 6.09 steps/s | VRAM=27.39 Go [2026-07-17 17:24:38] step 392550/1000000 | loss=2.3464 vocab=1.0115 token=1.6980 | 6.14 steps/s | VRAM=32.54 Go [2026-07-17 17:24:46] step 392600/1000000 | loss=2.2487 vocab=1.0305 token=1.0201 | 6.13 steps/s | VRAM=24.44 Go [2026-07-17 17:24:54] step 392650/1000000 | loss=2.3182 vocab=0.6140 token=1.1389 | 6.12 steps/s | VRAM=25.83 Go [2026-07-17 17:25:02] step 392700/1000000 | loss=2.2621 vocab=0.6880 token=1.2541 | 6.15 steps/s | VRAM=32.56 Go [2026-07-17 17:25:11] step 392750/1000000 | loss=2.3765 vocab=0.7191 token=1.0641 | 6.14 steps/s | VRAM=29.35 Go [2026-07-17 17:25:19] step 392800/1000000 | loss=2.2865 vocab=0.5768 token=1.6038 | 6.13 steps/s | VRAM=27.80 Go [2026-07-17 17:25:27] step 392850/1000000 | loss=2.2607 vocab=1.2699 token=1.2891 | 6.13 steps/s | VRAM=29.57 Go [2026-07-17 17:25:35] step 392900/1000000 | loss=2.3391 vocab=0.7409 token=1.5634 | 6.11 steps/s | VRAM=28.21 Go [2026-07-17 17:25:43] step 392950/1000000 | loss=2.2905 vocab=0.7757 token=1.4028 | 6.11 steps/s | VRAM=27.97 Go [2026-07-17 17:25:51] step 393000/1000000 | loss=2.1875 vocab=1.0005 token=1.4447 | 6.12 steps/s | VRAM=29.85 Go [2026-07-17 17:26:00] step 393050/1000000 | loss=2.3604 vocab=1.6043 token=1.2182 | 6.13 steps/s | VRAM=32.25 Go [2026-07-17 17:26:08] step 393100/1000000 | loss=2.1890 vocab=0.7832 token=1.3479 | 6.15 steps/s | VRAM=28.25 Go [2026-07-17 17:26:16] step 393150/1000000 | loss=2.3201 vocab=1.1178 token=1.3716 | 6.13 steps/s | VRAM=32.42 Go [2026-07-17 17:26:24] step 393200/1000000 | loss=2.3645 vocab=1.1177 token=1.3016 | 6.12 steps/s | VRAM=28.92 Go [2026-07-17 17:26:32] step 393250/1000000 | loss=2.3131 vocab=0.6863 token=0.8074 | 6.13 steps/s | VRAM=29.23 Go [2026-07-17 17:26:40] step 393300/1000000 | loss=2.2672 vocab=0.7309 token=1.3568 | 6.13 steps/s | VRAM=28.09 Go [2026-07-17 17:26:49] step 393350/1000000 | loss=2.2371 vocab=0.7463 token=1.0719 | 6.16 steps/s | VRAM=32.37 Go [2026-07-17 17:26:57] step 393400/1000000 | loss=2.2214 vocab=0.6960 token=0.9148 | 6.13 steps/s | VRAM=29.25 Go [2026-07-17 17:27:05] step 393450/1000000 | loss=2.3008 vocab=0.8196 token=1.2247 | 6.13 steps/s | VRAM=32.69 Go [2026-07-17 17:27:13] step 393500/1000000 | loss=2.2629 vocab=1.3029 token=1.9939 | 6.16 steps/s | VRAM=31.54 Go [2026-07-17 17:27:21] step 393550/1000000 | loss=2.2775 vocab=1.0250 token=1.0136 | 6.14 steps/s | VRAM=28.01 Go [2026-07-17 17:27:29] step 393600/1000000 | loss=2.3374 vocab=0.9389 token=0.9995 | 6.15 steps/s | VRAM=32.81 Go [2026-07-17 17:27:37] step 393650/1000000 | loss=2.1868 vocab=0.9801 token=1.2430 | 6.12 steps/s | VRAM=32.43 Go [2026-07-17 17:27:46] step 393700/1000000 | loss=2.3106 vocab=1.1616 token=1.3977 | 6.11 steps/s | VRAM=28.81 Go [2026-07-17 17:27:54] step 393750/1000000 | loss=2.2901 vocab=0.7542 token=1.3149 | 6.13 steps/s | VRAM=25.13 Go [2026-07-17 17:28:02] step 393800/1000000 | loss=2.3800 vocab=0.8425 token=1.6926 | 6.15 steps/s | VRAM=32.31 Go [2026-07-17 17:28:10] step 393850/1000000 | loss=2.2871 vocab=0.8383 token=1.2486 | 6.15 steps/s | VRAM=32.04 Go [2026-07-17 17:28:18] step 393900/1000000 | loss=2.2812 vocab=0.8472 token=1.6087 | 6.10 steps/s | VRAM=26.41 Go [2026-07-17 17:28:26] step 393950/1000000 | loss=2.3214 vocab=1.8462 token=1.7568 | 6.14 steps/s | VRAM=28.81 Go [2026-07-17 17:28:35] step 394000/1000000 | loss=2.2776 vocab=1.0869 token=1.2592 | 6.10 steps/s | VRAM=25.80 Go [2026-07-17 17:28:43] step 394050/1000000 | loss=2.2963 vocab=0.7794 token=1.0232 | 6.16 steps/s | VRAM=25.70 Go [2026-07-17 17:28:51] step 394100/1000000 | loss=2.3099 vocab=1.0035 token=0.9895 | 6.11 steps/s | VRAM=28.70 Go [2026-07-17 17:28:59] step 394150/1000000 | loss=2.1941 vocab=1.1284 token=1.6868 | 6.10 steps/s | VRAM=28.75 Go [2026-07-17 17:29:07] step 394200/1000000 | loss=2.2267 vocab=0.5474 token=0.8606 | 6.12 steps/s | VRAM=29.12 Go [2026-07-17 17:29:15] step 394250/1000000 | loss=2.3373 vocab=0.6051 token=1.2325 | 6.12 steps/s | VRAM=26.21 Go [2026-07-17 17:29:23] step 394300/1000000 | loss=2.2851 vocab=1.3403 token=1.4289 | 6.16 steps/s | VRAM=24.88 Go [2026-07-17 17:29:32] step 394350/1000000 | loss=2.2883 vocab=1.3168 token=1.4255 | 6.10 steps/s | VRAM=32.81 Go [2026-07-17 17:29:40] step 394400/1000000 | loss=2.1910 vocab=1.5900 token=1.5416 | 6.14 steps/s | VRAM=32.14 Go [2026-07-17 17:29:48] step 394450/1000000 | loss=2.3085 vocab=0.9089 token=1.2842 | 6.11 steps/s | VRAM=29.44 Go [2026-07-17 17:29:56] step 394500/1000000 | loss=2.2128 vocab=0.7889 token=1.4974 | 6.11 steps/s | VRAM=26.82 Go [2026-07-17 17:30:04] step 394550/1000000 | loss=2.2508 vocab=0.6220 token=1.1936 | 6.11 steps/s | VRAM=29.83 Go [2026-07-17 17:30:13] step 394600/1000000 | loss=2.3893 vocab=1.3481 token=1.5334 | 6.11 steps/s | VRAM=30.02 Go [2026-07-17 17:30:21] step 394650/1000000 | loss=2.2547 vocab=0.8280 token=0.9526 | 6.11 steps/s | VRAM=29.09 Go [2026-07-17 17:30:29] step 394700/1000000 | loss=2.1748 vocab=1.8404 token=1.5480 | 6.12 steps/s | VRAM=29.24 Go [2026-07-17 17:30:37] step 394750/1000000 | loss=2.3324 vocab=0.9324 token=1.3177 | 6.14 steps/s | VRAM=28.06 Go [2026-07-17 17:30:45] step 394800/1000000 | loss=2.2418 vocab=0.9718 token=1.0908 | 6.12 steps/s | VRAM=25.61 Go [2026-07-17 17:30:53] step 394850/1000000 | loss=2.2013 vocab=0.7466 token=1.3683 | 6.14 steps/s | VRAM=26.00 Go [2026-07-17 17:31:01] step 394900/1000000 | loss=2.3535 vocab=1.1513 token=1.2071 | 6.15 steps/s | VRAM=31.42 Go [2026-07-17 17:31:10] step 394950/1000000 | loss=2.3423 vocab=1.1043 token=1.5688 | 6.11 steps/s | VRAM=26.20 Go [2026-07-17 17:31:18] step 395000/1000000 | loss=2.2532 vocab=0.9830 token=1.4433 | 6.15 steps/s | VRAM=29.06 Go [2026-07-17 17:31:26] step 395050/1000000 | loss=2.1795 vocab=0.7258 token=1.4433 | 6.16 steps/s | VRAM=28.53 Go [2026-07-17 17:31:34] step 395100/1000000 | loss=2.2472 vocab=0.7331 token=1.1073 | 6.14 steps/s | VRAM=28.53 Go [2026-07-17 17:31:42] step 395150/1000000 | loss=2.3390 vocab=0.8332 token=1.2867 | 6.13 steps/s | VRAM=28.50 Go [2026-07-17 17:31:50] step 395200/1000000 | loss=2.3184 vocab=0.9494 token=1.0096 | 6.11 steps/s | VRAM=26.26 Go [2026-07-17 17:31:59] step 395250/1000000 | loss=2.2234 vocab=0.7087 token=1.4629 | 6.16 steps/s | VRAM=28.90 Go [2026-07-17 17:32:07] step 395300/1000000 | loss=2.2490 vocab=0.6926 token=1.3448 | 6.11 steps/s | VRAM=30.41 Go [2026-07-17 17:32:15] step 395350/1000000 | loss=2.3149 vocab=0.5741 token=1.2100 | 6.15 steps/s | VRAM=24.96 Go [2026-07-17 17:32:23] step 395400/1000000 | loss=2.2039 vocab=0.6072 token=1.3710 | 6.13 steps/s | VRAM=29.02 Go [2026-07-17 17:32:31] step 395450/1000000 | loss=2.2547 vocab=0.9978 token=1.1036 | 6.15 steps/s | VRAM=32.37 Go [2026-07-17 17:32:39] step 395500/1000000 | loss=2.2446 vocab=1.0283 token=1.0513 | 6.14 steps/s | VRAM=24.81 Go [2026-07-17 17:32:47] step 395550/1000000 | loss=2.3979 vocab=0.9695 token=1.5108 | 6.12 steps/s | VRAM=29.40 Go [2026-07-17 17:32:56] step 395600/1000000 | loss=2.2498 vocab=0.5249 token=1.2799 | 6.12 steps/s | VRAM=26.17 Go [2026-07-17 17:33:04] step 395650/1000000 | loss=2.2518 vocab=1.0579 token=1.3154 | 6.15 steps/s | VRAM=26.08 Go [2026-07-17 17:33:12] step 395700/1000000 | loss=2.2811 vocab=1.0947 token=1.2951 | 6.12 steps/s | VRAM=28.12 Go [2026-07-17 17:33:20] step 395750/1000000 | loss=2.3041 vocab=0.5824 token=0.8167 | 6.08 steps/s | VRAM=29.90 Go [2026-07-17 17:33:28] step 395800/1000000 | loss=2.2856 vocab=0.9115 token=1.0602 | 6.13 steps/s | VRAM=29.14 Go [2026-07-17 17:33:36] step 395850/1000000 | loss=2.1629 vocab=0.6887 token=1.4284 | 6.14 steps/s | VRAM=25.94 Go [2026-07-17 17:33:45] step 395900/1000000 | loss=2.2006 vocab=0.8043 token=1.1437 | 6.12 steps/s | VRAM=31.98 Go [2026-07-17 17:33:53] step 395950/1000000 | loss=2.3101 vocab=1.2688 token=1.3349 | 6.09 steps/s | VRAM=26.68 Go [2026-07-17 17:34:01] step 396000/1000000 | loss=2.3154 vocab=0.6395 token=1.3579 | 6.17 steps/s | VRAM=31.30 Go [2026-07-17 17:34:09] step 396050/1000000 | loss=2.1846 vocab=0.7972 token=1.3416 | 6.12 steps/s | VRAM=26.17 Go [2026-07-17 17:34:17] step 396100/1000000 | loss=2.3421 vocab=0.6284 token=1.0281 | 6.12 steps/s | VRAM=28.48 Go [2026-07-17 17:34:25] step 396150/1000000 | loss=2.2502 vocab=1.2238 token=1.3439 | 6.12 steps/s | VRAM=25.74 Go [2026-07-17 17:34:34] step 396200/1000000 | loss=2.3535 vocab=1.0525 token=1.4277 | 6.16 steps/s | VRAM=31.76 Go [2026-07-17 17:34:42] step 396250/1000000 | loss=2.2016 vocab=0.8890 token=1.0639 | 6.13 steps/s | VRAM=26.68 Go [2026-07-17 17:34:50] step 396300/1000000 | loss=2.2049 vocab=1.3184 token=1.3535 | 6.13 steps/s | VRAM=23.06 Go [2026-07-17 17:34:58] step 396350/1000000 | loss=2.3514 vocab=0.7395 token=1.0596 | 6.11 steps/s | VRAM=29.07 Go [2026-07-17 17:35:06] step 396400/1000000 | loss=2.3189 vocab=0.5856 token=1.2073 | 6.13 steps/s | VRAM=32.36 Go [2026-07-17 17:35:14] step 396450/1000000 | loss=2.3148 vocab=0.7315 token=1.0422 | 6.14 steps/s | VRAM=28.88 Go [2026-07-17 17:35:23] step 396500/1000000 | loss=2.2139 vocab=0.4948 token=0.6613 | 6.12 steps/s | VRAM=28.44 Go [2026-07-17 17:35:31] step 396550/1000000 | loss=2.2645 vocab=0.7074 token=1.1746 | 6.14 steps/s | VRAM=27.97 Go [2026-07-17 17:35:39] step 396600/1000000 | loss=2.2069 vocab=0.7441 token=1.2291 | 6.10 steps/s | VRAM=28.97 Go [2026-07-17 17:35:47] step 396650/1000000 | loss=2.3767 vocab=1.2060 token=1.3417 | 6.13 steps/s | VRAM=28.61 Go [2026-07-17 17:35:55] step 396700/1000000 | loss=2.1856 vocab=0.9372 token=1.5704 | 6.12 steps/s | VRAM=28.20 Go [2026-07-17 17:36:03] step 396750/1000000 | loss=2.2917 vocab=0.5968 token=1.4308 | 6.11 steps/s | VRAM=25.39 Go [2026-07-17 17:36:12] step 396800/1000000 | loss=2.2842 vocab=1.1729 token=1.5347 | 6.12 steps/s | VRAM=32.06 Go [2026-07-17 17:36:20] step 396850/1000000 | loss=2.2964 vocab=1.2186 token=1.7766 | 6.13 steps/s | VRAM=26.16 Go [2026-07-17 17:36:28] step 396900/1000000 | loss=2.3849 vocab=0.8832 token=1.3137 | 6.11 steps/s | VRAM=30.42 Go [2026-07-17 17:36:36] step 396950/1000000 | loss=2.1330 vocab=0.4296 token=1.4402 | 6.08 steps/s | VRAM=26.85 Go [2026-07-17 17:36:44] step 397000/1000000 | loss=2.2743 vocab=0.9478 token=1.0232 | 6.12 steps/s | VRAM=25.84 Go [2026-07-17 17:36:52] step 397050/1000000 | loss=2.2034 vocab=0.5494 token=1.0619 | 6.12 steps/s | VRAM=29.62 Go [2026-07-17 17:37:01] step 397100/1000000 | loss=2.2734 vocab=0.8805 token=1.2668 | 6.13 steps/s | VRAM=28.74 Go [2026-07-17 17:37:09] step 397150/1000000 | loss=2.3711 vocab=1.1830 token=1.5042 | 6.13 steps/s | VRAM=26.20 Go [2026-07-17 17:37:17] step 397200/1000000 | loss=2.1674 vocab=0.6408 token=1.4418 | 6.14 steps/s | VRAM=28.70 Go [2026-07-17 17:37:25] step 397250/1000000 | loss=2.2676 vocab=0.8273 token=1.0527 | 6.15 steps/s | VRAM=31.50 Go [2026-07-17 17:37:33] step 397300/1000000 | loss=2.2965 vocab=0.9068 token=1.2737 | 6.13 steps/s | VRAM=27.74 Go [2026-07-17 17:37:41] step 397350/1000000 | loss=2.2629 vocab=0.7179 token=1.1156 | 6.14 steps/s | VRAM=32.00 Go [2026-07-17 17:37:50] step 397400/1000000 | loss=2.2954 vocab=0.5931 token=1.1448 | 6.14 steps/s | VRAM=28.60 Go [2026-07-17 17:37:58] step 397450/1000000 | loss=2.2010 vocab=0.6196 token=1.3096 | 6.11 steps/s | VRAM=28.52 Go [2026-07-17 17:38:06] step 397500/1000000 | loss=2.4103 vocab=0.7580 token=1.0292 | 6.10 steps/s | VRAM=28.39 Go [2026-07-17 17:38:14] step 397550/1000000 | loss=2.3015 vocab=1.0943 token=1.7557 | 6.16 steps/s | VRAM=25.61 Go [2026-07-17 17:38:22] step 397600/1000000 | loss=2.2844 vocab=1.1024 token=1.3579 | 6.14 steps/s | VRAM=25.81 Go [2026-07-17 17:38:30] step 397650/1000000 | loss=2.2318 vocab=1.0508 token=1.3548 | 6.12 steps/s | VRAM=31.41 Go [2026-07-17 17:38:38] step 397700/1000000 | loss=2.3500 vocab=1.0965 token=1.5520 | 6.12 steps/s | VRAM=28.15 Go [2026-07-17 17:38:47] step 397750/1000000 | loss=2.2484 vocab=0.5054 token=1.2811 | 6.09 steps/s | VRAM=24.98 Go [2026-07-17 17:38:55] step 397800/1000000 | loss=2.3300 vocab=1.3161 token=1.4036 | 6.12 steps/s | VRAM=33.00 Go [2026-07-17 17:39:03] step 397850/1000000 | loss=2.4140 vocab=0.7485 token=1.3255 | 6.10 steps/s | VRAM=26.72 Go [2026-07-17 17:39:11] step 397900/1000000 | loss=2.2440 vocab=0.9699 token=1.5950 | 6.15 steps/s | VRAM=27.87 Go [2026-07-17 17:39:19] step 397950/1000000 | loss=2.3427 vocab=0.9535 token=1.1421 | 6.14 steps/s | VRAM=29.29 Go [2026-07-17 17:39:28] step 398000/1000000 | loss=2.3338 vocab=0.9415 token=1.6243 | 6.12 steps/s | VRAM=32.88 Go [2026-07-17 17:39:36] step 398050/1000000 | loss=2.3029 vocab=0.6421 token=0.9397 | 6.13 steps/s | VRAM=29.28 Go [2026-07-17 17:39:44] step 398100/1000000 | loss=2.2212 vocab=1.1067 token=1.9102 | 6.12 steps/s | VRAM=28.40 Go [2026-07-17 17:39:52] step 398150/1000000 | loss=2.2805 vocab=1.5369 token=1.5739 | 6.15 steps/s | VRAM=24.68 Go [2026-07-17 17:40:00] step 398200/1000000 | loss=2.2403 vocab=0.8039 token=1.2649 | 6.13 steps/s | VRAM=31.64 Go [2026-07-17 17:40:08] step 398250/1000000 | loss=2.2974 vocab=1.2104 token=1.3826 | 6.13 steps/s | VRAM=28.36 Go [2026-07-17 17:40:16] step 398300/1000000 | loss=2.2786 vocab=0.9449 token=1.1080 | 6.14 steps/s | VRAM=28.93 Go [2026-07-17 17:40:25] step 398350/1000000 | loss=2.2157 vocab=1.2045 token=1.4450 | 6.15 steps/s | VRAM=29.05 Go [2026-07-17 17:40:33] step 398400/1000000 | loss=2.3167 vocab=0.9209 token=1.0530 | 6.16 steps/s | VRAM=31.86 Go [2026-07-17 17:40:41] step 398450/1000000 | loss=2.3846 vocab=0.8307 token=1.3224 | 6.12 steps/s | VRAM=25.91 Go [2026-07-17 17:40:49] step 398500/1000000 | loss=2.2681 vocab=0.9076 token=1.3817 | 6.13 steps/s | VRAM=31.55 Go [2026-07-17 17:40:57] step 398550/1000000 | loss=2.1936 vocab=0.8352 token=1.6198 | 6.12 steps/s | VRAM=26.33 Go [2026-07-17 17:41:05] step 398600/1000000 | loss=2.3409 vocab=0.6767 token=1.7103 | 6.08 steps/s | VRAM=26.63 Go [2026-07-17 17:41:14] step 398650/1000000 | loss=2.3013 vocab=1.0599 token=1.3136 | 6.12 steps/s | VRAM=28.61 Go [2026-07-17 17:41:22] step 398700/1000000 | loss=2.3334 vocab=0.6877 token=1.3955 | 6.16 steps/s | VRAM=31.77 Go [2026-07-17 17:41:30] step 398750/1000000 | loss=2.2791 vocab=0.7527 token=1.1798 | 6.10 steps/s | VRAM=26.06 Go [2026-07-17 17:41:38] step 398800/1000000 | loss=2.2849 vocab=0.8760 token=1.1473 | 6.17 steps/s | VRAM=28.10 Go [2026-07-17 17:41:46] step 398850/1000000 | loss=2.3541 vocab=0.8132 token=1.6782 | 6.09 steps/s | VRAM=26.80 Go [2026-07-17 17:41:54] step 398900/1000000 | loss=2.3078 vocab=0.9616 token=1.5567 | 6.12 steps/s | VRAM=31.68 Go [2026-07-17 17:42:03] step 398950/1000000 | loss=2.2268 vocab=0.8416 token=1.4865 | 6.10 steps/s | VRAM=26.62 Go [2026-07-17 17:42:11] step 399000/1000000 | loss=2.2973 vocab=1.0765 token=1.3804 | 6.11 steps/s | VRAM=24.72 Go [2026-07-17 17:42:19] step 399050/1000000 | loss=2.2383 vocab=0.7500 token=0.9324 | 6.15 steps/s | VRAM=29.39 Go [2026-07-17 17:42:27] step 399100/1000000 | loss=2.2970 vocab=0.8919 token=1.5334 | 6.15 steps/s | VRAM=28.20 Go [2026-07-17 17:42:35] step 399150/1000000 | loss=2.3896 vocab=0.7814 token=1.5842 | 6.06 steps/s | VRAM=29.88 Go [2026-07-17 17:42:43] step 399200/1000000 | loss=2.2312 vocab=1.0391 token=1.5670 | 6.16 steps/s | VRAM=27.95 Go [2026-07-17 17:42:51] step 399250/1000000 | loss=2.3046 vocab=0.9159 token=1.3876 | 6.17 steps/s | VRAM=31.82 Go [2026-07-17 17:43:00] step 399300/1000000 | loss=2.1689 vocab=1.3410 token=1.5424 | 6.12 steps/s | VRAM=26.21 Go [2026-07-17 17:43:08] step 399350/1000000 | loss=2.2722 vocab=0.8101 token=1.2752 | 6.15 steps/s | VRAM=30.84 Go [2026-07-17 17:43:16] step 399400/1000000 | loss=2.3015 vocab=0.6998 token=1.1607 | 6.13 steps/s | VRAM=27.19 Go [2026-07-17 17:43:24] step 399450/1000000 | loss=2.2451 vocab=0.4773 token=1.3937 | 6.13 steps/s | VRAM=27.61 Go [2026-07-17 17:43:32] step 399500/1000000 | loss=2.2660 vocab=0.7430 token=1.3390 | 6.10 steps/s | VRAM=29.76 Go [2026-07-17 17:43:40] step 399550/1000000 | loss=2.1824 vocab=0.8741 token=1.2215 | 6.13 steps/s | VRAM=32.01 Go [2026-07-17 17:43:49] step 399600/1000000 | loss=2.1745 vocab=0.5575 token=0.9445 | 6.13 steps/s | VRAM=29.25 Go [2026-07-17 17:43:57] step 399650/1000000 | loss=2.2218 vocab=1.0064 token=1.1928 | 6.13 steps/s | VRAM=26.46 Go [2026-07-17 17:44:05] step 399700/1000000 | loss=2.3651 vocab=1.1744 token=1.1783 | 6.11 steps/s | VRAM=28.59 Go [2026-07-17 17:44:13] step 399750/1000000 | loss=2.4071 vocab=1.3975 token=1.4715 | 6.12 steps/s | VRAM=28.21 Go [2026-07-17 17:44:21] step 399800/1000000 | loss=2.1962 vocab=0.8854 token=1.1479 | 6.12 steps/s | VRAM=29.99 Go [2026-07-17 17:44:29] step 399850/1000000 | loss=2.2053 vocab=0.6667 token=1.1730 | 6.14 steps/s | VRAM=28.46 Go [2026-07-17 17:44:38] step 399900/1000000 | loss=2.3638 vocab=1.6941 token=1.3694 | 6.15 steps/s | VRAM=31.67 Go [2026-07-17 17:44:46] step 399950/1000000 | loss=2.3317 vocab=0.4608 token=1.1927 | 6.12 steps/s | VRAM=26.20 Go [2026-07-17 17:44:54] step 400000/1000000 | loss=2.2025 vocab=1.2492 token=1.6881 | 6.13 steps/s | VRAM=32.06 Go [step 400000] loss_token=1.6881 > min recent (1.0593) -> push saute [2026-07-17 17:45:02] step 400050/1000000 | loss=2.2750 vocab=0.9586 token=1.5021 | 6.09 steps/s | VRAM=26.57 Go [2026-07-17 17:45:10] step 400100/1000000 | loss=2.2335 vocab=0.5281 token=0.8871 | 6.13 steps/s | VRAM=28.32 Go [2026-07-17 17:45:18] step 400150/1000000 | loss=2.2619 vocab=1.0018 token=1.0265 | 6.14 steps/s | VRAM=31.78 Go [2026-07-17 17:45:27] step 400200/1000000 | loss=2.2612 vocab=0.9587 token=1.5040 | 6.15 steps/s | VRAM=26.09 Go [2026-07-17 17:45:35] step 400250/1000000 | loss=2.3001 vocab=1.1078 token=1.0702 | 6.15 steps/s | VRAM=32.42 Go [2026-07-17 17:45:43] step 400300/1000000 | loss=2.2845 vocab=1.1610 token=1.1660 | 6.09 steps/s | VRAM=29.27 Go [2026-07-17 17:45:51] step 400350/1000000 | loss=2.2938 vocab=1.5841 token=1.0632 | 6.17 steps/s | VRAM=31.32 Go [2026-07-17 17:45:59] step 400400/1000000 | loss=2.2925 vocab=0.8078 token=1.2673 | 6.15 steps/s | VRAM=27.81 Go [2026-07-17 17:46:07] step 400450/1000000 | loss=2.3516 vocab=0.8030 token=1.1439 | 6.13 steps/s | VRAM=27.87 Go [2026-07-17 17:46:15] step 400500/1000000 | loss=2.2445 vocab=0.7815 token=1.4583 | 6.12 steps/s | VRAM=24.11 Go [2026-07-17 17:46:24] step 400550/1000000 | loss=2.2333 vocab=0.6605 token=1.1347 | 6.13 steps/s | VRAM=28.76 Go [2026-07-17 17:46:32] step 400600/1000000 | loss=2.4590 vocab=1.5364 token=1.6665 | 6.12 steps/s | VRAM=29.43 Go [2026-07-17 17:46:40] step 400650/1000000 | loss=2.2753 vocab=1.1177 token=1.8187 | 6.11 steps/s | VRAM=29.09 Go [2026-07-17 17:46:48] step 400700/1000000 | loss=2.2285 vocab=0.6017 token=1.1429 | 6.14 steps/s | VRAM=28.81 Go [2026-07-17 17:46:56] step 400750/1000000 | loss=2.3152 vocab=1.1202 token=1.5643 | 6.13 steps/s | VRAM=32.76 Go [2026-07-17 17:47:04] step 400800/1000000 | loss=2.1798 vocab=1.0416 token=1.1381 | 6.14 steps/s | VRAM=32.48 Go [2026-07-17 17:47:13] step 400850/1000000 | loss=2.2864 vocab=1.0606 token=1.5497 | 6.14 steps/s | VRAM=28.18 Go [2026-07-17 17:47:21] step 400900/1000000 | loss=2.2338 vocab=1.1615 token=1.2712 | 6.16 steps/s | VRAM=24.40 Go [2026-07-17 17:47:29] step 400950/1000000 | loss=2.2539 vocab=0.7899 token=1.0595 | 6.12 steps/s | VRAM=26.66 Go [2026-07-17 17:47:37] step 401000/1000000 | loss=2.1946 vocab=0.8449 token=1.2118 | 6.16 steps/s | VRAM=28.30 Go [2026-07-17 17:47:45] step 401050/1000000 | loss=2.3334 vocab=0.8055 token=1.2312 | 6.12 steps/s | VRAM=27.08 Go [2026-07-17 17:47:53] step 401100/1000000 | loss=2.3244 vocab=0.9453 token=1.2590 | 6.15 steps/s | VRAM=32.13 Go [2026-07-17 17:48:01] step 401150/1000000 | loss=2.1956 vocab=0.8404 token=1.2134 | 6.13 steps/s | VRAM=25.34 Go [2026-07-17 17:48:09] step 401200/1000000 | loss=2.2514 vocab=1.2135 token=1.5813 | 6.16 steps/s | VRAM=32.88 Go [2026-07-17 17:48:18] step 401250/1000000 | loss=2.2376 vocab=0.8463 token=1.2553 | 6.11 steps/s | VRAM=32.07 Go [2026-07-17 17:48:26] step 401300/1000000 | loss=2.2633 vocab=1.0156 token=1.5350 | 6.12 steps/s | VRAM=31.17 Go [2026-07-17 17:48:34] step 401350/1000000 | loss=2.2328 vocab=0.9817 token=1.2055 | 6.12 steps/s | VRAM=32.90 Go [2026-07-17 17:48:42] step 401400/1000000 | loss=2.3884 vocab=1.3636 token=1.3974 | 6.15 steps/s | VRAM=31.83 Go [2026-07-17 17:48:50] step 401450/1000000 | loss=2.3405 vocab=0.8234 token=1.0495 | 6.11 steps/s | VRAM=29.09 Go [2026-07-17 17:48:59] step 401500/1000000 | loss=2.2484 vocab=0.4698 token=0.9226 | 6.10 steps/s | VRAM=29.63 Go [2026-07-17 17:49:07] step 401550/1000000 | loss=2.2457 vocab=1.2946 token=1.6719 | 6.14 steps/s | VRAM=29.17 Go [2026-07-17 17:49:15] step 401600/1000000 | loss=2.3047 vocab=0.7701 token=1.3316 | 6.14 steps/s | VRAM=27.54 Go [2026-07-17 17:49:23] step 401650/1000000 | loss=2.3160 vocab=0.5886 token=1.2612 | 6.14 steps/s | VRAM=32.14 Go [2026-07-17 17:49:31] step 401700/1000000 | loss=2.2602 vocab=1.0022 token=1.4157 | 6.11 steps/s | VRAM=32.05 Go [2026-07-17 17:49:39] step 401750/1000000 | loss=2.2689 vocab=0.4970 token=0.8887 | 6.15 steps/s | VRAM=31.40 Go [2026-07-17 17:49:47] step 401800/1000000 | loss=2.2986 vocab=0.8220 token=1.6206 | 6.13 steps/s | VRAM=26.74 Go [2026-07-17 17:49:56] step 401850/1000000 | loss=2.3509 vocab=0.9562 token=1.2464 | 6.12 steps/s | VRAM=32.31 Go [2026-07-17 17:50:04] step 401900/1000000 | loss=2.3158 vocab=0.9411 token=1.1292 | 6.16 steps/s | VRAM=31.45 Go [2026-07-17 17:50:12] step 401950/1000000 | loss=2.4126 vocab=1.6137 token=1.2980 | 6.12 steps/s | VRAM=32.17 Go [2026-07-17 17:50:20] step 402000/1000000 | loss=2.4382 vocab=1.0667 token=1.5144 | 6.12 steps/s | VRAM=28.48 Go [2026-07-17 17:50:28] step 402050/1000000 | loss=2.2343 vocab=0.8906 token=1.5336 | 6.15 steps/s | VRAM=31.83 Go [2026-07-17 17:50:36] step 402100/1000000 | loss=2.3182 vocab=0.9002 token=1.3688 | 6.13 steps/s | VRAM=28.39 Go [2026-07-17 17:50:45] step 402150/1000000 | loss=2.3184 vocab=0.8131 token=1.0551 | 6.11 steps/s | VRAM=26.36 Go [2026-07-17 17:50:53] step 402200/1000000 | loss=2.2611 vocab=1.4113 token=1.5185 | 6.13 steps/s | VRAM=28.97 Go [2026-07-17 17:51:01] step 402250/1000000 | loss=2.2648 vocab=0.4428 token=1.0730 | 6.14 steps/s | VRAM=28.48 Go [2026-07-17 17:51:09] step 402300/1000000 | loss=2.3466 vocab=1.0107 token=1.4308 | 6.13 steps/s | VRAM=28.58 Go [2026-07-17 17:51:17] step 402350/1000000 | loss=2.3354 vocab=0.9390 token=1.6232 | 6.13 steps/s | VRAM=27.88 Go [2026-07-17 17:51:25] step 402400/1000000 | loss=2.2315 vocab=1.2203 token=1.4675 | 6.12 steps/s | VRAM=29.44 Go [2026-07-17 17:51:33] step 402450/1000000 | loss=2.3617 vocab=0.9958 token=1.3927 | 6.12 steps/s | VRAM=25.25 Go [2026-07-17 17:51:42] step 402500/1000000 | loss=2.2987 vocab=1.4201 token=1.5851 | 6.14 steps/s | VRAM=24.76 Go [2026-07-17 17:51:50] step 402550/1000000 | loss=2.2808 vocab=0.6932 token=1.0528 | 6.12 steps/s | VRAM=29.23 Go [2026-07-17 17:51:58] step 402600/1000000 | loss=2.2407 vocab=0.7559 token=1.1556 | 6.15 steps/s | VRAM=28.93 Go [2026-07-17 17:52:06] step 402650/1000000 | loss=2.3603 vocab=0.5147 token=1.1742 | 6.15 steps/s | VRAM=28.28 Go [2026-07-17 17:52:14] step 402700/1000000 | loss=2.3360 vocab=1.2765 token=1.2646 | 6.14 steps/s | VRAM=32.91 Go [2026-07-17 17:52:22] step 402750/1000000 | loss=2.1981 vocab=0.8367 token=1.4550 | 6.14 steps/s | VRAM=32.10 Go [2026-07-17 17:52:30] step 402800/1000000 | loss=2.3174 vocab=0.7737 token=1.2576 | 6.14 steps/s | VRAM=28.48 Go [2026-07-17 17:52:39] step 402850/1000000 | loss=2.3517 vocab=0.8743 token=1.3097 | 6.13 steps/s | VRAM=30.09 Go [2026-07-17 17:52:47] step 402900/1000000 | loss=2.3119 vocab=1.2161 token=1.6540 | 6.14 steps/s | VRAM=31.72 Go [2026-07-17 17:52:55] step 402950/1000000 | loss=2.2588 vocab=0.8996 token=1.6395 | 6.13 steps/s | VRAM=26.46 Go [2026-07-17 17:53:03] step 403000/1000000 | loss=2.3758 vocab=0.8025 token=1.6186 | 6.15 steps/s | VRAM=27.35 Go [2026-07-17 17:53:11] step 403050/1000000 | loss=2.3218 vocab=0.8778 token=1.3612 | 6.13 steps/s | VRAM=32.74 Go [2026-07-17 17:53:19] step 403100/1000000 | loss=2.3498 vocab=0.7766 token=1.4481 | 6.11 steps/s | VRAM=28.54 Go [2026-07-17 17:53:28] step 403150/1000000 | loss=2.2047 vocab=1.3714 token=1.8100 | 6.13 steps/s | VRAM=28.81 Go [2026-07-17 17:53:36] step 403200/1000000 | loss=2.2276 vocab=0.7580 token=1.4229 | 6.14 steps/s | VRAM=32.24 Go [2026-07-17 17:53:44] step 403250/1000000 | loss=2.3139 vocab=0.9630 token=1.6217 | 6.13 steps/s | VRAM=27.84 Go [2026-07-17 17:53:52] step 403300/1000000 | loss=2.2695 vocab=0.9288 token=1.0423 | 6.12 steps/s | VRAM=32.11 Go [2026-07-17 17:54:00] step 403350/1000000 | loss=2.2780 vocab=0.7681 token=1.6414 | 6.09 steps/s | VRAM=26.88 Go [2026-07-17 17:54:08] step 403400/1000000 | loss=2.2860 vocab=0.8033 token=1.3890 | 6.13 steps/s | VRAM=28.63 Go [2026-07-17 17:54:17] step 403450/1000000 | loss=2.3484 vocab=1.0650 token=1.1968 | 6.09 steps/s | VRAM=28.30 Go [2026-07-17 17:54:25] step 403500/1000000 | loss=2.1957 vocab=1.3445 token=1.4580 | 6.14 steps/s | VRAM=25.90 Go [2026-07-17 17:54:33] step 403550/1000000 | loss=2.2296 vocab=1.2958 token=1.2704 | 6.15 steps/s | VRAM=29.00 Go [2026-07-17 17:54:41] step 403600/1000000 | loss=2.3640 vocab=0.8585 token=1.5586 | 6.13 steps/s | VRAM=32.04 Go [2026-07-17 17:54:49] step 403650/1000000 | loss=2.2623 vocab=0.9927 token=1.4055 | 6.13 steps/s | VRAM=26.42 Go [2026-07-17 17:54:57] step 403700/1000000 | loss=2.3868 vocab=0.6512 token=1.1208 | 6.12 steps/s | VRAM=29.36 Go [2026-07-17 17:55:06] step 403750/1000000 | loss=2.2103 vocab=1.0831 token=1.0041 | 6.11 steps/s | VRAM=28.80 Go [2026-07-17 17:55:14] step 403800/1000000 | loss=2.3316 vocab=1.0526 token=1.2420 | 6.12 steps/s | VRAM=25.25 Go [2026-07-17 17:55:22] step 403850/1000000 | loss=2.3422 vocab=1.4412 token=1.5353 | 6.11 steps/s | VRAM=29.11 Go [2026-07-17 17:55:30] step 403900/1000000 | loss=2.2785 vocab=0.6428 token=0.8877 | 6.11 steps/s | VRAM=26.43 Go [2026-07-17 17:55:38] step 403950/1000000 | loss=2.3932 vocab=0.6005 token=1.0063 | 6.14 steps/s | VRAM=29.65 Go [2026-07-17 17:55:46] step 404000/1000000 | loss=2.1629 vocab=1.1440 token=1.5117 | 6.13 steps/s | VRAM=28.12 Go [2026-07-17 17:55:55] step 404050/1000000 | loss=2.3058 vocab=0.6225 token=1.1134 | 6.12 steps/s | VRAM=31.61 Go [2026-07-17 17:56:03] step 404100/1000000 | loss=2.4817 vocab=0.9542 token=0.9588 | 6.11 steps/s | VRAM=29.45 Go [2026-07-17 17:56:11] step 404150/1000000 | loss=2.3388 vocab=1.7499 token=1.7077 | 6.12 steps/s | VRAM=25.75 Go [2026-07-17 17:56:19] step 404200/1000000 | loss=2.2876 vocab=0.8634 token=1.1666 | 6.14 steps/s | VRAM=26.81 Go [2026-07-17 17:56:27] step 404250/1000000 | loss=2.2771 vocab=1.0689 token=1.4374 | 6.17 steps/s | VRAM=31.98 Go [2026-07-17 17:56:35] step 404300/1000000 | loss=2.2524 vocab=0.8667 token=1.2229 | 6.14 steps/s | VRAM=27.90 Go [2026-07-17 17:56:43] step 404350/1000000 | loss=2.1896 vocab=1.0251 token=1.0173 | 6.16 steps/s | VRAM=29.36 Go [2026-07-17 17:56:52] step 404400/1000000 | loss=2.2239 vocab=0.8679 token=1.3416 | 6.16 steps/s | VRAM=28.85 Go [2026-07-17 17:57:00] step 404450/1000000 | loss=2.2667 vocab=0.7806 token=1.2490 | 6.12 steps/s | VRAM=28.04 Go [2026-07-17 17:57:08] step 404500/1000000 | loss=2.3230 vocab=0.9700 token=1.0954 | 6.16 steps/s | VRAM=28.01 Go [2026-07-17 17:57:16] step 404550/1000000 | loss=2.1709 vocab=0.7361 token=1.1353 | 6.15 steps/s | VRAM=28.00 Go [2026-07-17 17:57:24] step 404600/1000000 | loss=2.2099 vocab=1.3811 token=1.1427 | 6.13 steps/s | VRAM=28.78 Go [2026-07-17 17:57:32] step 404650/1000000 | loss=2.1280 vocab=1.0666 token=1.0095 | 6.15 steps/s | VRAM=28.24 Go [2026-07-17 17:57:40] step 404700/1000000 | loss=2.3216 vocab=0.6538 token=1.0831 | 6.15 steps/s | VRAM=32.16 Go [2026-07-17 17:57:49] step 404750/1000000 | loss=2.3186 vocab=0.7363 token=1.0419 | 6.13 steps/s | VRAM=28.06 Go [2026-07-17 17:57:57] step 404800/1000000 | loss=2.3241 vocab=0.9551 token=1.2843 | 6.15 steps/s | VRAM=28.23 Go [2026-07-17 17:58:05] step 404850/1000000 | loss=2.3594 vocab=1.3054 token=1.9209 | 6.14 steps/s | VRAM=25.97 Go [2026-07-17 17:58:13] step 404900/1000000 | loss=2.2706 vocab=0.9272 token=1.5407 | 6.18 steps/s | VRAM=24.94 Go [2026-07-17 17:58:21] step 404950/1000000 | loss=2.2448 vocab=1.2698 token=1.4174 | 6.16 steps/s | VRAM=26.40 Go [2026-07-17 17:58:29] step 405000/1000000 | loss=2.3508 vocab=0.7365 token=1.2293 | 6.18 steps/s | VRAM=31.84 Go [2026-07-17 17:58:37] step 405050/1000000 | loss=2.3034 vocab=0.8201 token=1.1565 | 6.13 steps/s | VRAM=32.83 Go [2026-07-17 17:58:45] step 405100/1000000 | loss=2.2137 vocab=1.5293 token=1.3933 | 6.13 steps/s | VRAM=28.43 Go [2026-07-17 17:58:54] step 405150/1000000 | loss=2.2819 vocab=0.8379 token=1.1136 | 6.14 steps/s | VRAM=29.14 Go [2026-07-17 17:59:02] step 405200/1000000 | loss=2.2588 vocab=1.1904 token=1.3615 | 6.13 steps/s | VRAM=28.07 Go [2026-07-17 17:59:10] step 405250/1000000 | loss=2.2418 vocab=0.8370 token=1.1905 | 6.12 steps/s | VRAM=26.69 Go [2026-07-17 17:59:18] step 405300/1000000 | loss=2.2777 vocab=0.7086 token=1.0243 | 6.11 steps/s | VRAM=26.61 Go [2026-07-17 17:59:26] step 405350/1000000 | loss=2.3345 vocab=0.7441 token=1.4404 | 6.15 steps/s | VRAM=27.88 Go [2026-07-17 17:59:34] step 405400/1000000 | loss=2.2825 vocab=0.8872 token=1.3454 | 6.15 steps/s | VRAM=26.17 Go [2026-07-17 17:59:42] step 405450/1000000 | loss=2.2788 vocab=0.9049 token=1.1245 | 6.16 steps/s | VRAM=28.40 Go [2026-07-17 17:59:51] step 405500/1000000 | loss=2.4387 vocab=1.3065 token=1.6827 | 6.15 steps/s | VRAM=26.19 Go [2026-07-17 17:59:59] step 405550/1000000 | loss=2.2904 vocab=1.2976 token=1.0614 | 6.16 steps/s | VRAM=32.83 Go [2026-07-17 18:00:07] step 405600/1000000 | loss=2.2495 vocab=0.5838 token=1.0156 | 6.17 steps/s | VRAM=32.36 Go [2026-07-17 18:00:15] step 405650/1000000 | loss=2.3217 vocab=0.5888 token=1.0830 | 6.14 steps/s | VRAM=29.06 Go [2026-07-17 18:00:23] step 405700/1000000 | loss=2.2376 vocab=1.0539 token=0.8820 | 6.13 steps/s | VRAM=29.39 Go [2026-07-17 18:00:31] step 405750/1000000 | loss=2.3821 vocab=1.0731 token=1.8575 | 6.11 steps/s | VRAM=26.24 Go [2026-07-17 18:00:39] step 405800/1000000 | loss=2.2884 vocab=0.9610 token=1.3760 | 6.14 steps/s | VRAM=32.78 Go [2026-07-17 18:00:48] step 405850/1000000 | loss=2.2990 vocab=0.9121 token=1.4487 | 6.13 steps/s | VRAM=32.52 Go [2026-07-17 18:00:56] step 405900/1000000 | loss=2.2133 vocab=1.0856 token=1.2388 | 6.13 steps/s | VRAM=28.91 Go [2026-07-17 18:01:04] step 405950/1000000 | loss=2.3674 vocab=0.7635 token=1.7302 | 6.16 steps/s | VRAM=25.85 Go [2026-07-17 18:01:12] step 406000/1000000 | loss=2.1906 vocab=1.3341 token=1.5288 | 6.11 steps/s | VRAM=30.05 Go [2026-07-17 18:01:20] step 406050/1000000 | loss=2.1655 vocab=0.7929 token=1.3371 | 6.16 steps/s | VRAM=25.14 Go [2026-07-17 18:01:28] step 406100/1000000 | loss=2.2328 vocab=0.5482 token=1.1726 | 6.15 steps/s | VRAM=29.30 Go [2026-07-17 18:01:37] step 406150/1000000 | loss=2.3134 vocab=0.8056 token=1.3512 | 6.13 steps/s | VRAM=28.73 Go [2026-07-17 18:01:45] step 406200/1000000 | loss=2.1891 vocab=0.8652 token=1.8538 | 6.14 steps/s | VRAM=25.56 Go [2026-07-17 18:01:53] step 406250/1000000 | loss=2.2391 vocab=1.0726 token=1.1982 | 6.14 steps/s | VRAM=29.21 Go [2026-07-17 18:02:01] step 406300/1000000 | loss=2.2674 vocab=0.8584 token=1.1257 | 6.15 steps/s | VRAM=28.82 Go [2026-07-17 18:02:09] step 406350/1000000 | loss=2.1892 vocab=0.7763 token=1.2528 | 6.14 steps/s | VRAM=28.53 Go [2026-07-17 18:02:17] step 406400/1000000 | loss=2.3088 vocab=0.9775 token=1.4612 | 6.17 steps/s | VRAM=32.40 Go [2026-07-17 18:02:25] step 406450/1000000 | loss=2.3065 vocab=1.2924 token=1.3847 | 6.13 steps/s | VRAM=26.31 Go [2026-07-17 18:02:33] step 406500/1000000 | loss=2.2794 vocab=0.9671 token=1.2154 | 6.14 steps/s | VRAM=28.79 Go [2026-07-17 18:02:42] step 406550/1000000 | loss=2.3044 vocab=0.5357 token=1.2764 | 6.15 steps/s | VRAM=32.97 Go [2026-07-17 18:02:50] step 406600/1000000 | loss=2.2036 vocab=1.0192 token=1.5834 | 6.13 steps/s | VRAM=29.30 Go [2026-07-17 18:02:58] step 406650/1000000 | loss=2.1532 vocab=1.0554 token=1.3234 | 6.15 steps/s | VRAM=26.38 Go [2026-07-17 18:03:06] step 406700/1000000 | loss=2.2196 vocab=0.6134 token=1.2569 | 6.17 steps/s | VRAM=31.63 Go [2026-07-17 18:03:14] step 406750/1000000 | loss=2.2942 vocab=1.0471 token=1.0099 | 6.14 steps/s | VRAM=32.87 Go [2026-07-17 18:03:22] step 406800/1000000 | loss=2.2272 vocab=1.2755 token=1.1948 | 6.13 steps/s | VRAM=28.36 Go [2026-07-17 18:03:30] step 406850/1000000 | loss=2.3752 vocab=1.2705 token=1.4091 | 6.13 steps/s | VRAM=32.91 Go [2026-07-17 18:03:39] step 406900/1000000 | loss=2.3157 vocab=1.0163 token=1.6594 | 6.10 steps/s | VRAM=28.59 Go [2026-07-17 18:03:47] step 406950/1000000 | loss=2.2519 vocab=1.2822 token=1.7107 | 6.16 steps/s | VRAM=32.54 Go [2026-07-17 18:03:55] step 407000/1000000 | loss=2.2276 vocab=0.8105 token=1.3552 | 6.14 steps/s | VRAM=28.81 Go [2026-07-17 18:04:03] step 407050/1000000 | loss=2.2894 vocab=0.9646 token=1.1905 | 6.12 steps/s | VRAM=28.59 Go [2026-07-17 18:04:11] step 407100/1000000 | loss=2.2776 vocab=0.9027 token=1.2085 | 6.10 steps/s | VRAM=26.88 Go [2026-07-17 18:04:19] step 407150/1000000 | loss=2.2993 vocab=0.8575 token=1.0630 | 6.16 steps/s | VRAM=24.67 Go [2026-07-17 18:04:28] step 407200/1000000 | loss=2.2517 vocab=1.0112 token=1.2481 | 6.11 steps/s | VRAM=29.16 Go [2026-07-17 18:04:36] step 407250/1000000 | loss=2.3070 vocab=0.9137 token=1.0294 | 6.09 steps/s | VRAM=28.80 Go [2026-07-17 18:04:44] step 407300/1000000 | loss=2.3193 vocab=1.3084 token=1.3870 | 6.17 steps/s | VRAM=28.66 Go [2026-07-17 18:04:52] step 407350/1000000 | loss=2.2168 vocab=0.3537 token=1.0640 | 6.13 steps/s | VRAM=25.91 Go [2026-07-17 18:05:00] step 407400/1000000 | loss=2.3451 vocab=1.1726 token=1.5783 | 6.15 steps/s | VRAM=28.74 Go [2026-07-17 18:05:08] step 407450/1000000 | loss=2.3530 vocab=0.6009 token=1.2115 | 6.10 steps/s | VRAM=28.50 Go [2026-07-17 18:05:17] step 407500/1000000 | loss=2.3313 vocab=0.7864 token=1.0624 | 6.15 steps/s | VRAM=31.69 Go [2026-07-17 18:05:25] step 407550/1000000 | loss=2.2281 vocab=1.1954 token=1.8818 | 6.11 steps/s | VRAM=30.11 Go [2026-07-17 18:05:33] step 407600/1000000 | loss=2.2576 vocab=0.5501 token=1.0674 | 6.15 steps/s | VRAM=27.62 Go [2026-07-17 18:05:41] step 407650/1000000 | loss=2.3261 vocab=0.7710 token=1.4934 | 6.08 steps/s | VRAM=30.27 Go [2026-07-17 18:05:49] step 407700/1000000 | loss=2.3480 vocab=0.8603 token=1.3728 | 6.12 steps/s | VRAM=32.30 Go [2026-07-17 18:05:57] step 407750/1000000 | loss=2.2516 vocab=1.0822 token=1.5408 | 6.14 steps/s | VRAM=32.21 Go [2026-07-17 18:06:06] step 407800/1000000 | loss=2.2607 vocab=0.4183 token=1.0414 | 6.12 steps/s | VRAM=32.38 Go [2026-07-17 18:06:14] step 407850/1000000 | loss=2.2676 vocab=0.6411 token=1.3060 | 6.12 steps/s | VRAM=29.01 Go [2026-07-17 18:06:22] step 407900/1000000 | loss=2.2740 vocab=1.0639 token=1.1814 | 6.14 steps/s | VRAM=28.18 Go [2026-07-17 18:06:30] step 407950/1000000 | loss=2.3665 vocab=1.1542 token=1.5867 | 6.10 steps/s | VRAM=29.15 Go [2026-07-17 18:06:38] step 408000/1000000 | loss=2.2442 vocab=0.7764 token=1.2568 | 6.14 steps/s | VRAM=25.17 Go [2026-07-17 18:06:46] step 408050/1000000 | loss=2.2406 vocab=0.8667 token=1.2368 | 6.14 steps/s | VRAM=28.66 Go [2026-07-17 18:06:54] step 408100/1000000 | loss=2.2346 vocab=0.3882 token=1.1867 | 6.16 steps/s | VRAM=24.99 Go [2026-07-17 18:07:03] step 408150/1000000 | loss=2.2001 vocab=0.7657 token=1.3125 | 6.14 steps/s | VRAM=26.21 Go [2026-07-17 18:07:11] step 408200/1000000 | loss=2.3530 vocab=1.7439 token=1.9726 | 6.12 steps/s | VRAM=29.13 Go [2026-07-17 18:07:19] step 408250/1000000 | loss=2.3316 vocab=0.7527 token=1.3952 | 6.12 steps/s | VRAM=26.03 Go [2026-07-17 18:07:27] step 408300/1000000 | loss=2.1981 vocab=0.7886 token=1.6377 | 6.17 steps/s | VRAM=30.84 Go [2026-07-17 18:07:35] step 408350/1000000 | loss=2.2823 vocab=0.7544 token=1.0752 | 6.09 steps/s | VRAM=26.88 Go [2026-07-17 18:07:43] step 408400/1000000 | loss=2.3403 vocab=0.7936 token=1.1851 | 6.15 steps/s | VRAM=24.72 Go [2026-07-17 18:07:52] step 408450/1000000 | loss=2.3266 vocab=1.1088 token=1.2903 | 6.13 steps/s | VRAM=28.50 Go [2026-07-17 18:08:00] step 408500/1000000 | loss=2.3348 vocab=0.9041 token=1.5899 | 6.15 steps/s | VRAM=25.39 Go [2026-07-17 18:08:08] step 408550/1000000 | loss=2.1368 vocab=1.1373 token=1.2458 | 6.15 steps/s | VRAM=32.54 Go [2026-07-17 18:08:16] step 408600/1000000 | loss=2.2733 vocab=0.7664 token=1.4528 | 6.11 steps/s | VRAM=29.40 Go [2026-07-17 18:08:24] step 408650/1000000 | loss=2.2679 vocab=0.6299 token=1.0818 | 6.12 steps/s | VRAM=27.96 Go [2026-07-17 18:08:32] step 408700/1000000 | loss=2.3137 vocab=1.1582 token=2.0576 | 6.15 steps/s | VRAM=28.68 Go [2026-07-17 18:08:40] step 408750/1000000 | loss=2.2146 vocab=0.9863 token=1.6333 | 6.12 steps/s | VRAM=26.55 Go [2026-07-17 18:08:49] step 408800/1000000 | loss=2.2455 vocab=1.2111 token=1.4934 | 6.13 steps/s | VRAM=27.62 Go [2026-07-17 18:08:57] step 408850/1000000 | loss=2.3029 vocab=0.6062 token=1.3289 | 6.15 steps/s | VRAM=24.89 Go [2026-07-17 18:09:05] step 408900/1000000 | loss=2.2471 vocab=1.0269 token=1.5210 | 6.15 steps/s | VRAM=28.78 Go [2026-07-17 18:09:13] step 408950/1000000 | loss=2.1550 vocab=0.9452 token=1.2385 | 6.15 steps/s | VRAM=31.24 Go [2026-07-17 18:09:21] step 409000/1000000 | loss=2.1965 vocab=0.8510 token=1.1750 | 6.13 steps/s | VRAM=28.37 Go [2026-07-17 18:09:29] step 409050/1000000 | loss=2.3610 vocab=1.3888 token=1.6378 | 6.15 steps/s | VRAM=32.25 Go [2026-07-17 18:09:37] step 409100/1000000 | loss=2.1672 vocab=1.2363 token=1.8733 | 6.13 steps/s | VRAM=28.36 Go [2026-07-17 18:09:46] step 409150/1000000 | loss=2.2687 vocab=0.7012 token=0.9073 | 6.09 steps/s | VRAM=26.65 Go [2026-07-17 18:09:54] step 409200/1000000 | loss=2.2582 vocab=0.8699 token=1.3963 | 6.13 steps/s | VRAM=28.31 Go [2026-07-17 18:10:02] step 409250/1000000 | loss=2.2664 vocab=0.7601 token=1.5427 | 6.13 steps/s | VRAM=28.73 Go [2026-07-17 18:10:10] step 409300/1000000 | loss=2.2577 vocab=0.9888 token=1.3952 | 6.13 steps/s | VRAM=28.74 Go [2026-07-17 18:10:18] step 409350/1000000 | loss=2.2878 vocab=1.3783 token=0.9714 | 6.11 steps/s | VRAM=32.96 Go [2026-07-17 18:10:26] step 409400/1000000 | loss=2.2459 vocab=0.7983 token=1.4920 | 6.16 steps/s | VRAM=28.28 Go [2026-07-17 18:10:35] step 409450/1000000 | loss=2.2599 vocab=0.6550 token=0.9080 | 6.16 steps/s | VRAM=26.04 Go [2026-07-17 18:10:43] step 409500/1000000 | loss=2.2757 vocab=1.5226 token=1.6059 | 6.15 steps/s | VRAM=28.25 Go [2026-07-17 18:10:51] step 409550/1000000 | loss=2.2898 vocab=0.8344 token=1.1751 | 6.15 steps/s | VRAM=25.17 Go [2026-07-17 18:10:59] step 409600/1000000 | loss=2.2217 vocab=0.5909 token=1.4399 | 6.14 steps/s | VRAM=25.86 Go [2026-07-17 18:11:07] step 409650/1000000 | loss=2.2591 vocab=1.0240 token=1.2369 | 6.11 steps/s | VRAM=28.76 Go [2026-07-17 18:11:15] step 409700/1000000 | loss=2.2929 vocab=0.3615 token=1.2994 | 6.12 steps/s | VRAM=29.82 Go [2026-07-17 18:11:24] step 409750/1000000 | loss=2.3060 vocab=0.7790 token=1.2805 | 6.10 steps/s | VRAM=26.64 Go [2026-07-17 18:11:32] step 409800/1000000 | loss=2.3144 vocab=0.7731 token=1.0194 | 6.13 steps/s | VRAM=28.37 Go [2026-07-17 18:11:40] step 409850/1000000 | loss=2.1624 vocab=0.5419 token=1.3574 | 6.14 steps/s | VRAM=32.85 Go [2026-07-17 18:11:48] step 409900/1000000 | loss=2.2169 vocab=1.0836 token=1.1585 | 6.10 steps/s | VRAM=28.28 Go [2026-07-17 18:11:56] step 409950/1000000 | loss=2.3146 vocab=0.8359 token=1.5052 | 6.16 steps/s | VRAM=29.05 Go [2026-07-17 18:12:04] step 410000/1000000 | loss=2.2244 vocab=0.9147 token=0.9583 | 6.15 steps/s | VRAM=32.05 Go [2026-07-17 18:12:12] step 410050/1000000 | loss=2.2180 vocab=0.8010 token=1.1740 | 6.12 steps/s | VRAM=28.31 Go [2026-07-17 18:12:21] step 410100/1000000 | loss=2.2177 vocab=0.7016 token=1.4130 | 6.11 steps/s | VRAM=26.21 Go [2026-07-17 18:12:29] step 410150/1000000 | loss=2.2178 vocab=0.5697 token=1.1755 | 6.15 steps/s | VRAM=31.32 Go [2026-07-17 18:12:37] step 410200/1000000 | loss=2.3061 vocab=0.5450 token=1.2711 | 6.13 steps/s | VRAM=32.57 Go [2026-07-17 18:12:45] step 410250/1000000 | loss=2.2731 vocab=0.4963 token=1.2641 | 6.16 steps/s | VRAM=32.42 Go [2026-07-17 18:12:53] step 410300/1000000 | loss=2.2317 vocab=1.0794 token=1.6690 | 6.15 steps/s | VRAM=32.42 Go [2026-07-17 18:13:01] step 410350/1000000 | loss=2.2790 vocab=0.8886 token=1.7780 | 6.12 steps/s | VRAM=26.22 Go [2026-07-17 18:13:09] step 410400/1000000 | loss=2.3469 vocab=0.9696 token=1.3198 | 6.13 steps/s | VRAM=25.08 Go [2026-07-17 18:13:18] step 410450/1000000 | loss=2.2909 vocab=1.5007 token=1.6781 | 6.15 steps/s | VRAM=31.70 Go [2026-07-17 18:13:26] step 410500/1000000 | loss=2.2068 vocab=1.2343 token=1.4522 | 6.12 steps/s | VRAM=23.68 Go [2026-07-17 18:13:34] step 410550/1000000 | loss=2.2601 vocab=1.3355 token=2.4410 | 6.18 steps/s | VRAM=31.52 Go [2026-07-17 18:13:42] step 410600/1000000 | loss=2.3306 vocab=1.1680 token=1.1364 | 6.12 steps/s | VRAM=26.21 Go [2026-07-17 18:13:50] step 410650/1000000 | loss=2.3655 vocab=0.9486 token=1.3273 | 6.14 steps/s | VRAM=31.32 Go [2026-07-17 18:13:58] step 410700/1000000 | loss=2.2358 vocab=1.0080 token=1.4909 | 6.12 steps/s | VRAM=32.99 Go [2026-07-17 18:14:07] step 410750/1000000 | loss=2.3528 vocab=1.0694 token=1.0574 | 6.11 steps/s | VRAM=26.44 Go [2026-07-17 18:14:15] step 410800/1000000 | loss=2.2639 vocab=0.9595 token=1.4503 | 6.16 steps/s | VRAM=27.94 Go [2026-07-17 18:14:23] step 410850/1000000 | loss=2.1906 vocab=0.8823 token=1.5331 | 6.11 steps/s | VRAM=29.49 Go [2026-07-17 18:14:31] step 410900/1000000 | loss=2.2598 vocab=0.7354 token=1.2208 | 6.12 steps/s | VRAM=28.51 Go [2026-07-17 18:14:39] step 410950/1000000 | loss=2.2345 vocab=0.9361 token=1.2704 | 6.13 steps/s | VRAM=29.15 Go [2026-07-17 18:14:47] step 411000/1000000 | loss=2.3548 vocab=0.5636 token=1.0603 | 6.12 steps/s | VRAM=26.25 Go [2026-07-17 18:14:55] step 411050/1000000 | loss=2.3047 vocab=0.5464 token=0.7900 | 6.13 steps/s | VRAM=28.91 Go [2026-07-17 18:15:04] step 411100/1000000 | loss=2.2763 vocab=0.8397 token=1.4265 | 6.14 steps/s | VRAM=26.35 Go [2026-07-17 18:15:12] step 411150/1000000 | loss=2.3238 vocab=1.5917 token=1.4391 | 6.13 steps/s | VRAM=28.18 Go [2026-07-17 18:15:20] step 411200/1000000 | loss=2.3324 vocab=0.5963 token=1.0351 | 6.14 steps/s | VRAM=32.44 Go [2026-07-17 18:15:28] step 411250/1000000 | loss=2.3749 vocab=1.3703 token=1.1269 | 6.16 steps/s | VRAM=31.89 Go [2026-07-17 18:15:36] step 411300/1000000 | loss=2.3621 vocab=0.9528 token=1.4505 | 6.11 steps/s | VRAM=26.36 Go [2026-07-17 18:15:44] step 411350/1000000 | loss=2.2941 vocab=0.9388 token=1.4533 | 6.17 steps/s | VRAM=31.56 Go [2026-07-17 18:15:52] step 411400/1000000 | loss=2.2648 vocab=0.3172 token=0.8851 | 6.16 steps/s | VRAM=31.26 Go [2026-07-17 18:16:01] step 411450/1000000 | loss=2.2798 vocab=0.7525 token=1.2806 | 6.11 steps/s | VRAM=27.16 Go [2026-07-17 18:16:09] step 411500/1000000 | loss=2.2551 vocab=0.7046 token=1.0474 | 6.15 steps/s | VRAM=29.22 Go [2026-07-17 18:16:17] step 411550/1000000 | loss=2.2542 vocab=0.5328 token=1.1832 | 6.16 steps/s | VRAM=31.98 Go [2026-07-17 18:16:25] step 411600/1000000 | loss=2.3226 vocab=0.9482 token=1.3528 | 6.14 steps/s | VRAM=25.57 Go [2026-07-17 18:16:33] step 411650/1000000 | loss=2.3027 vocab=0.9359 token=1.5956 | 6.15 steps/s | VRAM=28.12 Go [2026-07-17 18:16:41] step 411700/1000000 | loss=2.3262 vocab=1.4507 token=1.3568 | 6.15 steps/s | VRAM=28.09 Go [2026-07-17 18:16:49] step 411750/1000000 | loss=2.2017 vocab=0.7120 token=1.1390 | 6.14 steps/s | VRAM=28.64 Go [2026-07-17 18:16:58] step 411800/1000000 | loss=2.1987 vocab=0.7172 token=0.9098 | 6.11 steps/s | VRAM=29.12 Go [2026-07-17 18:17:06] step 411850/1000000 | loss=2.3216 vocab=0.8429 token=1.2500 | 6.13 steps/s | VRAM=28.35 Go [2026-07-17 18:17:14] step 411900/1000000 | loss=2.2489 vocab=1.0512 token=1.4293 | 6.13 steps/s | VRAM=28.54 Go [2026-07-17 18:17:22] step 411950/1000000 | loss=2.1994 vocab=0.8255 token=1.4543 | 6.11 steps/s | VRAM=28.98 Go [2026-07-17 18:17:30] step 412000/1000000 | loss=2.2346 vocab=0.8406 token=1.3085 | 6.08 steps/s | VRAM=26.73 Go [2026-07-17 18:17:38] step 412050/1000000 | loss=2.4135 vocab=0.9136 token=1.4266 | 6.14 steps/s | VRAM=24.32 Go [2026-07-17 18:17:47] step 412100/1000000 | loss=2.2351 vocab=0.7390 token=1.4353 | 6.10 steps/s | VRAM=27.61 Go [2026-07-17 18:17:55] step 412150/1000000 | loss=2.3345 vocab=0.5634 token=0.9634 | 6.14 steps/s | VRAM=26.67 Go [2026-07-17 18:18:03] step 412200/1000000 | loss=2.2364 vocab=0.6334 token=0.9249 | 6.15 steps/s | VRAM=31.56 Go [2026-07-17 18:18:11] step 412250/1000000 | loss=2.2865 vocab=0.7643 token=1.3842 | 6.12 steps/s | VRAM=25.93 Go [2026-07-17 18:18:19] step 412300/1000000 | loss=2.2804 vocab=1.1386 token=1.1711 | 6.15 steps/s | VRAM=29.63 Go [2026-07-17 18:18:27] step 412350/1000000 | loss=2.2919 vocab=0.7293 token=1.1302 | 6.10 steps/s | VRAM=28.74 Go [2026-07-17 18:18:36] step 412400/1000000 | loss=2.2484 vocab=0.7952 token=1.4563 | 6.16 steps/s | VRAM=28.01 Go [2026-07-17 18:18:44] step 412450/1000000 | loss=2.2544 vocab=1.3096 token=1.4462 | 6.13 steps/s | VRAM=28.47 Go [2026-07-17 18:18:52] step 412500/1000000 | loss=2.3132 vocab=0.7582 token=1.6275 | 6.08 steps/s | VRAM=25.34 Go [2026-07-17 18:19:00] step 412550/1000000 | loss=2.2466 vocab=0.7454 token=1.0155 | 6.11 steps/s | VRAM=28.11 Go [2026-07-17 18:19:08] step 412600/1000000 | loss=2.2479 vocab=0.8459 token=1.6405 | 6.11 steps/s | VRAM=29.67 Go [2026-07-17 18:19:16] step 412650/1000000 | loss=2.3527 vocab=0.8604 token=1.5347 | 6.15 steps/s | VRAM=27.89 Go [2026-07-17 18:19:25] step 412700/1000000 | loss=2.2183 vocab=1.3625 token=1.4489 | 6.15 steps/s | VRAM=32.49 Go [2026-07-17 18:19:33] step 412750/1000000 | loss=2.2414 vocab=1.3412 token=1.9908 | 6.09 steps/s | VRAM=26.87 Go [2026-07-17 18:19:41] step 412800/1000000 | loss=2.3433 vocab=1.0614 token=1.2312 | 6.14 steps/s | VRAM=28.95 Go [2026-07-17 18:19:49] step 412850/1000000 | loss=2.2338 vocab=1.3258 token=1.5806 | 6.16 steps/s | VRAM=32.21 Go [2026-07-17 18:19:57] step 412900/1000000 | loss=2.2757 vocab=0.5345 token=0.9726 | 6.13 steps/s | VRAM=27.82 Go [2026-07-17 18:20:05] step 412950/1000000 | loss=2.3300 vocab=0.9200 token=1.5422 | 6.13 steps/s | VRAM=26.24 Go [2026-07-17 18:20:13] step 413000/1000000 | loss=2.2667 vocab=0.7704 token=1.3958 | 6.14 steps/s | VRAM=25.32 Go [2026-07-17 18:20:22] step 413050/1000000 | loss=2.2011 vocab=0.8044 token=1.4468 | 6.15 steps/s | VRAM=32.13 Go [2026-07-17 18:20:30] step 413100/1000000 | loss=2.2593 vocab=1.0109 token=1.5525 | 6.12 steps/s | VRAM=27.99 Go [2026-07-17 18:20:38] step 413150/1000000 | loss=2.1503 vocab=0.8999 token=1.2659 | 6.11 steps/s | VRAM=28.51 Go [2026-07-17 18:20:46] step 413200/1000000 | loss=2.3482 vocab=1.4214 token=1.4800 | 6.13 steps/s | VRAM=28.67 Go [2026-07-17 18:20:54] step 413250/1000000 | loss=2.2263 vocab=1.1688 token=1.3333 | 6.06 steps/s | VRAM=28.46 Go [2026-07-17 18:21:03] step 413300/1000000 | loss=2.3235 vocab=0.8802 token=1.1968 | 6.14 steps/s | VRAM=29.18 Go [2026-07-17 18:21:11] step 413350/1000000 | loss=2.3367 vocab=0.6214 token=0.9851 | 6.11 steps/s | VRAM=27.00 Go [2026-07-17 18:21:19] step 413400/1000000 | loss=2.2318 vocab=0.9992 token=1.7968 | 6.13 steps/s | VRAM=25.37 Go [2026-07-17 18:21:27] step 413450/1000000 | loss=2.2878 vocab=0.6777 token=1.3231 | 6.09 steps/s | VRAM=29.31 Go [2026-07-17 18:21:35] step 413500/1000000 | loss=2.2724 vocab=0.9235 token=1.1110 | 6.10 steps/s | VRAM=28.57 Go [2026-07-17 18:21:43] step 413550/1000000 | loss=2.2397 vocab=0.5801 token=1.3878 | 6.09 steps/s | VRAM=25.61 Go [2026-07-17 18:21:52] step 413600/1000000 | loss=2.2955 vocab=0.7288 token=1.2927 | 6.15 steps/s | VRAM=31.20 Go [2026-07-17 18:22:00] step 413650/1000000 | loss=2.2385 vocab=0.9072 token=1.1675 | 6.11 steps/s | VRAM=32.08 Go [2026-07-17 18:22:08] step 413700/1000000 | loss=2.2469 vocab=0.8596 token=2.1073 | 6.12 steps/s | VRAM=29.07 Go [2026-07-17 18:22:16] step 413750/1000000 | loss=2.2717 vocab=0.9505 token=1.4042 | 6.10 steps/s | VRAM=27.74 Go [2026-07-17 18:22:24] step 413800/1000000 | loss=2.3313 vocab=0.7460 token=1.2928 | 6.12 steps/s | VRAM=29.67 Go [2026-07-17 18:22:33] step 413850/1000000 | loss=2.2831 vocab=1.1827 token=1.5811 | 6.12 steps/s | VRAM=28.21 Go [2026-07-17 18:22:41] step 413900/1000000 | loss=2.2542 vocab=1.1544 token=1.4867 | 6.11 steps/s | VRAM=32.95 Go [2026-07-17 18:22:49] step 413950/1000000 | loss=2.2373 vocab=0.8187 token=1.0529 | 6.12 steps/s | VRAM=28.77 Go [2026-07-17 18:22:57] step 414000/1000000 | loss=2.2517 vocab=1.0699 token=1.7562 | 6.10 steps/s | VRAM=25.15 Go [2026-07-17 18:23:05] step 414050/1000000 | loss=2.1972 vocab=0.8444 token=1.5251 | 6.15 steps/s | VRAM=27.76 Go [2026-07-17 18:23:13] step 414100/1000000 | loss=2.3053 vocab=1.1882 token=1.1718 | 6.13 steps/s | VRAM=32.33 Go [2026-07-17 18:23:22] step 414150/1000000 | loss=2.3188 vocab=1.2512 token=1.2922 | 6.11 steps/s | VRAM=28.07 Go [2026-07-17 18:23:30] step 414200/1000000 | loss=2.2321 vocab=0.9544 token=1.1787 | 6.13 steps/s | VRAM=30.21 Go [2026-07-17 18:23:38] step 414250/1000000 | loss=2.3926 vocab=1.0357 token=1.5247 | 6.17 steps/s | VRAM=31.53 Go [2026-07-17 18:23:46] step 414300/1000000 | loss=2.3105 vocab=1.0020 token=1.8166 | 6.13 steps/s | VRAM=28.28 Go [2026-07-17 18:23:54] step 414350/1000000 | loss=2.2811 vocab=0.6837 token=1.1203 | 6.09 steps/s | VRAM=26.51 Go [2026-07-17 18:24:02] step 414400/1000000 | loss=2.2098 vocab=0.8535 token=1.0133 | 6.14 steps/s | VRAM=28.45 Go [2026-07-17 18:24:10] step 414450/1000000 | loss=2.3090 vocab=0.6984 token=1.2089 | 6.12 steps/s | VRAM=28.24 Go [2026-07-17 18:24:19] step 414500/1000000 | loss=2.2214 vocab=0.5311 token=1.3288 | 6.14 steps/s | VRAM=29.83 Go [2026-07-17 18:24:27] step 414550/1000000 | loss=2.2917 vocab=1.1385 token=1.3189 | 6.13 steps/s | VRAM=32.87 Go [2026-07-17 18:24:35] step 414600/1000000 | loss=2.2173 vocab=0.6603 token=0.9726 | 6.11 steps/s | VRAM=32.11 Go [2026-07-17 18:24:43] step 414650/1000000 | loss=2.2867 vocab=1.3523 token=1.1450 | 6.12 steps/s | VRAM=32.36 Go [2026-07-17 18:24:51] step 414700/1000000 | loss=2.3732 vocab=0.8082 token=1.4239 | 6.10 steps/s | VRAM=29.89 Go [2026-07-17 18:24:59] step 414750/1000000 | loss=2.3198 vocab=1.0890 token=1.4854 | 6.15 steps/s | VRAM=32.36 Go [2026-07-17 18:25:08] step 414800/1000000 | loss=2.2999 vocab=0.4621 token=1.2102 | 6.14 steps/s | VRAM=32.33 Go [2026-07-17 18:25:16] step 414850/1000000 | loss=2.2698 vocab=0.6158 token=1.1903 | 6.16 steps/s | VRAM=28.95 Go [2026-07-17 18:25:24] step 414900/1000000 | loss=2.2621 vocab=0.9719 token=1.2635 | 6.13 steps/s | VRAM=32.83 Go [2026-07-17 18:25:32] step 414950/1000000 | loss=2.3104 vocab=1.1098 token=1.6606 | 6.12 steps/s | VRAM=29.31 Go [2026-07-17 18:25:40] step 415000/1000000 | loss=2.2169 vocab=1.0623 token=1.6961 | 6.13 steps/s | VRAM=32.25 Go [2026-07-17 18:25:48] step 415050/1000000 | loss=2.2315 vocab=1.0340 token=1.4977 | 6.14 steps/s | VRAM=32.27 Go [2026-07-17 18:25:56] step 415100/1000000 | loss=2.1902 vocab=1.1294 token=1.8382 | 6.15 steps/s | VRAM=29.17 Go [2026-07-17 18:26:05] step 415150/1000000 | loss=2.2640 vocab=1.0602 token=1.1749 | 6.11 steps/s | VRAM=26.35 Go [2026-07-17 18:26:13] step 415200/1000000 | loss=2.2112 vocab=1.4774 token=1.3770 | 6.14 steps/s | VRAM=31.74 Go [2026-07-17 18:26:21] step 415250/1000000 | loss=2.2810 vocab=1.1427 token=1.1404 | 6.15 steps/s | VRAM=32.74 Go [2026-07-17 18:26:29] step 415300/1000000 | loss=2.2556 vocab=0.6783 token=1.3398 | 6.12 steps/s | VRAM=28.29 Go [2026-07-17 18:26:37] step 415350/1000000 | loss=2.2521 vocab=1.0077 token=1.3536 | 6.16 steps/s | VRAM=25.34 Go [2026-07-17 18:26:45] step 415400/1000000 | loss=2.2016 vocab=1.5881 token=1.4897 | 6.13 steps/s | VRAM=28.05 Go [2026-07-17 18:26:53] step 415450/1000000 | loss=2.2428 vocab=0.5037 token=1.1226 | 6.16 steps/s | VRAM=25.85 Go [2026-07-17 18:27:02] step 415500/1000000 | loss=2.2368 vocab=0.9594 token=1.1119 | 6.15 steps/s | VRAM=32.57 Go [2026-07-17 18:27:10] step 415550/1000000 | loss=2.1970 vocab=0.8326 token=1.3022 | 6.10 steps/s | VRAM=26.00 Go [2026-07-17 18:27:18] step 415600/1000000 | loss=2.1859 vocab=1.1065 token=1.3782 | 6.13 steps/s | VRAM=27.88 Go [2026-07-17 18:27:26] step 415650/1000000 | loss=2.2743 vocab=0.7802 token=1.2975 | 6.11 steps/s | VRAM=29.29 Go [2026-07-17 18:27:34] step 415700/1000000 | loss=2.2017 vocab=1.1328 token=1.1760 | 6.11 steps/s | VRAM=29.28 Go [2026-07-17 18:27:42] step 415750/1000000 | loss=2.3070 vocab=0.7933 token=0.9889 | 6.17 steps/s | VRAM=27.76 Go [2026-07-17 18:27:51] step 415800/1000000 | loss=2.2631 vocab=1.0447 token=1.3245 | 6.15 steps/s | VRAM=31.27 Go [2026-07-17 18:27:59] step 415850/1000000 | loss=2.2786 vocab=0.7470 token=1.0943 | 6.11 steps/s | VRAM=30.86 Go [2026-07-17 18:28:07] step 415900/1000000 | loss=2.2749 vocab=1.1482 token=1.6832 | 6.13 steps/s | VRAM=28.42 Go [2026-07-17 18:28:15] step 415950/1000000 | loss=2.3375 vocab=0.7514 token=1.3724 | 6.13 steps/s | VRAM=32.14 Go [2026-07-17 18:28:23] step 416000/1000000 | loss=2.2597 vocab=0.9948 token=1.4418 | 6.14 steps/s | VRAM=32.32 Go [2026-07-17 18:28:31] step 416050/1000000 | loss=2.2215 vocab=0.7943 token=1.1714 | 6.15 steps/s | VRAM=27.34 Go [2026-07-17 18:28:39] step 416100/1000000 | loss=2.2790 vocab=0.8748 token=1.0755 | 6.15 steps/s | VRAM=28.61 Go [2026-07-17 18:28:48] step 416150/1000000 | loss=2.2945 vocab=1.0831 token=1.3968 | 6.13 steps/s | VRAM=29.42 Go [2026-07-17 18:28:56] step 416200/1000000 | loss=2.3777 vocab=1.5808 token=1.6980 | 6.15 steps/s | VRAM=31.69 Go [2026-07-17 18:29:04] step 416250/1000000 | loss=2.2321 vocab=0.7205 token=1.2945 | 6.15 steps/s | VRAM=25.07 Go [2026-07-17 18:29:12] step 416300/1000000 | loss=2.2675 vocab=0.8911 token=1.3486 | 6.14 steps/s | VRAM=31.80 Go [2026-07-17 18:29:20] step 416350/1000000 | loss=2.2450 vocab=1.0217 token=1.4386 | 6.12 steps/s | VRAM=25.89 Go [2026-07-17 18:29:28] step 416400/1000000 | loss=2.2806 vocab=0.8368 token=1.5952 | 6.12 steps/s | VRAM=26.11 Go [2026-07-17 18:29:36] step 416450/1000000 | loss=2.3213 vocab=1.0529 token=1.3108 | 6.13 steps/s | VRAM=28.56 Go [2026-07-17 18:29:45] step 416500/1000000 | loss=2.2685 vocab=1.3553 token=1.7284 | 6.12 steps/s | VRAM=32.24 Go [2026-07-17 18:29:53] step 416550/1000000 | loss=2.2566 vocab=1.1394 token=1.4967 | 6.12 steps/s | VRAM=29.52 Go [2026-07-17 18:30:01] step 416600/1000000 | loss=2.2822 vocab=0.7124 token=1.3518 | 6.11 steps/s | VRAM=31.76 Go [2026-07-17 18:30:09] step 416650/1000000 | loss=2.1998 vocab=1.3365 token=1.6941 | 6.15 steps/s | VRAM=25.53 Go [2026-07-17 18:30:17] step 416700/1000000 | loss=2.1769 vocab=1.0331 token=1.5168 | 6.13 steps/s | VRAM=29.05 Go [2026-07-17 18:30:25] step 416750/1000000 | loss=2.3409 vocab=0.7451 token=1.5585 | 6.10 steps/s | VRAM=29.15 Go [2026-07-17 18:30:34] step 416800/1000000 | loss=2.3078 vocab=1.2917 token=1.5082 | 6.10 steps/s | VRAM=26.24 Go [2026-07-17 18:30:42] step 416850/1000000 | loss=2.2839 vocab=1.4383 token=1.5341 | 6.12 steps/s | VRAM=29.15 Go [2026-07-17 18:30:50] step 416900/1000000 | loss=2.2880 vocab=0.7125 token=1.6645 | 6.13 steps/s | VRAM=26.24 Go [2026-07-17 18:30:58] step 416950/1000000 | loss=2.2080 vocab=0.7804 token=1.4369 | 6.13 steps/s | VRAM=30.95 Go [2026-07-17 18:31:06] step 417000/1000000 | loss=2.2939 vocab=0.6216 token=1.0650 | 6.13 steps/s | VRAM=26.41 Go [2026-07-17 18:31:14] step 417050/1000000 | loss=2.1244 vocab=0.8385 token=0.9145 | 6.15 steps/s | VRAM=25.96 Go [2026-07-17 18:31:23] step 417100/1000000 | loss=2.3158 vocab=1.1206 token=1.8491 | 6.14 steps/s | VRAM=28.55 Go [2026-07-17 18:31:31] step 417150/1000000 | loss=2.2656 vocab=1.0093 token=1.4434 | 6.11 steps/s | VRAM=32.68 Go [2026-07-17 18:31:39] step 417200/1000000 | loss=2.3621 vocab=1.0712 token=1.2126 | 6.13 steps/s | VRAM=28.87 Go [2026-07-17 18:31:47] step 417250/1000000 | loss=2.3126 vocab=0.6775 token=1.4298 | 6.10 steps/s | VRAM=29.66 Go [2026-07-17 18:31:55] step 417300/1000000 | loss=2.2844 vocab=0.6784 token=1.4432 | 6.13 steps/s | VRAM=28.50 Go [2026-07-17 18:32:03] step 417350/1000000 | loss=2.2592 vocab=1.3407 token=1.6871 | 6.14 steps/s | VRAM=28.26 Go [2026-07-17 18:32:12] step 417400/1000000 | loss=2.3104 vocab=1.1756 token=1.2179 | 6.10 steps/s | VRAM=28.45 Go [2026-07-17 18:32:20] step 417450/1000000 | loss=2.3087 vocab=1.0766 token=1.5747 | 6.15 steps/s | VRAM=27.48 Go [2026-07-17 18:32:28] step 417500/1000000 | loss=2.1432 vocab=0.9205 token=1.4688 | 6.12 steps/s | VRAM=31.95 Go [2026-07-17 18:32:36] step 417550/1000000 | loss=2.2407 vocab=0.7761 token=1.5879 | 6.14 steps/s | VRAM=26.34 Go [2026-07-17 18:32:44] step 417600/1000000 | loss=2.2419 vocab=0.6967 token=1.2989 | 6.10 steps/s | VRAM=28.71 Go [2026-07-17 18:32:52] step 417650/1000000 | loss=2.3523 vocab=0.9943 token=1.4072 | 6.13 steps/s | VRAM=28.46 Go [2026-07-17 18:33:01] step 417700/1000000 | loss=2.2991 vocab=0.9310 token=1.0419 | 6.12 steps/s | VRAM=29.66 Go [2026-07-17 18:33:09] step 417750/1000000 | loss=2.2209 vocab=0.8365 token=1.3257 | 6.13 steps/s | VRAM=25.14 Go [2026-07-17 18:33:17] step 417800/1000000 | loss=2.2458 vocab=0.5640 token=1.3816 | 6.11 steps/s | VRAM=27.03 Go [2026-07-17 18:33:25] step 417850/1000000 | loss=2.3408 vocab=0.7348 token=1.3556 | 6.15 steps/s | VRAM=31.68 Go [2026-07-17 18:33:33] step 417900/1000000 | loss=2.2952 vocab=0.7569 token=1.3598 | 6.09 steps/s | VRAM=29.34 Go [2026-07-17 18:33:41] step 417950/1000000 | loss=2.4227 vocab=0.7528 token=1.1700 | 6.14 steps/s | VRAM=32.13 Go [2026-07-17 18:33:50] step 418000/1000000 | loss=2.2040 vocab=0.5141 token=1.3379 | 6.11 steps/s | VRAM=26.84 Go [2026-07-17 18:33:58] step 418050/1000000 | loss=2.3096 vocab=1.5274 token=1.8205 | 6.15 steps/s | VRAM=24.36 Go [2026-07-17 18:34:06] step 418100/1000000 | loss=2.3062 vocab=0.9917 token=1.1184 | 6.14 steps/s | VRAM=32.58 Go [2026-07-17 18:34:14] step 418150/1000000 | loss=2.1659 vocab=0.9355 token=1.6601 | 6.12 steps/s | VRAM=25.78 Go [2026-07-17 18:34:22] step 418200/1000000 | loss=2.3068 vocab=0.8598 token=1.3578 | 6.12 steps/s | VRAM=26.65 Go [2026-07-17 18:34:30] step 418250/1000000 | loss=2.2862 vocab=0.7277 token=0.9522 | 6.15 steps/s | VRAM=28.06 Go [2026-07-17 18:34:39] step 418300/1000000 | loss=2.2778 vocab=0.4275 token=1.1087 | 6.13 steps/s | VRAM=32.60 Go [2026-07-17 18:34:47] step 418350/1000000 | loss=2.3162 vocab=0.6631 token=1.5281 | 6.09 steps/s | VRAM=28.09 Go [2026-07-17 18:34:55] step 418400/1000000 | loss=2.2632 vocab=0.5653 token=1.2422 | 6.14 steps/s | VRAM=32.06 Go [2026-07-17 18:35:03] step 418450/1000000 | loss=2.2040 vocab=1.2104 token=1.1063 | 6.17 steps/s | VRAM=32.63 Go [2026-07-17 18:35:11] step 418500/1000000 | loss=2.1447 vocab=1.2239 token=1.7901 | 6.16 steps/s | VRAM=30.76 Go [2026-07-17 18:35:19] step 418550/1000000 | loss=2.2673 vocab=1.9566 token=1.7044 | 6.12 steps/s | VRAM=29.13 Go [2026-07-17 18:35:27] step 418600/1000000 | loss=2.2686 vocab=1.9512 token=1.7297 | 6.13 steps/s | VRAM=25.78 Go [2026-07-17 18:35:36] step 418650/1000000 | loss=2.2750 vocab=0.8393 token=1.5114 | 6.14 steps/s | VRAM=24.75 Go [2026-07-17 18:35:44] step 418700/1000000 | loss=2.1763 vocab=0.5495 token=0.9954 | 6.14 steps/s | VRAM=26.56 Go [2026-07-17 18:35:52] step 418750/1000000 | loss=2.2531 vocab=1.2603 token=1.4738 | 6.12 steps/s | VRAM=26.29 Go [2026-07-17 18:36:00] step 418800/1000000 | loss=2.2713 vocab=0.7862 token=1.1905 | 6.14 steps/s | VRAM=29.07 Go [2026-07-17 18:36:08] step 418850/1000000 | loss=2.3097 vocab=0.8966 token=1.7062 | 6.15 steps/s | VRAM=32.00 Go [2026-07-17 18:36:16] step 418900/1000000 | loss=2.2122 vocab=0.7089 token=1.1518 | 6.16 steps/s | VRAM=28.50 Go [2026-07-17 18:36:24] step 418950/1000000 | loss=2.3511 vocab=0.7243 token=1.3733 | 6.12 steps/s | VRAM=27.99 Go [2026-07-17 18:36:33] step 419000/1000000 | loss=2.3052 vocab=1.3715 token=1.9681 | 6.15 steps/s | VRAM=25.87 Go [2026-07-17 18:36:41] step 419050/1000000 | loss=2.1951 vocab=0.9561 token=1.2733 | 6.15 steps/s | VRAM=29.66 Go [2026-07-17 18:36:49] step 419100/1000000 | loss=2.2000 vocab=0.5399 token=1.1017 | 6.13 steps/s | VRAM=29.53 Go [2026-07-17 18:36:57] step 419150/1000000 | loss=2.2663 vocab=0.8685 token=1.0023 | 6.16 steps/s | VRAM=27.89 Go [2026-07-17 18:37:05] step 419200/1000000 | loss=2.4039 vocab=0.5251 token=1.1400 | 6.14 steps/s | VRAM=32.79 Go [2026-07-17 18:37:13] step 419250/1000000 | loss=2.3564 vocab=1.2362 token=1.4797 | 6.11 steps/s | VRAM=28.76 Go [2026-07-17 18:37:21] step 419300/1000000 | loss=2.2651 vocab=0.6393 token=1.1483 | 6.19 steps/s | VRAM=30.92 Go [2026-07-17 18:37:30] step 419350/1000000 | loss=2.1901 vocab=0.6803 token=1.1013 | 6.14 steps/s | VRAM=24.52 Go [2026-07-17 18:37:38] step 419400/1000000 | loss=2.3134 vocab=1.5216 token=1.7724 | 6.12 steps/s | VRAM=27.93 Go [2026-07-17 18:37:46] step 419450/1000000 | loss=2.3054 vocab=0.9304 token=1.5738 | 6.15 steps/s | VRAM=29.44 Go [2026-07-17 18:37:54] step 419500/1000000 | loss=2.1801 vocab=0.7588 token=1.3946 | 6.15 steps/s | VRAM=30.94 Go [2026-07-17 18:38:02] step 419550/1000000 | loss=2.2588 vocab=0.5035 token=1.3864 | 6.15 steps/s | VRAM=31.30 Go [2026-07-17 18:38:10] step 419600/1000000 | loss=2.3849 vocab=0.5971 token=1.1405 | 6.16 steps/s | VRAM=32.35 Go [2026-07-17 18:38:18] step 419650/1000000 | loss=2.1767 vocab=0.7281 token=1.2492 | 6.10 steps/s | VRAM=26.62 Go [2026-07-17 18:38:27] step 419700/1000000 | loss=2.3955 vocab=1.3484 token=1.5549 | 6.12 steps/s | VRAM=28.06 Go [2026-07-17 18:38:35] step 419750/1000000 | loss=2.3634 vocab=1.3819 token=1.4181 | 6.16 steps/s | VRAM=27.73 Go [2026-07-17 18:38:43] step 419800/1000000 | loss=2.1844 vocab=1.0851 token=1.2831 | 6.13 steps/s | VRAM=27.01 Go [2026-07-17 18:38:51] step 419850/1000000 | loss=2.2555 vocab=0.7646 token=1.3980 | 6.15 steps/s | VRAM=32.06 Go [2026-07-17 18:38:59] step 419900/1000000 | loss=2.2602 vocab=0.9117 token=1.8247 | 6.12 steps/s | VRAM=26.33 Go [2026-07-17 18:39:07] step 419950/1000000 | loss=2.1643 vocab=0.7916 token=1.2688 | 6.15 steps/s | VRAM=27.45 Go [2026-07-17 18:39:15] step 420000/1000000 | loss=2.2081 vocab=0.7913 token=1.1611 | 6.15 steps/s | VRAM=28.82 Go [2026-07-17 18:39:23] step 420050/1000000 | loss=2.2509 vocab=0.9089 token=1.5292 | 6.16 steps/s | VRAM=31.49 Go [2026-07-17 18:39:32] step 420100/1000000 | loss=2.3328 vocab=0.9398 token=1.1102 | 6.13 steps/s | VRAM=32.20 Go [2026-07-17 18:39:40] step 420150/1000000 | loss=2.1438 vocab=0.9774 token=1.1898 | 6.13 steps/s | VRAM=26.67 Go [2026-07-17 18:39:48] step 420200/1000000 | loss=2.2773 vocab=1.4914 token=1.3689 | 6.16 steps/s | VRAM=32.07 Go [2026-07-17 18:39:56] step 420250/1000000 | loss=2.3370 vocab=1.0779 token=1.4051 | 6.15 steps/s | VRAM=25.98 Go [2026-07-17 18:40:04] step 420300/1000000 | loss=2.1957 vocab=1.5152 token=1.4758 | 6.15 steps/s | VRAM=28.54 Go [2026-07-17 18:40:12] step 420350/1000000 | loss=2.3190 vocab=0.8253 token=1.4573 | 6.15 steps/s | VRAM=24.45 Go [2026-07-17 18:40:20] step 420400/1000000 | loss=2.3697 vocab=0.6715 token=1.2169 | 6.15 steps/s | VRAM=29.13 Go [2026-07-17 18:40:29] step 420450/1000000 | loss=2.2341 vocab=0.8753 token=1.3155 | 6.12 steps/s | VRAM=28.94 Go [2026-07-17 18:40:37] step 420500/1000000 | loss=2.1960 vocab=0.6359 token=1.2009 | 6.12 steps/s | VRAM=28.19 Go [2026-07-17 18:40:45] step 420550/1000000 | loss=2.1282 vocab=0.4442 token=1.0898 | 6.16 steps/s | VRAM=28.83 Go [2026-07-17 18:40:53] step 420600/1000000 | loss=2.2077 vocab=0.8738 token=1.0505 | 6.14 steps/s | VRAM=29.71 Go [2026-07-17 18:41:01] step 420650/1000000 | loss=2.3284 vocab=1.0222 token=1.8207 | 6.12 steps/s | VRAM=26.90 Go [2026-07-17 18:41:09] step 420700/1000000 | loss=2.4133 vocab=0.8666 token=1.5850 | 6.16 steps/s | VRAM=25.33 Go [2026-07-17 18:41:18] step 420750/1000000 | loss=2.2227 vocab=1.4583 token=1.9037 | 6.13 steps/s | VRAM=32.87 Go [2026-07-17 18:41:26] step 420800/1000000 | loss=2.2998 vocab=1.4570 token=1.5788 | 6.11 steps/s | VRAM=26.16 Go [2026-07-17 18:41:34] step 420850/1000000 | loss=2.2828 vocab=0.6870 token=1.3532 | 6.13 steps/s | VRAM=29.06 Go [2026-07-17 18:41:42] step 420900/1000000 | loss=2.3194 vocab=0.5446 token=1.5126 | 6.12 steps/s | VRAM=28.46 Go [2026-07-17 18:41:50] step 420950/1000000 | loss=2.2458 vocab=1.1510 token=1.5589 | 6.15 steps/s | VRAM=31.67 Go [2026-07-17 18:41:58] step 421000/1000000 | loss=2.2477 vocab=1.1573 token=1.6257 | 6.12 steps/s | VRAM=26.30 Go [2026-07-17 18:42:06] step 421050/1000000 | loss=2.3535 vocab=1.3231 token=1.6456 | 6.13 steps/s | VRAM=29.27 Go [2026-07-17 18:42:15] step 421100/1000000 | loss=2.2459 vocab=0.9458 token=1.3103 | 6.14 steps/s | VRAM=28.57 Go [2026-07-17 18:42:23] step 421150/1000000 | loss=2.2702 vocab=1.0586 token=1.2937 | 6.12 steps/s | VRAM=26.13 Go [2026-07-17 18:42:31] step 421200/1000000 | loss=2.1986 vocab=0.9595 token=1.4562 | 6.14 steps/s | VRAM=28.13 Go [2026-07-17 18:42:39] step 421250/1000000 | loss=2.1575 vocab=0.7164 token=1.4933 | 6.14 steps/s | VRAM=29.51 Go [2026-07-17 18:42:47] step 421300/1000000 | loss=2.2888 vocab=1.8137 token=1.8307 | 6.17 steps/s | VRAM=30.51 Go [2026-07-17 18:42:55] step 421350/1000000 | loss=2.2615 vocab=0.9148 token=1.3724 | 6.17 steps/s | VRAM=31.07 Go [2026-07-17 18:43:03] step 421400/1000000 | loss=2.1812 vocab=0.6579 token=1.2869 | 6.12 steps/s | VRAM=28.70 Go [2026-07-17 18:43:12] step 421450/1000000 | loss=2.3025 vocab=0.9118 token=1.2120 | 6.18 steps/s | VRAM=28.20 Go [2026-07-17 18:43:20] step 421500/1000000 | loss=2.2895 vocab=1.0923 token=1.0444 | 6.13 steps/s | VRAM=29.02 Go [2026-07-17 18:43:28] step 421550/1000000 | loss=2.1514 vocab=0.6376 token=1.2580 | 6.13 steps/s | VRAM=29.71 Go [2026-07-17 18:43:36] step 421600/1000000 | loss=2.2482 vocab=0.8668 token=1.0427 | 6.12 steps/s | VRAM=29.04 Go [2026-07-17 18:43:44] step 421650/1000000 | loss=2.2797 vocab=0.8983 token=1.6508 | 6.15 steps/s | VRAM=31.79 Go [2026-07-17 18:43:52] step 421700/1000000 | loss=2.3683 vocab=1.0825 token=1.2123 | 6.12 steps/s | VRAM=32.39 Go [2026-07-17 18:44:00] step 421750/1000000 | loss=2.3274 vocab=0.9894 token=0.8222 | 6.12 steps/s | VRAM=26.58 Go [2026-07-17 18:44:09] step 421800/1000000 | loss=2.3344 vocab=1.1632 token=1.7154 | 6.13 steps/s | VRAM=28.86 Go [2026-07-17 18:44:17] step 421850/1000000 | loss=2.1257 vocab=1.4509 token=1.4698 | 6.12 steps/s | VRAM=29.48 Go [2026-07-17 18:44:25] step 421900/1000000 | loss=2.3223 vocab=0.7419 token=1.1498 | 6.15 steps/s | VRAM=29.24 Go [2026-07-17 18:44:33] step 421950/1000000 | loss=2.3762 vocab=2.1786 token=1.5594 | 6.12 steps/s | VRAM=29.85 Go [2026-07-17 18:44:41] step 422000/1000000 | loss=2.2506 vocab=0.7976 token=1.3802 | 6.14 steps/s | VRAM=26.61 Go [2026-07-17 18:44:49] step 422050/1000000 | loss=2.2325 vocab=0.6694 token=1.3342 | 6.15 steps/s | VRAM=30.90 Go [2026-07-17 18:44:58] step 422100/1000000 | loss=2.1893 vocab=0.8231 token=1.2587 | 6.14 steps/s | VRAM=32.54 Go [2026-07-17 18:45:06] step 422150/1000000 | loss=2.3232 vocab=0.6846 token=1.3352 | 6.08 steps/s | VRAM=26.88 Go [2026-07-17 18:45:14] step 422200/1000000 | loss=2.3349 vocab=1.2776 token=1.4581 | 6.13 steps/s | VRAM=27.92 Go [2026-07-17 18:45:22] step 422250/1000000 | loss=2.2081 vocab=0.7043 token=1.3907 | 6.15 steps/s | VRAM=32.20 Go [2026-07-17 18:45:30] step 422300/1000000 | loss=2.3099 vocab=0.8890 token=1.5476 | 6.13 steps/s | VRAM=32.90 Go [2026-07-17 18:45:38] step 422350/1000000 | loss=2.3078 vocab=0.8140 token=1.8748 | 6.11 steps/s | VRAM=28.62 Go [2026-07-17 18:45:47] step 422400/1000000 | loss=2.2503 vocab=1.4550 token=1.6404 | 6.15 steps/s | VRAM=28.60 Go [2026-07-17 18:45:55] step 422450/1000000 | loss=2.2789 vocab=0.8168 token=1.2246 | 6.15 steps/s | VRAM=26.91 Go [2026-07-17 18:46:03] step 422500/1000000 | loss=2.3355 vocab=1.3947 token=1.4476 | 6.15 steps/s | VRAM=28.23 Go [2026-07-17 18:46:11] step 422550/1000000 | loss=2.1373 vocab=1.4178 token=1.7053 | 6.14 steps/s | VRAM=28.32 Go [2026-07-17 18:46:19] step 422600/1000000 | loss=2.2432 vocab=0.6116 token=1.1949 | 6.14 steps/s | VRAM=31.44 Go [2026-07-17 18:46:27] step 422650/1000000 | loss=2.2442 vocab=0.8885 token=1.1669 | 6.09 steps/s | VRAM=28.94 Go [2026-07-17 18:46:35] step 422700/1000000 | loss=2.2706 vocab=0.9273 token=1.3669 | 6.13 steps/s | VRAM=29.12 Go [2026-07-17 18:46:44] step 422750/1000000 | loss=2.2975 vocab=0.9598 token=1.4371 | 6.16 steps/s | VRAM=28.02 Go [2026-07-17 18:46:52] step 422800/1000000 | loss=2.1785 vocab=0.7620 token=1.3077 | 6.11 steps/s | VRAM=32.32 Go [2026-07-17 18:47:00] step 422850/1000000 | loss=2.3789 vocab=1.1018 token=1.1303 | 6.14 steps/s | VRAM=25.31 Go [2026-07-17 18:47:08] step 422900/1000000 | loss=2.2333 vocab=0.9170 token=1.4743 | 6.11 steps/s | VRAM=29.14 Go [2026-07-17 18:47:16] step 422950/1000000 | loss=2.2904 vocab=0.8001 token=1.1568 | 6.16 steps/s | VRAM=31.48 Go [2026-07-17 18:47:24] step 423000/1000000 | loss=2.1928 vocab=0.9431 token=1.3805 | 6.11 steps/s | VRAM=25.73 Go [2026-07-17 18:47:33] step 423050/1000000 | loss=2.4124 vocab=0.8913 token=1.3082 | 6.14 steps/s | VRAM=28.72 Go [2026-07-17 18:47:41] step 423100/1000000 | loss=2.3319 vocab=0.7857 token=1.2305 | 6.13 steps/s | VRAM=26.05 Go [2026-07-17 18:47:49] step 423150/1000000 | loss=2.2520 vocab=0.8580 token=1.4025 | 6.16 steps/s | VRAM=26.43 Go [2026-07-17 18:47:57] step 423200/1000000 | loss=2.2487 vocab=0.6742 token=1.4783 | 6.14 steps/s | VRAM=28.54 Go [2026-07-17 18:48:05] step 423250/1000000 | loss=2.1659 vocab=0.8272 token=1.6102 | 6.13 steps/s | VRAM=29.03 Go [2026-07-17 18:48:13] step 423300/1000000 | loss=2.1952 vocab=0.9843 token=1.7765 | 6.12 steps/s | VRAM=26.12 Go [2026-07-17 18:48:21] step 423350/1000000 | loss=2.3694 vocab=1.3224 token=1.6860 | 6.12 steps/s | VRAM=29.25 Go [2026-07-17 18:48:30] step 423400/1000000 | loss=2.3287 vocab=0.8807 token=1.3483 | 6.13 steps/s | VRAM=26.28 Go [2026-07-17 18:48:38] step 423450/1000000 | loss=2.3000 vocab=1.4090 token=1.4737 | 6.12 steps/s | VRAM=28.95 Go [2026-07-17 18:48:46] step 423500/1000000 | loss=2.2070 vocab=0.6161 token=0.8964 | 6.12 steps/s | VRAM=28.29 Go [2026-07-17 18:48:54] step 423550/1000000 | loss=2.3218 vocab=0.6278 token=1.1175 | 6.13 steps/s | VRAM=26.11 Go [2026-07-17 18:49:02] step 423600/1000000 | loss=2.1773 vocab=1.3316 token=1.3937 | 6.11 steps/s | VRAM=29.27 Go [2026-07-17 18:49:10] step 423650/1000000 | loss=2.2701 vocab=0.8576 token=1.2768 | 6.10 steps/s | VRAM=29.31 Go [2026-07-17 18:49:19] step 423700/1000000 | loss=2.2635 vocab=0.9628 token=1.7386 | 6.16 steps/s | VRAM=28.97 Go [2026-07-17 18:49:27] step 423750/1000000 | loss=2.1829 vocab=0.6823 token=1.2572 | 6.13 steps/s | VRAM=26.06 Go [2026-07-17 18:49:35] step 423800/1000000 | loss=2.1744 vocab=0.8282 token=1.0869 | 6.17 steps/s | VRAM=31.65 Go [2026-07-17 18:49:43] step 423850/1000000 | loss=2.3575 vocab=1.0259 token=2.2060 | 6.15 steps/s | VRAM=29.02 Go [2026-07-17 18:49:51] step 423900/1000000 | loss=2.2677 vocab=0.6544 token=1.3871 | 6.15 steps/s | VRAM=24.90 Go [2026-07-17 18:49:59] step 423950/1000000 | loss=2.2284 vocab=1.1404 token=1.4826 | 6.17 steps/s | VRAM=28.08 Go [2026-07-17 18:50:07] step 424000/1000000 | loss=2.2736 vocab=0.7856 token=1.2978 | 6.16 steps/s | VRAM=31.83 Go [2026-07-17 18:50:15] step 424050/1000000 | loss=2.2164 vocab=0.9375 token=1.3829 | 6.14 steps/s | VRAM=31.30 Go [2026-07-17 18:50:24] step 424100/1000000 | loss=2.2431 vocab=0.7415 token=1.2931 | 6.16 steps/s | VRAM=28.40 Go [2026-07-17 18:50:32] step 424150/1000000 | loss=2.2246 vocab=0.8459 token=1.1210 | 6.15 steps/s | VRAM=25.32 Go [2026-07-17 18:50:40] step 424200/1000000 | loss=2.3023 vocab=0.8258 token=1.1616 | 6.16 steps/s | VRAM=32.67 Go [2026-07-17 18:50:48] step 424250/1000000 | loss=2.1689 vocab=0.7472 token=1.0064 | 6.13 steps/s | VRAM=26.19 Go [2026-07-17 18:50:56] step 424300/1000000 | loss=2.2969 vocab=1.1025 token=1.1964 | 6.16 steps/s | VRAM=29.11 Go [2026-07-17 18:51:04] step 424350/1000000 | loss=2.2012 vocab=1.2187 token=1.3646 | 6.15 steps/s | VRAM=23.48 Go [2026-07-17 18:51:12] step 424400/1000000 | loss=2.2673 vocab=0.9195 token=1.5861 | 6.17 steps/s | VRAM=32.21 Go [2026-07-17 18:51:20] step 424450/1000000 | loss=2.3205 vocab=0.7026 token=1.6584 | 6.14 steps/s | VRAM=32.26 Go [2026-07-17 18:51:29] step 424500/1000000 | loss=2.3068 vocab=1.3608 token=1.5820 | 6.14 steps/s | VRAM=27.43 Go [2026-07-17 18:51:37] step 424550/1000000 | loss=2.3177 vocab=1.5237 token=1.5650 | 6.16 steps/s | VRAM=29.37 Go [2026-07-17 18:51:45] step 424600/1000000 | loss=2.2341 vocab=0.7692 token=0.9246 | 6.14 steps/s | VRAM=31.17 Go [2026-07-17 18:51:53] step 424650/1000000 | loss=2.1855 vocab=1.0054 token=1.7881 | 6.16 steps/s | VRAM=32.28 Go [2026-07-17 18:52:01] step 424700/1000000 | loss=2.2193 vocab=0.7566 token=1.0617 | 6.14 steps/s | VRAM=32.72 Go [2026-07-17 18:52:09] step 424750/1000000 | loss=2.2214 vocab=1.0134 token=1.3152 | 6.17 steps/s | VRAM=31.94 Go [2026-07-17 18:52:17] step 424800/1000000 | loss=2.2535 vocab=0.4986 token=1.2294 | 6.11 steps/s | VRAM=26.06 Go [2026-07-17 18:52:26] step 424850/1000000 | loss=2.3202 vocab=0.7102 token=1.2377 | 6.16 steps/s | VRAM=32.10 Go [2026-07-17 18:52:34] step 424900/1000000 | loss=2.2349 vocab=0.9677 token=1.2242 | 6.15 steps/s | VRAM=24.73 Go [2026-07-17 18:52:42] step 424950/1000000 | loss=2.3037 vocab=1.7334 token=1.7798 | 6.11 steps/s | VRAM=29.95 Go [2026-07-17 18:52:50] step 425000/1000000 | loss=2.2426 vocab=1.1548 token=1.2640 | 6.13 steps/s | VRAM=28.72 Go [2026-07-17 18:52:58] step 425050/1000000 | loss=2.2527 vocab=0.6324 token=1.3251 | 6.11 steps/s | VRAM=32.29 Go [2026-07-17 18:53:06] step 425100/1000000 | loss=2.2245 vocab=0.9114 token=1.4227 | 6.11 steps/s | VRAM=26.28 Go [2026-07-17 18:53:15] step 425150/1000000 | loss=2.2328 vocab=0.7205 token=1.6075 | 6.15 steps/s | VRAM=28.40 Go [2026-07-17 18:53:23] step 425200/1000000 | loss=2.2235 vocab=0.9332 token=1.3264 | 6.15 steps/s | VRAM=28.72 Go [2026-07-17 18:53:31] step 425250/1000000 | loss=2.4031 vocab=1.2470 token=1.3354 | 6.15 steps/s | VRAM=31.43 Go [2026-07-17 18:53:39] step 425300/1000000 | loss=2.3721 vocab=1.0173 token=1.4784 | 6.13 steps/s | VRAM=29.19 Go [2026-07-17 18:53:47] step 425350/1000000 | loss=2.2962 vocab=1.3057 token=1.2624 | 6.15 steps/s | VRAM=31.85 Go [2026-07-17 18:53:55] step 425400/1000000 | loss=2.3386 vocab=1.0022 token=1.6683 | 6.14 steps/s | VRAM=29.27 Go [2026-07-17 18:54:03] step 425450/1000000 | loss=2.2651 vocab=0.8713 token=1.2806 | 6.10 steps/s | VRAM=32.12 Go [2026-07-17 18:54:12] step 425500/1000000 | loss=2.2854 vocab=0.7965 token=1.3015 | 6.13 steps/s | VRAM=31.64 Go [2026-07-17 18:54:20] step 425550/1000000 | loss=2.1705 vocab=0.5635 token=1.0400 | 6.12 steps/s | VRAM=25.48 Go [2026-07-17 18:54:28] step 425600/1000000 | loss=2.2559 vocab=0.8798 token=1.4674 | 6.14 steps/s | VRAM=26.11 Go [2026-07-17 18:54:36] step 425650/1000000 | loss=2.3686 vocab=1.2787 token=1.7938 | 6.15 steps/s | VRAM=27.94 Go [2026-07-17 18:54:44] step 425700/1000000 | loss=2.2600 vocab=0.9956 token=1.4257 | 6.14 steps/s | VRAM=29.99 Go [2026-07-17 18:54:52] step 425750/1000000 | loss=2.2466 vocab=1.3732 token=1.5851 | 6.14 steps/s | VRAM=31.19 Go [2026-07-17 18:55:00] step 425800/1000000 | loss=2.1722 vocab=0.8566 token=1.3991 | 6.13 steps/s | VRAM=29.34 Go [2026-07-17 18:55:09] step 425850/1000000 | loss=2.2907 vocab=0.9520 token=1.5813 | 6.15 steps/s | VRAM=28.98 Go [2026-07-17 18:55:17] step 425900/1000000 | loss=2.3034 vocab=1.1551 token=1.1354 | 6.16 steps/s | VRAM=28.31 Go [2026-07-17 18:55:25] step 425950/1000000 | loss=2.2153 vocab=1.4149 token=1.3484 | 6.13 steps/s | VRAM=27.50 Go [2026-07-17 18:55:33] step 426000/1000000 | loss=2.3188 vocab=1.7627 token=1.8681 | 6.13 steps/s | VRAM=28.70 Go [2026-07-17 18:55:41] step 426050/1000000 | loss=2.2820 vocab=0.7822 token=1.4923 | 6.12 steps/s | VRAM=29.03 Go [2026-07-17 18:55:49] step 426100/1000000 | loss=2.3313 vocab=0.8843 token=1.5849 | 6.12 steps/s | VRAM=29.24 Go [2026-07-17 18:55:57] step 426150/1000000 | loss=2.1964 vocab=0.8751 token=1.2379 | 6.14 steps/s | VRAM=28.34 Go [2026-07-17 18:56:06] step 426200/1000000 | loss=2.2922 vocab=0.5994 token=1.1695 | 6.14 steps/s | VRAM=28.07 Go [2026-07-17 18:56:14] step 426250/1000000 | loss=2.2687 vocab=1.3563 token=1.8104 | 6.13 steps/s | VRAM=28.79 Go [2026-07-17 18:56:22] step 426300/1000000 | loss=2.2707 vocab=0.8853 token=1.3853 | 6.10 steps/s | VRAM=30.23 Go [2026-07-17 18:56:30] step 426350/1000000 | loss=2.2695 vocab=0.3836 token=1.1627 | 6.13 steps/s | VRAM=30.37 Go [2026-07-17 18:56:38] step 426400/1000000 | loss=2.2783 vocab=1.6788 token=1.2180 | 6.09 steps/s | VRAM=32.76 Go [2026-07-17 18:56:47] step 426450/1000000 | loss=2.3115 vocab=1.0545 token=1.8321 | 6.12 steps/s | VRAM=32.26 Go [2026-07-17 18:56:55] step 426500/1000000 | loss=2.2653 vocab=1.2230 token=1.5145 | 6.12 steps/s | VRAM=25.82 Go [2026-07-17 18:57:03] step 426550/1000000 | loss=2.3898 vocab=0.8421 token=1.1349 | 6.12 steps/s | VRAM=32.51 Go [2026-07-17 18:57:11] step 426600/1000000 | loss=2.3243 vocab=0.6905 token=1.0307 | 6.16 steps/s | VRAM=26.08 Go [2026-07-17 18:57:19] step 426650/1000000 | loss=2.3080 vocab=1.0829 token=1.1938 | 6.13 steps/s | VRAM=29.32 Go [2026-07-17 18:57:27] step 426700/1000000 | loss=2.3286 vocab=1.0313 token=1.5598 | 6.13 steps/s | VRAM=29.38 Go [2026-07-17 18:57:35] step 426750/1000000 | loss=2.2838 vocab=1.5052 token=1.4095 | 6.12 steps/s | VRAM=28.99 Go [2026-07-17 18:57:44] step 426800/1000000 | loss=2.4068 vocab=1.0089 token=1.5561 | 6.15 steps/s | VRAM=27.68 Go [2026-07-17 18:57:52] step 426850/1000000 | loss=2.1596 vocab=0.9506 token=1.9073 | 6.13 steps/s | VRAM=26.17 Go [2026-07-17 18:58:00] step 426900/1000000 | loss=2.3139 vocab=0.7036 token=1.2764 | 6.15 steps/s | VRAM=27.81 Go [2026-07-17 18:58:08] step 426950/1000000 | loss=2.2855 vocab=1.2048 token=1.5630 | 6.13 steps/s | VRAM=26.23 Go [2026-07-17 18:58:16] step 427000/1000000 | loss=2.3566 vocab=0.6892 token=1.4325 | 6.14 steps/s | VRAM=28.17 Go [2026-07-17 18:58:24] step 427050/1000000 | loss=2.2394 vocab=1.0196 token=1.3609 | 6.13 steps/s | VRAM=26.37 Go [2026-07-17 18:58:32] step 427100/1000000 | loss=2.2402 vocab=0.8311 token=1.4460 | 6.13 steps/s | VRAM=29.70 Go [2026-07-17 18:58:41] step 427150/1000000 | loss=2.3047 vocab=0.9363 token=1.7135 | 6.13 steps/s | VRAM=27.88 Go [2026-07-17 18:58:49] step 427200/1000000 | loss=2.2218 vocab=1.2453 token=1.4698 | 6.09 steps/s | VRAM=26.14 Go [2026-07-17 18:58:57] step 427250/1000000 | loss=2.2938 vocab=0.7048 token=0.9995 | 6.13 steps/s | VRAM=28.71 Go [2026-07-17 18:59:05] step 427300/1000000 | loss=2.3624 vocab=0.8371 token=1.3861 | 6.11 steps/s | VRAM=29.58 Go [2026-07-17 18:59:13] step 427350/1000000 | loss=2.3075 vocab=0.7199 token=1.2735 | 6.13 steps/s | VRAM=29.05 Go [2026-07-17 18:59:21] step 427400/1000000 | loss=2.3584 vocab=1.0460 token=1.4434 | 6.15 steps/s | VRAM=28.20 Go [2026-07-17 18:59:30] step 427450/1000000 | loss=2.3221 vocab=1.0727 token=1.2176 | 6.15 steps/s | VRAM=28.55 Go [2026-07-17 18:59:38] step 427500/1000000 | loss=2.2963 vocab=0.4509 token=1.0434 | 6.12 steps/s | VRAM=28.74 Go [2026-07-17 18:59:46] step 427550/1000000 | loss=2.2748 vocab=0.7117 token=1.1932 | 6.15 steps/s | VRAM=32.32 Go [2026-07-17 18:59:54] step 427600/1000000 | loss=2.2964 vocab=0.8740 token=1.6359 | 6.09 steps/s | VRAM=30.17 Go [2026-07-17 19:00:02] step 427650/1000000 | loss=2.2290 vocab=0.7544 token=1.1355 | 6.13 steps/s | VRAM=26.75 Go [2026-07-17 19:00:10] step 427700/1000000 | loss=2.2979 vocab=0.8402 token=1.4180 | 6.16 steps/s | VRAM=25.70 Go [2026-07-17 19:00:19] step 427750/1000000 | loss=2.1882 vocab=1.0509 token=1.7941 | 6.15 steps/s | VRAM=27.57 Go [2026-07-17 19:00:27] step 427800/1000000 | loss=2.2899 vocab=1.3276 token=0.9629 | 6.10 steps/s | VRAM=28.55 Go [2026-07-17 19:00:35] step 427850/1000000 | loss=2.2924 vocab=1.3733 token=1.5544 | 6.10 steps/s | VRAM=26.22 Go [2026-07-17 19:00:43] step 427900/1000000 | loss=2.2330 vocab=1.0350 token=1.1439 | 6.14 steps/s | VRAM=24.73 Go [2026-07-17 19:00:51] step 427950/1000000 | loss=2.2258 vocab=0.5968 token=1.1608 | 6.13 steps/s | VRAM=28.71 Go [2026-07-17 19:00:59] step 428000/1000000 | loss=2.3401 vocab=1.1469 token=1.5368 | 6.11 steps/s | VRAM=28.49 Go [2026-07-17 19:01:08] step 428050/1000000 | loss=2.3391 vocab=0.4340 token=1.1589 | 6.12 steps/s | VRAM=31.67 Go [2026-07-17 19:01:16] step 428100/1000000 | loss=2.3740 vocab=0.4743 token=1.2163 | 6.13 steps/s | VRAM=26.43 Go [2026-07-17 19:01:24] step 428150/1000000 | loss=2.2614 vocab=1.0589 token=1.4095 | 6.12 steps/s | VRAM=28.75 Go [2026-07-17 19:01:32] step 428200/1000000 | loss=2.1535 vocab=1.2011 token=1.2269 | 6.14 steps/s | VRAM=32.57 Go [2026-07-17 19:01:40] step 428250/1000000 | loss=2.2424 vocab=1.0248 token=1.6244 | 6.11 steps/s | VRAM=26.03 Go [2026-07-17 19:01:48] step 428300/1000000 | loss=2.1964 vocab=1.0491 token=1.2142 | 6.12 steps/s | VRAM=32.67 Go [2026-07-17 19:01:57] step 428350/1000000 | loss=2.2937 vocab=1.2191 token=1.7847 | 6.13 steps/s | VRAM=28.43 Go [2026-07-17 19:02:05] step 428400/1000000 | loss=2.2592 vocab=1.0224 token=1.1040 | 6.13 steps/s | VRAM=28.51 Go [2026-07-17 19:02:13] step 428450/1000000 | loss=2.2433 vocab=0.9560 token=1.2582 | 6.14 steps/s | VRAM=24.97 Go [2026-07-17 19:02:21] step 428500/1000000 | loss=2.2094 vocab=0.3719 token=1.1390 | 6.16 steps/s | VRAM=32.53 Go [2026-07-17 19:02:29] step 428550/1000000 | loss=2.1782 vocab=0.4506 token=1.1931 | 6.14 steps/s | VRAM=28.17 Go [2026-07-17 19:02:37] step 428600/1000000 | loss=2.3420 vocab=1.2176 token=1.1123 | 6.15 steps/s | VRAM=25.59 Go [2026-07-17 19:02:45] step 428650/1000000 | loss=2.3398 vocab=1.0510 token=1.5058 | 6.13 steps/s | VRAM=28.58 Go [2026-07-17 19:02:54] step 428700/1000000 | loss=2.3347 vocab=0.7279 token=1.1624 | 6.16 steps/s | VRAM=26.92 Go [2026-07-17 19:03:02] step 428750/1000000 | loss=2.2252 vocab=1.1058 token=1.1140 | 6.15 steps/s | VRAM=24.12 Go [2026-07-17 19:03:10] step 428800/1000000 | loss=2.2505 vocab=0.7710 token=1.0953 | 6.13 steps/s | VRAM=31.67 Go [2026-07-17 19:03:18] step 428850/1000000 | loss=2.3578 vocab=1.0049 token=1.2810 | 6.13 steps/s | VRAM=25.74 Go [2026-07-17 19:03:26] step 428900/1000000 | loss=2.2177 vocab=1.4590 token=0.8499 | 6.14 steps/s | VRAM=31.82 Go [2026-07-17 19:03:34] step 428950/1000000 | loss=2.1846 vocab=0.8530 token=1.2224 | 6.16 steps/s | VRAM=28.12 Go [2026-07-17 19:03:42] step 429000/1000000 | loss=2.3017 vocab=1.5901 token=1.4637 | 6.12 steps/s | VRAM=32.79 Go [2026-07-17 19:03:51] step 429050/1000000 | loss=2.2865 vocab=0.8975 token=1.1801 | 6.14 steps/s | VRAM=28.07 Go [2026-07-17 19:03:59] step 429100/1000000 | loss=2.1493 vocab=1.1362 token=1.5153 | 6.14 steps/s | VRAM=28.84 Go [2026-07-17 19:04:07] step 429150/1000000 | loss=2.2540 vocab=0.5417 token=1.4611 | 6.17 steps/s | VRAM=27.72 Go [2026-07-17 19:04:15] step 429200/1000000 | loss=2.3088 vocab=0.8280 token=1.6520 | 6.17 steps/s | VRAM=32.21 Go [2026-07-17 19:04:23] step 429250/1000000 | loss=2.1852 vocab=0.7646 token=1.3052 | 6.16 steps/s | VRAM=25.22 Go [2026-07-17 19:04:31] step 429300/1000000 | loss=2.2066 vocab=0.8569 token=1.2200 | 6.14 steps/s | VRAM=29.21 Go [2026-07-17 19:04:39] step 429350/1000000 | loss=2.3358 vocab=0.9821 token=1.3769 | 6.13 steps/s | VRAM=32.59 Go [2026-07-17 19:04:47] step 429400/1000000 | loss=2.2147 vocab=0.9172 token=1.3702 | 6.12 steps/s | VRAM=24.53 Go [2026-07-17 19:04:56] step 429450/1000000 | loss=2.2862 vocab=0.8530 token=1.1987 | 6.15 steps/s | VRAM=25.20 Go [2026-07-17 19:05:04] step 429500/1000000 | loss=2.2266 vocab=0.8098 token=1.3671 | 6.14 steps/s | VRAM=28.01 Go [2026-07-17 19:05:12] step 429550/1000000 | loss=2.3176 vocab=1.2759 token=1.1390 | 6.14 steps/s | VRAM=32.23 Go [2026-07-17 19:05:20] step 429600/1000000 | loss=2.2496 vocab=0.6777 token=0.8843 | 6.13 steps/s | VRAM=28.60 Go [2026-07-17 19:05:28] step 429650/1000000 | loss=2.1140 vocab=1.0599 token=1.3056 | 6.13 steps/s | VRAM=29.05 Go [2026-07-17 19:05:36] step 429700/1000000 | loss=2.2125 vocab=0.8941 token=1.4162 | 6.12 steps/s | VRAM=27.75 Go [2026-07-17 19:05:45] step 429750/1000000 | loss=2.3363 vocab=1.3065 token=1.9811 | 6.14 steps/s | VRAM=26.37 Go [2026-07-17 19:05:53] step 429800/1000000 | loss=2.3098 vocab=0.7708 token=1.6095 | 6.14 steps/s | VRAM=28.99 Go [2026-07-17 19:06:01] step 429850/1000000 | loss=2.2436 vocab=1.1179 token=1.5245 | 6.13 steps/s | VRAM=27.98 Go [2026-07-17 19:06:09] step 429900/1000000 | loss=2.3289 vocab=0.9842 token=1.0799 | 6.14 steps/s | VRAM=32.64 Go [2026-07-17 19:06:17] step 429950/1000000 | loss=2.2799 vocab=1.3464 token=1.7419 | 6.15 steps/s | VRAM=29.44 Go [2026-07-17 19:06:25] step 430000/1000000 | loss=2.3121 vocab=0.8844 token=1.4624 | 6.13 steps/s | VRAM=25.81 Go [2026-07-17 19:06:33] step 430050/1000000 | loss=2.2002 vocab=1.1943 token=1.6116 | 6.13 steps/s | VRAM=27.24 Go [2026-07-17 19:06:42] step 430100/1000000 | loss=2.3495 vocab=1.1346 token=1.6781 | 6.16 steps/s | VRAM=31.54 Go [2026-07-17 19:06:50] step 430150/1000000 | loss=2.2612 vocab=0.7609 token=1.4263 | 6.12 steps/s | VRAM=28.55 Go [2026-07-17 19:06:58] step 430200/1000000 | loss=2.3084 vocab=1.0729 token=1.6036 | 6.15 steps/s | VRAM=26.06 Go [2026-07-17 19:07:06] step 430250/1000000 | loss=2.2913 vocab=1.1380 token=1.7017 | 6.14 steps/s | VRAM=29.33 Go [2026-07-17 19:07:14] step 430300/1000000 | loss=2.3317 vocab=0.7160 token=1.4768 | 6.12 steps/s | VRAM=25.81 Go [2026-07-17 19:07:22] step 430350/1000000 | loss=2.1736 vocab=0.8324 token=1.3310 | 6.16 steps/s | VRAM=28.15 Go [2026-07-17 19:07:30] step 430400/1000000 | loss=2.1832 vocab=0.6749 token=1.2056 | 6.12 steps/s | VRAM=28.36 Go [2026-07-17 19:07:39] step 430450/1000000 | loss=2.1500 vocab=0.7857 token=1.0659 | 6.10 steps/s | VRAM=32.76 Go [2026-07-17 19:07:47] step 430500/1000000 | loss=2.3832 vocab=1.0483 token=1.5239 | 6.10 steps/s | VRAM=26.50 Go [2026-07-17 19:07:55] step 430550/1000000 | loss=2.2414 vocab=0.7749 token=1.5861 | 6.16 steps/s | VRAM=27.96 Go [2026-07-17 19:08:03] step 430600/1000000 | loss=2.4005 vocab=0.9764 token=1.2119 | 6.13 steps/s | VRAM=27.08 Go [2026-07-17 19:08:11] step 430650/1000000 | loss=2.2928 vocab=0.9806 token=1.3293 | 6.10 steps/s | VRAM=29.36 Go [2026-07-17 19:08:19] step 430700/1000000 | loss=2.3744 vocab=0.9128 token=1.7922 | 6.16 steps/s | VRAM=24.60 Go [2026-07-17 19:08:28] step 430750/1000000 | loss=2.3476 vocab=0.7530 token=1.2913 | 6.14 steps/s | VRAM=25.82 Go [2026-07-17 19:08:36] step 430800/1000000 | loss=2.2845 vocab=1.0270 token=1.3418 | 6.13 steps/s | VRAM=29.46 Go [2026-07-17 19:08:44] step 430850/1000000 | loss=2.3873 vocab=0.9937 token=1.1543 | 6.12 steps/s | VRAM=28.73 Go [2026-07-17 19:08:52] step 430900/1000000 | loss=2.3308 vocab=1.0814 token=1.2988 | 6.12 steps/s | VRAM=28.89 Go [2026-07-17 19:09:00] step 430950/1000000 | loss=2.3372 vocab=1.3242 token=1.4053 | 6.14 steps/s | VRAM=31.59 Go [2026-07-17 19:09:08] step 431000/1000000 | loss=2.2727 vocab=1.0454 token=1.5212 | 6.11 steps/s | VRAM=31.79 Go [2026-07-17 19:09:17] step 431050/1000000 | loss=2.3233 vocab=0.6925 token=1.0158 | 6.13 steps/s | VRAM=32.10 Go [2026-07-17 19:09:25] step 431100/1000000 | loss=2.1785 vocab=1.4347 token=1.5217 | 6.08 steps/s | VRAM=28.63 Go [2026-07-17 19:09:33] step 431150/1000000 | loss=2.2261 vocab=0.6551 token=1.2588 | 6.13 steps/s | VRAM=29.17 Go [2026-07-17 19:09:41] step 431200/1000000 | loss=2.3023 vocab=0.7259 token=1.1377 | 6.11 steps/s | VRAM=25.65 Go [2026-07-17 19:09:49] step 431250/1000000 | loss=2.2655 vocab=1.1273 token=1.5100 | 6.09 steps/s | VRAM=26.52 Go [2026-07-17 19:09:57] step 431300/1000000 | loss=2.2709 vocab=2.3184 token=1.5570 | 6.14 steps/s | VRAM=32.23 Go [2026-07-17 19:10:06] step 431350/1000000 | loss=2.3081 vocab=0.6875 token=1.3577 | 6.14 steps/s | VRAM=31.43 Go [2026-07-17 19:10:14] step 431400/1000000 | loss=2.2540 vocab=1.0263 token=1.3105 | 6.12 steps/s | VRAM=30.94 Go [2026-07-17 19:10:22] step 431450/1000000 | loss=2.3394 vocab=0.8810 token=1.3588 | 6.12 steps/s | VRAM=26.88 Go [2026-07-17 19:10:30] step 431500/1000000 | loss=2.3725 vocab=0.8273 token=1.4174 | 6.13 steps/s | VRAM=28.40 Go [2026-07-17 19:10:38] step 431550/1000000 | loss=2.2470 vocab=1.6064 token=1.5002 | 6.12 steps/s | VRAM=29.60 Go [2026-07-17 19:10:46] step 431600/1000000 | loss=2.2878 vocab=0.8556 token=1.2460 | 6.14 steps/s | VRAM=28.38 Go [2026-07-17 19:10:55] step 431650/1000000 | loss=2.3086 vocab=0.6018 token=1.2839 | 6.14 steps/s | VRAM=32.33 Go [2026-07-17 19:11:03] step 431700/1000000 | loss=2.2926 vocab=0.8729 token=1.3146 | 6.13 steps/s | VRAM=32.66 Go [2026-07-17 19:11:11] step 431750/1000000 | loss=2.4045 vocab=0.7011 token=1.1207 | 6.12 steps/s | VRAM=32.39 Go [2026-07-17 19:11:19] step 431800/1000000 | loss=2.3296 vocab=0.4663 token=1.3242 | 6.14 steps/s | VRAM=29.67 Go [2026-07-17 19:11:27] step 431850/1000000 | loss=2.3667 vocab=1.4711 token=1.1376 | 6.15 steps/s | VRAM=28.67 Go [2026-07-17 19:11:35] step 431900/1000000 | loss=2.3477 vocab=0.5779 token=1.1316 | 6.12 steps/s | VRAM=31.53 Go [2026-07-17 19:11:43] step 431950/1000000 | loss=2.2889 vocab=0.5836 token=1.5501 | 6.12 steps/s | VRAM=28.63 Go [2026-07-17 19:11:52] step 432000/1000000 | loss=2.2036 vocab=0.7713 token=1.3853 | 6.11 steps/s | VRAM=27.51 Go [2026-07-17 19:12:00] step 432050/1000000 | loss=2.2180 vocab=0.8050 token=1.0371 | 6.09 steps/s | VRAM=29.00 Go [2026-07-17 19:12:08] step 432100/1000000 | loss=2.3193 vocab=0.6153 token=1.2865 | 6.10 steps/s | VRAM=28.81 Go [2026-07-17 19:12:16] step 432150/1000000 | loss=2.3831 vocab=1.3009 token=1.6418 | 6.12 steps/s | VRAM=30.16 Go [2026-07-17 19:12:24] step 432200/1000000 | loss=2.3659 vocab=1.4192 token=1.7994 | 6.11 steps/s | VRAM=25.58 Go [2026-07-17 19:12:33] step 432250/1000000 | loss=2.2901 vocab=0.8890 token=1.2524 | 6.15 steps/s | VRAM=28.18 Go [2026-07-17 19:12:41] step 432300/1000000 | loss=2.3019 vocab=0.7759 token=1.0727 | 6.13 steps/s | VRAM=28.87 Go [2026-07-17 19:12:49] step 432350/1000000 | loss=2.1603 vocab=0.9731 token=1.5135 | 6.15 steps/s | VRAM=28.83 Go [2026-07-17 19:12:57] step 432400/1000000 | loss=2.3397 vocab=1.1391 token=1.4210 | 6.09 steps/s | VRAM=30.08 Go [2026-07-17 19:13:05] step 432450/1000000 | loss=2.3084 vocab=1.1941 token=1.6858 | 6.15 steps/s | VRAM=32.49 Go [2026-07-17 19:13:13] step 432500/1000000 | loss=2.2408 vocab=1.0665 token=1.9307 | 6.13 steps/s | VRAM=27.81 Go [2026-07-17 19:13:21] step 432550/1000000 | loss=2.3279 vocab=0.9372 token=1.2522 | 6.16 steps/s | VRAM=31.41 Go [2026-07-17 19:13:30] step 432600/1000000 | loss=2.2251 vocab=1.0692 token=1.3780 | 6.14 steps/s | VRAM=28.65 Go [2026-07-17 19:13:38] step 432650/1000000 | loss=2.2934 vocab=0.9683 token=1.5077 | 6.15 steps/s | VRAM=29.34 Go [2026-07-17 19:13:46] step 432700/1000000 | loss=2.3177 vocab=0.9361 token=1.2969 | 6.15 steps/s | VRAM=26.83 Go [2026-07-17 19:13:54] step 432750/1000000 | loss=2.2646 vocab=0.8376 token=1.2826 | 6.12 steps/s | VRAM=28.76 Go [2026-07-17 19:14:02] step 432800/1000000 | loss=2.3815 vocab=1.2554 token=1.7349 | 6.13 steps/s | VRAM=29.56 Go [2026-07-17 19:14:10] step 432850/1000000 | loss=2.2006 vocab=0.9711 token=1.4138 | 6.12 steps/s | VRAM=29.77 Go [2026-07-17 19:14:19] step 432900/1000000 | loss=2.2398 vocab=0.6458 token=1.1813 | 6.12 steps/s | VRAM=25.35 Go [2026-07-17 19:14:27] step 432950/1000000 | loss=2.2918 vocab=1.2941 token=1.6072 | 6.17 steps/s | VRAM=31.75 Go [2026-07-17 19:14:35] step 433000/1000000 | loss=2.2541 vocab=0.7378 token=1.3080 | 6.15 steps/s | VRAM=29.15 Go [2026-07-17 19:14:43] step 433050/1000000 | loss=2.2389 vocab=0.9143 token=1.1102 | 6.14 steps/s | VRAM=28.85 Go [2026-07-17 19:14:51] step 433100/1000000 | loss=2.1942 vocab=0.7825 token=1.4269 | 6.16 steps/s | VRAM=28.58 Go [2026-07-17 19:14:59] step 433150/1000000 | loss=2.3443 vocab=0.7340 token=0.9534 | 6.12 steps/s | VRAM=26.19 Go [2026-07-17 19:15:07] step 433200/1000000 | loss=2.2295 vocab=0.4857 token=1.1006 | 6.15 steps/s | VRAM=28.44 Go [2026-07-17 19:15:16] step 433250/1000000 | loss=2.3968 vocab=1.2640 token=1.0428 | 6.13 steps/s | VRAM=32.55 Go [2026-07-17 19:15:24] step 433300/1000000 | loss=2.1216 vocab=0.5221 token=1.3260 | 6.13 steps/s | VRAM=28.86 Go [2026-07-17 19:15:32] step 433350/1000000 | loss=2.2803 vocab=0.9731 token=1.3973 | 6.14 steps/s | VRAM=28.38 Go [2026-07-17 19:15:40] step 433400/1000000 | loss=2.1764 vocab=0.5031 token=1.0705 | 6.10 steps/s | VRAM=26.19 Go [2026-07-17 19:15:48] step 433450/1000000 | loss=2.1670 vocab=0.9893 token=1.4134 | 6.14 steps/s | VRAM=28.59 Go [2026-07-17 19:15:56] step 433500/1000000 | loss=2.2151 vocab=1.2091 token=1.5353 | 6.12 steps/s | VRAM=29.48 Go [2026-07-17 19:16:05] step 433550/1000000 | loss=2.2123 vocab=0.8640 token=1.7489 | 6.13 steps/s | VRAM=27.72 Go [2026-07-17 19:16:13] step 433600/1000000 | loss=2.2305 vocab=0.9259 token=1.3986 | 6.13 steps/s | VRAM=29.85 Go [2026-07-17 19:16:21] step 433650/1000000 | loss=2.3110 vocab=0.9250 token=1.1799 | 6.16 steps/s | VRAM=31.08 Go [2026-07-17 19:16:29] step 433700/1000000 | loss=2.2887 vocab=1.2377 token=1.5189 | 6.14 steps/s | VRAM=28.06 Go [2026-07-17 19:16:37] step 433750/1000000 | loss=2.1457 vocab=0.6533 token=1.2565 | 6.14 steps/s | VRAM=28.02 Go [2026-07-17 19:16:45] step 433800/1000000 | loss=2.4104 vocab=0.7104 token=1.1421 | 6.13 steps/s | VRAM=26.34 Go [2026-07-17 19:16:53] step 433850/1000000 | loss=2.2252 vocab=1.2149 token=1.4787 | 6.15 steps/s | VRAM=25.87 Go [2026-07-17 19:17:01] step 433900/1000000 | loss=2.2401 vocab=0.8835 token=1.2539 | 6.17 steps/s | VRAM=32.38 Go [2026-07-17 19:17:10] step 433950/1000000 | loss=2.3620 vocab=1.3274 token=1.4702 | 6.13 steps/s | VRAM=25.95 Go [2026-07-17 19:17:18] step 434000/1000000 | loss=2.1572 vocab=0.6827 token=1.2425 | 6.13 steps/s | VRAM=28.96 Go [2026-07-17 19:17:26] step 434050/1000000 | loss=2.1530 vocab=0.8594 token=1.3302 | 6.11 steps/s | VRAM=26.12 Go [2026-07-17 19:17:34] step 434100/1000000 | loss=2.1715 vocab=0.6681 token=1.0048 | 6.15 steps/s | VRAM=31.93 Go [2026-07-17 19:17:42] step 434150/1000000 | loss=2.2254 vocab=0.8047 token=1.2431 | 6.14 steps/s | VRAM=25.60 Go [2026-07-17 19:17:50] step 434200/1000000 | loss=2.2363 vocab=0.8986 token=1.2458 | 6.14 steps/s | VRAM=32.08 Go [2026-07-17 19:17:59] step 434250/1000000 | loss=2.2218 vocab=0.9063 token=1.4470 | 6.13 steps/s | VRAM=32.32 Go [2026-07-17 19:18:07] step 434300/1000000 | loss=2.2727 vocab=0.9020 token=1.2078 | 6.13 steps/s | VRAM=29.30 Go [2026-07-17 19:18:15] step 434350/1000000 | loss=2.3108 vocab=1.0940 token=1.3941 | 6.15 steps/s | VRAM=32.01 Go [2026-07-17 19:18:23] step 434400/1000000 | loss=2.3293 vocab=0.8247 token=1.2800 | 6.12 steps/s | VRAM=31.12 Go [2026-07-17 19:18:31] step 434450/1000000 | loss=2.2225 vocab=0.6304 token=1.4719 | 6.13 steps/s | VRAM=26.20 Go [2026-07-17 19:18:39] step 434500/1000000 | loss=2.2323 vocab=0.8983 token=1.3422 | 6.13 steps/s | VRAM=32.42 Go [2026-07-17 19:18:47] step 434550/1000000 | loss=2.2327 vocab=1.1372 token=1.5155 | 6.15 steps/s | VRAM=28.20 Go [2026-07-17 19:18:56] step 434600/1000000 | loss=2.2344 vocab=1.1982 token=1.6710 | 6.16 steps/s | VRAM=27.49 Go [2026-07-17 19:19:04] step 434650/1000000 | loss=2.3241 vocab=0.9620 token=1.4480 | 6.12 steps/s | VRAM=26.71 Go [2026-07-17 19:19:12] step 434700/1000000 | loss=2.2071 vocab=1.0200 token=1.2129 | 6.15 steps/s | VRAM=31.57 Go [2026-07-17 19:19:20] step 434750/1000000 | loss=2.3338 vocab=1.0331 token=1.2549 | 6.16 steps/s | VRAM=28.55 Go [2026-07-17 19:19:28] step 434800/1000000 | loss=2.2768 vocab=0.8875 token=1.2069 | 6.12 steps/s | VRAM=30.05 Go [2026-07-17 19:19:36] step 434850/1000000 | loss=2.3196 vocab=0.6614 token=1.0919 | 6.14 steps/s | VRAM=27.82 Go [2026-07-17 19:19:44] step 434900/1000000 | loss=2.2200 vocab=1.3518 token=1.4073 | 6.14 steps/s | VRAM=29.48 Go [2026-07-17 19:19:53] step 434950/1000000 | loss=2.2657 vocab=0.7631 token=1.2452 | 6.19 steps/s | VRAM=31.15 Go [2026-07-17 19:20:01] step 435000/1000000 | loss=2.2799 vocab=0.6855 token=1.4147 | 6.16 steps/s | VRAM=28.23 Go [2026-07-17 19:20:09] step 435050/1000000 | loss=2.2677 vocab=0.8008 token=1.1734 | 6.17 steps/s | VRAM=25.80 Go [2026-07-17 19:20:17] step 435100/1000000 | loss=2.1939 vocab=1.1687 token=1.6350 | 6.15 steps/s | VRAM=31.53 Go [2026-07-17 19:20:25] step 435150/1000000 | loss=2.2335 vocab=1.4459 token=1.2471 | 6.11 steps/s | VRAM=27.67 Go [2026-07-17 19:20:33] step 435200/1000000 | loss=2.3035 vocab=0.8984 token=1.5140 | 6.18 steps/s | VRAM=28.80 Go [2026-07-17 19:20:41] step 435250/1000000 | loss=2.2972 vocab=1.1482 token=1.1514 | 6.12 steps/s | VRAM=25.97 Go [2026-07-17 19:20:49] step 435300/1000000 | loss=2.2557 vocab=0.5024 token=1.1020 | 6.13 steps/s | VRAM=29.59 Go [2026-07-17 19:20:58] step 435350/1000000 | loss=2.2589 vocab=0.9540 token=1.3083 | 6.14 steps/s | VRAM=31.30 Go [2026-07-17 19:21:06] step 435400/1000000 | loss=2.2802 vocab=1.1999 token=1.1791 | 6.15 steps/s | VRAM=26.03 Go [2026-07-17 19:21:14] step 435450/1000000 | loss=2.4055 vocab=1.0635 token=1.5749 | 6.16 steps/s | VRAM=31.90 Go [2026-07-17 19:21:22] step 435500/1000000 | loss=2.2940 vocab=0.8846 token=1.3780 | 6.14 steps/s | VRAM=29.43 Go [2026-07-17 19:21:30] step 435550/1000000 | loss=2.2273 vocab=0.6927 token=1.0046 | 6.17 steps/s | VRAM=24.69 Go [2026-07-17 19:21:38] step 435600/1000000 | loss=2.2757 vocab=0.7142 token=0.9894 | 6.15 steps/s | VRAM=28.19 Go [2026-07-17 19:21:46] step 435650/1000000 | loss=2.3489 vocab=0.9877 token=1.8652 | 6.15 steps/s | VRAM=25.01 Go [2026-07-17 19:21:54] step 435700/1000000 | loss=2.2169 vocab=0.6248 token=1.0636 | 6.17 steps/s | VRAM=31.79 Go [2026-07-17 19:22:03] step 435750/1000000 | loss=2.3650 vocab=1.4075 token=1.5589 | 6.10 steps/s | VRAM=23.78 Go [2026-07-17 19:22:11] step 435800/1000000 | loss=2.2972 vocab=1.1505 token=1.1231 | 6.17 steps/s | VRAM=31.20 Go [2026-07-17 19:22:19] step 435850/1000000 | loss=2.3006 vocab=0.7868 token=1.1303 | 6.15 steps/s | VRAM=29.71 Go [2026-07-17 19:22:27] step 435900/1000000 | loss=2.2703 vocab=1.0540 token=1.2478 | 6.13 steps/s | VRAM=29.03 Go [2026-07-17 19:22:35] step 435950/1000000 | loss=2.3109 vocab=1.0887 token=1.5098 | 6.14 steps/s | VRAM=27.08 Go [2026-07-17 19:22:43] step 436000/1000000 | loss=2.3244 vocab=0.7306 token=1.5111 | 6.17 steps/s | VRAM=30.93 Go [2026-07-17 19:22:51] step 436050/1000000 | loss=2.2031 vocab=0.4914 token=1.1693 | 6.15 steps/s | VRAM=28.31 Go [2026-07-17 19:23:00] step 436100/1000000 | loss=2.2325 vocab=1.0076 token=1.4578 | 6.16 steps/s | VRAM=25.95 Go [2026-07-17 19:23:08] step 436150/1000000 | loss=2.2909 vocab=0.6466 token=1.5062 | 6.20 steps/s | VRAM=30.36 Go [2026-07-17 19:23:16] step 436200/1000000 | loss=2.2263 vocab=1.0003 token=1.4780 | 6.13 steps/s | VRAM=30.93 Go [2026-07-17 19:23:24] step 436250/1000000 | loss=2.2934 vocab=0.7544 token=1.5047 | 6.11 steps/s | VRAM=28.41 Go [2026-07-17 19:23:32] step 436300/1000000 | loss=2.2596 vocab=0.4699 token=1.1796 | 6.12 steps/s | VRAM=27.32 Go [2026-07-17 19:23:40] step 436350/1000000 | loss=2.2139 vocab=0.4713 token=1.2490 | 6.12 steps/s | VRAM=28.20 Go [2026-07-17 19:23:48] step 436400/1000000 | loss=2.2912 vocab=0.9438 token=1.2221 | 6.16 steps/s | VRAM=27.79 Go [2026-07-17 19:23:57] step 436450/1000000 | loss=2.1888 vocab=0.6422 token=1.2189 | 6.14 steps/s | VRAM=28.28 Go [2026-07-17 19:24:05] step 436500/1000000 | loss=2.3182 vocab=1.1541 token=1.6305 | 6.15 steps/s | VRAM=25.65 Go [2026-07-17 19:24:13] step 436550/1000000 | loss=2.1732 vocab=0.7983 token=1.3894 | 6.12 steps/s | VRAM=26.09 Go [2026-07-17 19:24:21] step 436600/1000000 | loss=2.1766 vocab=1.4315 token=1.3875 | 6.16 steps/s | VRAM=32.91 Go [2026-07-17 19:24:29] step 436650/1000000 | loss=2.2448 vocab=1.3786 token=1.4195 | 6.13 steps/s | VRAM=31.60 Go [2026-07-17 19:24:37] step 436700/1000000 | loss=2.3045 vocab=1.1767 token=1.4165 | 6.12 steps/s | VRAM=28.96 Go [2026-07-17 19:24:45] step 436750/1000000 | loss=2.2618 vocab=0.8240 token=1.4594 | 6.13 steps/s | VRAM=29.23 Go [2026-07-17 19:24:54] step 436800/1000000 | loss=2.2033 vocab=0.7600 token=0.9226 | 6.13 steps/s | VRAM=26.09 Go [2026-07-17 19:25:02] step 436850/1000000 | loss=2.2495 vocab=1.2096 token=1.8673 | 6.12 steps/s | VRAM=28.43 Go [2026-07-17 19:25:10] step 436900/1000000 | loss=2.3140 vocab=0.9251 token=0.8576 | 6.10 steps/s | VRAM=30.79 Go [2026-07-17 19:25:18] step 436950/1000000 | loss=2.4071 vocab=0.8889 token=1.2589 | 6.14 steps/s | VRAM=24.72 Go [2026-07-17 19:25:26] step 437000/1000000 | loss=2.4292 vocab=0.9033 token=1.3121 | 6.14 steps/s | VRAM=27.90 Go [2026-07-17 19:25:34] step 437050/1000000 | loss=2.1954 vocab=0.8461 token=1.2838 | 6.16 steps/s | VRAM=27.45 Go [2026-07-17 19:25:43] step 437100/1000000 | loss=2.2349 vocab=0.9049 token=1.6218 | 6.17 steps/s | VRAM=31.68 Go [2026-07-17 19:25:51] step 437150/1000000 | loss=2.2321 vocab=0.4961 token=1.1524 | 6.15 steps/s | VRAM=32.94 Go [2026-07-17 19:25:59] step 437200/1000000 | loss=2.3189 vocab=0.6064 token=1.4844 | 6.11 steps/s | VRAM=24.99 Go [2026-07-17 19:26:07] step 437250/1000000 | loss=2.2999 vocab=0.5792 token=1.2690 | 6.19 steps/s | VRAM=28.84 Go [2026-07-17 19:26:15] step 437300/1000000 | loss=2.3194 vocab=0.9118 token=1.6808 | 6.14 steps/s | VRAM=31.05 Go [2026-07-17 19:26:23] step 437350/1000000 | loss=2.3109 vocab=0.6085 token=1.2174 | 6.12 steps/s | VRAM=28.57 Go [2026-07-17 19:26:31] step 437400/1000000 | loss=2.3616 vocab=0.6442 token=1.0626 | 6.14 steps/s | VRAM=28.11 Go [2026-07-17 19:26:40] step 437450/1000000 | loss=2.2618 vocab=0.7449 token=1.5510 | 6.13 steps/s | VRAM=32.87 Go [2026-07-17 19:26:48] step 437500/1000000 | loss=2.3307 vocab=0.8769 token=1.6371 | 6.18 steps/s | VRAM=31.30 Go [2026-07-17 19:26:56] step 437550/1000000 | loss=2.2426 vocab=0.8436 token=1.6382 | 6.15 steps/s | VRAM=25.48 Go [2026-07-17 19:27:04] step 437600/1000000 | loss=2.3137 vocab=0.8140 token=1.1267 | 6.14 steps/s | VRAM=28.28 Go [2026-07-17 19:27:12] step 437650/1000000 | loss=2.3851 vocab=1.1472 token=1.5436 | 6.15 steps/s | VRAM=32.13 Go [2026-07-17 19:27:20] step 437700/1000000 | loss=2.2261 vocab=0.7056 token=1.1599 | 6.12 steps/s | VRAM=28.45 Go [2026-07-17 19:27:28] step 437750/1000000 | loss=2.2637 vocab=0.7094 token=1.2391 | 6.11 steps/s | VRAM=26.64 Go [2026-07-17 19:27:37] step 437800/1000000 | loss=2.3438 vocab=0.8506 token=1.1316 | 6.15 steps/s | VRAM=29.23 Go [2026-07-17 19:27:45] step 437850/1000000 | loss=2.3740 vocab=1.3596 token=1.8070 | 6.15 steps/s | VRAM=26.63 Go [2026-07-17 19:27:53] step 437900/1000000 | loss=2.2782 vocab=0.6271 token=1.0743 | 6.13 steps/s | VRAM=28.59 Go [2026-07-17 19:28:01] step 437950/1000000 | loss=2.2114 vocab=0.6537 token=1.3434 | 6.15 steps/s | VRAM=25.60 Go [2026-07-17 19:28:09] step 438000/1000000 | loss=2.2866 vocab=1.1895 token=1.4741 | 6.15 steps/s | VRAM=29.11 Go [2026-07-17 19:28:17] step 438050/1000000 | loss=2.3118 vocab=1.0111 token=1.1548 | 6.14 steps/s | VRAM=28.49 Go [2026-07-17 19:28:25] step 438100/1000000 | loss=2.2562 vocab=0.5484 token=0.8790 | 6.13 steps/s | VRAM=27.04 Go [2026-07-17 19:28:33] step 438150/1000000 | loss=2.2439 vocab=0.7646 token=1.4572 | 6.14 steps/s | VRAM=33.00 Go [2026-07-17 19:28:42] step 438200/1000000 | loss=2.2320 vocab=1.0525 token=1.2802 | 6.14 steps/s | VRAM=28.91 Go [2026-07-17 19:28:50] step 438250/1000000 | loss=2.3528 vocab=0.8775 token=0.9392 | 6.13 steps/s | VRAM=28.23 Go [2026-07-17 19:28:58] step 438300/1000000 | loss=2.2954 vocab=1.0271 token=1.1260 | 6.17 steps/s | VRAM=31.93 Go [2026-07-17 19:29:06] step 438350/1000000 | loss=2.1546 vocab=1.3448 token=1.2466 | 6.17 steps/s | VRAM=32.89 Go [2026-07-17 19:29:14] step 438400/1000000 | loss=2.2700 vocab=1.0186 token=1.7994 | 6.15 steps/s | VRAM=29.38 Go [2026-07-17 19:29:22] step 438450/1000000 | loss=2.3761 vocab=1.3954 token=1.6113 | 6.15 steps/s | VRAM=24.96 Go [2026-07-17 19:29:30] step 438500/1000000 | loss=2.2742 vocab=0.8510 token=1.4723 | 6.12 steps/s | VRAM=29.96 Go [2026-07-17 19:29:39] step 438550/1000000 | loss=2.2256 vocab=0.5252 token=1.0184 | 6.14 steps/s | VRAM=28.48 Go [2026-07-17 19:29:47] step 438600/1000000 | loss=2.3844 vocab=0.9501 token=1.3932 | 6.15 steps/s | VRAM=27.91 Go [2026-07-17 19:29:55] step 438650/1000000 | loss=2.3546 vocab=1.6129 token=1.6473 | 6.15 steps/s | VRAM=29.91 Go [2026-07-17 19:30:03] step 438700/1000000 | loss=2.1809 vocab=1.0290 token=0.6700 | 6.13 steps/s | VRAM=32.98 Go [2026-07-17 19:30:11] step 438750/1000000 | loss=2.3848 vocab=0.8858 token=1.4153 | 6.13 steps/s | VRAM=29.83 Go [2026-07-17 19:30:19] step 438800/1000000 | loss=2.3097 vocab=0.8346 token=1.2189 | 6.15 steps/s | VRAM=32.63 Go [2026-07-17 19:30:27] step 438850/1000000 | loss=2.3304 vocab=1.0102 token=1.2994 | 6.15 steps/s | VRAM=29.75 Go [2026-07-17 19:30:36] step 438900/1000000 | loss=2.2924 vocab=1.0402 token=1.1512 | 6.16 steps/s | VRAM=32.77 Go [2026-07-17 19:30:44] step 438950/1000000 | loss=2.3219 vocab=0.8557 token=1.1168 | 6.16 steps/s | VRAM=29.06 Go [2026-07-17 19:30:52] step 439000/1000000 | loss=2.2983 vocab=0.7146 token=1.2013 | 6.18 steps/s | VRAM=24.79 Go [2026-07-17 19:31:00] step 439050/1000000 | loss=2.2877 vocab=0.8642 token=1.2317 | 6.16 steps/s | VRAM=25.81 Go [2026-07-17 19:31:08] step 439100/1000000 | loss=2.2233 vocab=0.7354 token=1.5394 | 6.14 steps/s | VRAM=29.09 Go [2026-07-17 19:31:16] step 439150/1000000 | loss=2.2519 vocab=1.0598 token=1.2814 | 6.11 steps/s | VRAM=28.95 Go [2026-07-17 19:31:24] step 439200/1000000 | loss=2.3234 vocab=0.8743 token=1.5383 | 6.17 steps/s | VRAM=31.43 Go [2026-07-17 19:31:32] step 439250/1000000 | loss=2.2158 vocab=0.7056 token=1.1962 | 6.15 steps/s | VRAM=31.41 Go [2026-07-17 19:31:41] step 439300/1000000 | loss=2.2897 vocab=0.8833 token=1.3530 | 6.12 steps/s | VRAM=29.93 Go [2026-07-17 19:31:49] step 439350/1000000 | loss=2.3049 vocab=1.0613 token=1.4096 | 6.12 steps/s | VRAM=29.22 Go [2026-07-17 19:31:57] step 439400/1000000 | loss=2.3006 vocab=1.0645 token=1.4514 | 6.15 steps/s | VRAM=27.99 Go [2026-07-17 19:32:05] step 439450/1000000 | loss=2.2043 vocab=0.6934 token=1.2780 | 6.14 steps/s | VRAM=28.37 Go [2026-07-17 19:32:13] step 439500/1000000 | loss=2.2369 vocab=0.7992 token=1.2140 | 6.11 steps/s | VRAM=32.03 Go [2026-07-17 19:32:21] step 439550/1000000 | loss=2.2806 vocab=1.5626 token=1.4979 | 6.11 steps/s | VRAM=28.50 Go [2026-07-17 19:32:30] step 439600/1000000 | loss=2.2443 vocab=0.8850 token=1.5120 | 6.15 steps/s | VRAM=31.93 Go [2026-07-17 19:32:38] step 439650/1000000 | loss=2.3305 vocab=1.0205 token=1.4534 | 6.14 steps/s | VRAM=24.71 Go [2026-07-17 19:32:46] step 439700/1000000 | loss=2.2553 vocab=1.3812 token=1.3119 | 6.11 steps/s | VRAM=29.36 Go [2026-07-17 19:32:54] step 439750/1000000 | loss=2.3094 vocab=0.5251 token=1.0739 | 6.15 steps/s | VRAM=28.03 Go [2026-07-17 19:33:02] step 439800/1000000 | loss=2.2708 vocab=0.6906 token=1.8211 | 6.13 steps/s | VRAM=29.47 Go [2026-07-17 19:33:10] step 439850/1000000 | loss=2.2615 vocab=0.6743 token=1.3271 | 6.14 steps/s | VRAM=25.92 Go [2026-07-17 19:33:18] step 439900/1000000 | loss=2.3073 vocab=0.9562 token=1.6253 | 6.10 steps/s | VRAM=27.24 Go [2026-07-17 19:33:27] step 439950/1000000 | loss=2.2642 vocab=1.0277 token=1.3909 | 6.12 steps/s | VRAM=28.30 Go [2026-07-17 19:33:35] step 440000/1000000 | loss=2.2459 vocab=1.1422 token=1.3023 | 6.12 steps/s | VRAM=26.06 Go [2026-07-17 19:33:43] step 440050/1000000 | loss=2.3390 vocab=0.5593 token=1.1499 | 6.15 steps/s | VRAM=28.49 Go [2026-07-17 19:33:51] step 440100/1000000 | loss=2.3365 vocab=0.9090 token=1.2062 | 6.13 steps/s | VRAM=26.19 Go [2026-07-17 19:33:59] step 440150/1000000 | loss=2.2982 vocab=0.5610 token=1.2444 | 6.10 steps/s | VRAM=27.37 Go [2026-07-17 19:34:07] step 440200/1000000 | loss=2.2586 vocab=0.5890 token=1.2359 | 6.12 steps/s | VRAM=32.07 Go [2026-07-17 19:34:16] step 440250/1000000 | loss=2.2612 vocab=1.0849 token=0.7934 | 6.13 steps/s | VRAM=31.83 Go [2026-07-17 19:34:24] step 440300/1000000 | loss=2.3044 vocab=0.8900 token=1.6854 | 6.14 steps/s | VRAM=27.83 Go [2026-07-17 19:34:32] step 440350/1000000 | loss=2.2741 vocab=1.0797 token=1.2562 | 6.14 steps/s | VRAM=32.62 Go [2026-07-17 19:34:40] step 440400/1000000 | loss=2.2328 vocab=0.6916 token=1.5404 | 6.12 steps/s | VRAM=29.88 Go [2026-07-17 19:34:48] step 440450/1000000 | loss=2.2438 vocab=1.2395 token=1.7399 | 6.13 steps/s | VRAM=29.03 Go [2026-07-17 19:34:56] step 440500/1000000 | loss=2.2639 vocab=0.9999 token=1.6828 | 6.14 steps/s | VRAM=28.37 Go [2026-07-17 19:35:05] step 440550/1000000 | loss=2.2558 vocab=0.7439 token=1.2763 | 6.13 steps/s | VRAM=25.73 Go [2026-07-17 19:35:13] step 440600/1000000 | loss=2.1592 vocab=1.1650 token=1.5180 | 6.13 steps/s | VRAM=29.62 Go [2026-07-17 19:35:21] step 440650/1000000 | loss=2.1514 vocab=0.8790 token=1.3718 | 6.14 steps/s | VRAM=32.78 Go [2026-07-17 19:35:29] step 440700/1000000 | loss=2.1989 vocab=0.9051 token=1.6703 | 6.13 steps/s | VRAM=25.91 Go [2026-07-17 19:35:37] step 440750/1000000 | loss=2.2161 vocab=1.1739 token=1.6371 | 6.16 steps/s | VRAM=32.17 Go [2026-07-17 19:35:45] step 440800/1000000 | loss=2.3809 vocab=0.6109 token=1.3543 | 6.14 steps/s | VRAM=28.49 Go [2026-07-17 19:35:53] step 440850/1000000 | loss=2.2416 vocab=1.1029 token=1.5381 | 6.14 steps/s | VRAM=29.13 Go [2026-07-17 19:36:02] step 440900/1000000 | loss=2.2536 vocab=0.9826 token=1.2739 | 6.12 steps/s | VRAM=30.07 Go [2026-07-17 19:36:10] step 440950/1000000 | loss=2.3687 vocab=1.6495 token=1.6571 | 6.16 steps/s | VRAM=29.44 Go [2026-07-17 19:36:18] step 441000/1000000 | loss=2.1945 vocab=0.6110 token=1.2350 | 6.15 steps/s | VRAM=31.23 Go [2026-07-17 19:36:26] step 441050/1000000 | loss=2.2950 vocab=0.8963 token=1.0027 | 6.12 steps/s | VRAM=29.03 Go [2026-07-17 19:36:34] step 441100/1000000 | loss=2.1557 vocab=0.9725 token=1.1568 | 6.14 steps/s | VRAM=25.53 Go [2026-07-17 19:36:42] step 441150/1000000 | loss=2.1954 vocab=0.9145 token=1.3101 | 6.14 steps/s | VRAM=31.81 Go [2026-07-17 19:36:50] step 441200/1000000 | loss=2.2814 vocab=0.9033 token=1.4829 | 6.11 steps/s | VRAM=32.81 Go [2026-07-17 19:36:59] step 441250/1000000 | loss=2.1505 vocab=0.7013 token=0.9430 | 6.13 steps/s | VRAM=24.92 Go [2026-07-17 19:37:07] step 441300/1000000 | loss=2.3528 vocab=0.9047 token=1.5278 | 6.14 steps/s | VRAM=29.07 Go [2026-07-17 19:37:15] step 441350/1000000 | loss=2.2549 vocab=0.7855 token=1.2609 | 6.12 steps/s | VRAM=26.69 Go [2026-07-17 19:37:23] step 441400/1000000 | loss=2.2346 vocab=0.5158 token=1.4099 | 6.11 steps/s | VRAM=25.08 Go [2026-07-17 19:37:31] step 441450/1000000 | loss=2.2398 vocab=0.8004 token=1.1891 | 6.12 steps/s | VRAM=28.96 Go [2026-07-17 19:37:39] step 441500/1000000 | loss=2.2177 vocab=1.0686 token=1.5506 | 6.15 steps/s | VRAM=31.87 Go [2026-07-17 19:37:48] step 441550/1000000 | loss=2.2416 vocab=1.0421 token=1.1230 | 6.16 steps/s | VRAM=28.80 Go [2026-07-17 19:37:56] step 441600/1000000 | loss=2.2599 vocab=0.5218 token=1.3069 | 6.12 steps/s | VRAM=26.82 Go [2026-07-17 19:38:04] step 441650/1000000 | loss=2.2475 vocab=1.5520 token=1.2810 | 6.13 steps/s | VRAM=32.25 Go [2026-07-17 19:38:12] step 441700/1000000 | loss=2.2505 vocab=0.8437 token=1.1586 | 6.14 steps/s | VRAM=27.89 Go [2026-07-17 19:38:20] step 441750/1000000 | loss=2.1951 vocab=0.8117 token=1.2051 | 6.12 steps/s | VRAM=29.12 Go [2026-07-17 19:38:28] step 441800/1000000 | loss=2.2074 vocab=0.5826 token=1.7799 | 6.10 steps/s | VRAM=32.28 Go [2026-07-17 19:38:37] step 441850/1000000 | loss=2.2201 vocab=1.3559 token=0.9527 | 6.12 steps/s | VRAM=29.67 Go [2026-07-17 19:38:45] step 441900/1000000 | loss=2.3516 vocab=1.1245 token=1.6149 | 6.13 steps/s | VRAM=29.62 Go [2026-07-17 19:38:53] step 441950/1000000 | loss=2.3047 vocab=0.6698 token=1.5444 | 6.13 steps/s | VRAM=28.51 Go [2026-07-17 19:39:01] step 442000/1000000 | loss=2.3364 vocab=0.8274 token=1.4403 | 6.13 steps/s | VRAM=26.73 Go [2026-07-17 19:39:09] step 442050/1000000 | loss=2.2972 vocab=0.6912 token=1.2661 | 6.14 steps/s | VRAM=32.12 Go [2026-07-17 19:39:17] step 442100/1000000 | loss=2.2324 vocab=0.8798 token=1.5547 | 6.12 steps/s | VRAM=29.58 Go [2026-07-17 19:39:26] step 442150/1000000 | loss=2.1671 vocab=0.9948 token=1.2967 | 6.12 steps/s | VRAM=25.78 Go [2026-07-17 19:39:34] step 442200/1000000 | loss=2.2282 vocab=0.6752 token=1.0477 | 6.14 steps/s | VRAM=32.10 Go [2026-07-17 19:39:42] step 442250/1000000 | loss=2.2603 vocab=0.5442 token=1.0435 | 6.17 steps/s | VRAM=27.82 Go [2026-07-17 19:39:50] step 442300/1000000 | loss=2.4224 vocab=0.7487 token=1.5841 | 6.12 steps/s | VRAM=26.18 Go [2026-07-17 19:39:58] step 442350/1000000 | loss=2.2586 vocab=1.1587 token=1.4839 | 6.14 steps/s | VRAM=26.17 Go [2026-07-17 19:40:06] step 442400/1000000 | loss=2.3086 vocab=0.8517 token=1.2237 | 6.16 steps/s | VRAM=25.24 Go [2026-07-17 19:40:14] step 442450/1000000 | loss=2.2389 vocab=0.8227 token=1.3811 | 6.14 steps/s | VRAM=28.08 Go [2026-07-17 19:40:22] step 442500/1000000 | loss=2.1968 vocab=0.7647 token=1.3942 | 6.15 steps/s | VRAM=31.14 Go [2026-07-17 19:40:31] step 442550/1000000 | loss=2.3216 vocab=0.6839 token=1.3837 | 6.13 steps/s | VRAM=26.56 Go [2026-07-17 19:40:39] step 442600/1000000 | loss=2.2015 vocab=0.7758 token=1.1334 | 6.12 steps/s | VRAM=28.63 Go [2026-07-17 19:40:47] step 442650/1000000 | loss=2.1875 vocab=0.7329 token=1.0924 | 6.12 steps/s | VRAM=25.81 Go [2026-07-17 19:40:55] step 442700/1000000 | loss=2.2024 vocab=0.8594 token=1.1043 | 6.15 steps/s | VRAM=28.86 Go [2026-07-17 19:41:03] step 442750/1000000 | loss=2.2841 vocab=0.6112 token=1.3604 | 6.16 steps/s | VRAM=31.81 Go [2026-07-17 19:41:11] step 442800/1000000 | loss=2.2172 vocab=0.8639 token=1.0203 | 6.16 steps/s | VRAM=32.45 Go [2026-07-17 19:41:19] step 442850/1000000 | loss=2.2708 vocab=0.8484 token=1.2908 | 6.13 steps/s | VRAM=32.73 Go [2026-07-17 19:41:28] step 442900/1000000 | loss=2.2693 vocab=0.7898 token=1.3611 | 6.13 steps/s | VRAM=28.73 Go [2026-07-17 19:41:36] step 442950/1000000 | loss=2.3155 vocab=1.0738 token=0.7978 | 6.12 steps/s | VRAM=27.01 Go [2026-07-17 19:41:44] step 443000/1000000 | loss=2.3499 vocab=0.5204 token=1.2090 | 6.12 steps/s | VRAM=29.29 Go [2026-07-17 19:41:52] step 443050/1000000 | loss=2.2777 vocab=1.0478 token=1.6756 | 6.13 steps/s | VRAM=24.90 Go [2026-07-17 19:42:00] step 443100/1000000 | loss=2.3237 vocab=1.0057 token=1.5314 | 6.15 steps/s | VRAM=26.52 Go [2026-07-17 19:42:08] step 443150/1000000 | loss=2.3291 vocab=1.2126 token=1.6592 | 6.15 steps/s | VRAM=31.47 Go [2026-07-17 19:42:17] step 443200/1000000 | loss=2.3244 vocab=0.9580 token=1.4297 | 6.11 steps/s | VRAM=28.71 Go [2026-07-17 19:42:25] step 443250/1000000 | loss=2.3091 vocab=0.9851 token=1.4805 | 6.14 steps/s | VRAM=28.25 Go [2026-07-17 19:42:33] step 443300/1000000 | loss=2.2611 vocab=1.1786 token=1.1871 | 6.12 steps/s | VRAM=32.69 Go [2026-07-17 19:42:41] step 443350/1000000 | loss=2.2288 vocab=0.5221 token=1.2787 | 6.14 steps/s | VRAM=29.39 Go [2026-07-17 19:42:49] step 443400/1000000 | loss=2.3089 vocab=0.7164 token=1.7252 | 6.13 steps/s | VRAM=29.68 Go [2026-07-17 19:42:57] step 443450/1000000 | loss=2.2734 vocab=1.1394 token=1.8000 | 6.14 steps/s | VRAM=28.93 Go [2026-07-17 19:43:06] step 443500/1000000 | loss=2.2662 vocab=0.6095 token=1.2493 | 6.12 steps/s | VRAM=29.38 Go [2026-07-17 19:43:14] step 443550/1000000 | loss=2.3669 vocab=1.1249 token=1.7322 | 6.14 steps/s | VRAM=31.68 Go [2026-07-17 19:43:22] step 443600/1000000 | loss=2.2392 vocab=1.5256 token=1.5199 | 6.15 steps/s | VRAM=32.07 Go [2026-07-17 19:43:30] step 443650/1000000 | loss=2.1625 vocab=0.8053 token=1.2477 | 6.14 steps/s | VRAM=32.09 Go [2026-07-17 19:43:38] step 443700/1000000 | loss=2.2211 vocab=0.6677 token=1.3500 | 6.14 steps/s | VRAM=31.31 Go [2026-07-17 19:43:46] step 443750/1000000 | loss=2.2149 vocab=0.4643 token=1.1882 | 6.14 steps/s | VRAM=26.61 Go [2026-07-17 19:43:54] step 443800/1000000 | loss=2.2124 vocab=0.7240 token=1.1684 | 6.11 steps/s | VRAM=29.84 Go [2026-07-17 19:44:03] step 443850/1000000 | loss=2.2380 vocab=0.9851 token=1.0804 | 6.13 steps/s | VRAM=29.04 Go [2026-07-17 19:44:11] step 443900/1000000 | loss=2.2780 vocab=1.1521 token=1.4639 | 6.12 steps/s | VRAM=26.70 Go [2026-07-17 19:44:19] step 443950/1000000 | loss=2.2148 vocab=0.8719 token=1.4811 | 6.15 steps/s | VRAM=31.76 Go [2026-07-17 19:44:27] step 444000/1000000 | loss=2.2081 vocab=1.2421 token=1.5327 | 6.12 steps/s | VRAM=32.61 Go [2026-07-17 19:44:35] step 444050/1000000 | loss=2.3253 vocab=0.9110 token=1.2197 | 6.16 steps/s | VRAM=24.78 Go [2026-07-17 19:44:43] step 444100/1000000 | loss=2.2780 vocab=1.2155 token=1.5700 | 6.16 steps/s | VRAM=25.97 Go [2026-07-17 19:44:51] step 444150/1000000 | loss=2.2273 vocab=0.3887 token=1.0184 | 6.13 steps/s | VRAM=30.16 Go [2026-07-17 19:45:00] step 444200/1000000 | loss=2.2952 vocab=0.7701 token=1.6218 | 6.16 steps/s | VRAM=28.62 Go [2026-07-17 19:45:08] step 444250/1000000 | loss=2.2613 vocab=0.5816 token=1.2113 | 6.13 steps/s | VRAM=28.56 Go [2026-07-17 19:45:16] step 444300/1000000 | loss=2.3436 vocab=0.9157 token=1.7061 | 6.11 steps/s | VRAM=29.60 Go [2026-07-17 19:45:24] step 444350/1000000 | loss=2.2773 vocab=1.0402 token=1.1457 | 6.13 steps/s | VRAM=31.07 Go [2026-07-17 19:45:32] step 444400/1000000 | loss=2.2041 vocab=1.0459 token=1.2775 | 6.15 steps/s | VRAM=24.66 Go [2026-07-17 19:45:40] step 444450/1000000 | loss=2.3626 vocab=0.9282 token=1.4172 | 6.15 steps/s | VRAM=28.65 Go [2026-07-17 19:45:48] step 444500/1000000 | loss=2.2844 vocab=1.3018 token=1.4079 | 6.15 steps/s | VRAM=26.40 Go [2026-07-17 19:45:57] step 444550/1000000 | loss=2.2655 vocab=0.5067 token=0.9642 | 6.12 steps/s | VRAM=30.80 Go [2026-07-17 19:46:05] step 444600/1000000 | loss=2.1978 vocab=0.7920 token=1.4632 | 6.11 steps/s | VRAM=27.11 Go [2026-07-17 19:46:13] step 444650/1000000 | loss=2.3197 vocab=1.1079 token=1.4310 | 6.15 steps/s | VRAM=28.80 Go [2026-07-17 19:46:21] step 444700/1000000 | loss=2.2003 vocab=0.9861 token=1.3203 | 6.14 steps/s | VRAM=28.85 Go [2026-07-17 19:46:29] step 444750/1000000 | loss=2.3142 vocab=1.0662 token=1.7684 | 6.12 steps/s | VRAM=29.51 Go [2026-07-17 19:46:37] step 444800/1000000 | loss=2.2792 vocab=0.8898 token=1.2961 | 6.12 steps/s | VRAM=24.71 Go [2026-07-17 19:46:46] step 444850/1000000 | loss=2.3670 vocab=0.9791 token=1.5409 | 6.14 steps/s | VRAM=26.02 Go [2026-07-17 19:46:54] step 444900/1000000 | loss=2.3269 vocab=0.8585 token=1.7926 | 6.15 steps/s | VRAM=31.92 Go [2026-07-17 19:47:02] step 444950/1000000 | loss=2.2257 vocab=0.6092 token=1.1612 | 6.09 steps/s | VRAM=32.95 Go [2026-07-17 19:47:10] step 445000/1000000 | loss=2.2497 vocab=1.1479 token=1.5133 | 6.13 steps/s | VRAM=31.44 Go [2026-07-17 19:47:18] step 445050/1000000 | loss=2.3142 vocab=0.9590 token=1.5249 | 6.14 steps/s | VRAM=28.46 Go [2026-07-17 19:47:26] step 445100/1000000 | loss=2.3611 vocab=0.7273 token=1.1640 | 6.15 steps/s | VRAM=29.24 Go [2026-07-17 19:47:35] step 445150/1000000 | loss=2.2828 vocab=0.7104 token=1.2159 | 6.11 steps/s | VRAM=28.29 Go [2026-07-17 19:47:43] step 445200/1000000 | loss=2.2978 vocab=0.5736 token=0.8607 | 6.14 steps/s | VRAM=31.91 Go [2026-07-17 19:47:51] step 445250/1000000 | loss=2.2744 vocab=1.1203 token=1.3823 | 6.10 steps/s | VRAM=28.94 Go [2026-07-17 19:47:59] step 445300/1000000 | loss=2.2832 vocab=0.9528 token=1.3439 | 6.12 steps/s | VRAM=26.03 Go [2026-07-17 19:48:07] step 445350/1000000 | loss=2.2389 vocab=0.8152 token=1.4168 | 6.16 steps/s | VRAM=25.87 Go [2026-07-17 19:48:15] step 445400/1000000 | loss=2.3093 vocab=0.8657 token=1.3624 | 6.15 steps/s | VRAM=31.93 Go [2026-07-17 19:48:23] step 445450/1000000 | loss=2.2551 vocab=0.9011 token=1.4571 | 6.16 steps/s | VRAM=24.91 Go [2026-07-17 19:48:32] step 445500/1000000 | loss=2.1964 vocab=0.7777 token=1.1883 | 6.11 steps/s | VRAM=30.41 Go [2026-07-17 19:48:40] step 445550/1000000 | loss=2.2902 vocab=0.8044 token=1.2548 | 6.09 steps/s | VRAM=26.40 Go [2026-07-17 19:48:48] step 445600/1000000 | loss=2.3220 vocab=0.7028 token=1.6710 | 6.12 steps/s | VRAM=27.95 Go [2026-07-17 19:48:56] step 445650/1000000 | loss=2.2232 vocab=1.0456 token=1.2941 | 6.15 steps/s | VRAM=32.83 Go [2026-07-17 19:49:04] step 445700/1000000 | loss=2.3640 vocab=0.6543 token=1.0170 | 6.15 steps/s | VRAM=32.61 Go [2026-07-17 19:49:12] step 445750/1000000 | loss=2.2172 vocab=0.8397 token=1.2481 | 6.13 steps/s | VRAM=32.12 Go [2026-07-17 19:49:21] step 445800/1000000 | loss=2.3990 vocab=0.9420 token=1.4228 | 6.10 steps/s | VRAM=26.47 Go [2026-07-17 19:49:29] step 445850/1000000 | loss=2.3019 vocab=1.1411 token=1.2191 | 6.16 steps/s | VRAM=32.03 Go [2026-07-17 19:49:37] step 445900/1000000 | loss=2.1945 vocab=0.9980 token=1.1797 | 6.12 steps/s | VRAM=28.51 Go [2026-07-17 19:49:45] step 445950/1000000 | loss=2.2882 vocab=0.9288 token=1.5348 | 6.10 steps/s | VRAM=27.16 Go [2026-07-17 19:49:53] step 446000/1000000 | loss=2.2470 vocab=0.7913 token=1.3278 | 6.14 steps/s | VRAM=32.61 Go [2026-07-17 19:50:01] step 446050/1000000 | loss=2.2212 vocab=0.9125 token=1.3294 | 6.17 steps/s | VRAM=31.59 Go [2026-07-17 19:50:09] step 446100/1000000 | loss=2.2135 vocab=0.8302 token=1.2934 | 6.13 steps/s | VRAM=29.87 Go [2026-07-17 19:50:18] step 446150/1000000 | loss=2.2047 vocab=0.4651 token=0.9669 | 6.14 steps/s | VRAM=31.53 Go [2026-07-17 19:50:26] step 446200/1000000 | loss=2.3096 vocab=0.7245 token=1.7103 | 6.14 steps/s | VRAM=27.90 Go [2026-07-17 19:50:34] step 446250/1000000 | loss=2.1891 vocab=0.8792 token=1.3211 | 6.16 steps/s | VRAM=31.80 Go [2026-07-17 19:50:42] step 446300/1000000 | loss=2.2663 vocab=0.9238 token=1.2363 | 6.13 steps/s | VRAM=26.38 Go [2026-07-17 19:50:50] step 446350/1000000 | loss=2.2332 vocab=1.1134 token=1.6310 | 6.11 steps/s | VRAM=29.65 Go [2026-07-17 19:50:58] step 446400/1000000 | loss=2.1862 vocab=0.8004 token=1.2421 | 6.14 steps/s | VRAM=32.99 Go [2026-07-17 19:51:07] step 446450/1000000 | loss=2.2903 vocab=0.8029 token=1.3276 | 6.12 steps/s | VRAM=32.73 Go [2026-07-17 19:51:15] step 446500/1000000 | loss=2.1583 vocab=1.0347 token=1.4903 | 6.15 steps/s | VRAM=28.71 Go [2026-07-17 19:51:23] step 446550/1000000 | loss=2.2381 vocab=1.2434 token=1.4510 | 6.16 steps/s | VRAM=31.45 Go [2026-07-17 19:51:31] step 446600/1000000 | loss=2.3531 vocab=0.5205 token=1.1398 | 6.13 steps/s | VRAM=28.87 Go [2026-07-17 19:51:39] step 446650/1000000 | loss=2.2473 vocab=1.2744 token=1.7146 | 6.11 steps/s | VRAM=25.34 Go [2026-07-17 19:51:47] step 446700/1000000 | loss=2.2992 vocab=0.6422 token=1.0811 | 6.14 steps/s | VRAM=26.68 Go [2026-07-17 19:51:55] step 446750/1000000 | loss=2.2344 vocab=0.6810 token=0.8693 | 6.16 steps/s | VRAM=28.81 Go [2026-07-17 19:52:04] step 446800/1000000 | loss=2.2444 vocab=1.1537 token=1.3024 | 6.13 steps/s | VRAM=29.39 Go [2026-07-17 19:52:12] step 446850/1000000 | loss=2.2182 vocab=0.6233 token=0.9687 | 6.12 steps/s | VRAM=28.75 Go [2026-07-17 19:52:20] step 446900/1000000 | loss=2.1779 vocab=0.6094 token=1.2788 | 6.14 steps/s | VRAM=24.75 Go [2026-07-17 19:52:28] step 446950/1000000 | loss=2.2400 vocab=0.5484 token=0.9450 | 6.11 steps/s | VRAM=27.28 Go [2026-07-17 19:52:36] step 447000/1000000 | loss=2.2658 vocab=0.9276 token=1.1303 | 6.14 steps/s | VRAM=25.31 Go [2026-07-17 19:52:44] step 447050/1000000 | loss=2.2069 vocab=1.2194 token=1.4842 | 6.13 steps/s | VRAM=26.10 Go [2026-07-17 19:52:52] step 447100/1000000 | loss=2.2796 vocab=0.8964 token=1.7995 | 6.14 steps/s | VRAM=25.33 Go [2026-07-17 19:53:01] step 447150/1000000 | loss=2.2197 vocab=0.6386 token=1.3056 | 6.15 steps/s | VRAM=28.11 Go [2026-07-17 19:53:09] step 447200/1000000 | loss=2.2472 vocab=0.9961 token=1.6057 | 6.13 steps/s | VRAM=26.02 Go [2026-07-17 19:53:17] step 447250/1000000 | loss=2.3280 vocab=0.4703 token=0.9931 | 6.13 steps/s | VRAM=28.46 Go [2026-07-17 19:53:25] step 447300/1000000 | loss=2.3401 vocab=1.1699 token=1.2266 | 6.12 steps/s | VRAM=26.22 Go [2026-07-17 19:53:33] step 447350/1000000 | loss=2.2876 vocab=0.9573 token=1.3580 | 6.11 steps/s | VRAM=28.76 Go [2026-07-17 19:53:41] step 447400/1000000 | loss=2.3470 vocab=0.8317 token=1.1326 | 6.10 steps/s | VRAM=28.85 Go [2026-07-17 19:53:50] step 447450/1000000 | loss=2.4106 vocab=0.8470 token=0.8844 | 6.11 steps/s | VRAM=31.94 Go [2026-07-17 19:53:58] step 447500/1000000 | loss=2.2706 vocab=0.5584 token=1.7047 | 6.12 steps/s | VRAM=24.31 Go [2026-07-17 19:54:06] step 447550/1000000 | loss=2.2713 vocab=0.8821 token=1.1319 | 6.12 steps/s | VRAM=30.32 Go [2026-07-17 19:54:14] step 447600/1000000 | loss=2.2959 vocab=0.9925 token=1.9078 | 6.13 steps/s | VRAM=32.37 Go [2026-07-17 19:54:22] step 447650/1000000 | loss=2.2112 vocab=1.0066 token=1.2322 | 6.11 steps/s | VRAM=25.21 Go [2026-07-17 19:54:31] step 447700/1000000 | loss=2.2706 vocab=1.4472 token=1.5716 | 6.12 steps/s | VRAM=29.28 Go [2026-07-17 19:54:39] step 447750/1000000 | loss=2.3379 vocab=1.1514 token=1.1224 | 6.13 steps/s | VRAM=29.05 Go [2026-07-17 19:54:47] step 447800/1000000 | loss=2.2158 vocab=1.1942 token=1.2055 | 6.14 steps/s | VRAM=27.97 Go [2026-07-17 19:54:55] step 447850/1000000 | loss=2.1898 vocab=0.7449 token=1.3270 | 6.14 steps/s | VRAM=26.04 Go [2026-07-17 19:55:03] step 447900/1000000 | loss=2.1117 vocab=0.7294 token=1.2788 | 6.16 steps/s | VRAM=28.11 Go [2026-07-17 19:55:11] step 447950/1000000 | loss=2.2465 vocab=0.9823 token=1.6361 | 6.13 steps/s | VRAM=26.80 Go [2026-07-17 19:55:19] step 448000/1000000 | loss=2.3390 vocab=0.3834 token=1.0328 | 6.16 steps/s | VRAM=32.09 Go [2026-07-17 19:55:27] step 448050/1000000 | loss=2.3096 vocab=1.1242 token=1.3000 | 6.13 steps/s | VRAM=28.21 Go [2026-07-17 19:55:36] step 448100/1000000 | loss=2.2285 vocab=0.9596 token=1.5319 | 6.11 steps/s | VRAM=29.07 Go [2026-07-17 19:55:44] step 448150/1000000 | loss=2.2706 vocab=0.7687 token=1.2719 | 6.12 steps/s | VRAM=26.25 Go [2026-07-17 19:55:52] step 448200/1000000 | loss=2.2658 vocab=0.9400 token=1.2883 | 6.16 steps/s | VRAM=31.54 Go [2026-07-17 19:56:00] step 448250/1000000 | loss=2.2407 vocab=1.0034 token=1.4568 | 6.15 steps/s | VRAM=28.24 Go [2026-07-17 19:56:08] step 448300/1000000 | loss=2.2291 vocab=1.3074 token=1.6404 | 6.14 steps/s | VRAM=27.99 Go [2026-07-17 19:56:16] step 448350/1000000 | loss=2.2140 vocab=0.9590 token=1.5383 | 6.12 steps/s | VRAM=27.92 Go [2026-07-17 19:56:25] step 448400/1000000 | loss=2.3218 vocab=0.7954 token=1.3404 | 6.12 steps/s | VRAM=29.05 Go [2026-07-17 19:56:33] step 448450/1000000 | loss=2.2001 vocab=0.9865 token=0.9469 | 6.14 steps/s | VRAM=28.59 Go [2026-07-17 19:56:41] step 448500/1000000 | loss=2.1977 vocab=1.0577 token=1.5600 | 6.10 steps/s | VRAM=27.11 Go [2026-07-17 19:56:49] step 448550/1000000 | loss=2.1764 vocab=1.1524 token=1.7459 | 6.17 steps/s | VRAM=27.79 Go [2026-07-17 19:56:57] step 448600/1000000 | loss=2.2209 vocab=1.2411 token=1.6562 | 6.15 steps/s | VRAM=28.29 Go [2026-07-17 19:57:05] step 448650/1000000 | loss=2.2707 vocab=0.7169 token=1.3718 | 6.12 steps/s | VRAM=29.16 Go [2026-07-17 19:57:13] step 448700/1000000 | loss=2.2854 vocab=0.9819 token=1.2805 | 6.14 steps/s | VRAM=32.08 Go [2026-07-17 19:57:22] step 448750/1000000 | loss=2.2423 vocab=1.2062 token=1.5622 | 6.14 steps/s | VRAM=29.01 Go [2026-07-17 19:57:30] step 448800/1000000 | loss=2.1654 vocab=1.1356 token=1.6922 | 6.13 steps/s | VRAM=28.41 Go [2026-07-17 19:57:38] step 448850/1000000 | loss=2.2304 vocab=0.8559 token=1.0467 | 6.15 steps/s | VRAM=31.57 Go [2026-07-17 19:57:46] step 448900/1000000 | loss=2.3631 vocab=0.8611 token=1.3417 | 6.14 steps/s | VRAM=27.72 Go [2026-07-17 19:57:54] step 448950/1000000 | loss=2.3402 vocab=1.0550 token=1.4617 | 6.14 steps/s | VRAM=27.86 Go [2026-07-17 19:58:02] step 449000/1000000 | loss=2.2791 vocab=0.9531 token=1.1110 | 6.15 steps/s | VRAM=29.09 Go [2026-07-17 19:58:10] step 449050/1000000 | loss=2.3054 vocab=0.8414 token=1.0743 | 6.13 steps/s | VRAM=32.50 Go [2026-07-17 19:58:19] step 449100/1000000 | loss=2.2433 vocab=1.2075 token=1.3313 | 6.12 steps/s | VRAM=32.15 Go [2026-07-17 19:58:27] step 449150/1000000 | loss=2.4129 vocab=1.4635 token=1.5894 | 6.16 steps/s | VRAM=31.00 Go [2026-07-17 19:58:35] step 449200/1000000 | loss=2.2022 vocab=0.5509 token=0.9399 | 6.14 steps/s | VRAM=29.00 Go [2026-07-17 19:58:43] step 449250/1000000 | loss=2.1601 vocab=0.7644 token=1.3660 | 6.13 steps/s | VRAM=26.53 Go [2026-07-17 19:58:51] step 449300/1000000 | loss=2.2739 vocab=0.9119 token=1.4852 | 6.15 steps/s | VRAM=32.04 Go [2026-07-17 19:58:59] step 449350/1000000 | loss=2.1649 vocab=1.0448 token=1.1705 | 6.15 steps/s | VRAM=31.90 Go [2026-07-17 19:59:07] step 449400/1000000 | loss=2.3035 vocab=0.7098 token=1.1740 | 6.13 steps/s | VRAM=29.30 Go [2026-07-17 19:59:16] step 449450/1000000 | loss=2.1979 vocab=0.8922 token=1.4380 | 6.14 steps/s | VRAM=28.95 Go [2026-07-17 19:59:24] step 449500/1000000 | loss=2.3536 vocab=1.0742 token=1.4038 | 6.15 steps/s | VRAM=32.74 Go [2026-07-17 20:12:42] step 449550/1000000 | loss=0.4996 vocab=0.7648 token=1.2749 | 18.75 steps/s | VRAM=32.31 Go [2026-07-17 20:12:50] step 449600/1000000 | loss=2.3084 vocab=0.8764 token=1.2596 | 6.26 steps/s | VRAM=26.72 Go [2026-07-17 20:12:58] step 449650/1000000 | loss=2.1980 vocab=1.0419 token=1.2707 | 6.22 steps/s | VRAM=28.31 Go [2026-07-17 20:13:06] step 449700/1000000 | loss=2.2676 vocab=0.9182 token=1.3409 | 6.27 steps/s | VRAM=31.35 Go [2026-07-17 20:13:14] step 449750/1000000 | loss=2.1409 vocab=0.9304 token=1.3946 | 6.24 steps/s | VRAM=30.67 Go [2026-07-17 20:13:22] step 449800/1000000 | loss=2.1940 vocab=0.9804 token=1.6237 | 6.23 steps/s | VRAM=31.73 Go [2026-07-17 20:13:30] step 449850/1000000 | loss=2.2629 vocab=0.9982 token=1.5492 | 6.20 steps/s | VRAM=31.08 Go [2026-07-17 20:13:38] step 449900/1000000 | loss=2.2848 vocab=1.4735 token=1.5812 | 6.21 steps/s | VRAM=24.90 Go [2026-07-17 20:13:46] step 449950/1000000 | loss=2.2504 vocab=1.4832 token=1.1557 | 6.18 steps/s | VRAM=24.45 Go [2026-07-17 20:13:54] step 450000/1000000 | loss=2.3243 vocab=1.1222 token=1.2654 | 6.21 steps/s | VRAM=32.15 Go [push] step 450000 OK (best-of-2, loss_token=1.2654) [2026-07-17 20:15:18] step 450050/1000000 | loss=2.2353 vocab=0.9515 token=1.6389 | 0.59 steps/s | VRAM=31.69 Go [2026-07-17 20:15:26] step 450100/1000000 | loss=2.3620 vocab=0.9091 token=1.3172 | 6.19 steps/s | VRAM=29.91 Go [2026-07-17 20:15:34] step 450150/1000000 | loss=2.3327 vocab=0.8375 token=1.8292 | 6.19 steps/s | VRAM=31.08 Go [2026-07-17 20:15:42] step 450200/1000000 | loss=2.1819 vocab=0.6597 token=0.9730 | 6.19 steps/s | VRAM=30.58 Go [2026-07-17 20:15:50] step 450250/1000000 | loss=2.1750 vocab=0.6516 token=1.2842 | 6.17 steps/s | VRAM=29.98 Go [2026-07-17 20:15:59] step 450300/1000000 | loss=2.2721 vocab=1.2634 token=1.6067 | 6.15 steps/s | VRAM=29.98 Go [2026-07-17 20:16:07] step 450350/1000000 | loss=2.3708 vocab=1.2687 token=1.5255 | 6.21 steps/s | VRAM=28.44 Go [2026-07-17 20:16:15] step 450400/1000000 | loss=2.2617 vocab=0.7316 token=1.0059 | 6.15 steps/s | VRAM=31.29 Go [2026-07-17 20:16:23] step 450450/1000000 | loss=2.1397 vocab=1.5296 token=1.0418 | 6.12 steps/s | VRAM=26.58 Go [2026-07-17 20:16:31] step 450500/1000000 | loss=2.2289 vocab=0.5999 token=1.3704 | 6.13 steps/s | VRAM=26.40 Go [2026-07-17 20:16:39] step 450550/1000000 | loss=2.1689 vocab=0.6858 token=1.1420 | 6.13 steps/s | VRAM=28.74 Go [2026-07-17 20:16:47] step 450600/1000000 | loss=2.2316 vocab=0.7641 token=1.5050 | 6.16 steps/s | VRAM=31.37 Go [2026-07-17 20:16:56] step 450650/1000000 | loss=2.3039 vocab=0.7991 token=1.4388 | 6.16 steps/s | VRAM=31.11 Go [2026-07-17 20:17:04] step 450700/1000000 | loss=2.3175 vocab=0.9861 token=1.2352 | 6.14 steps/s | VRAM=26.72 Go [2026-07-17 20:17:12] step 450750/1000000 | loss=2.2250 vocab=1.0022 token=1.7338 | 6.14 steps/s | VRAM=27.98 Go [2026-07-17 20:17:20] step 450800/1000000 | loss=2.2843 vocab=1.3997 token=1.4754 | 6.13 steps/s | VRAM=32.66 Go [2026-07-17 20:17:28] step 450850/1000000 | loss=2.2547 vocab=1.0800 token=1.4314 | 6.12 steps/s | VRAM=32.89 Go [2026-07-17 20:17:36] step 450900/1000000 | loss=2.1360 vocab=0.8149 token=1.1624 | 6.12 steps/s | VRAM=32.68 Go [2026-07-17 20:17:44] step 450950/1000000 | loss=2.2546 vocab=0.8006 token=1.2967 | 6.13 steps/s | VRAM=32.17 Go [2026-07-17 20:17:53] step 451000/1000000 | loss=2.3369 vocab=0.7313 token=1.5306 | 6.14 steps/s | VRAM=32.46 Go [2026-07-17 20:18:01] step 451050/1000000 | loss=2.2349 vocab=0.1846 token=1.0139 | 6.14 steps/s | VRAM=28.48 Go [2026-07-17 20:18:09] step 451100/1000000 | loss=2.3499 vocab=1.6470 token=1.4624 | 6.12 steps/s | VRAM=25.61 Go [2026-07-17 20:18:17] step 451150/1000000 | loss=2.1715 vocab=0.7633 token=1.2327 | 6.11 steps/s | VRAM=28.86 Go [2026-07-17 20:18:25] step 451200/1000000 | loss=2.3322 vocab=0.7397 token=1.3343 | 6.13 steps/s | VRAM=28.36 Go [2026-07-17 20:18:33] step 451250/1000000 | loss=2.3032 vocab=1.0947 token=1.4569 | 6.15 steps/s | VRAM=32.88 Go [2026-07-17 20:18:42] step 451300/1000000 | loss=2.1976 vocab=0.6348 token=1.1570 | 6.11 steps/s | VRAM=28.81 Go [2026-07-17 20:18:50] step 451350/1000000 | loss=2.2904 vocab=0.9927 token=1.3582 | 6.14 steps/s | VRAM=28.48 Go [2026-07-17 20:18:58] step 451400/1000000 | loss=2.1972 vocab=0.8185 token=1.2083 | 6.12 steps/s | VRAM=29.25 Go [2026-07-17 20:19:06] step 451450/1000000 | loss=2.2901 vocab=0.7324 token=1.0245 | 6.15 steps/s | VRAM=32.28 Go [2026-07-17 20:19:14] step 451500/1000000 | loss=2.2138 vocab=0.8487 token=1.2852 | 6.11 steps/s | VRAM=26.57 Go [2026-07-17 20:19:22] step 451550/1000000 | loss=2.3052 vocab=0.6166 token=1.3083 | 6.15 steps/s | VRAM=31.63 Go [2026-07-17 20:19:30] step 451600/1000000 | loss=2.2459 vocab=0.7680 token=1.2839 | 6.15 steps/s | VRAM=31.44 Go [2026-07-17 20:19:39] step 451650/1000000 | loss=2.2419 vocab=1.3540 token=1.1713 | 6.15 steps/s | VRAM=31.88 Go [2026-07-17 20:19:47] step 451700/1000000 | loss=2.1632 vocab=0.8570 token=1.1302 | 6.12 steps/s | VRAM=26.61 Go [2026-07-17 20:19:55] step 451750/1000000 | loss=2.2416 vocab=0.4253 token=0.8828 | 6.12 steps/s | VRAM=25.78 Go [2026-07-17 20:20:03] step 451800/1000000 | loss=2.3393 vocab=1.0688 token=1.2367 | 6.14 steps/s | VRAM=29.63 Go [2026-07-17 20:20:11] step 451850/1000000 | loss=2.2315 vocab=1.0343 token=1.4111 | 6.12 steps/s | VRAM=31.88 Go [2026-07-17 20:20:19] step 451900/1000000 | loss=2.1916 vocab=0.8947 token=1.3085 | 6.12 steps/s | VRAM=29.05 Go [2026-07-17 20:20:28] step 451950/1000000 | loss=2.2540 vocab=0.9808 token=1.4967 | 6.14 steps/s | VRAM=28.95 Go [2026-07-17 20:20:36] step 452000/1000000 | loss=2.2061 vocab=0.8081 token=1.2295 | 6.13 steps/s | VRAM=31.11 Go [2026-07-17 20:20:44] step 452050/1000000 | loss=2.1888 vocab=0.8536 token=1.4039 | 6.09 steps/s | VRAM=28.55 Go [2026-07-17 20:20:52] step 452100/1000000 | loss=2.3433 vocab=1.1853 token=1.6771 | 6.09 steps/s | VRAM=22.97 Go [2026-07-17 20:21:00] step 452150/1000000 | loss=2.2729 vocab=1.1720 token=1.3315 | 6.12 steps/s | VRAM=26.37 Go [2026-07-17 20:21:08] step 452200/1000000 | loss=2.3101 vocab=1.3842 token=1.4479 | 6.10 steps/s | VRAM=29.41 Go [2026-07-17 20:21:17] step 452250/1000000 | loss=2.2398 vocab=1.6563 token=1.4561 | 6.10 steps/s | VRAM=25.36 Go [2026-07-17 20:21:25] step 452300/1000000 | loss=2.2784 vocab=0.8467 token=1.6052 | 6.13 steps/s | VRAM=31.77 Go [2026-07-17 20:21:33] step 452350/1000000 | loss=2.2111 vocab=1.0725 token=2.1442 | 6.12 steps/s | VRAM=30.71 Go [2026-07-17 20:21:41] step 452400/1000000 | loss=2.2528 vocab=0.8121 token=1.1107 | 6.12 steps/s | VRAM=29.07 Go [2026-07-17 20:21:49] step 452450/1000000 | loss=2.2321 vocab=0.7460 token=1.5812 | 6.15 steps/s | VRAM=28.74 Go [2026-07-17 20:21:58] step 452500/1000000 | loss=2.2182 vocab=0.6104 token=1.0445 | 6.10 steps/s | VRAM=32.62 Go [2026-07-17 20:22:06] step 452550/1000000 | loss=2.2863 vocab=0.7993 token=1.3024 | 6.14 steps/s | VRAM=30.87 Go [2026-07-17 20:22:14] step 452600/1000000 | loss=2.3651 vocab=1.1702 token=1.5637 | 6.13 steps/s | VRAM=26.27 Go [2026-07-17 20:22:22] step 452650/1000000 | loss=2.2249 vocab=0.9972 token=1.3930 | 6.14 steps/s | VRAM=28.71 Go [2026-07-17 20:22:30] step 452700/1000000 | loss=2.2488 vocab=0.8014 token=1.3846 | 6.14 steps/s | VRAM=25.41 Go [2026-07-17 20:22:38] step 452750/1000000 | loss=2.1763 vocab=1.2371 token=1.3750 | 6.12 steps/s | VRAM=28.18 Go [2026-07-17 20:22:46] step 452800/1000000 | loss=2.2705 vocab=1.0111 token=1.3148 | 6.14 steps/s | VRAM=26.46 Go [2026-07-17 20:22:55] step 452850/1000000 | loss=2.2395 vocab=0.9778 token=1.4693 | 6.15 steps/s | VRAM=31.68 Go [2026-07-17 20:23:03] step 452900/1000000 | loss=2.1326 vocab=1.0077 token=1.4315 | 6.13 steps/s | VRAM=28.94 Go [2026-07-17 20:23:11] step 452950/1000000 | loss=2.3159 vocab=0.8120 token=1.4879 | 6.09 steps/s | VRAM=28.42 Go [2026-07-17 20:23:19] step 453000/1000000 | loss=2.2824 vocab=1.0563 token=1.4736 | 6.12 steps/s | VRAM=32.56 Go [2026-07-17 20:23:27] step 453050/1000000 | loss=2.3445 vocab=0.8447 token=1.4848 | 6.12 steps/s | VRAM=27.08 Go [2026-07-17 20:23:35] step 453100/1000000 | loss=2.1606 vocab=1.2886 token=1.8088 | 6.13 steps/s | VRAM=29.23 Go [2026-07-17 20:23:44] step 453150/1000000 | loss=2.1645 vocab=0.9823 token=1.3389 | 6.15 steps/s | VRAM=28.18 Go [2026-07-17 20:23:52] step 453200/1000000 | loss=2.2505 vocab=0.8095 token=1.4287 | 6.15 steps/s | VRAM=31.36 Go [2026-07-17 20:24:00] step 453250/1000000 | loss=2.3717 vocab=0.8238 token=1.2934 | 6.18 steps/s | VRAM=31.34 Go [2026-07-17 20:24:08] step 453300/1000000 | loss=2.1990 vocab=1.0155 token=1.5087 | 6.11 steps/s | VRAM=26.41 Go [2026-07-17 20:24:16] step 453350/1000000 | loss=2.1983 vocab=0.4768 token=1.0112 | 6.14 steps/s | VRAM=28.19 Go [2026-07-17 20:24:24] step 453400/1000000 | loss=2.2698 vocab=1.1596 token=1.5041 | 6.12 steps/s | VRAM=28.73 Go [2026-07-17 20:24:32] step 453450/1000000 | loss=2.3056 vocab=0.8941 token=1.6830 | 6.16 steps/s | VRAM=31.18 Go [2026-07-17 20:24:41] step 453500/1000000 | loss=2.2421 vocab=1.3553 token=1.7111 | 6.15 steps/s | VRAM=31.51 Go [2026-07-17 20:24:49] step 453550/1000000 | loss=2.1763 vocab=0.8340 token=0.8798 | 6.15 steps/s | VRAM=32.78 Go [2026-07-17 20:24:57] step 453600/1000000 | loss=2.2504 vocab=0.8250 token=1.2798 | 6.10 steps/s | VRAM=29.24 Go [2026-07-17 20:25:05] step 453650/1000000 | loss=2.2179 vocab=1.0301 token=1.2116 | 6.15 steps/s | VRAM=31.79 Go [2026-07-17 20:25:13] step 453700/1000000 | loss=2.1749 vocab=0.5900 token=1.2441 | 6.14 steps/s | VRAM=32.51 Go [2026-07-17 20:25:21] step 453750/1000000 | loss=2.2614 vocab=0.9761 token=1.7868 | 6.10 steps/s | VRAM=28.75 Go [2026-07-17 20:25:29] step 453800/1000000 | loss=2.1058 vocab=0.9998 token=1.6320 | 6.13 steps/s | VRAM=29.40 Go [2026-07-17 20:25:38] step 453850/1000000 | loss=2.2892 vocab=0.9662 token=1.5536 | 6.15 steps/s | VRAM=31.50 Go [2026-07-17 20:25:46] step 453900/1000000 | loss=2.2256 vocab=1.2006 token=1.4206 | 6.12 steps/s | VRAM=26.99 Go [2026-07-17 20:25:54] step 453950/1000000 | loss=2.2753 vocab=0.9283 token=1.6451 | 6.11 steps/s | VRAM=29.99 Go [2026-07-17 20:26:02] step 454000/1000000 | loss=2.2509 vocab=1.0431 token=1.3661 | 6.15 steps/s | VRAM=32.63 Go [2026-07-17 20:26:10] step 454050/1000000 | loss=2.2482 vocab=0.7680 token=1.2440 | 6.14 steps/s | VRAM=28.63 Go [2026-07-17 20:26:18] step 454100/1000000 | loss=2.2130 vocab=0.9271 token=1.3679 | 6.11 steps/s | VRAM=25.86 Go [2026-07-17 20:26:27] step 454150/1000000 | loss=2.2086 vocab=0.8482 token=1.3578 | 6.13 steps/s | VRAM=28.87 Go [2026-07-17 20:26:35] step 454200/1000000 | loss=2.2358 vocab=1.2092 token=1.4744 | 6.14 steps/s | VRAM=32.22 Go [2026-07-17 20:26:43] step 454250/1000000 | loss=2.2090 vocab=1.2188 token=1.3572 | 6.14 steps/s | VRAM=27.90 Go [2026-07-17 20:26:51] step 454300/1000000 | loss=2.3252 vocab=1.1036 token=1.2354 | 6.16 steps/s | VRAM=25.75 Go [2026-07-17 20:26:59] step 454350/1000000 | loss=2.2958 vocab=1.1724 token=1.5525 | 6.12 steps/s | VRAM=31.35 Go [2026-07-17 20:27:07] step 454400/1000000 | loss=2.2380 vocab=0.6428 token=1.0821 | 6.14 steps/s | VRAM=28.45 Go [2026-07-17 20:27:15] step 454450/1000000 | loss=2.1889 vocab=1.0014 token=1.3928 | 6.13 steps/s | VRAM=25.49 Go [2026-07-17 20:27:24] step 454500/1000000 | loss=2.2882 vocab=0.8589 token=0.9277 | 6.16 steps/s | VRAM=31.36 Go [2026-07-17 20:27:32] step 454550/1000000 | loss=2.0797 vocab=0.5517 token=0.8359 | 6.11 steps/s | VRAM=29.16 Go [2026-07-17 20:27:40] step 454600/1000000 | loss=2.3180 vocab=1.1959 token=1.4270 | 6.11 steps/s | VRAM=26.03 Go [2026-07-17 20:27:48] step 454650/1000000 | loss=2.1637 vocab=1.1024 token=1.6868 | 6.13 steps/s | VRAM=31.91 Go [2026-07-17 20:27:56] step 454700/1000000 | loss=2.3176 vocab=1.2254 token=1.6510 | 6.13 steps/s | VRAM=29.02 Go [2026-07-17 20:28:04] step 454750/1000000 | loss=2.2615 vocab=0.9368 token=1.3935 | 6.13 steps/s | VRAM=31.90 Go [2026-07-17 20:28:13] step 454800/1000000 | loss=2.2176 vocab=0.9064 token=1.2569 | 6.13 steps/s | VRAM=32.12 Go [2026-07-17 20:28:21] step 454850/1000000 | loss=2.2583 vocab=1.0865 token=1.5912 | 6.13 steps/s | VRAM=25.34 Go [2026-07-17 20:28:29] step 454900/1000000 | loss=2.2079 vocab=1.6749 token=1.2780 | 6.15 steps/s | VRAM=29.22 Go [2026-07-17 20:28:37] step 454950/1000000 | loss=2.2263 vocab=0.6785 token=1.6392 | 6.13 steps/s | VRAM=28.66 Go [2026-07-17 20:28:45] step 455000/1000000 | loss=2.2441 vocab=1.0094 token=1.0907 | 6.14 steps/s | VRAM=32.79 Go [2026-07-17 20:28:53] step 455050/1000000 | loss=2.2075 vocab=1.3044 token=1.5016 | 6.12 steps/s | VRAM=25.72 Go [2026-07-17 20:29:01] step 455100/1000000 | loss=2.3362 vocab=0.9474 token=1.3309 | 6.14 steps/s | VRAM=26.15 Go [2026-07-17 20:29:10] step 455150/1000000 | loss=2.1482 vocab=1.2136 token=1.5776 | 6.17 steps/s | VRAM=28.48 Go [2026-07-17 20:29:18] step 455200/1000000 | loss=2.2517 vocab=1.2245 token=1.5745 | 6.14 steps/s | VRAM=28.96 Go [2026-07-17 20:29:26] step 455250/1000000 | loss=2.2428 vocab=0.8901 token=1.1728 | 6.12 steps/s | VRAM=25.95 Go [2026-07-17 20:29:34] step 455300/1000000 | loss=2.3895 vocab=1.1076 token=1.5241 | 6.16 steps/s | VRAM=28.06 Go [2026-07-17 20:29:42] step 455350/1000000 | loss=2.2890 vocab=1.0799 token=1.4797 | 6.11 steps/s | VRAM=26.43 Go [2026-07-17 20:29:50] step 455400/1000000 | loss=2.4042 vocab=0.9138 token=1.3421 | 6.12 steps/s | VRAM=26.08 Go [2026-07-17 20:29:58] step 455450/1000000 | loss=2.1576 vocab=0.4473 token=0.9322 | 6.13 steps/s | VRAM=29.14 Go [2026-07-17 20:30:07] step 455500/1000000 | loss=2.2126 vocab=1.2991 token=1.6676 | 6.13 steps/s | VRAM=32.78 Go [2026-07-17 20:30:15] step 455550/1000000 | loss=2.2143 vocab=1.2899 token=1.0449 | 6.12 steps/s | VRAM=28.45 Go [2026-07-17 20:30:23] step 455600/1000000 | loss=2.2254 vocab=0.8926 token=1.8153 | 6.15 steps/s | VRAM=32.03 Go [2026-07-17 20:30:31] step 455650/1000000 | loss=2.2870 vocab=1.0052 token=1.3825 | 6.14 steps/s | VRAM=29.60 Go [2026-07-17 20:30:39] step 455700/1000000 | loss=2.2866 vocab=0.6861 token=1.3185 | 6.12 steps/s | VRAM=28.17 Go [2026-07-17 20:30:47] step 455750/1000000 | loss=2.2728 vocab=0.6202 token=0.8984 | 6.13 steps/s | VRAM=26.56 Go [2026-07-17 20:30:56] step 455800/1000000 | loss=2.2536 vocab=0.3971 token=1.0041 | 6.18 steps/s | VRAM=31.80 Go [2026-07-17 20:31:04] step 455850/1000000 | loss=2.1163 vocab=0.7795 token=1.3775 | 6.15 steps/s | VRAM=32.59 Go [2026-07-17 20:31:12] step 455900/1000000 | loss=2.1735 vocab=0.8159 token=1.6303 | 6.15 steps/s | VRAM=32.36 Go [2026-07-17 20:31:20] step 455950/1000000 | loss=2.2057 vocab=0.5814 token=1.2906 | 6.12 steps/s | VRAM=26.23 Go [2026-07-17 20:31:28] step 456000/1000000 | loss=2.2423 vocab=1.0617 token=1.7348 | 6.16 steps/s | VRAM=29.07 Go [2026-07-17 20:31:36] step 456050/1000000 | loss=2.2438 vocab=0.6137 token=1.1138 | 6.14 steps/s | VRAM=29.36 Go [2026-07-17 20:31:44] step 456100/1000000 | loss=2.1906 vocab=0.9100 token=1.3189 | 6.15 steps/s | VRAM=27.93 Go [2026-07-17 20:31:52] step 456150/1000000 | loss=2.1351 vocab=0.9211 token=1.3453 | 6.16 steps/s | VRAM=25.79 Go [2026-07-17 20:32:01] step 456200/1000000 | loss=2.3218 vocab=0.5565 token=1.3472 | 6.14 steps/s | VRAM=28.70 Go [2026-07-17 20:32:09] step 456250/1000000 | loss=2.3213 vocab=0.4082 token=1.0494 | 6.14 steps/s | VRAM=32.83 Go [2026-07-17 20:32:17] step 456300/1000000 | loss=2.2623 vocab=0.9021 token=1.4471 | 6.16 steps/s | VRAM=32.50 Go [2026-07-17 20:32:25] step 456350/1000000 | loss=2.2627 vocab=0.6327 token=1.5183 | 6.14 steps/s | VRAM=27.52 Go [2026-07-17 20:32:33] step 456400/1000000 | loss=2.2463 vocab=0.8299 token=0.9211 | 6.17 steps/s | VRAM=31.63 Go [2026-07-17 20:32:41] step 456450/1000000 | loss=2.1563 vocab=1.0642 token=1.2499 | 6.14 steps/s | VRAM=32.06 Go [2026-07-17 20:32:49] step 456500/1000000 | loss=2.2943 vocab=1.2447 token=1.6649 | 6.16 steps/s | VRAM=29.12 Go [2026-07-17 20:32:58] step 456550/1000000 | loss=2.2754 vocab=1.3804 token=1.6413 | 6.16 steps/s | VRAM=28.81 Go [2026-07-17 20:33:06] step 456600/1000000 | loss=2.2502 vocab=0.8211 token=1.2700 | 6.14 steps/s | VRAM=28.15 Go [2026-07-17 20:33:14] step 456650/1000000 | loss=2.2452 vocab=0.5878 token=1.3911 | 6.13 steps/s | VRAM=26.93 Go [2026-07-17 20:33:22] step 456700/1000000 | loss=2.2525 vocab=0.5725 token=0.8875 | 6.15 steps/s | VRAM=29.04 Go [2026-07-17 20:33:30] step 456750/1000000 | loss=2.1727 vocab=0.9951 token=1.4273 | 6.15 steps/s | VRAM=26.63 Go [2026-07-17 20:33:38] step 456800/1000000 | loss=2.2532 vocab=0.9209 token=1.4046 | 6.15 steps/s | VRAM=27.43 Go [2026-07-17 20:33:46] step 456850/1000000 | loss=2.2218 vocab=0.7564 token=1.1703 | 6.13 steps/s | VRAM=25.93 Go [2026-07-17 20:33:54] step 456900/1000000 | loss=2.2604 vocab=0.3899 token=1.1058 | 6.15 steps/s | VRAM=25.76 Go [2026-07-17 20:34:03] step 456950/1000000 | loss=2.2165 vocab=0.6734 token=1.4187 | 6.12 steps/s | VRAM=26.96 Go [2026-07-17 20:34:11] step 457000/1000000 | loss=2.2589 vocab=1.0759 token=1.6240 | 6.19 steps/s | VRAM=31.39 Go [2026-07-17 20:34:19] step 457050/1000000 | loss=2.3286 vocab=1.3644 token=1.7115 | 6.12 steps/s | VRAM=29.99 Go [2026-07-17 20:34:27] step 457100/1000000 | loss=2.2439 vocab=1.0661 token=1.1475 | 6.14 steps/s | VRAM=23.34 Go [2026-07-17 20:34:35] step 457150/1000000 | loss=2.3329 vocab=0.7641 token=1.3379 | 6.17 steps/s | VRAM=31.98 Go [2026-07-17 20:34:43] step 457200/1000000 | loss=2.1692 vocab=1.1428 token=1.1150 | 6.15 steps/s | VRAM=27.70 Go [2026-07-17 20:34:51] step 457250/1000000 | loss=2.2559 vocab=0.8052 token=1.3967 | 6.13 steps/s | VRAM=26.04 Go [2026-07-17 20:35:00] step 457300/1000000 | loss=2.2152 vocab=1.3145 token=1.2717 | 6.14 steps/s | VRAM=28.12 Go [2026-07-17 20:35:08] step 457350/1000000 | loss=2.1350 vocab=0.7416 token=1.4328 | 6.13 steps/s | VRAM=28.13 Go [2026-07-17 20:35:16] step 457400/1000000 | loss=2.2144 vocab=0.7377 token=1.4910 | 6.14 steps/s | VRAM=26.26 Go [2026-07-17 20:35:24] step 457450/1000000 | loss=2.3099 vocab=0.8367 token=0.8756 | 6.17 steps/s | VRAM=32.04 Go [2026-07-17 20:35:32] step 457500/1000000 | loss=2.1918 vocab=1.1720 token=1.6838 | 6.13 steps/s | VRAM=24.47 Go [2026-07-17 20:35:40] step 457550/1000000 | loss=2.1447 vocab=0.6476 token=1.3935 | 6.15 steps/s | VRAM=32.39 Go [2026-07-17 20:35:48] step 457600/1000000 | loss=2.2457 vocab=1.0363 token=1.4000 | 6.10 steps/s | VRAM=28.66 Go [2026-07-17 20:35:57] step 457650/1000000 | loss=2.3334 vocab=0.7669 token=1.1942 | 6.12 steps/s | VRAM=28.76 Go [2026-07-17 20:36:05] step 457700/1000000 | loss=2.3066 vocab=0.7301 token=1.3559 | 6.13 steps/s | VRAM=26.10 Go [2026-07-17 20:36:13] step 457750/1000000 | loss=2.1137 vocab=0.7592 token=1.1421 | 6.13 steps/s | VRAM=28.93 Go [2026-07-17 20:36:21] step 457800/1000000 | loss=2.2888 vocab=1.6752 token=1.4548 | 6.13 steps/s | VRAM=30.16 Go [2026-07-17 20:36:29] step 457850/1000000 | loss=2.2292 vocab=0.9041 token=1.3936 | 6.12 steps/s | VRAM=28.38 Go [2026-07-17 20:36:37] step 457900/1000000 | loss=2.2485 vocab=1.0187 token=1.4239 | 6.10 steps/s | VRAM=26.10 Go [2026-07-17 20:36:46] step 457950/1000000 | loss=2.1705 vocab=0.5972 token=1.3841 | 6.14 steps/s | VRAM=26.32 Go [2026-07-17 20:36:54] step 458000/1000000 | loss=2.3304 vocab=0.8979 token=1.0625 | 6.08 steps/s | VRAM=28.80 Go [2026-07-17 20:37:02] step 458050/1000000 | loss=2.1955 vocab=1.2814 token=1.4300 | 6.15 steps/s | VRAM=31.76 Go [2026-07-17 20:37:10] step 458100/1000000 | loss=2.2455 vocab=0.8011 token=1.5101 | 6.12 steps/s | VRAM=25.92 Go [2026-07-17 20:37:18] step 458150/1000000 | loss=2.1993 vocab=0.6852 token=1.0577 | 6.15 steps/s | VRAM=25.88 Go [2026-07-17 20:37:26] step 458200/1000000 | loss=2.3483 vocab=0.7727 token=1.3993 | 6.10 steps/s | VRAM=26.90 Go [2026-07-17 20:37:35] step 458250/1000000 | loss=2.2860 vocab=1.1473 token=1.3613 | 6.12 steps/s | VRAM=26.67 Go [2026-07-17 20:37:43] step 458300/1000000 | loss=2.1619 vocab=0.8934 token=1.3060 | 6.18 steps/s | VRAM=30.17 Go [2026-07-17 20:37:51] step 458350/1000000 | loss=2.2915 vocab=0.7060 token=1.4218 | 6.14 steps/s | VRAM=29.06 Go [2026-07-17 20:37:59] step 458400/1000000 | loss=2.2217 vocab=0.7316 token=1.4889 | 6.11 steps/s | VRAM=29.73 Go [2026-07-17 20:38:07] step 458450/1000000 | loss=2.1886 vocab=1.1892 token=1.4369 | 6.12 steps/s | VRAM=32.88 Go [2026-07-17 20:38:15] step 458500/1000000 | loss=2.2204 vocab=0.8755 token=1.3053 | 6.14 steps/s | VRAM=29.21 Go [2026-07-17 20:38:24] step 458550/1000000 | loss=2.1809 vocab=1.2966 token=1.1665 | 6.13 steps/s | VRAM=27.80 Go [2026-07-17 20:38:32] step 458600/1000000 | loss=2.2412 vocab=0.7801 token=1.6751 | 6.11 steps/s | VRAM=28.82 Go [2026-07-17 20:38:40] step 458650/1000000 | loss=2.1611 vocab=0.5090 token=1.2613 | 6.11 steps/s | VRAM=28.94 Go [2026-07-17 20:38:48] step 458700/1000000 | loss=2.3399 vocab=1.1516 token=1.2663 | 6.14 steps/s | VRAM=29.21 Go [2026-07-17 20:38:56] step 458750/1000000 | loss=2.2574 vocab=0.7628 token=1.7752 | 6.16 steps/s | VRAM=29.30 Go [2026-07-17 20:39:04] step 458800/1000000 | loss=2.1675 vocab=0.7244 token=1.1935 | 6.17 steps/s | VRAM=27.33 Go [2026-07-17 20:39:12] step 458850/1000000 | loss=2.2440 vocab=0.6930 token=1.0886 | 6.12 steps/s | VRAM=26.25 Go [2026-07-17 20:39:21] step 458900/1000000 | loss=2.3184 vocab=0.6231 token=1.0281 | 6.13 steps/s | VRAM=32.72 Go [2026-07-17 20:39:29] step 458950/1000000 | loss=2.2604 vocab=0.4990 token=1.1339 | 6.13 steps/s | VRAM=28.15 Go [2026-07-17 20:39:37] step 459000/1000000 | loss=2.1974 vocab=0.7756 token=0.7756 | 6.11 steps/s | VRAM=26.33 Go [2026-07-17 20:39:45] step 459050/1000000 | loss=2.2249 vocab=1.0377 token=1.0686 | 6.12 steps/s | VRAM=28.26 Go [2026-07-17 20:39:53] step 459100/1000000 | loss=2.3186 vocab=0.9765 token=1.2794 | 6.10 steps/s | VRAM=30.61 Go [2026-07-17 20:40:01] step 459150/1000000 | loss=2.1780 vocab=0.6302 token=1.0389 | 6.13 steps/s | VRAM=28.88 Go [2026-07-17 20:40:10] step 459200/1000000 | loss=2.2939 vocab=0.8004 token=1.6006 | 6.13 steps/s | VRAM=26.44 Go [2026-07-17 20:40:18] step 459250/1000000 | loss=2.1746 vocab=1.4161 token=1.1059 | 6.15 steps/s | VRAM=31.90 Go [2026-07-17 20:40:26] step 459300/1000000 | loss=2.1131 vocab=0.5424 token=1.2889 | 6.14 steps/s | VRAM=26.29 Go [2026-07-17 20:40:34] step 459350/1000000 | loss=2.2556 vocab=0.9822 token=0.9708 | 6.11 steps/s | VRAM=29.19 Go [2026-07-17 20:40:42] step 459400/1000000 | loss=2.2078 vocab=0.6639 token=0.8884 | 6.15 steps/s | VRAM=28.47 Go [2026-07-17 20:40:50] step 459450/1000000 | loss=2.2793 vocab=0.5778 token=1.0244 | 6.13 steps/s | VRAM=27.84 Go [2026-07-17 20:40:59] step 459500/1000000 | loss=2.2456 vocab=0.7073 token=1.2646 | 6.14 steps/s | VRAM=27.56 Go [2026-07-17 20:41:07] step 459550/1000000 | loss=2.2246 vocab=0.6688 token=1.1492 | 6.14 steps/s | VRAM=32.30 Go [2026-07-17 20:41:15] step 459600/1000000 | loss=2.2807 vocab=1.0540 token=1.3169 | 6.12 steps/s | VRAM=26.19 Go [2026-07-17 20:41:23] step 459650/1000000 | loss=2.1607 vocab=1.4235 token=1.6503 | 6.13 steps/s | VRAM=32.56 Go [2026-07-17 20:41:31] step 459700/1000000 | loss=2.2482 vocab=1.1712 token=1.2051 | 6.16 steps/s | VRAM=31.60 Go [2026-07-17 20:41:39] step 459750/1000000 | loss=2.2092 vocab=0.6179 token=1.7205 | 6.12 steps/s | VRAM=26.04 Go [2026-07-17 20:41:47] step 459800/1000000 | loss=2.2610 vocab=0.8451 token=1.3018 | 6.15 steps/s | VRAM=32.40 Go [2026-07-17 20:41:56] step 459850/1000000 | loss=2.2482 vocab=1.0183 token=1.7067 | 6.14 steps/s | VRAM=25.88 Go [2026-07-17 20:42:04] step 459900/1000000 | loss=2.3004 vocab=1.2535 token=1.9825 | 6.15 steps/s | VRAM=29.47 Go [2026-07-17 20:42:12] step 459950/1000000 | loss=2.2124 vocab=0.6806 token=0.8798 | 6.15 steps/s | VRAM=31.46 Go [2026-07-17 20:42:20] step 460000/1000000 | loss=2.2401 vocab=0.5554 token=1.0585 | 6.15 steps/s | VRAM=25.80 Go [2026-07-17 20:42:28] step 460050/1000000 | loss=2.2587 vocab=1.4413 token=1.4964 | 6.17 steps/s | VRAM=32.78 Go [2026-07-17 20:42:36] step 460100/1000000 | loss=2.2387 vocab=1.0115 token=1.4460 | 6.13 steps/s | VRAM=27.67 Go [2026-07-17 20:42:44] step 460150/1000000 | loss=2.1987 vocab=1.1043 token=1.5052 | 6.14 steps/s | VRAM=26.28 Go [2026-07-17 20:42:53] step 460200/1000000 | loss=2.2813 vocab=0.6187 token=1.4135 | 6.15 steps/s | VRAM=32.75 Go [2026-07-17 20:43:01] step 460250/1000000 | loss=2.3407 vocab=0.7529 token=1.3559 | 6.15 steps/s | VRAM=29.03 Go [2026-07-17 20:43:09] step 460300/1000000 | loss=2.2780 vocab=1.5769 token=1.5381 | 6.13 steps/s | VRAM=28.95 Go [2026-07-17 20:43:17] step 460350/1000000 | loss=2.2227 vocab=0.6468 token=1.3892 | 6.12 steps/s | VRAM=29.05 Go [2026-07-17 20:43:25] step 460400/1000000 | loss=2.1990 vocab=0.8388 token=1.2272 | 6.14 steps/s | VRAM=28.84 Go [2026-07-17 20:43:33] step 460450/1000000 | loss=2.1940 vocab=0.6582 token=0.9087 | 6.14 steps/s | VRAM=26.82 Go [2026-07-17 20:43:41] step 460500/1000000 | loss=2.3151 vocab=0.4610 token=0.8596 | 6.13 steps/s | VRAM=28.63 Go [2026-07-17 20:43:50] step 460550/1000000 | loss=2.3251 vocab=0.7150 token=1.7954 | 6.16 steps/s | VRAM=29.35 Go [2026-07-17 20:43:58] step 460600/1000000 | loss=2.3174 vocab=0.8972 token=1.5239 | 6.14 steps/s | VRAM=29.65 Go [2026-07-17 20:44:06] step 460650/1000000 | loss=2.2836 vocab=0.9377 token=1.3469 | 6.13 steps/s | VRAM=29.64 Go [2026-07-17 20:44:14] step 460700/1000000 | loss=2.2320 vocab=1.3843 token=1.2820 | 6.14 steps/s | VRAM=28.05 Go [2026-07-17 20:44:22] step 460750/1000000 | loss=2.2316 vocab=0.5703 token=1.1440 | 6.15 steps/s | VRAM=28.02 Go [2026-07-17 20:44:30] step 460800/1000000 | loss=2.2590 vocab=0.9766 token=1.2081 | 6.10 steps/s | VRAM=30.03 Go [2026-07-17 20:44:38] step 460850/1000000 | loss=2.1999 vocab=1.4324 token=1.6762 | 6.16 steps/s | VRAM=32.13 Go [2026-07-17 20:44:47] step 460900/1000000 | loss=2.2735 vocab=0.9776 token=1.0623 | 6.14 steps/s | VRAM=32.79 Go [2026-07-17 20:44:55] step 460950/1000000 | loss=2.1698 vocab=0.9804 token=1.0686 | 6.10 steps/s | VRAM=28.11 Go [2026-07-17 20:45:03] step 461000/1000000 | loss=2.2376 vocab=1.2030 token=1.3623 | 6.15 steps/s | VRAM=27.86 Go [2026-07-17 20:45:11] step 461050/1000000 | loss=2.2514 vocab=0.6515 token=1.2741 | 6.14 steps/s | VRAM=25.51 Go [2026-07-17 20:45:19] step 461100/1000000 | loss=2.2096 vocab=0.6437 token=1.3168 | 6.16 steps/s | VRAM=27.90 Go [2026-07-17 20:45:27] step 461150/1000000 | loss=2.3099 vocab=1.0574 token=1.6472 | 6.14 steps/s | VRAM=32.35 Go [2026-07-17 20:45:35] step 461200/1000000 | loss=2.2704 vocab=0.8856 token=1.4140 | 6.14 steps/s | VRAM=29.74 Go [2026-07-17 20:45:44] step 461250/1000000 | loss=2.1732 vocab=1.1688 token=0.9996 | 6.15 steps/s | VRAM=29.16 Go [2026-07-17 20:45:52] step 461300/1000000 | loss=2.3015 vocab=0.8248 token=1.5149 | 6.15 steps/s | VRAM=31.86 Go [2026-07-17 20:46:00] step 461350/1000000 | loss=2.1973 vocab=0.9298 token=1.4475 | 6.10 steps/s | VRAM=26.66 Go [2026-07-17 20:46:08] step 461400/1000000 | loss=2.2552 vocab=0.9339 token=1.5287 | 6.15 steps/s | VRAM=29.21 Go [2026-07-17 20:46:16] step 461450/1000000 | loss=2.1573 vocab=0.5433 token=1.1774 | 6.18 steps/s | VRAM=31.25 Go [2026-07-17 20:46:24] step 461500/1000000 | loss=2.2055 vocab=1.0465 token=1.5252 | 6.16 steps/s | VRAM=31.98 Go [2026-07-17 20:46:32] step 461550/1000000 | loss=2.2290 vocab=0.8257 token=1.1164 | 6.12 steps/s | VRAM=28.34 Go [2026-07-17 20:46:40] step 461600/1000000 | loss=2.1821 vocab=1.3132 token=1.6941 | 6.17 steps/s | VRAM=24.96 Go [2026-07-17 20:46:49] step 461650/1000000 | loss=2.2689 vocab=0.6226 token=1.0136 | 6.16 steps/s | VRAM=28.46 Go [2026-07-17 20:46:57] step 461700/1000000 | loss=2.3937 vocab=0.4618 token=0.8131 | 6.12 steps/s | VRAM=26.45 Go [2026-07-17 20:47:05] step 461750/1000000 | loss=2.2600 vocab=1.0954 token=1.4698 | 6.17 steps/s | VRAM=32.58 Go [2026-07-17 20:47:13] step 461800/1000000 | loss=2.2901 vocab=1.1615 token=1.4846 | 6.13 steps/s | VRAM=29.03 Go [2026-07-17 20:47:21] step 461850/1000000 | loss=2.1299 vocab=0.7728 token=1.2187 | 6.16 steps/s | VRAM=31.11 Go [2026-07-17 20:47:29] step 461900/1000000 | loss=2.1358 vocab=0.6641 token=1.3976 | 6.15 steps/s | VRAM=26.23 Go [2026-07-17 20:47:37] step 461950/1000000 | loss=2.2718 vocab=0.8824 token=1.2774 | 6.15 steps/s | VRAM=32.22 Go [2026-07-17 20:47:46] step 462000/1000000 | loss=2.3543 vocab=0.8532 token=1.2253 | 6.13 steps/s | VRAM=32.64 Go [2026-07-17 20:47:54] step 462050/1000000 | loss=2.2718 vocab=0.9091 token=1.2174 | 6.14 steps/s | VRAM=31.58 Go [2026-07-17 20:48:02] step 462100/1000000 | loss=2.2556 vocab=1.1755 token=1.3175 | 6.16 steps/s | VRAM=25.74 Go [2026-07-17 20:48:10] step 462150/1000000 | loss=2.2595 vocab=1.2489 token=1.7424 | 6.14 steps/s | VRAM=29.22 Go [2026-07-17 20:48:18] step 462200/1000000 | loss=2.1943 vocab=0.9479 token=1.3187 | 6.16 steps/s | VRAM=28.88 Go [2026-07-17 20:48:26] step 462250/1000000 | loss=2.3054 vocab=0.9374 token=1.5879 | 6.18 steps/s | VRAM=27.90 Go [2026-07-17 20:48:34] step 462300/1000000 | loss=2.0729 vocab=0.8761 token=1.1155 | 6.17 steps/s | VRAM=32.66 Go [2026-07-17 20:48:42] step 462350/1000000 | loss=2.3869 vocab=0.7706 token=1.2716 | 6.13 steps/s | VRAM=30.69 Go [2026-07-17 20:48:51] step 462400/1000000 | loss=2.2768 vocab=0.6945 token=1.0667 | 6.16 steps/s | VRAM=28.57 Go [2026-07-17 20:48:59] step 462450/1000000 | loss=2.3005 vocab=0.9532 token=1.0906 | 6.17 steps/s | VRAM=26.12 Go [2026-07-17 20:49:07] step 462500/1000000 | loss=2.3151 vocab=0.9208 token=1.3780 | 6.13 steps/s | VRAM=28.97 Go [2026-07-17 20:49:15] step 462550/1000000 | loss=2.2915 vocab=1.1528 token=1.7586 | 6.14 steps/s | VRAM=29.16 Go [2026-07-17 20:49:23] step 462600/1000000 | loss=2.2476 vocab=0.8861 token=1.2713 | 6.14 steps/s | VRAM=32.19 Go [2026-07-17 20:49:31] step 462650/1000000 | loss=2.3560 vocab=0.8467 token=1.1639 | 6.16 steps/s | VRAM=28.08 Go [2026-07-17 20:49:39] step 462700/1000000 | loss=2.1487 vocab=0.6452 token=1.3903 | 6.13 steps/s | VRAM=25.34 Go [2026-07-17 20:49:48] step 462750/1000000 | loss=2.2001 vocab=0.8558 token=1.1723 | 6.13 steps/s | VRAM=27.99 Go [2026-07-17 20:49:56] step 462800/1000000 | loss=2.1707 vocab=0.6141 token=1.0366 | 6.15 steps/s | VRAM=32.34 Go [2026-07-17 20:50:04] step 462850/1000000 | loss=2.3556 vocab=0.8274 token=1.4566 | 6.13 steps/s | VRAM=29.34 Go [2026-07-17 20:50:12] step 462900/1000000 | loss=2.2320 vocab=0.9462 token=1.2604 | 6.15 steps/s | VRAM=29.73 Go [2026-07-17 20:50:20] step 462950/1000000 | loss=2.2166 vocab=0.5632 token=1.1522 | 6.17 steps/s | VRAM=31.14 Go [2026-07-17 20:50:28] step 463000/1000000 | loss=2.3824 vocab=1.2619 token=1.3974 | 6.15 steps/s | VRAM=31.68 Go [2026-07-17 20:50:36] step 463050/1000000 | loss=2.2882 vocab=1.1266 token=1.3672 | 6.14 steps/s | VRAM=31.70 Go [2026-07-17 20:50:45] step 463100/1000000 | loss=2.1582 vocab=0.7130 token=1.2806 | 6.13 steps/s | VRAM=29.67 Go [2026-07-17 20:50:53] step 463150/1000000 | loss=2.3279 vocab=1.1341 token=1.7487 | 6.18 steps/s | VRAM=28.27 Go [2026-07-17 20:51:01] step 463200/1000000 | loss=2.1900 vocab=1.2072 token=1.2155 | 6.17 steps/s | VRAM=28.32 Go [2026-07-17 20:51:09] step 463250/1000000 | loss=2.1359 vocab=0.8589 token=1.5054 | 6.14 steps/s | VRAM=30.09 Go [2026-07-17 20:51:17] step 463300/1000000 | loss=2.3243 vocab=1.2491 token=1.8186 | 6.12 steps/s | VRAM=28.43 Go [2026-07-17 20:51:25] step 463350/1000000 | loss=2.2220 vocab=0.6987 token=0.8543 | 6.16 steps/s | VRAM=31.42 Go [2026-07-17 20:51:33] step 463400/1000000 | loss=2.3077 vocab=1.3298 token=1.1850 | 6.15 steps/s | VRAM=26.10 Go [2026-07-17 20:51:41] step 463450/1000000 | loss=2.2247 vocab=1.2225 token=1.4938 | 6.13 steps/s | VRAM=31.72 Go [2026-07-17 20:51:50] step 463500/1000000 | loss=2.2971 vocab=0.8219 token=1.4951 | 6.15 steps/s | VRAM=28.58 Go [2026-07-17 20:51:58] step 463550/1000000 | loss=2.2622 vocab=0.6321 token=1.2991 | 6.16 steps/s | VRAM=26.05 Go [2026-07-17 20:52:06] step 463600/1000000 | loss=2.2504 vocab=0.5802 token=1.2278 | 6.13 steps/s | VRAM=29.19 Go [2026-07-17 20:52:14] step 463650/1000000 | loss=2.1800 vocab=1.2445 token=1.8349 | 6.13 steps/s | VRAM=31.84 Go [2026-07-17 20:52:22] step 463700/1000000 | loss=2.3019 vocab=1.0665 token=1.8158 | 6.14 steps/s | VRAM=28.67 Go [2026-07-17 20:52:30] step 463750/1000000 | loss=2.2581 vocab=0.5086 token=1.1250 | 6.12 steps/s | VRAM=29.72 Go [2026-07-17 20:52:38] step 463800/1000000 | loss=2.1985 vocab=0.7275 token=1.3865 | 6.16 steps/s | VRAM=27.97 Go [2026-07-17 20:52:47] step 463850/1000000 | loss=2.2077 vocab=0.8173 token=1.4637 | 6.15 steps/s | VRAM=31.97 Go [2026-07-17 20:52:55] step 463900/1000000 | loss=2.1852 vocab=0.7823 token=1.4873 | 6.15 steps/s | VRAM=28.36 Go [2026-07-17 20:53:03] step 463950/1000000 | loss=2.3444 vocab=1.1212 token=1.3673 | 6.11 steps/s | VRAM=24.90 Go [2026-07-17 20:53:11] step 464000/1000000 | loss=2.2360 vocab=1.0009 token=1.5545 | 6.16 steps/s | VRAM=32.69 Go [2026-07-17 20:53:19] step 464050/1000000 | loss=2.2146 vocab=1.2449 token=1.5380 | 6.16 steps/s | VRAM=28.52 Go [2026-07-17 20:53:27] step 464100/1000000 | loss=2.2995 vocab=0.7048 token=1.0508 | 6.14 steps/s | VRAM=24.96 Go [2026-07-17 20:53:35] step 464150/1000000 | loss=2.2185 vocab=0.6795 token=1.4251 | 6.12 steps/s | VRAM=27.71 Go [2026-07-17 20:53:44] step 464200/1000000 | loss=2.3668 vocab=0.7486 token=1.5901 | 6.18 steps/s | VRAM=32.05 Go [2026-07-17 20:53:52] step 464250/1000000 | loss=2.2113 vocab=0.8394 token=1.1107 | 6.15 steps/s | VRAM=25.26 Go [2026-07-17 20:54:00] step 464300/1000000 | loss=2.2147 vocab=1.1046 token=1.4315 | 6.14 steps/s | VRAM=29.24 Go [2026-07-17 20:54:08] step 464350/1000000 | loss=2.1946 vocab=1.1824 token=1.5401 | 6.14 steps/s | VRAM=32.82 Go [2026-07-17 20:54:16] step 464400/1000000 | loss=2.1663 vocab=0.7187 token=1.4582 | 6.16 steps/s | VRAM=25.49 Go [2026-07-17 20:54:24] step 464450/1000000 | loss=2.2371 vocab=0.7598 token=1.0788 | 6.14 steps/s | VRAM=27.65 Go [2026-07-17 20:54:32] step 464500/1000000 | loss=2.2233 vocab=0.9419 token=1.2158 | 6.11 steps/s | VRAM=32.80 Go [2026-07-17 20:54:40] step 464550/1000000 | loss=2.2219 vocab=0.7283 token=1.4531 | 6.18 steps/s | VRAM=28.17 Go [2026-07-17 20:54:49] step 464600/1000000 | loss=2.3775 vocab=1.1359 token=1.6419 | 6.09 steps/s | VRAM=26.41 Go [2026-07-17 20:54:57] step 464650/1000000 | loss=2.2738 vocab=0.8269 token=1.4500 | 6.18 steps/s | VRAM=28.18 Go [2026-07-17 20:55:05] step 464700/1000000 | loss=2.1226 vocab=1.2865 token=1.2745 | 6.16 steps/s | VRAM=23.04 Go [2026-07-17 20:55:13] step 464750/1000000 | loss=2.2768 vocab=1.2925 token=1.5456 | 6.17 steps/s | VRAM=31.51 Go [2026-07-17 20:55:21] step 464800/1000000 | loss=2.2025 vocab=1.0560 token=1.0342 | 6.13 steps/s | VRAM=24.65 Go [2026-07-17 20:55:29] step 464850/1000000 | loss=2.3664 vocab=1.1937 token=1.4757 | 6.12 steps/s | VRAM=32.71 Go [2026-07-17 20:55:37] step 464900/1000000 | loss=2.1934 vocab=0.6542 token=1.4459 | 6.12 steps/s | VRAM=27.17 Go [2026-07-17 20:55:46] step 464950/1000000 | loss=2.2876 vocab=0.5832 token=1.2325 | 6.15 steps/s | VRAM=29.76 Go [2026-07-17 20:55:54] step 465000/1000000 | loss=2.3109 vocab=0.6548 token=1.4993 | 6.16 steps/s | VRAM=27.60 Go [2026-07-17 20:56:02] step 465050/1000000 | loss=2.2411 vocab=0.8044 token=1.1546 | 6.17 steps/s | VRAM=29.45 Go [2026-07-17 20:56:10] step 465100/1000000 | loss=2.1890 vocab=0.6816 token=1.0057 | 6.12 steps/s | VRAM=29.12 Go [2026-07-17 20:56:18] step 465150/1000000 | loss=2.1971 vocab=1.0156 token=1.4644 | 6.14 steps/s | VRAM=32.50 Go [2026-07-17 20:56:26] step 465200/1000000 | loss=2.2215 vocab=0.8231 token=1.0372 | 6.13 steps/s | VRAM=32.49 Go [2026-07-17 20:56:35] step 465250/1000000 | loss=2.1776 vocab=0.8751 token=1.2542 | 6.10 steps/s | VRAM=28.76 Go [2026-07-17 20:56:43] step 465300/1000000 | loss=2.2082 vocab=0.7705 token=1.2379 | 6.13 steps/s | VRAM=31.22 Go [2026-07-17 20:56:51] step 465350/1000000 | loss=2.3012 vocab=0.9519 token=1.4447 | 6.12 steps/s | VRAM=28.08 Go [2026-07-17 20:56:59] step 465400/1000000 | loss=2.1092 vocab=0.8909 token=1.5964 | 6.11 steps/s | VRAM=26.10 Go [2026-07-17 20:57:07] step 465450/1000000 | loss=2.2684 vocab=0.5413 token=1.1591 | 6.15 steps/s | VRAM=32.96 Go [2026-07-17 20:57:15] step 465500/1000000 | loss=2.2064 vocab=0.6970 token=1.2759 | 6.11 steps/s | VRAM=26.52 Go [2026-07-17 20:57:23] step 465550/1000000 | loss=2.3635 vocab=0.9268 token=1.2752 | 6.15 steps/s | VRAM=32.45 Go [2026-07-17 20:57:32] step 465600/1000000 | loss=2.1987 vocab=0.8096 token=1.2942 | 6.12 steps/s | VRAM=28.32 Go [2026-07-17 20:57:40] step 465650/1000000 | loss=2.1951 vocab=0.7448 token=1.3239 | 6.11 steps/s | VRAM=29.06 Go [2026-07-17 20:57:48] step 465700/1000000 | loss=2.1149 vocab=0.5752 token=0.8741 | 6.14 steps/s | VRAM=28.23 Go [2026-07-17 20:57:56] step 465750/1000000 | loss=2.1421 vocab=1.0891 token=1.3105 | 6.13 steps/s | VRAM=26.55 Go [2026-07-17 20:58:04] step 465800/1000000 | loss=2.1945 vocab=1.1363 token=1.6458 | 6.15 steps/s | VRAM=28.39 Go [2026-07-17 20:58:12] step 465850/1000000 | loss=2.2205 vocab=0.7465 token=1.1648 | 6.15 steps/s | VRAM=26.11 Go [2026-07-17 20:58:21] step 465900/1000000 | loss=2.2881 vocab=0.4032 token=1.3734 | 6.15 steps/s | VRAM=26.04 Go [2026-07-17 20:58:29] step 465950/1000000 | loss=2.2036 vocab=1.1794 token=1.5064 | 6.13 steps/s | VRAM=28.59 Go [2026-07-17 20:58:37] step 466000/1000000 | loss=2.3083 vocab=1.0060 token=1.4799 | 6.13 steps/s | VRAM=25.32 Go [2026-07-17 20:58:45] step 466050/1000000 | loss=2.1809 vocab=1.1347 token=1.2955 | 6.12 steps/s | VRAM=31.96 Go [2026-07-17 20:58:53] step 466100/1000000 | loss=2.2364 vocab=0.7114 token=1.2937 | 6.13 steps/s | VRAM=28.87 Go [2026-07-17 20:59:01] step 466150/1000000 | loss=2.2535 vocab=0.9139 token=1.2005 | 6.15 steps/s | VRAM=31.30 Go [2026-07-17 20:59:09] step 466200/1000000 | loss=2.1888 vocab=0.8512 token=1.2761 | 6.12 steps/s | VRAM=28.99 Go [2026-07-17 20:59:18] step 466250/1000000 | loss=2.1768 vocab=0.8607 token=1.1273 | 6.12 steps/s | VRAM=26.59 Go [2026-07-17 20:59:26] step 466300/1000000 | loss=2.2422 vocab=0.5346 token=1.1694 | 6.15 steps/s | VRAM=31.67 Go [2026-07-17 20:59:34] step 466350/1000000 | loss=2.1211 vocab=0.9418 token=1.7625 | 6.13 steps/s | VRAM=28.09 Go [2026-07-17 20:59:42] step 466400/1000000 | loss=2.1912 vocab=0.8848 token=1.3235 | 6.13 steps/s | VRAM=28.30 Go [2026-07-17 20:59:50] step 466450/1000000 | loss=2.3119 vocab=1.2985 token=1.1942 | 6.09 steps/s | VRAM=26.61 Go [2026-07-17 20:59:58] step 466500/1000000 | loss=2.2588 vocab=0.5118 token=1.0956 | 6.14 steps/s | VRAM=28.01 Go [2026-07-17 21:00:07] step 466550/1000000 | loss=2.2850 vocab=1.0441 token=1.1483 | 6.14 steps/s | VRAM=28.37 Go [2026-07-17 21:00:15] step 466600/1000000 | loss=2.2306 vocab=0.9720 token=1.5516 | 6.14 steps/s | VRAM=26.26 Go [2026-07-17 21:00:23] step 466650/1000000 | loss=2.2735 vocab=0.9434 token=1.2584 | 6.13 steps/s | VRAM=32.72 Go [2026-07-17 21:00:31] step 466700/1000000 | loss=2.1572 vocab=0.6209 token=1.2742 | 6.16 steps/s | VRAM=26.03 Go [2026-07-17 21:00:39] step 466750/1000000 | loss=2.2094 vocab=1.1686 token=1.5354 | 6.12 steps/s | VRAM=26.50 Go [2026-07-17 21:00:47] step 466800/1000000 | loss=2.2372 vocab=0.8191 token=1.1200 | 6.16 steps/s | VRAM=28.89 Go [2026-07-17 21:00:56] step 466850/1000000 | loss=2.1572 vocab=0.9262 token=1.2257 | 6.09 steps/s | VRAM=26.52 Go [2026-07-17 21:01:04] step 466900/1000000 | loss=2.3012 vocab=0.9866 token=1.2414 | 6.14 steps/s | VRAM=28.63 Go [2026-07-17 21:01:12] step 466950/1000000 | loss=2.2528 vocab=1.3929 token=1.2413 | 6.12 steps/s | VRAM=29.10 Go [2026-07-17 21:01:20] step 467000/1000000 | loss=2.3242 vocab=0.3672 token=0.9598 | 6.16 steps/s | VRAM=31.34 Go [2026-07-17 21:01:28] step 467050/1000000 | loss=2.2374 vocab=0.7472 token=1.3212 | 6.12 steps/s | VRAM=28.86 Go [2026-07-17 21:01:36] step 467100/1000000 | loss=2.2965 vocab=1.1836 token=1.6763 | 6.14 steps/s | VRAM=31.72 Go [2026-07-17 21:01:44] step 467150/1000000 | loss=2.2858 vocab=1.0480 token=1.3113 | 6.13 steps/s | VRAM=29.53 Go [2026-07-17 21:01:53] step 467200/1000000 | loss=2.3032 vocab=1.2669 token=1.6797 | 6.13 steps/s | VRAM=29.69 Go [2026-07-17 21:02:01] step 467250/1000000 | loss=2.2063 vocab=0.5641 token=1.5083 | 6.13 steps/s | VRAM=29.04 Go [2026-07-17 21:02:09] step 467300/1000000 | loss=2.1774 vocab=1.1702 token=1.4400 | 6.13 steps/s | VRAM=31.78 Go [2026-07-17 21:02:17] step 467350/1000000 | loss=2.2323 vocab=1.3192 token=1.6456 | 6.14 steps/s | VRAM=27.07 Go [2026-07-17 21:02:25] step 467400/1000000 | loss=2.2466 vocab=0.5735 token=1.2030 | 6.16 steps/s | VRAM=27.86 Go [2026-07-17 21:02:33] step 467450/1000000 | loss=2.2121 vocab=0.8775 token=1.0670 | 6.12 steps/s | VRAM=29.44 Go [2026-07-17 21:02:41] step 467500/1000000 | loss=2.3417 vocab=0.7226 token=1.1298 | 6.15 steps/s | VRAM=28.88 Go [2026-07-17 21:02:50] step 467550/1000000 | loss=2.3636 vocab=0.9100 token=1.4431 | 6.13 steps/s | VRAM=29.45 Go [2026-07-17 21:02:58] step 467600/1000000 | loss=2.2888 vocab=0.8160 token=1.1153 | 6.13 steps/s | VRAM=29.39 Go [2026-07-17 21:03:06] step 467650/1000000 | loss=2.2738 vocab=1.3835 token=1.8153 | 6.15 steps/s | VRAM=31.52 Go [2026-07-17 21:03:14] step 467700/1000000 | loss=2.2873 vocab=0.5977 token=1.1833 | 6.14 steps/s | VRAM=28.74 Go [2026-07-17 21:03:22] step 467750/1000000 | loss=2.2216 vocab=0.7689 token=1.2644 | 6.13 steps/s | VRAM=28.22 Go [2026-07-17 21:03:30] step 467800/1000000 | loss=2.2976 vocab=1.6481 token=1.4636 | 6.12 steps/s | VRAM=26.10 Go [2026-07-17 21:03:38] step 467850/1000000 | loss=2.2359 vocab=0.4806 token=0.7771 | 6.14 steps/s | VRAM=32.52 Go [2026-07-17 21:03:47] step 467900/1000000 | loss=2.1809 vocab=1.2998 token=1.5556 | 6.14 steps/s | VRAM=28.31 Go [2026-07-17 21:03:55] step 467950/1000000 | loss=2.2223 vocab=0.6854 token=1.4714 | 6.15 steps/s | VRAM=25.94 Go [2026-07-17 21:04:03] step 468000/1000000 | loss=2.2760 vocab=1.1719 token=1.7998 | 6.11 steps/s | VRAM=29.90 Go [2026-07-17 21:04:11] step 468050/1000000 | loss=2.3825 vocab=1.1804 token=1.6873 | 6.12 steps/s | VRAM=27.87 Go [2026-07-17 21:04:19] step 468100/1000000 | loss=2.0890 vocab=0.9295 token=1.4345 | 6.13 steps/s | VRAM=28.80 Go [2026-07-17 21:04:27] step 468150/1000000 | loss=2.2852 vocab=0.7719 token=1.3583 | 6.09 steps/s | VRAM=27.55 Go [2026-07-17 21:04:36] step 468200/1000000 | loss=2.3115 vocab=1.0982 token=1.6868 | 6.14 steps/s | VRAM=31.56 Go [2026-07-17 21:04:44] step 468250/1000000 | loss=2.2925 vocab=0.8261 token=1.2248 | 6.14 steps/s | VRAM=32.62 Go [2026-07-17 21:04:52] step 468300/1000000 | loss=2.2397 vocab=0.4786 token=1.0060 | 6.14 steps/s | VRAM=32.35 Go [2026-07-17 21:05:00] step 468350/1000000 | loss=2.2319 vocab=0.7642 token=1.3327 | 6.12 steps/s | VRAM=32.76 Go [2026-07-17 21:05:08] step 468400/1000000 | loss=2.1593 vocab=1.1885 token=2.1738 | 6.10 steps/s | VRAM=28.57 Go [2026-07-17 21:05:16] step 468450/1000000 | loss=2.2859 vocab=1.4710 token=1.3680 | 6.12 steps/s | VRAM=28.82 Go [2026-07-17 21:05:25] step 468500/1000000 | loss=2.2022 vocab=1.4246 token=1.6451 | 6.11 steps/s | VRAM=32.85 Go [2026-07-17 21:05:33] step 468550/1000000 | loss=2.2525 vocab=0.8336 token=1.4749 | 6.12 steps/s | VRAM=28.14 Go [2026-07-17 21:05:41] step 468600/1000000 | loss=2.2218 vocab=0.8664 token=1.2674 | 6.14 steps/s | VRAM=31.99 Go [2026-07-17 21:05:49] step 468650/1000000 | loss=2.2408 vocab=0.6938 token=1.3468 | 6.07 steps/s | VRAM=26.82 Go [2026-07-17 21:05:57] step 468700/1000000 | loss=2.1775 vocab=1.0652 token=1.3534 | 6.13 steps/s | VRAM=28.44 Go [2026-07-17 21:06:05] step 468750/1000000 | loss=2.2596 vocab=1.7502 token=1.1133 | 6.13 steps/s | VRAM=31.37 Go [2026-07-17 21:06:14] step 468800/1000000 | loss=2.2219 vocab=1.1585 token=1.1652 | 6.13 steps/s | VRAM=29.27 Go [2026-07-17 21:06:22] step 468850/1000000 | loss=2.2582 vocab=0.9674 token=1.8622 | 6.16 steps/s | VRAM=32.27 Go [2026-07-17 21:06:30] step 468900/1000000 | loss=2.3537 vocab=0.9031 token=1.4290 | 6.13 steps/s | VRAM=30.07 Go [2026-07-17 21:06:38] step 468950/1000000 | loss=2.1725 vocab=0.7039 token=1.1875 | 6.17 steps/s | VRAM=24.39 Go [2026-07-17 21:06:46] step 469000/1000000 | loss=2.3039 vocab=0.4977 token=1.4657 | 6.17 steps/s | VRAM=31.56 Go [2026-07-17 21:06:54] step 469050/1000000 | loss=2.2103 vocab=0.5992 token=1.3392 | 6.12 steps/s | VRAM=32.79 Go [2026-07-17 21:07:02] step 469100/1000000 | loss=2.2353 vocab=0.7144 token=1.2835 | 6.15 steps/s | VRAM=25.61 Go [2026-07-17 21:07:11] step 469150/1000000 | loss=2.2720 vocab=0.7409 token=1.2815 | 6.15 steps/s | VRAM=26.15 Go [2026-07-17 21:07:19] step 469200/1000000 | loss=2.2030 vocab=1.1454 token=1.4329 | 6.14 steps/s | VRAM=26.28 Go [2026-07-17 21:07:27] step 469250/1000000 | loss=2.2124 vocab=1.4220 token=1.3543 | 6.15 steps/s | VRAM=25.76 Go [2026-07-17 21:07:35] step 469300/1000000 | loss=2.2436 vocab=0.6837 token=1.1427 | 6.16 steps/s | VRAM=32.50 Go [2026-07-17 21:07:43] step 469350/1000000 | loss=2.3118 vocab=0.8882 token=1.1885 | 6.14 steps/s | VRAM=26.01 Go [2026-07-17 21:07:51] step 469400/1000000 | loss=2.2674 vocab=0.8025 token=1.3245 | 6.13 steps/s | VRAM=28.10 Go [2026-07-17 21:07:59] step 469450/1000000 | loss=2.2905 vocab=0.9499 token=1.2748 | 6.12 steps/s | VRAM=29.03 Go [2026-07-17 21:08:08] step 469500/1000000 | loss=2.3480 vocab=1.1861 token=1.3241 | 6.14 steps/s | VRAM=28.91 Go [2026-07-17 21:08:16] step 469550/1000000 | loss=2.2582 vocab=0.6935 token=1.5301 | 6.14 steps/s | VRAM=28.76 Go [2026-07-17 21:08:24] step 469600/1000000 | loss=2.2232 vocab=1.1366 token=1.3092 | 6.10 steps/s | VRAM=26.62 Go [2026-07-17 21:08:32] step 469650/1000000 | loss=2.1521 vocab=1.1228 token=1.5537 | 6.15 steps/s | VRAM=29.06 Go [2026-07-17 21:08:40] step 469700/1000000 | loss=2.2448 vocab=0.5279 token=1.5207 | 6.12 steps/s | VRAM=26.42 Go [2026-07-17 21:08:48] step 469750/1000000 | loss=2.2608 vocab=1.0928 token=1.2973 | 6.12 steps/s | VRAM=26.45 Go [2026-07-17 21:08:57] step 469800/1000000 | loss=2.2951 vocab=0.7186 token=1.4868 | 6.17 steps/s | VRAM=32.24 Go [2026-07-17 21:09:05] step 469850/1000000 | loss=2.1839 vocab=0.6554 token=0.9906 | 6.16 steps/s | VRAM=28.23 Go [2026-07-17 21:09:13] step 469900/1000000 | loss=2.3075 vocab=0.9297 token=1.6249 | 6.15 steps/s | VRAM=31.64 Go [2026-07-17 21:09:21] step 469950/1000000 | loss=2.1531 vocab=0.7969 token=1.4556 | 6.12 steps/s | VRAM=26.01 Go [2026-07-17 21:09:29] step 470000/1000000 | loss=2.2330 vocab=0.5881 token=1.3244 | 6.12 steps/s | VRAM=32.51 Go [2026-07-17 21:09:37] step 470050/1000000 | loss=2.2628 vocab=0.7473 token=1.4666 | 6.11 steps/s | VRAM=26.34 Go [2026-07-17 21:09:45] step 470100/1000000 | loss=2.2596 vocab=1.2727 token=1.4481 | 6.14 steps/s | VRAM=32.52 Go [2026-07-17 21:09:54] step 470150/1000000 | loss=2.2269 vocab=0.9671 token=1.4974 | 6.14 steps/s | VRAM=28.71 Go [2026-07-17 21:10:02] step 470200/1000000 | loss=2.2169 vocab=1.0411 token=1.2342 | 6.12 steps/s | VRAM=26.81 Go [2026-07-17 21:10:10] step 470250/1000000 | loss=2.2114 vocab=0.8199 token=1.4346 | 6.15 steps/s | VRAM=29.43 Go [2026-07-17 21:10:18] step 470300/1000000 | loss=2.1933 vocab=0.6493 token=1.3225 | 6.16 steps/s | VRAM=26.77 Go [2026-07-17 21:10:26] step 470350/1000000 | loss=2.3811 vocab=1.3510 token=1.6386 | 6.13 steps/s | VRAM=26.50 Go [2026-07-17 21:10:34] step 470400/1000000 | loss=2.2272 vocab=1.7629 token=1.5723 | 6.16 steps/s | VRAM=28.45 Go [2026-07-17 21:10:42] step 470450/1000000 | loss=2.2436 vocab=1.0304 token=1.2796 | 6.14 steps/s | VRAM=27.82 Go [2026-07-17 21:10:51] step 470500/1000000 | loss=2.2063 vocab=0.7690 token=1.3425 | 6.14 steps/s | VRAM=28.49 Go [2026-07-17 21:10:59] step 470550/1000000 | loss=2.2512 vocab=0.7545 token=1.3923 | 6.13 steps/s | VRAM=25.49 Go [2026-07-17 21:11:07] step 470600/1000000 | loss=2.2802 vocab=0.6839 token=0.9526 | 6.11 steps/s | VRAM=26.24 Go [2026-07-17 21:11:15] step 470650/1000000 | loss=2.1931 vocab=0.6514 token=1.0827 | 6.15 steps/s | VRAM=25.73 Go [2026-07-17 21:11:23] step 470700/1000000 | loss=2.2684 vocab=0.5330 token=1.2623 | 6.15 steps/s | VRAM=28.44 Go [2026-07-17 21:11:31] step 470750/1000000 | loss=2.2191 vocab=1.3009 token=1.5357 | 6.13 steps/s | VRAM=32.56 Go [2026-07-17 21:11:39] step 470800/1000000 | loss=2.2308 vocab=0.5246 token=1.2762 | 6.17 steps/s | VRAM=30.58 Go [2026-07-17 21:11:48] step 470850/1000000 | loss=2.3737 vocab=1.2555 token=1.3713 | 6.15 steps/s | VRAM=25.72 Go [2026-07-17 21:11:56] step 470900/1000000 | loss=2.2337 vocab=0.5266 token=1.0855 | 6.15 steps/s | VRAM=28.55 Go [2026-07-17 21:12:04] step 470950/1000000 | loss=2.2639 vocab=1.0601 token=1.2594 | 6.13 steps/s | VRAM=26.61 Go [2026-07-17 21:12:12] step 471000/1000000 | loss=2.1790 vocab=0.6687 token=1.1142 | 6.14 steps/s | VRAM=32.75 Go [2026-07-17 21:12:20] step 471050/1000000 | loss=2.2720 vocab=1.1172 token=1.5362 | 6.11 steps/s | VRAM=29.00 Go [2026-07-17 21:12:28] step 471100/1000000 | loss=2.2254 vocab=0.7122 token=1.1534 | 6.13 steps/s | VRAM=26.98 Go [2026-07-17 21:12:36] step 471150/1000000 | loss=2.2958 vocab=0.9408 token=1.5721 | 6.14 steps/s | VRAM=28.75 Go [2026-07-17 21:12:45] step 471200/1000000 | loss=2.1910 vocab=0.9608 token=1.5838 | 6.16 steps/s | VRAM=31.71 Go [2026-07-17 21:12:53] step 471250/1000000 | loss=2.3437 vocab=1.1790 token=1.7110 | 6.11 steps/s | VRAM=29.22 Go [2026-07-17 21:13:01] step 471300/1000000 | loss=2.2613 vocab=1.0406 token=1.7160 | 6.13 steps/s | VRAM=28.85 Go [2026-07-17 21:13:09] step 471350/1000000 | loss=2.2877 vocab=0.5014 token=0.9301 | 6.08 steps/s | VRAM=26.99 Go [2026-07-17 21:13:17] step 471400/1000000 | loss=2.2023 vocab=0.6150 token=1.1118 | 6.16 steps/s | VRAM=32.42 Go [2026-07-17 21:13:25] step 471450/1000000 | loss=2.2418 vocab=1.6258 token=1.6095 | 6.15 steps/s | VRAM=24.84 Go [2026-07-17 21:13:34] step 471500/1000000 | loss=2.3292 vocab=1.3156 token=1.4742 | 6.11 steps/s | VRAM=28.97 Go [2026-07-17 21:13:42] step 471550/1000000 | loss=2.1559 vocab=0.8838 token=1.5310 | 6.14 steps/s | VRAM=32.84 Go [2026-07-17 21:13:50] step 471600/1000000 | loss=2.2461 vocab=0.8322 token=1.2461 | 6.12 steps/s | VRAM=28.12 Go [2026-07-17 21:13:58] step 471650/1000000 | loss=2.2535 vocab=0.3498 token=1.1503 | 6.12 steps/s | VRAM=25.41 Go [2026-07-17 21:14:06] step 471700/1000000 | loss=2.3579 vocab=1.5267 token=1.5273 | 6.08 steps/s | VRAM=29.22 Go [2026-07-17 21:14:14] step 471750/1000000 | loss=2.3078 vocab=1.0565 token=1.0191 | 6.12 steps/s | VRAM=25.86 Go [2026-07-17 21:14:23] step 471800/1000000 | loss=2.1372 vocab=0.8398 token=1.0663 | 6.18 steps/s | VRAM=27.70 Go [2026-07-17 21:14:31] step 471850/1000000 | loss=2.3601 vocab=1.1850 token=1.2276 | 6.16 steps/s | VRAM=24.80 Go [2026-07-17 21:14:39] step 471900/1000000 | loss=2.1865 vocab=1.2993 token=1.6229 | 6.15 steps/s | VRAM=29.36 Go [2026-07-17 21:14:47] step 471950/1000000 | loss=2.1817 vocab=1.0991 token=1.3053 | 6.16 steps/s | VRAM=29.31 Go [2026-07-17 21:14:55] step 472000/1000000 | loss=2.2301 vocab=0.7907 token=0.9486 | 6.14 steps/s | VRAM=30.13 Go [2026-07-17 21:15:03] step 472050/1000000 | loss=2.2411 vocab=0.5691 token=1.3053 | 6.15 steps/s | VRAM=27.37 Go [2026-07-17 21:15:11] step 472100/1000000 | loss=2.2480 vocab=0.3997 token=1.5030 | 6.16 steps/s | VRAM=28.94 Go [2026-07-17 21:15:19] step 472150/1000000 | loss=2.2604 vocab=0.8474 token=1.1536 | 6.16 steps/s | VRAM=28.51 Go [2026-07-17 21:15:28] step 472200/1000000 | loss=2.3201 vocab=1.6653 token=1.9579 | 6.15 steps/s | VRAM=27.00 Go [2026-07-17 21:15:36] step 472250/1000000 | loss=2.2303 vocab=0.9693 token=1.6686 | 6.13 steps/s | VRAM=25.16 Go [2026-07-17 21:15:44] step 472300/1000000 | loss=2.2822 vocab=1.3854 token=1.5947 | 6.13 steps/s | VRAM=26.46 Go [2026-07-17 21:15:52] step 472350/1000000 | loss=2.2453 vocab=0.8432 token=1.3489 | 6.15 steps/s | VRAM=27.88 Go [2026-07-17 21:16:00] step 472400/1000000 | loss=2.3138 vocab=0.7600 token=1.3294 | 6.14 steps/s | VRAM=31.30 Go [2026-07-17 21:16:08] step 472450/1000000 | loss=2.2883 vocab=1.2381 token=1.5303 | 6.12 steps/s | VRAM=26.66 Go [2026-07-17 21:16:16] step 472500/1000000 | loss=2.3038 vocab=0.8771 token=1.2508 | 6.15 steps/s | VRAM=27.82 Go [2026-07-17 21:16:25] step 472550/1000000 | loss=2.2722 vocab=0.9276 token=1.4099 | 6.14 steps/s | VRAM=31.95 Go [2026-07-17 21:16:33] step 472600/1000000 | loss=2.2635 vocab=0.7183 token=1.0307 | 6.14 steps/s | VRAM=26.08 Go [2026-07-17 21:16:41] step 472650/1000000 | loss=2.2471 vocab=0.9055 token=1.3725 | 6.16 steps/s | VRAM=28.49 Go [2026-07-17 21:16:49] step 472700/1000000 | loss=2.4934 vocab=1.4497 token=1.5236 | 6.15 steps/s | VRAM=24.83 Go [2026-07-17 21:16:57] step 472750/1000000 | loss=2.3184 vocab=0.6655 token=1.3329 | 6.14 steps/s | VRAM=28.31 Go [2026-07-17 21:17:05] step 472800/1000000 | loss=2.3271 vocab=0.8571 token=1.0860 | 6.13 steps/s | VRAM=28.56 Go [2026-07-17 21:17:13] step 472850/1000000 | loss=2.2057 vocab=1.1300 token=1.3244 | 6.11 steps/s | VRAM=29.15 Go [2026-07-17 21:17:22] step 472900/1000000 | loss=2.1169 vocab=1.1972 token=1.0036 | 6.14 steps/s | VRAM=33.02 Go [2026-07-17 21:17:30] step 472950/1000000 | loss=2.1596 vocab=0.7845 token=1.2406 | 6.09 steps/s | VRAM=26.37 Go [2026-07-17 21:17:38] step 473000/1000000 | loss=2.1991 vocab=0.6422 token=1.2548 | 6.17 steps/s | VRAM=24.53 Go [2026-07-17 21:17:46] step 473050/1000000 | loss=2.2358 vocab=0.5733 token=0.9541 | 6.17 steps/s | VRAM=32.25 Go [2026-07-17 21:17:54] step 473100/1000000 | loss=2.2784 vocab=1.4296 token=1.3988 | 6.13 steps/s | VRAM=29.08 Go [2026-07-17 21:18:02] step 473150/1000000 | loss=2.2335 vocab=0.7497 token=1.0955 | 6.12 steps/s | VRAM=28.11 Go [2026-07-17 21:18:10] step 473200/1000000 | loss=2.1355 vocab=1.0206 token=1.7863 | 6.16 steps/s | VRAM=31.85 Go [2026-07-17 21:18:19] step 473250/1000000 | loss=2.2164 vocab=1.4659 token=1.7736 | 6.10 steps/s | VRAM=26.37 Go [2026-07-17 21:18:27] step 473300/1000000 | loss=2.2384 vocab=0.9600 token=1.7441 | 6.14 steps/s | VRAM=27.58 Go [2026-07-17 21:18:35] step 473350/1000000 | loss=2.2071 vocab=0.6929 token=0.9843 | 6.15 steps/s | VRAM=27.97 Go [2026-07-17 21:18:43] step 473400/1000000 | loss=2.3882 vocab=1.1844 token=1.2660 | 6.13 steps/s | VRAM=28.83 Go [2026-07-17 21:18:51] step 473450/1000000 | loss=2.2051 vocab=1.3387 token=1.2503 | 6.14 steps/s | VRAM=25.18 Go [2026-07-17 21:18:59] step 473500/1000000 | loss=2.2872 vocab=0.6982 token=1.1053 | 6.14 steps/s | VRAM=32.63 Go [2026-07-17 21:19:08] step 473550/1000000 | loss=2.1539 vocab=0.9501 token=1.5448 | 6.12 steps/s | VRAM=25.92 Go [2026-07-17 21:19:16] step 473600/1000000 | loss=2.2164 vocab=0.9498 token=1.1045 | 6.15 steps/s | VRAM=32.07 Go [2026-07-17 21:19:24] step 473650/1000000 | loss=2.2825 vocab=0.8000 token=1.4479 | 6.14 steps/s | VRAM=28.95 Go [2026-07-17 21:19:32] step 473700/1000000 | loss=2.2601 vocab=1.9465 token=1.3744 | 6.14 steps/s | VRAM=29.25 Go [2026-07-17 21:19:40] step 473750/1000000 | loss=2.3299 vocab=1.4317 token=0.9725 | 6.13 steps/s | VRAM=32.53 Go [2026-07-17 21:19:48] step 473800/1000000 | loss=2.3402 vocab=0.7445 token=1.2210 | 6.12 steps/s | VRAM=26.80 Go [2026-07-17 21:19:56] step 473850/1000000 | loss=2.2522 vocab=1.5599 token=1.4768 | 6.15 steps/s | VRAM=32.93 Go [2026-07-17 21:20:05] step 473900/1000000 | loss=2.2788 vocab=0.8543 token=1.4755 | 6.15 steps/s | VRAM=32.04 Go [2026-07-17 21:20:13] step 473950/1000000 | loss=2.2572 vocab=0.6100 token=1.2492 | 6.12 steps/s | VRAM=32.22 Go [2026-07-17 21:20:21] step 474000/1000000 | loss=2.2299 vocab=0.9528 token=1.2344 | 6.14 steps/s | VRAM=31.40 Go [2026-07-17 21:20:29] step 474050/1000000 | loss=2.2117 vocab=0.9927 token=1.4024 | 6.16 steps/s | VRAM=31.73 Go [2026-07-17 21:20:37] step 474100/1000000 | loss=2.3155 vocab=0.9078 token=1.3355 | 6.12 steps/s | VRAM=27.55 Go [2026-07-17 21:20:45] step 474150/1000000 | loss=2.3293 vocab=1.4291 token=1.3325 | 6.12 steps/s | VRAM=26.86 Go [2026-07-17 21:20:53] step 474200/1000000 | loss=2.2235 vocab=1.3933 token=1.5483 | 6.16 steps/s | VRAM=31.62 Go [2026-07-17 21:21:02] step 474250/1000000 | loss=2.3080 vocab=1.1060 token=1.1751 | 6.14 steps/s | VRAM=28.63 Go [2026-07-17 21:21:10] step 474300/1000000 | loss=2.2610 vocab=1.3164 token=1.7674 | 6.15 steps/s | VRAM=28.93 Go [2026-07-17 21:21:18] step 474350/1000000 | loss=2.2504 vocab=0.5210 token=1.4445 | 6.15 steps/s | VRAM=29.35 Go [2026-07-17 21:21:26] step 474400/1000000 | loss=2.1923 vocab=0.8960 token=1.4896 | 6.18 steps/s | VRAM=30.47 Go [2026-07-17 21:21:34] step 474450/1000000 | loss=2.3197 vocab=1.0462 token=1.6366 | 6.14 steps/s | VRAM=25.60 Go [2026-07-17 21:21:42] step 474500/1000000 | loss=2.2934 vocab=0.6417 token=1.2717 | 6.15 steps/s | VRAM=32.01 Go [2026-07-17 21:21:50] step 474550/1000000 | loss=2.1811 vocab=0.6524 token=0.9214 | 6.17 steps/s | VRAM=28.60 Go [2026-07-17 21:21:58] step 474600/1000000 | loss=2.3506 vocab=0.9651 token=1.6131 | 6.14 steps/s | VRAM=32.52 Go [2026-07-17 21:22:07] step 474650/1000000 | loss=2.2182 vocab=0.7917 token=1.2043 | 6.14 steps/s | VRAM=31.41 Go [2026-07-17 21:22:15] step 474700/1000000 | loss=2.3098 vocab=1.5792 token=1.6719 | 6.11 steps/s | VRAM=26.15 Go [2026-07-17 21:22:23] step 474750/1000000 | loss=2.1174 vocab=0.6506 token=1.3758 | 6.16 steps/s | VRAM=25.80 Go [2026-07-17 21:22:31] step 474800/1000000 | loss=2.1246 vocab=0.9855 token=1.2661 | 6.13 steps/s | VRAM=28.33 Go [2026-07-17 21:22:39] step 474850/1000000 | loss=2.1678 vocab=0.7057 token=1.1407 | 6.16 steps/s | VRAM=29.31 Go [2026-07-17 21:22:47] step 474900/1000000 | loss=2.3060 vocab=0.9451 token=1.2759 | 6.13 steps/s | VRAM=29.88 Go [2026-07-17 21:22:55] step 474950/1000000 | loss=2.3097 vocab=1.3163 token=1.1633 | 6.14 steps/s | VRAM=32.14 Go [2026-07-17 21:23:04] step 475000/1000000 | loss=2.2752 vocab=1.2618 token=1.5098 | 6.13 steps/s | VRAM=27.88 Go [2026-07-17 21:23:12] step 475050/1000000 | loss=2.3065 vocab=0.9649 token=0.9221 | 6.16 steps/s | VRAM=32.11 Go [2026-07-17 21:23:20] step 475100/1000000 | loss=2.2213 vocab=1.3629 token=1.4081 | 6.16 steps/s | VRAM=32.28 Go [2026-07-17 21:23:28] step 475150/1000000 | loss=2.2449 vocab=0.3307 token=1.2271 | 6.13 steps/s | VRAM=26.10 Go [2026-07-17 21:23:36] step 475200/1000000 | loss=2.2967 vocab=0.9051 token=1.3779 | 6.12 steps/s | VRAM=29.27 Go [2026-07-17 21:23:44] step 475250/1000000 | loss=2.2471 vocab=0.6897 token=1.1725 | 6.14 steps/s | VRAM=32.17 Go [2026-07-17 21:23:52] step 475300/1000000 | loss=2.2431 vocab=0.8333 token=1.0586 | 6.13 steps/s | VRAM=29.00 Go [2026-07-17 21:24:01] step 475350/1000000 | loss=2.3502 vocab=1.7350 token=1.5844 | 6.13 steps/s | VRAM=25.13 Go [2026-07-17 21:24:09] step 475400/1000000 | loss=2.3111 vocab=0.9657 token=1.1128 | 6.12 steps/s | VRAM=26.05 Go [2026-07-17 21:24:17] step 475450/1000000 | loss=2.3107 vocab=1.0026 token=1.2686 | 6.15 steps/s | VRAM=32.38 Go [2026-07-17 21:24:25] step 475500/1000000 | loss=2.2238 vocab=0.9345 token=1.2263 | 6.14 steps/s | VRAM=25.59 Go [2026-07-17 21:24:33] step 475550/1000000 | loss=2.2947 vocab=0.8128 token=0.8777 | 6.14 steps/s | VRAM=28.42 Go [2026-07-17 21:24:41] step 475600/1000000 | loss=2.1955 vocab=0.6325 token=1.3782 | 6.14 steps/s | VRAM=29.36 Go [2026-07-17 21:24:50] step 475650/1000000 | loss=2.1604 vocab=1.0847 token=1.5685 | 6.13 steps/s | VRAM=30.09 Go [2026-07-17 21:24:58] step 475700/1000000 | loss=2.2899 vocab=0.6598 token=1.3477 | 6.14 steps/s | VRAM=29.88 Go [2026-07-17 21:25:06] step 475750/1000000 | loss=2.2678 vocab=0.8773 token=1.4432 | 6.17 steps/s | VRAM=31.02 Go [2026-07-17 21:25:14] step 475800/1000000 | loss=2.1987 vocab=0.6701 token=1.0716 | 6.14 steps/s | VRAM=28.80 Go [2026-07-17 21:25:22] step 475850/1000000 | loss=2.2804 vocab=1.2191 token=1.6350 | 6.17 steps/s | VRAM=28.52 Go [2026-07-17 21:25:30] step 475900/1000000 | loss=2.2566 vocab=1.0319 token=1.3893 | 6.17 steps/s | VRAM=32.67 Go [2026-07-17 21:25:38] step 475950/1000000 | loss=2.3488 vocab=0.6069 token=1.3638 | 6.11 steps/s | VRAM=30.01 Go [2026-07-17 21:25:46] step 476000/1000000 | loss=2.2981 vocab=0.8689 token=1.2833 | 6.13 steps/s | VRAM=29.39 Go [2026-07-17 21:25:55] step 476050/1000000 | loss=2.1955 vocab=1.1419 token=1.5486 | 6.14 steps/s | VRAM=27.99 Go [2026-07-17 21:26:03] step 476100/1000000 | loss=2.2124 vocab=0.8404 token=1.0918 | 6.18 steps/s | VRAM=27.83 Go [2026-07-17 21:26:11] step 476150/1000000 | loss=2.2768 vocab=0.5363 token=1.0214 | 6.13 steps/s | VRAM=26.53 Go [2026-07-17 21:26:19] step 476200/1000000 | loss=2.1721 vocab=0.8217 token=1.2832 | 6.17 steps/s | VRAM=32.18 Go [2026-07-17 21:26:27] step 476250/1000000 | loss=2.2585 vocab=0.8988 token=1.3757 | 6.18 steps/s | VRAM=25.05 Go [2026-07-17 21:26:35] step 476300/1000000 | loss=2.2741 vocab=0.7437 token=1.5953 | 6.15 steps/s | VRAM=27.91 Go [2026-07-17 21:26:43] step 476350/1000000 | loss=2.2255 vocab=1.1495 token=1.5224 | 6.13 steps/s | VRAM=28.57 Go [2026-07-17 21:26:51] step 476400/1000000 | loss=2.3214 vocab=0.7264 token=1.2815 | 6.15 steps/s | VRAM=27.36 Go [2026-07-17 21:27:00] step 476450/1000000 | loss=2.3591 vocab=0.9392 token=1.3626 | 6.14 steps/s | VRAM=26.42 Go [2026-07-17 21:27:08] step 476500/1000000 | loss=2.2507 vocab=0.3941 token=0.7735 | 6.15 steps/s | VRAM=29.23 Go [2026-07-17 21:27:16] step 476550/1000000 | loss=2.1766 vocab=1.2169 token=1.2846 | 6.13 steps/s | VRAM=26.29 Go [2026-07-17 21:27:24] step 476600/1000000 | loss=2.2115 vocab=1.0863 token=1.3235 | 6.18 steps/s | VRAM=30.64 Go [2026-07-17 21:27:32] step 476650/1000000 | loss=2.3673 vocab=1.2017 token=1.4593 | 6.16 steps/s | VRAM=31.96 Go [2026-07-17 21:27:40] step 476700/1000000 | loss=2.2703 vocab=0.7353 token=1.5771 | 6.12 steps/s | VRAM=29.50 Go [2026-07-17 21:27:48] step 476750/1000000 | loss=2.2555 vocab=0.6623 token=1.3975 | 6.14 steps/s | VRAM=32.23 Go [2026-07-17 21:27:57] step 476800/1000000 | loss=2.3103 vocab=0.7197 token=1.1084 | 6.12 steps/s | VRAM=32.63 Go [2026-07-17 21:28:05] step 476850/1000000 | loss=2.2231 vocab=1.1022 token=1.5740 | 6.15 steps/s | VRAM=32.39 Go [2026-07-17 21:28:13] step 476900/1000000 | loss=2.2957 vocab=0.9200 token=0.8888 | 6.16 steps/s | VRAM=31.64 Go [2026-07-17 21:28:21] step 476950/1000000 | loss=2.2257 vocab=0.9167 token=1.0622 | 6.12 steps/s | VRAM=25.55 Go [2026-07-17 21:28:29] step 477000/1000000 | loss=2.2814 vocab=0.5875 token=1.3032 | 6.16 steps/s | VRAM=28.19 Go [2026-07-17 21:28:37] step 477050/1000000 | loss=2.2229 vocab=0.9205 token=1.0713 | 6.14 steps/s | VRAM=32.61 Go [2026-07-17 21:28:45] step 477100/1000000 | loss=2.2256 vocab=1.4526 token=1.7812 | 6.14 steps/s | VRAM=28.36 Go [2026-07-17 21:28:54] step 477150/1000000 | loss=2.2564 vocab=0.4667 token=1.0230 | 6.14 steps/s | VRAM=26.99 Go [2026-07-17 21:29:02] step 477200/1000000 | loss=2.2887 vocab=1.5258 token=1.2894 | 6.17 steps/s | VRAM=29.24 Go [2026-07-17 21:29:10] step 477250/1000000 | loss=2.2220 vocab=0.5606 token=1.0255 | 6.17 steps/s | VRAM=31.49 Go [2026-07-17 21:29:18] step 477300/1000000 | loss=2.3179 vocab=0.8566 token=1.5365 | 6.14 steps/s | VRAM=31.30 Go [2026-07-17 21:29:26] step 477350/1000000 | loss=2.0939 vocab=0.8689 token=1.1429 | 6.12 steps/s | VRAM=25.84 Go [2026-07-17 21:29:34] step 477400/1000000 | loss=2.2258 vocab=1.0546 token=1.3322 | 6.14 steps/s | VRAM=28.68 Go [2026-07-17 21:29:42] step 477450/1000000 | loss=2.1811 vocab=1.0411 token=1.3169 | 6.13 steps/s | VRAM=32.75 Go [2026-07-17 21:29:51] step 477500/1000000 | loss=2.1793 vocab=0.8198 token=1.4149 | 6.11 steps/s | VRAM=29.32 Go [2026-07-17 21:29:59] step 477550/1000000 | loss=2.2373 vocab=0.9998 token=1.3258 | 6.12 steps/s | VRAM=25.14 Go [2026-07-17 21:30:07] step 477600/1000000 | loss=2.2540 vocab=0.8792 token=1.4844 | 6.14 steps/s | VRAM=28.61 Go [2026-07-17 21:30:15] step 477650/1000000 | loss=2.1905 vocab=1.2283 token=1.4766 | 6.14 steps/s | VRAM=32.73 Go [2026-07-17 21:30:23] step 477700/1000000 | loss=2.2820 vocab=0.5391 token=1.3404 | 6.15 steps/s | VRAM=28.59 Go [2026-07-17 21:30:31] step 477750/1000000 | loss=2.2325 vocab=0.4772 token=1.3183 | 6.15 steps/s | VRAM=29.30 Go [2026-07-17 21:30:39] step 477800/1000000 | loss=2.2916 vocab=0.7196 token=1.2849 | 6.17 steps/s | VRAM=33.01 Go [2026-07-17 21:30:48] step 477850/1000000 | loss=2.1719 vocab=0.6231 token=0.7425 | 6.12 steps/s | VRAM=28.73 Go [2026-07-17 21:30:56] step 477900/1000000 | loss=2.1020 vocab=1.3324 token=1.5814 | 6.17 steps/s | VRAM=32.23 Go [2026-07-17 21:31:04] step 477950/1000000 | loss=2.2664 vocab=1.1419 token=1.4100 | 6.15 steps/s | VRAM=28.78 Go [2026-07-17 21:31:12] step 478000/1000000 | loss=2.2649 vocab=0.6351 token=1.1311 | 6.16 steps/s | VRAM=26.39 Go [2026-07-17 21:31:20] step 478050/1000000 | loss=2.2019 vocab=0.8966 token=1.4994 | 6.19 steps/s | VRAM=32.08 Go [2026-07-17 21:31:28] step 478100/1000000 | loss=2.2414 vocab=0.8362 token=1.4813 | 6.16 steps/s | VRAM=28.65 Go [2026-07-17 21:31:36] step 478150/1000000 | loss=2.1610 vocab=1.0773 token=0.9103 | 6.15 steps/s | VRAM=25.78 Go [2026-07-17 21:31:44] step 478200/1000000 | loss=2.2790 vocab=0.9840 token=1.6040 | 6.18 steps/s | VRAM=28.45 Go [2026-07-17 21:31:52] step 478250/1000000 | loss=2.2209 vocab=0.8594 token=1.7720 | 6.14 steps/s | VRAM=26.36 Go [2026-07-17 21:32:01] step 478300/1000000 | loss=2.2341 vocab=0.8268 token=1.3100 | 6.10 steps/s | VRAM=28.70 Go [2026-07-17 21:32:09] step 478350/1000000 | loss=2.2753 vocab=0.8908 token=1.5055 | 6.13 steps/s | VRAM=28.43 Go [2026-07-17 21:32:17] step 478400/1000000 | loss=2.2659 vocab=0.7183 token=1.2561 | 6.16 steps/s | VRAM=32.11 Go [2026-07-17 21:32:25] step 478450/1000000 | loss=2.3380 vocab=0.5368 token=1.3966 | 6.15 steps/s | VRAM=27.67 Go [2026-07-17 21:32:33] step 478500/1000000 | loss=2.1768 vocab=0.9251 token=1.5954 | 6.15 steps/s | VRAM=28.44 Go [2026-07-17 21:32:41] step 478550/1000000 | loss=2.2413 vocab=0.9696 token=1.4712 | 6.17 steps/s | VRAM=28.61 Go [2026-07-17 21:32:49] step 478600/1000000 | loss=2.1835 vocab=0.7694 token=1.3453 | 6.17 steps/s | VRAM=29.64 Go [2026-07-17 21:32:58] step 478650/1000000 | loss=2.3079 vocab=0.7774 token=1.1601 | 6.17 steps/s | VRAM=31.97 Go [2026-07-17 21:33:06] step 478700/1000000 | loss=2.2604 vocab=1.0170 token=1.6608 | 6.17 steps/s | VRAM=29.59 Go [2026-07-17 21:33:14] step 478750/1000000 | loss=2.2586 vocab=1.7332 token=1.8280 | 6.14 steps/s | VRAM=29.69 Go [2026-07-17 21:33:22] step 478800/1000000 | loss=2.2822 vocab=0.9742 token=1.8336 | 6.14 steps/s | VRAM=26.22 Go [2026-07-17 21:33:30] step 478850/1000000 | loss=2.2251 vocab=0.9305 token=1.5054 | 6.16 steps/s | VRAM=32.47 Go [2026-07-17 21:33:38] step 478900/1000000 | loss=2.2154 vocab=0.7994 token=1.0701 | 6.13 steps/s | VRAM=26.02 Go [2026-07-17 21:33:46] step 478950/1000000 | loss=2.2386 vocab=0.8379 token=1.0514 | 6.15 steps/s | VRAM=26.46 Go [2026-07-17 21:33:54] step 479000/1000000 | loss=2.1989 vocab=1.1640 token=1.4659 | 6.12 steps/s | VRAM=29.24 Go [2026-07-17 21:34:03] step 479050/1000000 | loss=2.2732 vocab=1.0930 token=1.4158 | 6.13 steps/s | VRAM=29.97 Go [2026-07-17 21:34:11] step 479100/1000000 | loss=2.2052 vocab=0.5853 token=1.1962 | 6.15 steps/s | VRAM=29.78 Go [2026-07-17 21:34:19] step 479150/1000000 | loss=2.2067 vocab=1.2036 token=1.3707 | 6.18 steps/s | VRAM=32.16 Go [2026-07-17 21:34:27] step 479200/1000000 | loss=2.2851 vocab=0.8193 token=1.5089 | 6.17 steps/s | VRAM=28.93 Go [2026-07-17 21:34:35] step 479250/1000000 | loss=2.2534 vocab=0.8043 token=1.4784 | 6.17 steps/s | VRAM=29.42 Go [2026-07-17 21:34:43] step 479300/1000000 | loss=2.3494 vocab=0.9216 token=1.0984 | 6.15 steps/s | VRAM=28.73 Go [2026-07-17 21:34:51] step 479350/1000000 | loss=2.2444 vocab=0.9219 token=1.1475 | 6.16 steps/s | VRAM=29.24 Go [2026-07-17 21:34:59] step 479400/1000000 | loss=2.1554 vocab=1.2416 token=1.5103 | 6.16 steps/s | VRAM=25.60 Go [2026-07-17 21:35:08] step 479450/1000000 | loss=2.1405 vocab=0.7087 token=1.2652 | 6.17 steps/s | VRAM=32.61 Go [2026-07-17 21:35:16] step 479500/1000000 | loss=2.1813 vocab=0.8230 token=1.5296 | 6.16 steps/s | VRAM=23.00 Go [2026-07-17 21:35:24] step 479550/1000000 | loss=2.1880 vocab=1.0180 token=1.1672 | 6.21 steps/s | VRAM=31.62 Go [2026-07-17 21:35:32] step 479600/1000000 | loss=2.2342 vocab=0.7410 token=0.7247 | 6.14 steps/s | VRAM=26.29 Go [2026-07-17 21:35:40] step 479650/1000000 | loss=2.1749 vocab=0.5162 token=1.2451 | 6.18 steps/s | VRAM=26.34 Go [2026-07-17 21:35:48] step 479700/1000000 | loss=2.3302 vocab=0.5276 token=0.9704 | 6.17 steps/s | VRAM=28.93 Go [2026-07-17 21:35:56] step 479750/1000000 | loss=2.2706 vocab=1.1705 token=1.4811 | 6.18 steps/s | VRAM=26.12 Go [2026-07-17 21:36:04] step 479800/1000000 | loss=2.1574 vocab=1.4552 token=1.2288 | 6.12 steps/s | VRAM=27.50 Go [2026-07-17 21:36:12] step 479850/1000000 | loss=2.1817 vocab=1.0181 token=1.2507 | 6.15 steps/s | VRAM=26.77 Go [2026-07-17 21:36:21] step 479900/1000000 | loss=2.2328 vocab=0.7766 token=1.3988 | 6.17 steps/s | VRAM=29.00 Go [2026-07-17 21:36:29] step 479950/1000000 | loss=2.3416 vocab=1.1336 token=1.5212 | 6.12 steps/s | VRAM=32.83 Go [2026-07-17 21:36:37] step 480000/1000000 | loss=2.2714 vocab=0.8454 token=1.5930 | 6.16 steps/s | VRAM=32.62 Go [2026-07-17 21:36:45] step 480050/1000000 | loss=2.1757 vocab=0.6077 token=1.0922 | 6.14 steps/s | VRAM=28.83 Go [2026-07-17 21:36:53] step 480100/1000000 | loss=2.1952 vocab=0.8596 token=1.4183 | 6.10 steps/s | VRAM=29.82 Go [2026-07-17 21:37:01] step 480150/1000000 | loss=2.2201 vocab=0.7566 token=1.3720 | 6.14 steps/s | VRAM=29.45 Go [2026-07-17 21:37:09] step 480200/1000000 | loss=2.3263 vocab=1.0151 token=1.2208 | 6.14 steps/s | VRAM=29.16 Go [2026-07-17 21:37:18] step 480250/1000000 | loss=2.2580 vocab=0.5840 token=1.4768 | 6.12 steps/s | VRAM=27.86 Go [2026-07-17 21:37:26] step 480300/1000000 | loss=2.1444 vocab=1.0185 token=1.3080 | 6.13 steps/s | VRAM=32.54 Go [2026-07-17 21:37:34] step 480350/1000000 | loss=2.2852 vocab=0.5678 token=1.1408 | 6.18 steps/s | VRAM=31.36 Go [2026-07-17 21:37:42] step 480400/1000000 | loss=2.2333 vocab=1.2822 token=1.5029 | 6.15 steps/s | VRAM=30.61 Go [2026-07-17 21:37:50] step 480450/1000000 | loss=2.1582 vocab=0.7635 token=1.2833 | 6.14 steps/s | VRAM=25.31 Go [2026-07-17 21:37:58] step 480500/1000000 | loss=2.2395 vocab=1.0091 token=1.6725 | 6.14 steps/s | VRAM=25.99 Go [2026-07-17 21:38:06] step 480550/1000000 | loss=2.2230 vocab=0.6484 token=0.9816 | 6.13 steps/s | VRAM=25.94 Go [2026-07-17 21:38:15] step 480600/1000000 | loss=2.2847 vocab=0.6578 token=1.4160 | 6.11 steps/s | VRAM=26.81 Go [2026-07-17 21:38:23] step 480650/1000000 | loss=2.1973 vocab=0.6910 token=1.4936 | 6.17 steps/s | VRAM=32.16 Go [2026-07-17 21:38:31] step 480700/1000000 | loss=2.2238 vocab=0.7964 token=1.2047 | 6.14 steps/s | VRAM=28.48 Go [2026-07-17 21:38:39] step 480750/1000000 | loss=2.2584 vocab=1.0409 token=1.4515 | 6.14 steps/s | VRAM=28.76 Go [2026-07-17 21:38:47] step 480800/1000000 | loss=2.2124 vocab=1.0987 token=1.0753 | 6.14 steps/s | VRAM=24.84 Go [2026-07-17 21:38:55] step 480850/1000000 | loss=2.3659 vocab=0.9756 token=1.2259 | 6.15 steps/s | VRAM=27.00 Go [2026-07-17 21:39:03] step 480900/1000000 | loss=2.2627 vocab=1.1787 token=1.3781 | 6.14 steps/s | VRAM=28.88 Go [2026-07-17 21:39:12] step 480950/1000000 | loss=2.1611 vocab=0.6942 token=1.0464 | 6.15 steps/s | VRAM=28.77 Go [2026-07-17 21:39:20] step 481000/1000000 | loss=2.1500 vocab=1.4487 token=1.3813 | 6.16 steps/s | VRAM=29.64 Go [2026-07-17 21:39:28] step 481050/1000000 | loss=2.2082 vocab=0.8019 token=1.5734 | 6.15 steps/s | VRAM=28.82 Go [2026-07-17 21:39:36] step 481100/1000000 | loss=2.3246 vocab=0.9149 token=1.3760 | 6.14 steps/s | VRAM=25.81 Go [2026-07-17 21:39:44] step 481150/1000000 | loss=2.3316 vocab=0.9528 token=1.0545 | 6.16 steps/s | VRAM=31.50 Go [2026-07-17 21:39:52] step 481200/1000000 | loss=2.2199 vocab=0.8526 token=1.3862 | 6.14 steps/s | VRAM=29.45 Go [2026-07-17 21:40:00] step 481250/1000000 | loss=2.2102 vocab=1.0778 token=1.4930 | 6.15 steps/s | VRAM=28.28 Go [2026-07-17 21:40:08] step 481300/1000000 | loss=2.2504 vocab=0.7036 token=0.9252 | 6.16 steps/s | VRAM=25.97 Go [2026-07-17 21:40:17] step 481350/1000000 | loss=2.2664 vocab=1.2495 token=1.2858 | 6.18 steps/s | VRAM=32.77 Go [2026-07-17 21:40:25] step 481400/1000000 | loss=2.3316 vocab=0.9233 token=1.3249 | 6.15 steps/s | VRAM=29.13 Go [2026-07-17 21:40:33] step 481450/1000000 | loss=2.3352 vocab=1.7705 token=1.4070 | 6.14 steps/s | VRAM=28.88 Go [2026-07-17 21:40:41] step 481500/1000000 | loss=2.2345 vocab=0.7407 token=1.3885 | 6.11 steps/s | VRAM=28.24 Go [2026-07-17 21:40:49] step 481550/1000000 | loss=2.2305 vocab=1.5255 token=1.3557 | 6.16 steps/s | VRAM=31.76 Go [2026-07-17 21:40:57] step 481600/1000000 | loss=2.2586 vocab=0.7502 token=1.4964 | 6.14 steps/s | VRAM=30.03 Go [2026-07-17 21:41:05] step 481650/1000000 | loss=2.1862 vocab=1.0030 token=1.3086 | 6.15 steps/s | VRAM=29.05 Go [2026-07-17 21:41:14] step 481700/1000000 | loss=2.2084 vocab=0.7342 token=1.1474 | 6.15 steps/s | VRAM=32.38 Go [2026-07-17 21:41:22] step 481750/1000000 | loss=2.2569 vocab=0.8184 token=1.7396 | 6.15 steps/s | VRAM=31.49 Go [2026-07-17 21:41:30] step 481800/1000000 | loss=2.2490 vocab=0.6389 token=1.4121 | 6.11 steps/s | VRAM=26.87 Go [2026-07-17 21:41:38] step 481850/1000000 | loss=2.2379 vocab=0.9506 token=1.4940 | 6.15 steps/s | VRAM=32.88 Go [2026-07-17 21:41:46] step 481900/1000000 | loss=2.2394 vocab=0.7045 token=1.3052 | 6.09 steps/s | VRAM=29.33 Go [2026-07-17 21:41:54] step 481950/1000000 | loss=2.2112 vocab=0.5775 token=1.2977 | 6.13 steps/s | VRAM=26.70 Go [2026-07-17 21:42:02] step 482000/1000000 | loss=2.2529 vocab=1.2041 token=1.5342 | 6.17 steps/s | VRAM=31.87 Go [2026-07-17 21:42:11] step 482050/1000000 | loss=2.1694 vocab=0.7677 token=1.0386 | 6.17 steps/s | VRAM=32.34 Go [2026-07-17 21:42:19] step 482100/1000000 | loss=2.2589 vocab=1.0463 token=1.3449 | 6.12 steps/s | VRAM=25.85 Go [2026-07-17 21:42:27] step 482150/1000000 | loss=2.3506 vocab=0.7056 token=1.3442 | 6.16 steps/s | VRAM=24.76 Go [2026-07-17 21:42:35] step 482200/1000000 | loss=2.2231 vocab=0.7850 token=1.1619 | 6.17 steps/s | VRAM=29.71 Go [2026-07-17 21:42:43] step 482250/1000000 | loss=2.3105 vocab=0.4572 token=1.2754 | 6.16 steps/s | VRAM=26.43 Go [2026-07-17 21:42:51] step 482300/1000000 | loss=2.2757 vocab=1.0329 token=0.9387 | 6.16 steps/s | VRAM=31.29 Go [2026-07-17 21:42:59] step 482350/1000000 | loss=2.2636 vocab=1.1176 token=1.3796 | 6.17 steps/s | VRAM=27.28 Go [2026-07-17 21:43:07] step 482400/1000000 | loss=2.2513 vocab=0.4400 token=1.1748 | 6.16 steps/s | VRAM=31.25 Go [2026-07-17 21:43:16] step 482450/1000000 | loss=2.1744 vocab=0.5285 token=1.7577 | 6.15 steps/s | VRAM=28.11 Go [2026-07-17 21:43:24] step 482500/1000000 | loss=2.3112 vocab=1.9927 token=1.6500 | 6.10 steps/s | VRAM=24.14 Go [2026-07-17 21:43:32] step 482550/1000000 | loss=2.1790 vocab=0.8239 token=1.3057 | 6.14 steps/s | VRAM=28.91 Go [2026-07-17 21:43:40] step 482600/1000000 | loss=2.1884 vocab=0.4947 token=1.2259 | 6.16 steps/s | VRAM=28.08 Go [2026-07-17 21:43:48] step 482650/1000000 | loss=2.2348 vocab=1.1294 token=1.2525 | 6.15 steps/s | VRAM=27.88 Go [2026-07-17 21:43:56] step 482700/1000000 | loss=2.1196 vocab=0.9889 token=1.4838 | 6.18 steps/s | VRAM=31.69 Go [2026-07-17 21:44:04] step 482750/1000000 | loss=2.3094 vocab=1.1937 token=1.7229 | 6.15 steps/s | VRAM=24.55 Go [2026-07-17 21:44:12] step 482800/1000000 | loss=2.2138 vocab=0.7910 token=1.0856 | 6.15 steps/s | VRAM=28.32 Go [2026-07-17 21:44:21] step 482850/1000000 | loss=2.3231 vocab=0.9668 token=1.4269 | 6.15 steps/s | VRAM=28.95 Go [2026-07-17 21:44:29] step 482900/1000000 | loss=2.2714 vocab=0.8070 token=1.1750 | 6.17 steps/s | VRAM=31.32 Go [2026-07-17 21:44:37] step 482950/1000000 | loss=2.1662 vocab=0.9738 token=0.7667 | 6.14 steps/s | VRAM=28.25 Go [2026-07-17 21:44:45] step 483000/1000000 | loss=2.3329 vocab=1.4690 token=1.4307 | 6.15 steps/s | VRAM=32.38 Go [2026-07-17 21:44:53] step 483050/1000000 | loss=2.3135 vocab=1.0599 token=1.5181 | 6.13 steps/s | VRAM=29.69 Go [2026-07-17 21:45:01] step 483100/1000000 | loss=2.1331 vocab=1.0959 token=1.5529 | 6.17 steps/s | VRAM=32.06 Go [2026-07-17 21:45:09] step 483150/1000000 | loss=2.2848 vocab=1.3840 token=1.5632 | 6.12 steps/s | VRAM=26.08 Go [2026-07-17 21:45:18] step 483200/1000000 | loss=2.2170 vocab=1.1849 token=1.4035 | 6.14 steps/s | VRAM=23.63 Go [2026-07-17 21:45:26] step 483250/1000000 | loss=2.2052 vocab=1.1459 token=1.6891 | 6.18 steps/s | VRAM=25.33 Go [2026-07-17 21:45:34] step 483300/1000000 | loss=2.1148 vocab=0.7885 token=1.1704 | 6.14 steps/s | VRAM=26.24 Go [2026-07-17 21:45:42] step 483350/1000000 | loss=2.2129 vocab=1.2211 token=1.1314 | 6.14 steps/s | VRAM=32.61 Go [2026-07-17 21:45:50] step 483400/1000000 | loss=2.1669 vocab=1.2493 token=1.2322 | 6.15 steps/s | VRAM=29.23 Go [2026-07-17 21:45:58] step 483450/1000000 | loss=2.3697 vocab=1.4057 token=1.1008 | 6.12 steps/s | VRAM=25.71 Go [2026-07-17 21:46:06] step 483500/1000000 | loss=2.3608 vocab=1.0014 token=1.8406 | 6.18 steps/s | VRAM=31.85 Go [2026-07-17 21:46:14] step 483550/1000000 | loss=2.2757 vocab=0.9918 token=1.4564 | 6.17 steps/s | VRAM=30.86 Go [2026-07-17 21:46:23] step 483600/1000000 | loss=2.2957 vocab=1.4867 token=1.6610 | 6.16 steps/s | VRAM=28.44 Go [2026-07-17 21:46:31] step 483650/1000000 | loss=2.1845 vocab=0.9143 token=1.5593 | 6.15 steps/s | VRAM=24.84 Go [2026-07-17 21:46:39] step 483700/1000000 | loss=2.3004 vocab=1.1690 token=0.9246 | 6.10 steps/s | VRAM=26.64 Go [2026-07-17 21:46:47] step 483750/1000000 | loss=2.2178 vocab=1.1421 token=1.7682 | 6.17 steps/s | VRAM=25.05 Go [2026-07-17 21:46:55] step 483800/1000000 | loss=2.2553 vocab=1.1910 token=1.3243 | 6.18 steps/s | VRAM=27.90 Go [2026-07-17 21:47:03] step 483850/1000000 | loss=2.1747 vocab=0.8859 token=1.6723 | 6.18 steps/s | VRAM=29.30 Go [2026-07-17 21:47:11] step 483900/1000000 | loss=2.1926 vocab=1.0628 token=1.2030 | 6.16 steps/s | VRAM=31.88 Go [2026-07-17 21:47:19] step 483950/1000000 | loss=2.3013 vocab=0.7197 token=1.4606 | 6.19 steps/s | VRAM=31.89 Go [2026-07-17 21:47:28] step 484000/1000000 | loss=2.2328 vocab=1.1230 token=1.4139 | 6.16 steps/s | VRAM=28.74 Go [2026-07-17 21:47:36] step 484050/1000000 | loss=2.2896 vocab=1.0674 token=1.4443 | 6.08 steps/s | VRAM=27.05 Go [2026-07-17 21:47:44] step 484100/1000000 | loss=2.2993 vocab=0.6287 token=1.3605 | 6.16 steps/s | VRAM=31.34 Go [2026-07-17 21:47:52] step 484150/1000000 | loss=2.3073 vocab=0.8521 token=1.0421 | 6.12 steps/s | VRAM=28.93 Go [2026-07-17 21:48:00] step 484200/1000000 | loss=2.2356 vocab=0.7159 token=1.0661 | 6.15 steps/s | VRAM=30.40 Go [2026-07-17 21:48:08] step 484250/1000000 | loss=2.2891 vocab=0.3359 token=1.2847 | 6.15 steps/s | VRAM=31.35 Go [2026-07-17 21:48:16] step 484300/1000000 | loss=2.2376 vocab=0.9408 token=1.2280 | 6.14 steps/s | VRAM=25.75 Go [2026-07-17 21:48:25] step 484350/1000000 | loss=2.2375 vocab=1.0685 token=1.1994 | 6.15 steps/s | VRAM=29.27 Go [2026-07-17 21:48:33] step 484400/1000000 | loss=2.2069 vocab=0.7345 token=1.2865 | 6.16 steps/s | VRAM=32.70 Go [2026-07-17 21:48:41] step 484450/1000000 | loss=2.2854 vocab=1.0837 token=1.3795 | 6.12 steps/s | VRAM=28.53 Go [2026-07-17 21:48:49] step 484500/1000000 | loss=2.2480 vocab=0.8540 token=1.2415 | 6.16 steps/s | VRAM=32.16 Go [2026-07-17 21:48:57] step 484550/1000000 | loss=2.1415 vocab=0.8385 token=1.1145 | 6.17 steps/s | VRAM=32.42 Go [2026-07-17 21:49:05] step 484600/1000000 | loss=2.2461 vocab=0.6550 token=0.9580 | 6.16 steps/s | VRAM=29.28 Go [2026-07-17 21:49:13] step 484650/1000000 | loss=2.2212 vocab=1.0031 token=1.1291 | 6.14 steps/s | VRAM=27.74 Go [2026-07-17 21:49:21] step 484700/1000000 | loss=2.1725 vocab=0.8938 token=1.4872 | 6.21 steps/s | VRAM=27.11 Go [2026-07-17 21:49:29] step 484750/1000000 | loss=2.2821 vocab=0.9361 token=1.1418 | 6.17 steps/s | VRAM=24.09 Go [2026-07-17 21:49:38] step 484800/1000000 | loss=2.2415 vocab=0.9325 token=1.3333 | 6.16 steps/s | VRAM=26.89 Go [2026-07-17 21:49:46] step 484850/1000000 | loss=2.3054 vocab=0.7350 token=1.0349 | 6.16 steps/s | VRAM=29.09 Go [2026-07-17 21:49:54] step 484900/1000000 | loss=2.3135 vocab=0.9043 token=0.9851 | 6.15 steps/s | VRAM=26.60 Go [2026-07-17 21:50:02] step 484950/1000000 | loss=2.2279 vocab=0.9726 token=0.9993 | 6.14 steps/s | VRAM=28.31 Go [2026-07-17 21:50:10] step 485000/1000000 | loss=2.2926 vocab=1.0910 token=1.2619 | 6.15 steps/s | VRAM=26.33 Go [2026-07-17 21:50:18] step 485050/1000000 | loss=2.3285 vocab=1.1704 token=1.8304 | 6.15 steps/s | VRAM=24.47 Go [2026-07-17 21:50:26] step 485100/1000000 | loss=2.2269 vocab=0.7542 token=1.0348 | 6.15 steps/s | VRAM=26.69 Go [2026-07-17 21:50:34] step 485150/1000000 | loss=2.2323 vocab=1.4414 token=1.3868 | 6.16 steps/s | VRAM=31.11 Go [2026-07-17 21:50:43] step 485200/1000000 | loss=2.3135 vocab=1.0812 token=1.2328 | 6.15 steps/s | VRAM=32.23 Go [2026-07-17 21:50:51] step 485250/1000000 | loss=2.3407 vocab=1.2725 token=1.6495 | 6.18 steps/s | VRAM=31.44 Go [2026-07-17 21:50:59] step 485300/1000000 | loss=2.1879 vocab=0.4966 token=1.1803 | 6.18 steps/s | VRAM=31.58 Go [2026-07-17 21:51:07] step 485350/1000000 | loss=2.1571 vocab=1.0805 token=1.5221 | 6.17 steps/s | VRAM=25.78 Go [2026-07-17 21:51:15] step 485400/1000000 | loss=2.2036 vocab=1.5693 token=1.4786 | 6.14 steps/s | VRAM=31.43 Go [2026-07-17 21:51:23] step 485450/1000000 | loss=2.2160 vocab=1.2259 token=1.3020 | 6.12 steps/s | VRAM=32.47 Go [2026-07-17 21:51:31] step 485500/1000000 | loss=2.2721 vocab=0.9675 token=1.0218 | 6.14 steps/s | VRAM=29.38 Go [2026-07-17 21:51:39] step 485550/1000000 | loss=2.1582 vocab=1.1853 token=1.2136 | 6.17 steps/s | VRAM=32.29 Go [2026-07-17 21:51:48] step 485600/1000000 | loss=2.2504 vocab=0.7407 token=1.1741 | 6.12 steps/s | VRAM=29.11 Go [2026-07-17 21:51:56] step 485650/1000000 | loss=2.2254 vocab=1.4836 token=1.4648 | 6.12 steps/s | VRAM=28.71 Go [2026-07-17 21:52:04] step 485700/1000000 | loss=2.2096 vocab=1.3833 token=1.6684 | 6.15 steps/s | VRAM=32.47 Go [2026-07-17 21:52:12] step 485750/1000000 | loss=2.2531 vocab=0.9834 token=1.3705 | 6.17 steps/s | VRAM=31.52 Go [2026-07-17 21:52:20] step 485800/1000000 | loss=2.2116 vocab=0.6168 token=1.3474 | 6.13 steps/s | VRAM=26.26 Go [2026-07-17 21:52:28] step 485850/1000000 | loss=2.2582 vocab=0.7624 token=0.8226 | 6.17 steps/s | VRAM=32.21 Go [2026-07-17 21:52:36] step 485900/1000000 | loss=2.2788 vocab=1.4063 token=1.3431 | 6.16 steps/s | VRAM=31.91 Go [2026-07-17 21:52:45] step 485950/1000000 | loss=2.1921 vocab=1.2445 token=1.4857 | 6.17 steps/s | VRAM=32.48 Go [2026-07-17 21:52:53] step 486000/1000000 | loss=2.3235 vocab=0.7771 token=1.0401 | 6.14 steps/s | VRAM=28.07 Go [2026-07-17 21:53:01] step 486050/1000000 | loss=2.2381 vocab=0.8637 token=1.6088 | 6.14 steps/s | VRAM=29.05 Go [2026-07-17 21:53:09] step 486100/1000000 | loss=2.2364 vocab=1.2480 token=1.8806 | 6.14 steps/s | VRAM=28.42 Go [2026-07-17 21:53:17] step 486150/1000000 | loss=2.2761 vocab=0.8569 token=1.3807 | 6.13 steps/s | VRAM=32.14 Go [2026-07-17 21:53:25] step 486200/1000000 | loss=2.2599 vocab=0.7038 token=1.3449 | 6.11 steps/s | VRAM=28.28 Go [2026-07-17 21:53:33] step 486250/1000000 | loss=2.3680 vocab=1.1117 token=1.7584 | 6.16 steps/s | VRAM=25.95 Go [2026-07-17 21:53:42] step 486300/1000000 | loss=2.2327 vocab=0.7385 token=1.2412 | 6.15 steps/s | VRAM=32.33 Go [2026-07-17 21:53:50] step 486350/1000000 | loss=2.2888 vocab=0.9018 token=1.3309 | 6.12 steps/s | VRAM=28.93 Go [2026-07-17 21:53:58] step 486400/1000000 | loss=2.2051 vocab=0.9994 token=1.4560 | 6.13 steps/s | VRAM=29.40 Go [2026-07-17 21:54:06] step 486450/1000000 | loss=2.2614 vocab=0.7453 token=1.2962 | 6.15 steps/s | VRAM=29.56 Go [2026-07-17 21:54:14] step 486500/1000000 | loss=2.2385 vocab=1.2093 token=1.6142 | 6.11 steps/s | VRAM=29.39 Go [2026-07-17 21:54:22] step 486550/1000000 | loss=2.2289 vocab=0.9255 token=1.1733 | 6.16 steps/s | VRAM=31.92 Go [2026-07-17 21:54:30] step 486600/1000000 | loss=2.2458 vocab=1.0533 token=1.4412 | 6.09 steps/s | VRAM=29.14 Go [2026-07-17 21:54:39] step 486650/1000000 | loss=2.2554 vocab=0.6674 token=1.4068 | 6.16 steps/s | VRAM=31.86 Go [2026-07-17 21:54:47] step 486700/1000000 | loss=2.2147 vocab=1.1276 token=1.4428 | 6.13 steps/s | VRAM=29.80 Go [2026-07-17 21:54:55] step 486750/1000000 | loss=2.3665 vocab=1.0623 token=1.5442 | 6.16 steps/s | VRAM=27.75 Go [2026-07-17 21:55:03] step 486800/1000000 | loss=2.2360 vocab=0.6900 token=1.1805 | 6.11 steps/s | VRAM=25.03 Go [2026-07-17 21:55:11] step 486850/1000000 | loss=2.3177 vocab=0.9076 token=1.4912 | 6.13 steps/s | VRAM=26.97 Go [2026-07-17 21:55:19] step 486900/1000000 | loss=2.2402 vocab=0.6687 token=1.1649 | 6.15 steps/s | VRAM=28.50 Go [2026-07-17 21:55:27] step 486950/1000000 | loss=2.1552 vocab=1.3993 token=1.4314 | 6.13 steps/s | VRAM=29.37 Go [2026-07-17 21:55:36] step 487000/1000000 | loss=2.2962 vocab=0.4600 token=1.1221 | 6.14 steps/s | VRAM=27.63 Go [2026-07-17 21:55:44] step 487050/1000000 | loss=2.1954 vocab=0.6356 token=1.1830 | 6.13 steps/s | VRAM=29.13 Go [2026-07-17 21:55:52] step 487100/1000000 | loss=2.1901 vocab=1.0765 token=1.1268 | 6.14 steps/s | VRAM=25.52 Go [2026-07-17 21:56:00] step 487150/1000000 | loss=2.2957 vocab=1.0007 token=1.4143 | 6.13 steps/s | VRAM=32.45 Go [2026-07-17 21:56:08] step 487200/1000000 | loss=2.2244 vocab=0.9150 token=1.3487 | 6.15 steps/s | VRAM=28.81 Go [2026-07-17 21:56:16] step 487250/1000000 | loss=2.1755 vocab=0.6729 token=1.2836 | 6.14 steps/s | VRAM=28.12 Go [2026-07-17 21:56:25] step 487300/1000000 | loss=2.1943 vocab=1.1381 token=1.6022 | 6.12 steps/s | VRAM=32.39 Go [2026-07-17 21:56:33] step 487350/1000000 | loss=2.2920 vocab=1.3136 token=1.0861 | 6.10 steps/s | VRAM=28.16 Go [2026-07-17 21:56:41] step 487400/1000000 | loss=2.1486 vocab=1.0174 token=1.4315 | 6.14 steps/s | VRAM=31.92 Go [2026-07-17 21:56:49] step 487450/1000000 | loss=2.1641 vocab=1.0953 token=1.3668 | 6.12 steps/s | VRAM=28.45 Go [2026-07-17 21:56:57] step 487500/1000000 | loss=2.2823 vocab=0.6201 token=1.2999 | 6.10 steps/s | VRAM=27.38 Go [2026-07-17 21:57:05] step 487550/1000000 | loss=2.3208 vocab=0.4737 token=1.2346 | 6.14 steps/s | VRAM=26.79 Go [2026-07-17 21:57:14] step 487600/1000000 | loss=2.1585 vocab=0.9242 token=1.3691 | 6.10 steps/s | VRAM=28.13 Go [2026-07-17 21:57:22] step 487650/1000000 | loss=2.2581 vocab=0.8827 token=1.1274 | 6.11 steps/s | VRAM=23.73 Go [2026-07-17 21:57:30] step 487700/1000000 | loss=2.2717 vocab=0.6206 token=1.3692 | 6.16 steps/s | VRAM=32.82 Go [2026-07-17 21:57:38] step 487750/1000000 | loss=2.3040 vocab=0.7116 token=1.2432 | 6.10 steps/s | VRAM=28.83 Go [2026-07-17 21:57:46] step 487800/1000000 | loss=2.1502 vocab=0.8372 token=1.3644 | 6.15 steps/s | VRAM=26.50 Go [2026-07-17 21:57:54] step 487850/1000000 | loss=2.1649 vocab=1.1117 token=0.9725 | 6.15 steps/s | VRAM=28.37 Go [2026-07-17 21:58:02] step 487900/1000000 | loss=2.1764 vocab=0.8204 token=1.3023 | 6.13 steps/s | VRAM=26.39 Go [2026-07-17 21:58:11] step 487950/1000000 | loss=2.1483 vocab=0.5355 token=0.9416 | 6.15 steps/s | VRAM=32.93 Go [2026-07-17 21:58:19] step 488000/1000000 | loss=2.2783 vocab=0.5184 token=1.2929 | 6.12 steps/s | VRAM=25.63 Go [2026-07-17 21:58:27] step 488050/1000000 | loss=2.2880 vocab=0.9886 token=1.1995 | 6.14 steps/s | VRAM=28.01 Go [2026-07-17 21:58:35] step 488100/1000000 | loss=2.3000 vocab=1.0070 token=1.5330 | 6.13 steps/s | VRAM=32.07 Go [2026-07-17 21:58:43] step 488150/1000000 | loss=2.2787 vocab=1.1146 token=1.4510 | 6.14 steps/s | VRAM=30.13 Go [2026-07-17 21:58:51] step 488200/1000000 | loss=2.3149 vocab=1.2113 token=1.4818 | 6.14 steps/s | VRAM=28.33 Go [2026-07-17 21:59:00] step 488250/1000000 | loss=2.2236 vocab=0.9764 token=1.4842 | 6.12 steps/s | VRAM=29.09 Go [2026-07-17 21:59:08] step 488300/1000000 | loss=2.2156 vocab=0.9265 token=1.8030 | 6.15 steps/s | VRAM=31.38 Go [2026-07-17 21:59:16] step 488350/1000000 | loss=2.2284 vocab=1.3392 token=1.0832 | 6.11 steps/s | VRAM=24.89 Go [2026-07-17 21:59:24] step 488400/1000000 | loss=2.3246 vocab=0.8339 token=1.6561 | 6.12 steps/s | VRAM=29.96 Go [2026-07-17 21:59:32] step 488450/1000000 | loss=2.2606 vocab=0.6154 token=1.0618 | 6.13 steps/s | VRAM=26.21 Go [2026-07-17 21:59:40] step 488500/1000000 | loss=2.2986 vocab=0.7833 token=1.4270 | 6.08 steps/s | VRAM=28.27 Go [2026-07-17 21:59:49] step 488550/1000000 | loss=2.2462 vocab=0.8492 token=1.1987 | 6.13 steps/s | VRAM=25.35 Go [2026-07-17 21:59:57] step 488600/1000000 | loss=2.2347 vocab=1.0858 token=1.3306 | 6.16 steps/s | VRAM=25.73 Go [2026-07-17 22:00:05] step 488650/1000000 | loss=2.3142 vocab=0.8224 token=0.9702 | 6.14 steps/s | VRAM=28.87 Go [2026-07-17 22:00:13] step 488700/1000000 | loss=2.2660 vocab=1.4498 token=0.9804 | 6.14 steps/s | VRAM=25.28 Go [2026-07-17 22:00:21] step 488750/1000000 | loss=2.1893 vocab=1.6375 token=1.5072 | 6.14 steps/s | VRAM=32.26 Go [2026-07-17 22:00:29] step 488800/1000000 | loss=2.2042 vocab=0.7545 token=1.3585 | 6.12 steps/s | VRAM=29.74 Go [2026-07-17 22:00:37] step 488850/1000000 | loss=2.1714 vocab=0.6202 token=1.3462 | 6.13 steps/s | VRAM=28.64 Go [2026-07-17 22:00:46] step 488900/1000000 | loss=2.2895 vocab=0.7790 token=1.1620 | 6.14 steps/s | VRAM=26.97 Go [2026-07-17 22:00:54] step 488950/1000000 | loss=2.1514 vocab=0.8940 token=1.3715 | 6.13 steps/s | VRAM=32.94 Go [2026-07-17 22:01:02] step 489000/1000000 | loss=2.1525 vocab=0.7955 token=0.9369 | 6.14 steps/s | VRAM=25.48 Go [2026-07-17 22:01:10] step 489050/1000000 | loss=2.2037 vocab=0.8391 token=0.8599 | 6.17 steps/s | VRAM=28.00 Go [2026-07-17 22:01:18] step 489100/1000000 | loss=2.3749 vocab=1.0710 token=1.5778 | 6.15 steps/s | VRAM=25.26 Go [2026-07-17 22:01:26] step 489150/1000000 | loss=2.2228 vocab=0.6223 token=1.2956 | 6.15 steps/s | VRAM=28.53 Go [2026-07-17 22:01:34] step 489200/1000000 | loss=2.1371 vocab=1.3891 token=1.1493 | 6.13 steps/s | VRAM=28.96 Go [2026-07-17 22:01:43] step 489250/1000000 | loss=2.2037 vocab=1.0840 token=1.5762 | 6.11 steps/s | VRAM=25.87 Go [2026-07-17 22:01:51] step 489300/1000000 | loss=2.2264 vocab=0.6304 token=1.1586 | 6.15 steps/s | VRAM=25.62 Go [2026-07-17 22:01:59] step 489350/1000000 | loss=2.2998 vocab=0.7517 token=1.2207 | 6.12 steps/s | VRAM=29.49 Go [2026-07-17 22:02:07] step 489400/1000000 | loss=2.2387 vocab=0.7885 token=1.1051 | 6.14 steps/s | VRAM=29.36 Go [2026-07-17 22:02:15] step 489450/1000000 | loss=2.1239 vocab=0.8131 token=1.4700 | 6.13 steps/s | VRAM=28.74 Go [2026-07-17 22:02:23] step 489500/1000000 | loss=2.2183 vocab=0.6229 token=0.7991 | 6.14 steps/s | VRAM=29.14 Go [2026-07-17 22:02:31] step 489550/1000000 | loss=2.1199 vocab=0.5341 token=0.9818 | 6.14 steps/s | VRAM=25.16 Go [2026-07-17 22:02:40] step 489600/1000000 | loss=2.3966 vocab=0.9625 token=1.3213 | 6.13 steps/s | VRAM=32.82 Go [2026-07-17 22:02:48] step 489650/1000000 | loss=2.2270 vocab=0.9557 token=1.4794 | 6.14 steps/s | VRAM=31.58 Go [2026-07-17 22:02:56] step 489700/1000000 | loss=2.1945 vocab=1.0588 token=1.4114 | 6.12 steps/s | VRAM=28.10 Go [2026-07-17 22:03:04] step 489750/1000000 | loss=2.2016 vocab=0.9693 token=1.1323 | 6.13 steps/s | VRAM=29.62 Go [2026-07-17 22:03:12] step 489800/1000000 | loss=2.2510 vocab=0.7560 token=1.2060 | 6.13 steps/s | VRAM=24.09 Go [2026-07-17 22:03:20] step 489850/1000000 | loss=2.2365 vocab=0.8654 token=1.1274 | 6.13 steps/s | VRAM=27.05 Go [2026-07-17 22:03:29] step 489900/1000000 | loss=2.1657 vocab=0.8742 token=1.4506 | 6.17 steps/s | VRAM=30.60 Go [2026-07-17 22:03:37] step 489950/1000000 | loss=2.2633 vocab=0.9874 token=1.3924 | 6.15 steps/s | VRAM=28.68 Go [2026-07-17 22:03:45] step 490000/1000000 | loss=2.1066 vocab=0.4509 token=1.2826 | 6.13 steps/s | VRAM=28.65 Go [2026-07-17 22:03:53] step 490050/1000000 | loss=2.2501 vocab=0.5794 token=0.9458 | 6.10 steps/s | VRAM=29.11 Go [2026-07-17 22:04:01] step 490100/1000000 | loss=2.3113 vocab=0.6438 token=1.5392 | 6.10 steps/s | VRAM=28.53 Go [2026-07-17 22:04:09] step 490150/1000000 | loss=2.2495 vocab=0.7506 token=1.5536 | 6.13 steps/s | VRAM=32.20 Go [2026-07-17 22:04:18] step 490200/1000000 | loss=2.2677 vocab=0.9924 token=1.3208 | 6.14 steps/s | VRAM=29.45 Go [2026-07-17 22:04:26] step 490250/1000000 | loss=2.3119 vocab=0.6502 token=1.6341 | 6.18 steps/s | VRAM=31.36 Go [2026-07-17 22:04:34] step 490300/1000000 | loss=2.2897 vocab=0.8877 token=1.4351 | 6.17 steps/s | VRAM=32.11 Go [2026-07-17 22:04:42] step 490350/1000000 | loss=2.2399 vocab=0.8970 token=1.4492 | 6.15 steps/s | VRAM=28.55 Go [2026-07-17 22:04:50] step 490400/1000000 | loss=2.1694 vocab=0.7714 token=1.2435 | 6.18 steps/s | VRAM=30.21 Go [2026-07-17 22:04:58] step 490450/1000000 | loss=2.2084 vocab=1.4054 token=1.5616 | 6.14 steps/s | VRAM=29.64 Go [2026-07-17 22:05:06] step 490500/1000000 | loss=2.3717 vocab=0.7705 token=1.2080 | 6.15 steps/s | VRAM=29.24 Go [2026-07-17 22:05:14] step 490550/1000000 | loss=2.1625 vocab=1.2880 token=1.2954 | 6.13 steps/s | VRAM=26.52 Go [2026-07-17 22:05:22] step 490600/1000000 | loss=2.4028 vocab=1.0949 token=1.4446 | 6.15 steps/s | VRAM=29.73 Go [2026-07-17 22:05:31] step 490650/1000000 | loss=2.3297 vocab=0.9485 token=1.3402 | 6.14 steps/s | VRAM=32.80 Go [2026-07-17 22:05:39] step 490700/1000000 | loss=2.2705 vocab=0.9840 token=1.4207 | 6.13 steps/s | VRAM=27.75 Go [2026-07-17 22:05:47] step 490750/1000000 | loss=2.2737 vocab=0.8132 token=1.5176 | 6.10 steps/s | VRAM=28.62 Go [2026-07-17 22:05:55] step 490800/1000000 | loss=2.3299 vocab=0.6522 token=1.1810 | 6.17 steps/s | VRAM=25.07 Go [2026-07-17 22:06:03] step 490850/1000000 | loss=2.2894 vocab=1.3329 token=1.5858 | 6.11 steps/s | VRAM=29.94 Go [2026-07-17 22:06:11] step 490900/1000000 | loss=2.2612 vocab=0.7368 token=1.4047 | 6.14 steps/s | VRAM=27.81 Go [2026-07-17 22:06:20] step 490950/1000000 | loss=2.3089 vocab=0.7391 token=1.4081 | 6.15 steps/s | VRAM=31.26 Go [2026-07-17 22:06:28] step 491000/1000000 | loss=2.1725 vocab=0.6914 token=1.1183 | 6.17 steps/s | VRAM=24.90 Go [2026-07-17 22:06:36] step 491050/1000000 | loss=2.1435 vocab=0.7582 token=1.2374 | 6.14 steps/s | VRAM=27.78 Go [2026-07-17 22:06:44] step 491100/1000000 | loss=2.1953 vocab=1.0021 token=1.3472 | 6.18 steps/s | VRAM=23.66 Go [2026-07-17 22:06:52] step 491150/1000000 | loss=2.1920 vocab=1.1766 token=1.4253 | 6.17 steps/s | VRAM=28.09 Go [2026-07-17 22:07:00] step 491200/1000000 | loss=2.2277 vocab=0.8611 token=1.2239 | 6.17 steps/s | VRAM=31.20 Go [2026-07-17 22:07:08] step 491250/1000000 | loss=2.2045 vocab=0.7092 token=1.5837 | 6.16 steps/s | VRAM=29.47 Go [2026-07-17 22:07:16] step 491300/1000000 | loss=2.2394 vocab=0.8287 token=1.3003 | 6.19 steps/s | VRAM=25.12 Go [2026-07-17 22:07:24] step 491350/1000000 | loss=2.2123 vocab=0.7369 token=0.9712 | 6.18 steps/s | VRAM=26.98 Go [2026-07-17 22:07:33] step 491400/1000000 | loss=2.1891 vocab=0.9628 token=1.1865 | 6.13 steps/s | VRAM=29.41 Go [2026-07-17 22:07:41] step 491450/1000000 | loss=2.2283 vocab=0.6926 token=1.4514 | 6.15 steps/s | VRAM=28.94 Go [2026-07-17 22:07:49] step 491500/1000000 | loss=2.2648 vocab=0.9617 token=1.4273 | 6.12 steps/s | VRAM=29.33 Go [2026-07-17 22:07:57] step 491550/1000000 | loss=2.1613 vocab=0.8724 token=1.2256 | 6.15 steps/s | VRAM=31.89 Go [2026-07-17 22:08:05] step 491600/1000000 | loss=2.2666 vocab=0.5507 token=1.4983 | 6.14 steps/s | VRAM=26.00 Go [2026-07-17 22:08:13] step 491650/1000000 | loss=2.2622 vocab=0.4634 token=1.1139 | 6.16 steps/s | VRAM=25.37 Go [2026-07-17 22:08:21] step 491700/1000000 | loss=2.2395 vocab=0.9998 token=1.0535 | 6.15 steps/s | VRAM=32.13 Go [2026-07-17 22:08:30] step 491750/1000000 | loss=2.2493 vocab=0.7749 token=1.2916 | 6.10 steps/s | VRAM=26.53 Go [2026-07-17 22:08:38] step 491800/1000000 | loss=2.2354 vocab=0.8088 token=1.3256 | 6.13 steps/s | VRAM=28.37 Go [2026-07-17 22:08:46] step 491850/1000000 | loss=2.3143 vocab=0.9077 token=1.1744 | 6.16 steps/s | VRAM=32.93 Go [2026-07-17 22:08:54] step 491900/1000000 | loss=2.1467 vocab=0.6515 token=0.7691 | 6.14 steps/s | VRAM=28.58 Go [2026-07-17 22:09:02] step 491950/1000000 | loss=2.3041 vocab=1.1735 token=1.3932 | 6.13 steps/s | VRAM=31.81 Go [2026-07-17 22:09:10] step 492000/1000000 | loss=2.2434 vocab=0.5618 token=0.9000 | 6.12 steps/s | VRAM=32.65 Go [2026-07-17 22:09:18] step 492050/1000000 | loss=2.2790 vocab=1.1439 token=1.3928 | 6.15 steps/s | VRAM=26.23 Go [2026-07-17 22:09:27] step 492100/1000000 | loss=2.2591 vocab=0.8017 token=1.1714 | 6.13 steps/s | VRAM=28.81 Go [2026-07-17 22:09:35] step 492150/1000000 | loss=2.3627 vocab=1.6085 token=1.5137 | 6.14 steps/s | VRAM=30.96 Go [2026-07-17 22:09:43] step 492200/1000000 | loss=2.2255 vocab=0.9417 token=1.2416 | 6.13 steps/s | VRAM=23.90 Go [2026-07-17 22:09:51] step 492250/1000000 | loss=2.2603 vocab=0.9020 token=1.3263 | 6.14 steps/s | VRAM=24.94 Go [2026-07-17 22:09:59] step 492300/1000000 | loss=2.1576 vocab=0.5617 token=0.8313 | 6.11 steps/s | VRAM=27.77 Go [2026-07-17 22:10:07] step 492350/1000000 | loss=2.1727 vocab=0.7416 token=1.4623 | 6.13 steps/s | VRAM=29.15 Go [2026-07-17 22:10:15] step 492400/1000000 | loss=2.2418 vocab=1.2652 token=1.0872 | 6.15 steps/s | VRAM=27.33 Go [2026-07-17 22:10:24] step 492450/1000000 | loss=2.2742 vocab=2.1492 token=1.5442 | 6.13 steps/s | VRAM=28.66 Go [2026-07-17 22:10:32] step 492500/1000000 | loss=2.2411 vocab=0.4121 token=0.9870 | 6.12 steps/s | VRAM=31.47 Go [2026-07-17 22:10:40] step 492550/1000000 | loss=2.2226 vocab=0.6544 token=1.2112 | 6.13 steps/s | VRAM=29.88 Go [2026-07-17 22:10:48] step 492600/1000000 | loss=2.2702 vocab=0.8385 token=1.5738 | 6.10 steps/s | VRAM=28.81 Go [2026-07-17 22:10:56] step 492650/1000000 | loss=2.2521 vocab=0.6180 token=1.3353 | 6.13 steps/s | VRAM=28.34 Go [2026-07-17 22:11:04] step 492700/1000000 | loss=2.2698 vocab=0.6093 token=1.2475 | 6.14 steps/s | VRAM=28.03 Go [2026-07-17 22:11:13] step 492750/1000000 | loss=2.2641 vocab=1.0054 token=1.3633 | 6.12 steps/s | VRAM=27.84 Go [2026-07-17 22:11:21] step 492800/1000000 | loss=2.2341 vocab=1.3439 token=1.7060 | 6.14 steps/s | VRAM=27.07 Go [2026-07-17 22:11:29] step 492850/1000000 | loss=2.1907 vocab=1.0895 token=1.2672 | 6.13 steps/s | VRAM=32.61 Go [2026-07-17 22:11:37] step 492900/1000000 | loss=2.1580 vocab=0.9557 token=1.0565 | 6.09 steps/s | VRAM=28.37 Go [2026-07-17 22:11:45] step 492950/1000000 | loss=2.1234 vocab=0.5585 token=1.0604 | 6.12 steps/s | VRAM=26.24 Go [2026-07-17 22:11:53] step 493000/1000000 | loss=2.1520 vocab=1.0605 token=1.4664 | 6.11 steps/s | VRAM=26.24 Go [2026-07-17 22:12:02] step 493050/1000000 | loss=2.2549 vocab=1.2350 token=1.5059 | 6.12 steps/s | VRAM=25.60 Go [2026-07-17 22:12:10] step 493100/1000000 | loss=2.1932 vocab=0.6901 token=0.9101 | 6.16 steps/s | VRAM=32.11 Go [2026-07-17 22:12:18] step 493150/1000000 | loss=2.2167 vocab=0.4650 token=1.1732 | 6.16 steps/s | VRAM=32.18 Go [2026-07-17 22:12:26] step 493200/1000000 | loss=2.2862 vocab=0.9244 token=1.1311 | 6.13 steps/s | VRAM=32.33 Go [2026-07-17 22:12:34] step 493250/1000000 | loss=2.2681 vocab=0.5886 token=1.1067 | 6.16 steps/s | VRAM=31.60 Go [2026-07-17 22:12:42] step 493300/1000000 | loss=2.2208 vocab=0.9062 token=1.7519 | 6.13 steps/s | VRAM=24.96 Go [2026-07-17 22:12:50] step 493350/1000000 | loss=2.2657 vocab=0.6873 token=1.3411 | 6.14 steps/s | VRAM=31.34 Go [2026-07-17 22:12:59] step 493400/1000000 | loss=2.2017 vocab=1.1073 token=1.5141 | 6.13 steps/s | VRAM=28.73 Go [2026-07-17 22:13:07] step 493450/1000000 | loss=2.2533 vocab=0.9933 token=1.5083 | 6.10 steps/s | VRAM=26.88 Go [2026-07-17 22:13:15] step 493500/1000000 | loss=2.1953 vocab=0.8899 token=1.1871 | 6.16 steps/s | VRAM=24.58 Go [2026-07-17 22:13:23] step 493550/1000000 | loss=2.2087 vocab=0.9555 token=1.5023 | 6.12 steps/s | VRAM=25.97 Go [2026-07-17 22:13:31] step 493600/1000000 | loss=2.2195 vocab=0.9548 token=1.5453 | 6.11 steps/s | VRAM=29.12 Go [2026-07-17 22:13:39] step 493650/1000000 | loss=2.2020 vocab=0.7979 token=0.9101 | 6.11 steps/s | VRAM=29.22 Go [2026-07-17 22:13:48] step 493700/1000000 | loss=2.2424 vocab=0.8606 token=1.4634 | 6.10 steps/s | VRAM=29.21 Go [2026-07-17 22:13:56] step 493750/1000000 | loss=2.2220 vocab=0.7553 token=2.0666 | 6.11 steps/s | VRAM=28.62 Go [2026-07-17 22:14:04] step 493800/1000000 | loss=2.2298 vocab=0.9129 token=1.4413 | 6.15 steps/s | VRAM=31.31 Go [2026-07-17 22:14:12] step 493850/1000000 | loss=2.2446 vocab=0.9366 token=1.4808 | 6.12 steps/s | VRAM=28.35 Go [2026-07-17 22:14:20] step 493900/1000000 | loss=2.2609 vocab=0.9918 token=0.9961 | 6.13 steps/s | VRAM=31.66 Go [2026-07-17 22:14:28] step 493950/1000000 | loss=2.2669 vocab=0.6260 token=1.7627 | 6.11 steps/s | VRAM=26.54 Go [2026-07-17 22:14:37] step 494000/1000000 | loss=2.1366 vocab=0.5610 token=1.1744 | 6.13 steps/s | VRAM=28.05 Go [2026-07-17 22:14:45] step 494050/1000000 | loss=2.3233 vocab=1.8809 token=1.7976 | 6.12 steps/s | VRAM=26.83 Go [2026-07-17 22:14:53] step 494100/1000000 | loss=2.3318 vocab=1.1851 token=1.6412 | 6.14 steps/s | VRAM=29.11 Go [2026-07-17 22:15:01] step 494150/1000000 | loss=2.2802 vocab=1.0927 token=1.7684 | 6.14 steps/s | VRAM=29.31 Go [2026-07-17 22:15:09] step 494200/1000000 | loss=2.1847 vocab=1.1099 token=1.5522 | 6.14 steps/s | VRAM=28.53 Go [2026-07-17 22:15:17] step 494250/1000000 | loss=2.3663 vocab=0.8415 token=1.5692 | 6.10 steps/s | VRAM=29.55 Go [2026-07-17 22:15:26] step 494300/1000000 | loss=2.2722 vocab=0.6807 token=1.3461 | 6.15 steps/s | VRAM=25.22 Go [2026-07-17 22:15:34] step 494350/1000000 | loss=2.2165 vocab=0.7571 token=1.3405 | 6.14 steps/s | VRAM=31.40 Go [2026-07-17 22:15:42] step 494400/1000000 | loss=2.1953 vocab=0.8168 token=1.1166 | 6.12 steps/s | VRAM=29.15 Go [2026-07-17 22:15:50] step 494450/1000000 | loss=2.2324 vocab=0.5369 token=1.1846 | 6.11 steps/s | VRAM=30.03 Go [2026-07-17 22:15:58] step 494500/1000000 | loss=2.2081 vocab=1.3149 token=1.4128 | 6.15 steps/s | VRAM=32.59 Go [2026-07-17 22:16:06] step 494550/1000000 | loss=2.1970 vocab=0.6482 token=1.2817 | 6.11 steps/s | VRAM=26.93 Go [2026-07-17 22:16:14] step 494600/1000000 | loss=2.2559 vocab=0.6733 token=1.4490 | 6.16 steps/s | VRAM=24.25 Go [2026-07-17 22:16:23] step 494650/1000000 | loss=2.2771 vocab=0.6555 token=1.3084 | 6.17 steps/s | VRAM=31.99 Go [2026-07-17 22:16:31] step 494700/1000000 | loss=2.1636 vocab=1.0523 token=1.1799 | 6.13 steps/s | VRAM=29.70 Go [2026-07-17 22:16:39] step 494750/1000000 | loss=2.2129 vocab=0.8609 token=1.4584 | 6.12 steps/s | VRAM=27.27 Go [2026-07-17 22:16:47] step 494800/1000000 | loss=2.3576 vocab=0.5564 token=1.0725 | 6.13 steps/s | VRAM=26.07 Go [2026-07-17 22:16:55] step 494850/1000000 | loss=2.3329 vocab=0.9630 token=1.4027 | 6.16 steps/s | VRAM=31.38 Go [2026-07-17 22:17:03] step 494900/1000000 | loss=2.2757 vocab=0.6675 token=0.9723 | 6.15 steps/s | VRAM=25.56 Go [2026-07-17 22:17:12] step 494950/1000000 | loss=2.1795 vocab=0.9842 token=1.0055 | 6.10 steps/s | VRAM=29.28 Go [2026-07-17 22:17:20] step 495000/1000000 | loss=2.2863 vocab=0.7160 token=1.3274 | 6.13 steps/s | VRAM=32.17 Go [2026-07-17 22:17:28] step 495050/1000000 | loss=2.2648 vocab=0.5939 token=1.0241 | 6.13 steps/s | VRAM=27.53 Go [2026-07-17 22:17:36] step 495100/1000000 | loss=2.1334 vocab=1.0216 token=0.5833 | 6.15 steps/s | VRAM=26.78 Go [2026-07-17 22:17:44] step 495150/1000000 | loss=2.2979 vocab=0.5827 token=1.2062 | 6.15 steps/s | VRAM=24.86 Go [2026-07-17 22:17:52] step 495200/1000000 | loss=2.3161 vocab=1.1642 token=1.1285 | 6.15 steps/s | VRAM=31.67 Go [2026-07-17 22:18:00] step 495250/1000000 | loss=2.2824 vocab=1.0127 token=1.8019 | 6.14 steps/s | VRAM=29.27 Go [2026-07-17 22:18:09] step 495300/1000000 | loss=2.1736 vocab=0.7334 token=0.9566 | 6.14 steps/s | VRAM=25.61 Go [2026-07-17 22:18:17] step 495350/1000000 | loss=2.2509 vocab=1.0130 token=1.2962 | 6.13 steps/s | VRAM=28.75 Go [2026-07-17 22:18:25] step 495400/1000000 | loss=2.1775 vocab=0.7151 token=1.1433 | 6.12 steps/s | VRAM=26.29 Go [2026-07-17 22:18:33] step 495450/1000000 | loss=2.1729 vocab=0.8379 token=1.3033 | 6.13 steps/s | VRAM=28.37 Go [2026-07-17 22:18:41] step 495500/1000000 | loss=2.2557 vocab=0.8374 token=1.6891 | 6.15 steps/s | VRAM=32.76 Go [2026-07-17 22:18:49] step 495550/1000000 | loss=2.2782 vocab=1.1279 token=1.4115 | 6.14 steps/s | VRAM=28.35 Go [2026-07-17 22:18:57] step 495600/1000000 | loss=2.1311 vocab=0.8134 token=1.1491 | 6.16 steps/s | VRAM=32.16 Go [2026-07-17 22:19:06] step 495650/1000000 | loss=2.1736 vocab=0.9342 token=1.6609 | 6.15 steps/s | VRAM=29.28 Go [2026-07-17 22:19:14] step 495700/1000000 | loss=2.2016 vocab=0.6004 token=1.6130 | 6.13 steps/s | VRAM=28.71 Go [2026-07-17 22:19:22] step 495750/1000000 | loss=2.2848 vocab=0.9860 token=1.2610 | 6.13 steps/s | VRAM=29.58 Go [2026-07-17 22:19:30] step 495800/1000000 | loss=2.1430 vocab=0.5897 token=1.0845 | 6.13 steps/s | VRAM=26.02 Go [2026-07-17 22:19:38] step 495850/1000000 | loss=2.2705 vocab=0.6609 token=1.5325 | 6.15 steps/s | VRAM=28.62 Go [2026-07-17 22:19:46] step 495900/1000000 | loss=2.1468 vocab=1.2996 token=1.3018 | 6.16 steps/s | VRAM=31.11 Go [2026-07-17 22:19:54] step 495950/1000000 | loss=2.2504 vocab=0.8353 token=1.3439 | 6.08 steps/s | VRAM=27.10 Go [2026-07-17 22:20:03] step 496000/1000000 | loss=2.2817 vocab=1.9207 token=2.0463 | 6.15 steps/s | VRAM=29.08 Go [2026-07-17 22:20:11] step 496050/1000000 | loss=2.2107 vocab=0.7049 token=1.2783 | 6.12 steps/s | VRAM=32.51 Go [2026-07-17 22:20:19] step 496100/1000000 | loss=2.3180 vocab=0.8618 token=1.5428 | 6.12 steps/s | VRAM=23.13 Go [2026-07-17 22:20:27] step 496150/1000000 | loss=2.2840 vocab=1.1943 token=1.5056 | 6.13 steps/s | VRAM=23.28 Go [2026-07-17 22:20:35] step 496200/1000000 | loss=2.3069 vocab=1.0396 token=1.5256 | 6.16 steps/s | VRAM=25.01 Go [2026-07-17 22:20:43] step 496250/1000000 | loss=2.2790 vocab=0.6464 token=1.1872 | 6.18 steps/s | VRAM=32.05 Go [2026-07-17 22:20:51] step 496300/1000000 | loss=2.2583 vocab=1.9352 token=1.4242 | 6.14 steps/s | VRAM=31.36 Go [2026-07-17 22:21:00] step 496350/1000000 | loss=2.2159 vocab=0.9295 token=1.2945 | 6.15 steps/s | VRAM=32.62 Go [2026-07-17 22:21:08] step 496400/1000000 | loss=2.2818 vocab=1.2368 token=1.3524 | 6.17 steps/s | VRAM=32.30 Go [2026-07-17 22:21:16] step 496450/1000000 | loss=2.2188 vocab=0.9082 token=1.2378 | 6.14 steps/s | VRAM=28.94 Go [2026-07-17 22:21:24] step 496500/1000000 | loss=2.2577 vocab=0.7416 token=1.2345 | 6.15 steps/s | VRAM=29.02 Go [2026-07-17 22:21:32] step 496550/1000000 | loss=2.2782 vocab=0.8991 token=0.9704 | 6.11 steps/s | VRAM=26.33 Go [2026-07-17 22:21:40] step 496600/1000000 | loss=2.1673 vocab=1.1609 token=1.5191 | 6.17 steps/s | VRAM=31.89 Go [2026-07-17 22:21:48] step 496650/1000000 | loss=2.3209 vocab=0.9385 token=1.4152 | 6.10 steps/s | VRAM=31.17 Go [2026-07-17 22:21:57] step 496700/1000000 | loss=2.2188 vocab=0.7256 token=1.3101 | 6.11 steps/s | VRAM=28.20 Go [2026-07-17 22:22:05] step 496750/1000000 | loss=2.2100 vocab=1.0516 token=1.7134 | 6.14 steps/s | VRAM=28.14 Go [2026-07-17 22:22:13] step 496800/1000000 | loss=2.3087 vocab=1.4332 token=1.7128 | 6.14 steps/s | VRAM=25.71 Go [2026-07-17 22:22:21] step 496850/1000000 | loss=2.2254 vocab=1.3835 token=1.1797 | 6.13 steps/s | VRAM=28.81 Go [2026-07-17 22:22:29] step 496900/1000000 | loss=2.1822 vocab=0.7258 token=1.5750 | 6.15 steps/s | VRAM=29.43 Go [2026-07-17 22:22:37] step 496950/1000000 | loss=2.2328 vocab=0.7195 token=1.5490 | 6.16 steps/s | VRAM=29.08 Go [2026-07-17 22:22:45] step 497000/1000000 | loss=2.2351 vocab=0.7099 token=1.6190 | 6.13 steps/s | VRAM=28.80 Go [2026-07-17 22:22:54] step 497050/1000000 | loss=2.3679 vocab=1.0764 token=1.2679 | 6.13 steps/s | VRAM=32.76 Go [2026-07-17 22:23:02] step 497100/1000000 | loss=2.1599 vocab=0.9744 token=1.5131 | 6.13 steps/s | VRAM=28.48 Go [2026-07-17 22:23:10] step 497150/1000000 | loss=2.2230 vocab=0.4111 token=1.0605 | 6.11 steps/s | VRAM=28.32 Go [2026-07-17 22:23:18] step 497200/1000000 | loss=2.3310 vocab=1.1685 token=1.6867 | 6.10 steps/s | VRAM=23.85 Go [2026-07-17 22:23:26] step 497250/1000000 | loss=2.1964 vocab=1.3472 token=1.3038 | 6.16 steps/s | VRAM=28.87 Go [2026-07-17 22:23:34] step 497300/1000000 | loss=2.1359 vocab=0.8271 token=1.2353 | 6.12 steps/s | VRAM=27.94 Go [2026-07-17 22:23:43] step 497350/1000000 | loss=2.1837 vocab=1.1940 token=1.5177 | 6.16 steps/s | VRAM=28.57 Go [2026-07-17 22:23:51] step 497400/1000000 | loss=2.2564 vocab=0.9064 token=1.3520 | 6.16 steps/s | VRAM=29.08 Go [2026-07-17 22:23:59] step 497450/1000000 | loss=2.1623 vocab=0.9578 token=1.2274 | 6.14 steps/s | VRAM=29.37 Go [2026-07-17 22:24:07] step 497500/1000000 | loss=2.3511 vocab=0.8012 token=1.5289 | 6.13 steps/s | VRAM=29.10 Go [2026-07-17 22:24:15] step 497550/1000000 | loss=2.2235 vocab=1.5121 token=1.7130 | 6.12 steps/s | VRAM=25.08 Go [2026-07-17 22:24:23] step 497600/1000000 | loss=2.2229 vocab=1.4534 token=1.3462 | 6.14 steps/s | VRAM=32.13 Go [2026-07-17 22:24:31] step 497650/1000000 | loss=2.3329 vocab=1.0557 token=1.2447 | 6.13 steps/s | VRAM=31.78 Go [2026-07-17 22:24:40] step 497700/1000000 | loss=2.2689 vocab=0.9013 token=1.4692 | 6.15 steps/s | VRAM=28.63 Go [2026-07-17 22:24:48] step 497750/1000000 | loss=2.3351 vocab=1.1898 token=1.5975 | 6.12 steps/s | VRAM=28.99 Go [2026-07-17 22:24:56] step 497800/1000000 | loss=2.2435 vocab=0.5453 token=0.9505 | 6.10 steps/s | VRAM=29.25 Go [2026-07-17 22:25:04] step 497850/1000000 | loss=2.2007 vocab=1.6013 token=1.5074 | 6.14 steps/s | VRAM=31.47 Go [2026-07-17 22:25:12] step 497900/1000000 | loss=2.2690 vocab=2.2600 token=1.4711 | 6.12 steps/s | VRAM=32.68 Go [2026-07-17 22:25:20] step 497950/1000000 | loss=2.2265 vocab=1.3102 token=1.0682 | 6.16 steps/s | VRAM=32.50 Go [2026-07-17 22:25:29] step 498000/1000000 | loss=2.1652 vocab=0.7444 token=1.2898 | 6.11 steps/s | VRAM=28.35 Go [2026-07-17 22:25:37] step 498050/1000000 | loss=2.3915 vocab=0.7791 token=1.1382 | 6.11 steps/s | VRAM=30.08 Go [2026-07-17 22:25:45] step 498100/1000000 | loss=2.2753 vocab=0.7526 token=1.2729 | 6.12 steps/s | VRAM=32.91 Go [2026-07-17 22:25:53] step 498150/1000000 | loss=2.2357 vocab=0.7650 token=1.4672 | 6.12 steps/s | VRAM=30.30 Go [2026-07-17 22:26:01] step 498200/1000000 | loss=2.2355 vocab=0.6712 token=1.2963 | 6.12 steps/s | VRAM=28.52 Go [2026-07-17 22:26:09] step 498250/1000000 | loss=2.3518 vocab=0.9741 token=1.2492 | 6.10 steps/s | VRAM=26.45 Go [2026-07-17 22:26:18] step 498300/1000000 | loss=2.2979 vocab=0.5888 token=1.2587 | 6.14 steps/s | VRAM=29.02 Go [2026-07-17 22:26:26] step 498350/1000000 | loss=2.1578 vocab=1.1702 token=1.3121 | 6.11 steps/s | VRAM=32.56 Go [2026-07-17 22:26:34] step 498400/1000000 | loss=2.1467 vocab=0.8417 token=1.4164 | 6.11 steps/s | VRAM=26.38 Go [2026-07-17 22:26:42] step 498450/1000000 | loss=2.2793 vocab=1.1020 token=1.3002 | 6.14 steps/s | VRAM=31.54 Go [2026-07-17 22:26:50] step 498500/1000000 | loss=2.3064 vocab=0.5330 token=1.6189 | 6.11 steps/s | VRAM=31.82 Go [2026-07-17 22:26:58] step 498550/1000000 | loss=2.2366 vocab=0.5336 token=0.7605 | 6.12 steps/s | VRAM=29.20 Go [2026-07-17 22:27:07] step 498600/1000000 | loss=2.2603 vocab=0.7077 token=1.2058 | 6.12 steps/s | VRAM=31.58 Go [2026-07-17 22:27:15] step 498650/1000000 | loss=2.2600 vocab=0.5866 token=1.2157 | 6.10 steps/s | VRAM=25.83 Go [2026-07-17 22:27:23] step 498700/1000000 | loss=2.2537 vocab=0.6382 token=1.2221 | 6.13 steps/s | VRAM=32.83 Go [2026-07-17 22:27:31] step 498750/1000000 | loss=2.2972 vocab=1.3419 token=1.4084 | 6.08 steps/s | VRAM=25.95 Go [2026-07-17 22:27:39] step 498800/1000000 | loss=2.2075 vocab=1.3366 token=1.2400 | 6.12 steps/s | VRAM=26.14 Go [2026-07-17 22:27:48] step 498850/1000000 | loss=2.2440 vocab=1.4328 token=1.3143 | 6.14 steps/s | VRAM=29.23 Go [2026-07-17 22:27:56] step 498900/1000000 | loss=2.2854 vocab=0.9534 token=1.2817 | 6.12 steps/s | VRAM=31.78 Go [2026-07-17 22:28:04] step 498950/1000000 | loss=2.2256 vocab=0.8502 token=1.2480 | 6.14 steps/s | VRAM=29.31 Go [2026-07-17 22:28:12] step 499000/1000000 | loss=2.1638 vocab=1.2471 token=1.4860 | 6.12 steps/s | VRAM=28.39 Go [2026-07-17 22:28:20] step 499050/1000000 | loss=2.2095 vocab=1.1715 token=1.0739 | 6.15 steps/s | VRAM=32.36 Go [2026-07-17 22:28:28] step 499100/1000000 | loss=2.3313 vocab=1.1354 token=1.4592 | 6.12 steps/s | VRAM=26.35 Go [2026-07-17 22:28:36] step 499150/1000000 | loss=2.1975 vocab=1.2357 token=1.5882 | 6.14 steps/s | VRAM=31.98 Go [2026-07-17 22:28:45] step 499200/1000000 | loss=2.3549 vocab=1.0159 token=1.6637 | 6.13 steps/s | VRAM=25.02 Go [2026-07-17 22:28:53] step 499250/1000000 | loss=2.3428 vocab=0.7238 token=1.4602 | 6.13 steps/s | VRAM=29.29 Go [2026-07-17 22:29:01] step 499300/1000000 | loss=2.2326 vocab=0.7138 token=1.3438 | 6.15 steps/s | VRAM=31.98 Go [2026-07-17 22:29:09] step 499350/1000000 | loss=2.2521 vocab=0.8254 token=1.4068 | 6.16 steps/s | VRAM=31.58 Go [2026-07-17 22:29:17] step 499400/1000000 | loss=2.2688 vocab=1.0614 token=1.6447 | 6.11 steps/s | VRAM=24.45 Go [2026-07-17 22:29:25] step 499450/1000000 | loss=2.2509 vocab=0.6939 token=1.2220 | 6.12 steps/s | VRAM=27.89 Go [2026-07-17 22:29:33] step 499500/1000000 | loss=2.2305 vocab=0.9056 token=1.4233 | 6.17 steps/s | VRAM=28.76 Go [2026-07-17 22:29:42] step 499550/1000000 | loss=2.1901 vocab=1.1611 token=1.7334 | 6.14 steps/s | VRAM=27.23 Go [2026-07-17 22:29:50] step 499600/1000000 | loss=2.1850 vocab=1.2816 token=1.4486 | 6.16 steps/s | VRAM=26.55 Go [2026-07-17 22:29:58] step 499650/1000000 | loss=2.3503 vocab=0.8538 token=1.6823 | 6.14 steps/s | VRAM=24.26 Go [2026-07-17 22:30:06] step 499700/1000000 | loss=2.2107 vocab=0.8195 token=1.4479 | 6.16 steps/s | VRAM=26.89 Go [2026-07-17 22:30:14] step 499750/1000000 | loss=2.2343 vocab=0.8734 token=1.5089 | 6.16 steps/s | VRAM=28.39 Go [2026-07-17 22:30:22] step 499800/1000000 | loss=2.1483 vocab=1.3452 token=1.5633 | 6.16 steps/s | VRAM=27.96 Go [2026-07-17 22:30:30] step 499850/1000000 | loss=2.3111 vocab=0.9959 token=1.4580 | 6.15 steps/s | VRAM=24.89 Go [2026-07-17 22:30:39] step 499900/1000000 | loss=2.1799 vocab=0.4607 token=0.9492 | 6.14 steps/s | VRAM=32.31 Go [2026-07-17 22:30:47] step 499950/1000000 | loss=2.2855 vocab=1.0704 token=1.0452 | 6.14 steps/s | VRAM=32.68 Go [2026-07-17 22:30:55] step 500000/1000000 | loss=2.1442 vocab=1.3606 token=1.2255 | 6.15 steps/s | VRAM=30.95 Go [push] step 500000 OK (best-of-2, loss_token=1.2255) [2026-07-17 22:32:26] step 500050/1000000 | loss=2.3041 vocab=1.0388 token=1.4289 | 0.55 steps/s | VRAM=28.43 Go [2026-07-17 22:32:34] step 500100/1000000 | loss=2.2936 vocab=0.8570 token=1.7228 | 6.20 steps/s | VRAM=32.21 Go [2026-07-17 22:32:42] step 500150/1000000 | loss=2.0766 vocab=0.7289 token=1.3785 | 6.22 steps/s | VRAM=32.58 Go [2026-07-17 22:32:51] step 500200/1000000 | loss=2.2921 vocab=1.3046 token=1.5457 | 6.21 steps/s | VRAM=29.21 Go [2026-07-17 22:32:59] step 500250/1000000 | loss=2.2874 vocab=1.0064 token=1.7115 | 6.18 steps/s | VRAM=28.37 Go [2026-07-17 22:33:07] step 500300/1000000 | loss=2.2631 vocab=0.5944 token=1.0480 | 6.16 steps/s | VRAM=28.80 Go [2026-07-17 22:33:15] step 500350/1000000 | loss=2.2393 vocab=0.9726 token=1.4217 | 6.14 steps/s | VRAM=28.42 Go [2026-07-17 22:33:23] step 500400/1000000 | loss=2.1055 vocab=0.4760 token=1.1235 | 6.19 steps/s | VRAM=27.39 Go [2026-07-17 22:33:31] step 500450/1000000 | loss=2.3400 vocab=0.7052 token=1.5160 | 6.15 steps/s | VRAM=29.54 Go [2026-07-17 22:33:39] step 500500/1000000 | loss=2.2286 vocab=1.1213 token=1.0881 | 6.16 steps/s | VRAM=25.45 Go [2026-07-17 22:33:47] step 500550/1000000 | loss=2.1909 vocab=0.5780 token=1.3334 | 6.16 steps/s | VRAM=25.98 Go [2026-07-17 22:33:55] step 500600/1000000 | loss=2.3458 vocab=0.8265 token=1.4730 | 6.11 steps/s | VRAM=29.90 Go [2026-07-17 22:34:04] step 500650/1000000 | loss=2.2536 vocab=1.1034 token=1.3978 | 6.13 steps/s | VRAM=26.45 Go [2026-07-17 22:34:12] step 500700/1000000 | loss=2.1612 vocab=0.7687 token=1.0441 | 6.13 steps/s | VRAM=32.83 Go [2026-07-17 22:34:20] step 500750/1000000 | loss=2.1711 vocab=0.9126 token=1.0429 | 6.10 steps/s | VRAM=25.66 Go [2026-07-17 22:34:28] step 500800/1000000 | loss=2.3579 vocab=0.8202 token=1.2026 | 6.15 steps/s | VRAM=31.67 Go [2026-07-17 22:34:36] step 500850/1000000 | loss=2.2468 vocab=1.0672 token=1.2163 | 6.14 steps/s | VRAM=31.55 Go [2026-07-17 22:34:44] step 500900/1000000 | loss=2.2483 vocab=0.8274 token=1.6693 | 6.11 steps/s | VRAM=28.43 Go [2026-07-17 22:34:53] step 500950/1000000 | loss=2.2255 vocab=0.9149 token=1.4759 | 6.12 steps/s | VRAM=26.76 Go [2026-07-17 22:35:01] step 501000/1000000 | loss=2.2511 vocab=1.0133 token=1.2613 | 6.15 steps/s | VRAM=26.39 Go [2026-07-17 22:35:09] step 501050/1000000 | loss=2.3050 vocab=1.0875 token=1.2708 | 6.15 steps/s | VRAM=28.11 Go [2026-07-17 22:35:17] step 501100/1000000 | loss=2.2859 vocab=0.9252 token=1.1401 | 6.13 steps/s | VRAM=29.40 Go [2026-07-17 22:35:25] step 501150/1000000 | loss=2.1839 vocab=0.7254 token=1.1274 | 6.16 steps/s | VRAM=32.29 Go [2026-07-17 22:35:33] step 501200/1000000 | loss=2.2616 vocab=1.4465 token=1.6293 | 6.12 steps/s | VRAM=28.18 Go [2026-07-17 22:35:41] step 501250/1000000 | loss=2.2038 vocab=1.1076 token=1.4946 | 6.14 steps/s | VRAM=32.86 Go [2026-07-17 22:35:50] step 501300/1000000 | loss=2.2520 vocab=0.5762 token=1.2277 | 6.16 steps/s | VRAM=27.95 Go [2026-07-17 22:35:58] step 501350/1000000 | loss=2.1519 vocab=0.8147 token=1.0800 | 6.14 steps/s | VRAM=28.57 Go [2026-07-17 22:36:06] step 501400/1000000 | loss=2.2919 vocab=1.1244 token=1.0065 | 6.15 steps/s | VRAM=32.23 Go [2026-07-17 22:36:14] step 501450/1000000 | loss=2.2074 vocab=0.8774 token=1.0639 | 6.14 steps/s | VRAM=32.43 Go [2026-07-17 22:36:22] step 501500/1000000 | loss=2.2307 vocab=1.0818 token=1.6286 | 6.14 steps/s | VRAM=31.75 Go [2026-07-17 22:36:30] step 501550/1000000 | loss=2.2762 vocab=0.8242 token=1.3123 | 6.16 steps/s | VRAM=31.46 Go [2026-07-17 22:36:38] step 501600/1000000 | loss=2.3228 vocab=0.4918 token=1.2838 | 6.10 steps/s | VRAM=24.62 Go [2026-07-17 22:36:47] step 501650/1000000 | loss=2.2891 vocab=0.8288 token=1.7619 | 6.16 steps/s | VRAM=28.06 Go [2026-07-17 22:36:55] step 501700/1000000 | loss=2.4445 vocab=1.1783 token=1.4437 | 6.12 steps/s | VRAM=29.01 Go [2026-07-17 22:37:03] step 501750/1000000 | loss=2.1659 vocab=0.8254 token=1.3439 | 6.13 steps/s | VRAM=26.00 Go [2026-07-17 22:37:11] step 501800/1000000 | loss=2.2360 vocab=0.4702 token=1.4930 | 6.14 steps/s | VRAM=29.43 Go [2026-07-17 22:37:19] step 501850/1000000 | loss=2.1590 vocab=1.1778 token=1.0378 | 6.11 steps/s | VRAM=28.28 Go [2026-07-17 22:37:27] step 501900/1000000 | loss=2.2516 vocab=1.0694 token=1.0938 | 6.13 steps/s | VRAM=26.55 Go [2026-07-17 22:37:36] step 501950/1000000 | loss=2.2121 vocab=1.4127 token=1.9742 | 6.15 steps/s | VRAM=32.31 Go [2026-07-17 22:37:44] step 502000/1000000 | loss=2.2001 vocab=0.7245 token=1.4558 | 6.12 steps/s | VRAM=29.33 Go [2026-07-17 22:37:52] step 502050/1000000 | loss=2.3206 vocab=0.9331 token=1.8714 | 6.13 steps/s | VRAM=32.35 Go [2026-07-17 22:38:00] step 502100/1000000 | loss=2.2671 vocab=1.1564 token=1.2018 | 6.13 steps/s | VRAM=32.38 Go [2026-07-17 22:38:08] step 502150/1000000 | loss=2.2704 vocab=1.2641 token=1.7901 | 6.13 steps/s | VRAM=29.34 Go [2026-07-17 22:38:16] step 502200/1000000 | loss=2.1471 vocab=1.0956 token=1.3463 | 6.17 steps/s | VRAM=29.44 Go [2026-07-17 22:38:24] step 502250/1000000 | loss=2.2841 vocab=1.2565 token=1.4176 | 6.17 steps/s | VRAM=31.19 Go [2026-07-17 22:38:33] step 502300/1000000 | loss=2.2663 vocab=1.3998 token=1.4767 | 6.15 steps/s | VRAM=25.79 Go [2026-07-17 22:38:41] step 502350/1000000 | loss=2.2244 vocab=0.5645 token=1.4021 | 6.15 steps/s | VRAM=26.44 Go [2026-07-17 22:38:49] step 502400/1000000 | loss=2.2626 vocab=0.9628 token=1.2601 | 6.16 steps/s | VRAM=32.20 Go [2026-07-17 22:38:57] step 502450/1000000 | loss=2.2682 vocab=0.5094 token=1.1725 | 6.16 steps/s | VRAM=28.49 Go [2026-07-17 22:39:05] step 502500/1000000 | loss=2.3233 vocab=0.6419 token=1.0133 | 6.17 steps/s | VRAM=28.21 Go [2026-07-17 22:39:13] step 502550/1000000 | loss=2.3338 vocab=0.6682 token=1.4387 | 6.18 steps/s | VRAM=31.31 Go [2026-07-17 22:39:21] step 502600/1000000 | loss=2.2773 vocab=1.6294 token=1.6566 | 6.11 steps/s | VRAM=25.90 Go [2026-07-17 22:39:29] step 502650/1000000 | loss=2.1833 vocab=0.7265 token=1.2890 | 6.17 steps/s | VRAM=28.22 Go [2026-07-17 22:39:38] step 502700/1000000 | loss=2.2385 vocab=0.8250 token=1.2707 | 6.15 steps/s | VRAM=29.15 Go [2026-07-17 22:39:46] step 502750/1000000 | loss=2.1022 vocab=0.7307 token=1.1511 | 6.15 steps/s | VRAM=32.29 Go [2026-07-17 22:39:54] step 502800/1000000 | loss=2.2576 vocab=1.4140 token=1.5744 | 6.11 steps/s | VRAM=26.27 Go [2026-07-17 22:40:02] step 502850/1000000 | loss=2.1836 vocab=0.7668 token=1.4157 | 6.18 steps/s | VRAM=32.37 Go [2026-07-17 22:40:10] step 502900/1000000 | loss=2.2855 vocab=0.8562 token=1.6478 | 6.14 steps/s | VRAM=29.23 Go [2026-07-17 22:40:18] step 502950/1000000 | loss=2.2189 vocab=0.4689 token=0.8838 | 6.16 steps/s | VRAM=31.90 Go [2026-07-17 22:40:26] step 503000/1000000 | loss=2.2092 vocab=0.6420 token=1.6188 | 6.15 steps/s | VRAM=23.31 Go [2026-07-17 22:40:34] step 503050/1000000 | loss=2.2929 vocab=2.2992 token=1.3380 | 6.14 steps/s | VRAM=32.61 Go [2026-07-17 22:40:43] step 503100/1000000 | loss=2.2828 vocab=1.1201 token=1.3342 | 6.11 steps/s | VRAM=29.81 Go [2026-07-17 22:40:51] step 503150/1000000 | loss=2.1469 vocab=1.1520 token=1.3304 | 6.15 steps/s | VRAM=31.75 Go [2026-07-17 22:40:59] step 503200/1000000 | loss=2.2266 vocab=0.6384 token=1.4363 | 6.14 steps/s | VRAM=29.17 Go [2026-07-17 22:41:07] step 503250/1000000 | loss=2.2167 vocab=0.9241 token=1.6061 | 6.15 steps/s | VRAM=28.74 Go [2026-07-17 22:41:15] step 503300/1000000 | loss=2.1231 vocab=0.8803 token=0.9729 | 6.12 steps/s | VRAM=26.35 Go [2026-07-17 22:41:23] step 503350/1000000 | loss=2.1971 vocab=0.7848 token=1.5212 | 6.12 steps/s | VRAM=29.82 Go [2026-07-17 22:41:32] step 503400/1000000 | loss=2.3063 vocab=1.5037 token=1.7294 | 6.13 steps/s | VRAM=25.51 Go [2026-07-17 22:41:40] step 503450/1000000 | loss=2.2643 vocab=1.0780 token=1.2075 | 6.13 steps/s | VRAM=28.79 Go [2026-07-17 22:41:48] step 503500/1000000 | loss=2.2613 vocab=0.9678 token=1.4855 | 6.14 steps/s | VRAM=29.24 Go [2026-07-17 22:41:56] step 503550/1000000 | loss=2.2401 vocab=1.1075 token=1.1630 | 6.12 steps/s | VRAM=28.73 Go [2026-07-17 22:42:04] step 503600/1000000 | loss=2.2095 vocab=0.8088 token=1.1087 | 6.14 steps/s | VRAM=32.72 Go [2026-07-17 22:42:12] step 503650/1000000 | loss=2.2587 vocab=0.5956 token=1.2304 | 6.15 steps/s | VRAM=31.92 Go [2026-07-17 22:42:21] step 503700/1000000 | loss=2.3377 vocab=1.0339 token=1.1947 | 6.09 steps/s | VRAM=25.90 Go [2026-07-17 22:42:29] step 503750/1000000 | loss=2.2241 vocab=1.5876 token=1.4791 | 6.16 steps/s | VRAM=31.63 Go [2026-07-17 22:42:37] step 503800/1000000 | loss=2.2025 vocab=0.5683 token=0.9789 | 6.14 steps/s | VRAM=28.57 Go [2026-07-17 22:42:45] step 503850/1000000 | loss=2.2582 vocab=1.0919 token=1.4505 | 6.16 steps/s | VRAM=31.14 Go [2026-07-17 22:42:53] step 503900/1000000 | loss=2.1919 vocab=0.9934 token=1.6008 | 6.15 steps/s | VRAM=24.89 Go [2026-07-17 22:43:01] step 503950/1000000 | loss=2.2720 vocab=0.6439 token=1.1720 | 6.17 steps/s | VRAM=25.16 Go [2026-07-17 22:43:09] step 504000/1000000 | loss=2.1947 vocab=0.9929 token=1.5948 | 6.14 steps/s | VRAM=31.68 Go [2026-07-17 22:43:17] step 504050/1000000 | loss=2.1868 vocab=1.1007 token=1.4448 | 6.14 steps/s | VRAM=28.64 Go [2026-07-17 22:43:26] step 504100/1000000 | loss=2.2676 vocab=0.9638 token=1.1402 | 6.14 steps/s | VRAM=28.45 Go [2026-07-17 22:43:34] step 504150/1000000 | loss=2.2015 vocab=0.7207 token=1.2366 | 6.12 steps/s | VRAM=26.39 Go [2026-07-17 22:43:42] step 504200/1000000 | loss=2.1688 vocab=0.8983 token=0.8280 | 6.14 steps/s | VRAM=32.11 Go [2026-07-17 22:43:50] step 504250/1000000 | loss=2.2652 vocab=1.1207 token=1.2862 | 6.14 steps/s | VRAM=33.02 Go [2026-07-17 22:43:58] step 504300/1000000 | loss=2.2425 vocab=0.8885 token=1.4156 | 6.15 steps/s | VRAM=32.18 Go [2026-07-17 22:44:06] step 504350/1000000 | loss=2.1289 vocab=1.1123 token=1.6288 | 6.15 steps/s | VRAM=31.78 Go [2026-07-17 22:44:14] step 504400/1000000 | loss=2.3474 vocab=1.2430 token=1.5635 | 6.15 steps/s | VRAM=28.48 Go [2026-07-17 22:44:23] step 504450/1000000 | loss=2.2695 vocab=0.9822 token=2.0062 | 6.12 steps/s | VRAM=26.25 Go [2026-07-17 22:44:31] step 504500/1000000 | loss=2.2510 vocab=1.1381 token=1.1480 | 6.14 steps/s | VRAM=28.97 Go [2026-07-17 22:44:39] step 504550/1000000 | loss=2.2511 vocab=0.7132 token=1.1605 | 6.16 steps/s | VRAM=32.39 Go [2026-07-17 22:44:47] step 504600/1000000 | loss=2.2830 vocab=1.0576 token=1.5865 | 6.15 steps/s | VRAM=28.85 Go [2026-07-17 22:44:55] step 504650/1000000 | loss=2.3277 vocab=1.0408 token=1.7119 | 6.15 steps/s | VRAM=24.78 Go [2026-07-17 22:45:03] step 504700/1000000 | loss=2.3655 vocab=1.1218 token=1.6606 | 6.15 steps/s | VRAM=31.66 Go [2026-07-17 22:45:11] step 504750/1000000 | loss=2.2613 vocab=0.8388 token=1.3221 | 6.12 steps/s | VRAM=29.40 Go [2026-07-17 22:45:20] step 504800/1000000 | loss=2.2209 vocab=0.9684 token=1.3144 | 6.14 steps/s | VRAM=25.58 Go [2026-07-17 22:45:28] step 504850/1000000 | loss=2.2209 vocab=0.5791 token=1.5010 | 6.15 steps/s | VRAM=29.07 Go [2026-07-17 22:45:36] step 504900/1000000 | loss=2.2647 vocab=1.1469 token=1.4352 | 6.15 steps/s | VRAM=28.79 Go [2026-07-17 22:45:44] step 504950/1000000 | loss=2.1960 vocab=0.5999 token=1.3313 | 6.11 steps/s | VRAM=26.67 Go [2026-07-17 22:45:52] step 505000/1000000 | loss=2.2724 vocab=0.8336 token=1.4459 | 6.13 steps/s | VRAM=26.14 Go [2026-07-17 22:46:00] step 505050/1000000 | loss=2.1384 vocab=0.5980 token=1.3401 | 6.18 steps/s | VRAM=28.49 Go [2026-07-17 22:46:08] step 505100/1000000 | loss=2.2763 vocab=0.5710 token=1.8794 | 6.12 steps/s | VRAM=26.33 Go [2026-07-17 22:46:17] step 505150/1000000 | loss=2.2489 vocab=0.5885 token=1.1376 | 6.15 steps/s | VRAM=28.16 Go [2026-07-17 22:46:25] step 505200/1000000 | loss=2.1592 vocab=0.5513 token=1.1815 | 6.13 steps/s | VRAM=32.41 Go [2026-07-17 22:46:33] step 505250/1000000 | loss=2.2730 vocab=1.1387 token=1.6589 | 6.15 steps/s | VRAM=28.18 Go [2026-07-17 22:46:41] step 505300/1000000 | loss=2.2323 vocab=0.7862 token=1.1359 | 6.13 steps/s | VRAM=25.37 Go [2026-07-17 22:46:49] step 505350/1000000 | loss=2.2146 vocab=0.7916 token=1.1472 | 6.15 steps/s | VRAM=28.64 Go [2026-07-17 22:46:57] step 505400/1000000 | loss=2.2368 vocab=1.1343 token=1.3708 | 6.13 steps/s | VRAM=25.79 Go [2026-07-17 22:47:05] step 505450/1000000 | loss=2.2317 vocab=0.9408 token=1.3741 | 6.15 steps/s | VRAM=31.57 Go [2026-07-17 22:47:14] step 505500/1000000 | loss=2.2272 vocab=0.7045 token=1.1594 | 6.13 steps/s | VRAM=28.66 Go [2026-07-17 22:47:22] step 505550/1000000 | loss=2.2776 vocab=0.6928 token=1.2044 | 6.15 steps/s | VRAM=31.18 Go [2026-07-17 22:47:30] step 505600/1000000 | loss=2.3088 vocab=1.0025 token=1.4976 | 6.15 steps/s | VRAM=28.03 Go [2026-07-17 22:47:38] step 505650/1000000 | loss=2.4632 vocab=0.6700 token=1.0198 | 6.14 steps/s | VRAM=28.98 Go [2026-07-17 22:47:46] step 505700/1000000 | loss=2.2646 vocab=0.8587 token=1.3296 | 6.14 steps/s | VRAM=25.78 Go [2026-07-17 22:47:54] step 505750/1000000 | loss=2.2922 vocab=0.5834 token=1.1263 | 6.13 steps/s | VRAM=28.59 Go [2026-07-17 22:48:02] step 505800/1000000 | loss=2.3150 vocab=1.1663 token=1.2012 | 6.12 steps/s | VRAM=28.26 Go [2026-07-17 22:48:11] step 505850/1000000 | loss=2.2683 vocab=1.5393 token=1.4862 | 6.14 steps/s | VRAM=29.20 Go [2026-07-17 22:48:19] step 505900/1000000 | loss=2.1893 vocab=0.6762 token=1.5847 | 6.15 steps/s | VRAM=26.10 Go [2026-07-17 22:48:27] step 505950/1000000 | loss=2.3609 vocab=1.0613 token=1.5448 | 6.13 steps/s | VRAM=25.49 Go [2026-07-17 22:48:35] step 506000/1000000 | loss=2.1644 vocab=0.9349 token=1.4838 | 6.13 steps/s | VRAM=29.25 Go [2026-07-17 22:48:43] step 506050/1000000 | loss=2.2145 vocab=0.8937 token=1.0339 | 6.12 steps/s | VRAM=26.58 Go [2026-07-17 22:48:51] step 506100/1000000 | loss=2.3110 vocab=1.4106 token=1.8182 | 6.15 steps/s | VRAM=28.73 Go [2026-07-17 22:48:59] step 506150/1000000 | loss=2.1409 vocab=0.6726 token=1.1741 | 6.13 steps/s | VRAM=25.64 Go [2026-07-17 22:49:08] step 506200/1000000 | loss=2.2082 vocab=1.3235 token=1.0585 | 6.14 steps/s | VRAM=31.96 Go [2026-07-17 22:49:16] step 506250/1000000 | loss=2.1766 vocab=0.9817 token=1.2916 | 6.13 steps/s | VRAM=32.63 Go [2026-07-17 22:49:24] step 506300/1000000 | loss=2.2542 vocab=1.2000 token=1.6286 | 6.14 steps/s | VRAM=32.67 Go [2026-07-17 22:49:32] step 506350/1000000 | loss=2.3146 vocab=1.3105 token=1.7136 | 6.10 steps/s | VRAM=28.48 Go [2026-07-17 22:49:40] step 506400/1000000 | loss=2.1450 vocab=0.6312 token=1.3140 | 6.15 steps/s | VRAM=29.24 Go [2026-07-17 22:49:48] step 506450/1000000 | loss=2.2812 vocab=1.0474 token=1.3616 | 6.13 steps/s | VRAM=26.81 Go [2026-07-17 22:49:57] step 506500/1000000 | loss=2.2089 vocab=1.3437 token=1.1344 | 6.15 steps/s | VRAM=31.74 Go [2026-07-17 22:50:05] step 506550/1000000 | loss=2.2402 vocab=1.2286 token=1.6351 | 6.14 steps/s | VRAM=31.48 Go [2026-07-17 22:50:13] step 506600/1000000 | loss=2.2877 vocab=0.6641 token=0.9594 | 6.12 steps/s | VRAM=25.41 Go [2026-07-17 22:50:21] step 506650/1000000 | loss=2.2317 vocab=0.9199 token=1.3054 | 6.14 steps/s | VRAM=28.45 Go [2026-07-17 22:50:29] step 506700/1000000 | loss=2.2495 vocab=0.3756 token=0.6325 | 6.17 steps/s | VRAM=30.58 Go [2026-07-17 22:50:37] step 506750/1000000 | loss=2.3005 vocab=0.4981 token=1.0349 | 6.10 steps/s | VRAM=29.36 Go [2026-07-17 22:50:45] step 506800/1000000 | loss=2.2548 vocab=1.1613 token=1.0873 | 6.16 steps/s | VRAM=32.66 Go [2026-07-17 22:50:54] step 506850/1000000 | loss=2.2457 vocab=0.5226 token=1.1570 | 6.12 steps/s | VRAM=28.56 Go [2026-07-17 22:51:02] step 506900/1000000 | loss=2.2165 vocab=2.0429 token=1.3121 | 6.15 steps/s | VRAM=25.15 Go [2026-07-17 22:51:10] step 506950/1000000 | loss=2.2175 vocab=0.9641 token=1.0328 | 6.15 steps/s | VRAM=26.13 Go [2026-07-17 22:51:18] step 507000/1000000 | loss=2.1847 vocab=1.3408 token=1.3121 | 6.13 steps/s | VRAM=32.39 Go [2026-07-17 22:51:26] step 507050/1000000 | loss=2.2037 vocab=1.0160 token=1.2762 | 6.11 steps/s | VRAM=28.93 Go [2026-07-17 22:51:34] step 507100/1000000 | loss=2.2395 vocab=1.0698 token=1.3860 | 6.14 steps/s | VRAM=32.58 Go [2026-07-17 22:51:42] step 507150/1000000 | loss=2.2802 vocab=0.7476 token=1.1889 | 6.14 steps/s | VRAM=32.66 Go [2026-07-17 22:51:51] step 507200/1000000 | loss=2.2613 vocab=0.9929 token=1.1077 | 6.12 steps/s | VRAM=26.52 Go [2026-07-17 22:51:59] step 507250/1000000 | loss=2.1676 vocab=1.0215 token=1.3399 | 6.14 steps/s | VRAM=31.68 Go [2026-07-17 22:52:07] step 507300/1000000 | loss=2.2163 vocab=0.9960 token=1.2670 | 6.17 steps/s | VRAM=32.05 Go [2026-07-17 22:52:15] step 507350/1000000 | loss=2.3213 vocab=0.6506 token=1.5873 | 6.17 steps/s | VRAM=24.64 Go [2026-07-17 22:52:23] step 507400/1000000 | loss=2.2568 vocab=1.2292 token=1.3903 | 6.17 steps/s | VRAM=32.73 Go [2026-07-17 22:52:31] step 507450/1000000 | loss=2.2286 vocab=1.4070 token=1.3844 | 6.12 steps/s | VRAM=32.75 Go [2026-07-17 22:52:39] step 507500/1000000 | loss=2.2928 vocab=0.8585 token=1.3495 | 6.16 steps/s | VRAM=28.14 Go [2026-07-17 22:52:47] step 507550/1000000 | loss=2.1487 vocab=0.8570 token=1.2244 | 6.14 steps/s | VRAM=28.87 Go [2026-07-17 22:52:56] step 507600/1000000 | loss=2.2697 vocab=0.8977 token=1.3174 | 6.17 steps/s | VRAM=32.61 Go [2026-07-17 22:53:04] step 507650/1000000 | loss=2.3166 vocab=1.3333 token=1.2767 | 6.14 steps/s | VRAM=26.19 Go [2026-07-17 22:53:12] step 507700/1000000 | loss=2.3530 vocab=0.8268 token=1.2511 | 6.17 steps/s | VRAM=32.59 Go [2026-07-17 22:53:20] step 507750/1000000 | loss=2.2665 vocab=1.2537 token=1.4158 | 6.11 steps/s | VRAM=26.50 Go [2026-07-17 22:53:28] step 507800/1000000 | loss=2.2774 vocab=0.9511 token=1.5142 | 6.16 steps/s | VRAM=32.13 Go [2026-07-17 22:53:36] step 507850/1000000 | loss=2.3062 vocab=1.2071 token=1.2824 | 6.15 steps/s | VRAM=32.31 Go [2026-07-17 22:53:44] step 507900/1000000 | loss=2.2135 vocab=0.6434 token=1.5008 | 6.17 steps/s | VRAM=32.40 Go [2026-07-17 22:53:53] step 507950/1000000 | loss=2.3822 vocab=0.8298 token=0.9575 | 6.15 steps/s | VRAM=30.92 Go [2026-07-17 22:54:01] step 508000/1000000 | loss=2.3641 vocab=0.5843 token=0.8345 | 6.15 steps/s | VRAM=31.85 Go [2026-07-17 22:54:09] step 508050/1000000 | loss=2.2579 vocab=1.0054 token=1.7285 | 6.15 steps/s | VRAM=31.51 Go [2026-07-17 22:54:17] step 508100/1000000 | loss=2.2174 vocab=1.0392 token=1.3439 | 6.14 steps/s | VRAM=23.63 Go [2026-07-17 22:54:25] step 508150/1000000 | loss=2.2734 vocab=0.6947 token=1.0502 | 6.16 steps/s | VRAM=29.43 Go [2026-07-17 22:54:33] step 508200/1000000 | loss=2.3056 vocab=1.1460 token=1.4009 | 6.13 steps/s | VRAM=26.00 Go [2026-07-17 22:54:41] step 508250/1000000 | loss=2.2766 vocab=0.8009 token=1.3332 | 6.18 steps/s | VRAM=32.06 Go [2026-07-17 22:54:49] step 508300/1000000 | loss=2.2223 vocab=1.0364 token=1.7034 | 6.13 steps/s | VRAM=26.32 Go [2026-07-17 22:54:58] step 508350/1000000 | loss=2.3119 vocab=0.5402 token=0.8670 | 6.15 steps/s | VRAM=28.16 Go [2026-07-17 22:55:06] step 508400/1000000 | loss=2.2433 vocab=1.3473 token=1.8786 | 6.16 steps/s | VRAM=29.58 Go [2026-07-17 22:55:14] step 508450/1000000 | loss=2.3728 vocab=0.8761 token=1.5015 | 6.14 steps/s | VRAM=28.59 Go [2026-07-17 22:55:22] step 508500/1000000 | loss=2.2005 vocab=0.6986 token=1.5373 | 6.14 steps/s | VRAM=25.75 Go [2026-07-17 22:55:30] step 508550/1000000 | loss=2.2748 vocab=0.5705 token=1.2218 | 6.15 steps/s | VRAM=28.85 Go [2026-07-17 22:55:38] step 508600/1000000 | loss=2.1686 vocab=1.3590 token=1.1871 | 6.18 steps/s | VRAM=30.57 Go [2026-07-17 22:55:46] step 508650/1000000 | loss=2.1487 vocab=0.9124 token=1.3030 | 6.14 steps/s | VRAM=26.40 Go [2026-07-17 22:55:54] step 508700/1000000 | loss=2.2240 vocab=1.1858 token=1.4000 | 6.17 steps/s | VRAM=28.32 Go [2026-07-17 22:56:03] step 508750/1000000 | loss=2.2110 vocab=0.8998 token=1.2213 | 6.14 steps/s | VRAM=32.76 Go [2026-07-17 22:56:11] step 508800/1000000 | loss=2.2357 vocab=0.8260 token=1.2186 | 6.11 steps/s | VRAM=29.73 Go [2026-07-17 22:56:19] step 508850/1000000 | loss=2.2753 vocab=1.2151 token=1.6100 | 6.15 steps/s | VRAM=26.52 Go [2026-07-17 22:56:27] step 508900/1000000 | loss=2.2655 vocab=0.6042 token=1.2625 | 6.16 steps/s | VRAM=28.92 Go [2026-07-17 22:56:35] step 508950/1000000 | loss=2.2595 vocab=0.8859 token=1.3675 | 6.17 steps/s | VRAM=29.04 Go [2026-07-17 22:56:43] step 509000/1000000 | loss=2.1765 vocab=0.9961 token=1.5238 | 6.15 steps/s | VRAM=28.66 Go [2026-07-17 22:56:51] step 509050/1000000 | loss=2.2058 vocab=0.4036 token=0.8611 | 6.14 steps/s | VRAM=32.25 Go [2026-07-17 22:57:00] step 509100/1000000 | loss=2.2176 vocab=1.2811 token=1.2911 | 6.12 steps/s | VRAM=29.29 Go [2026-07-17 22:57:08] step 509150/1000000 | loss=2.1592 vocab=0.6223 token=0.9341 | 6.19 steps/s | VRAM=31.89 Go [2026-07-17 22:57:16] step 509200/1000000 | loss=2.2754 vocab=0.5602 token=1.0073 | 6.15 steps/s | VRAM=32.23 Go [2026-07-17 22:57:24] step 509250/1000000 | loss=2.1668 vocab=0.7837 token=1.2864 | 6.15 steps/s | VRAM=26.59 Go [2026-07-17 22:57:32] step 509300/1000000 | loss=2.1814 vocab=0.5177 token=1.0371 | 6.19 steps/s | VRAM=31.48 Go [2026-07-17 22:57:40] step 509350/1000000 | loss=2.1223 vocab=0.6649 token=0.8861 | 6.15 steps/s | VRAM=29.05 Go [2026-07-17 22:57:48] step 509400/1000000 | loss=2.3878 vocab=0.8085 token=1.0164 | 6.13 steps/s | VRAM=30.19 Go [2026-07-17 22:57:56] step 509450/1000000 | loss=2.3594 vocab=0.9911 token=1.3995 | 6.20 steps/s | VRAM=25.82 Go [2026-07-17 22:58:04] step 509500/1000000 | loss=2.2126 vocab=0.8457 token=1.0857 | 6.14 steps/s | VRAM=28.38 Go [2026-07-17 22:58:13] step 509550/1000000 | loss=2.3647 vocab=0.6030 token=1.5525 | 6.13 steps/s | VRAM=28.11 Go [2026-07-17 22:58:21] step 509600/1000000 | loss=2.1843 vocab=1.2076 token=1.3630 | 6.16 steps/s | VRAM=25.75 Go [2026-07-17 22:58:29] step 509650/1000000 | loss=2.2214 vocab=0.6912 token=1.2803 | 6.14 steps/s | VRAM=26.42 Go [2026-07-17 22:58:37] step 509700/1000000 | loss=2.2900 vocab=1.7482 token=1.8073 | 6.16 steps/s | VRAM=28.85 Go [2026-07-17 22:58:45] step 509750/1000000 | loss=2.2619 vocab=1.3748 token=1.5601 | 6.14 steps/s | VRAM=26.80 Go [2026-07-17 22:58:53] step 509800/1000000 | loss=2.2459 vocab=1.3029 token=1.2032 | 6.17 steps/s | VRAM=25.03 Go [2026-07-17 22:59:01] step 509850/1000000 | loss=2.1832 vocab=0.8183 token=1.1064 | 6.14 steps/s | VRAM=25.53 Go [2026-07-17 22:59:10] step 509900/1000000 | loss=2.1322 vocab=1.0190 token=1.5550 | 6.11 steps/s | VRAM=29.44 Go [2026-07-17 22:59:18] step 509950/1000000 | loss=2.1967 vocab=0.6963 token=1.2187 | 6.13 steps/s | VRAM=29.11 Go [2026-07-17 22:59:26] step 510000/1000000 | loss=2.2847 vocab=0.7977 token=1.2850 | 6.13 steps/s | VRAM=25.77 Go [2026-07-17 22:59:34] step 510050/1000000 | loss=2.1457 vocab=0.5092 token=1.0703 | 6.15 steps/s | VRAM=26.70 Go [2026-07-17 22:59:42] step 510100/1000000 | loss=2.1931 vocab=0.7850 token=1.4590 | 6.13 steps/s | VRAM=29.25 Go [2026-07-17 22:59:50] step 510150/1000000 | loss=2.2724 vocab=0.8478 token=1.1633 | 6.17 steps/s | VRAM=32.02 Go [2026-07-17 22:59:58] step 510200/1000000 | loss=2.2198 vocab=0.5818 token=1.4720 | 6.15 steps/s | VRAM=28.34 Go [2026-07-17 23:00:07] step 510250/1000000 | loss=2.3374 vocab=0.9674 token=1.3425 | 6.14 steps/s | VRAM=28.61 Go [2026-07-17 23:00:15] step 510300/1000000 | loss=2.3192 vocab=0.9870 token=1.1214 | 6.12 steps/s | VRAM=29.09 Go [2026-07-17 23:00:23] step 510350/1000000 | loss=2.2395 vocab=0.9176 token=1.2947 | 6.15 steps/s | VRAM=29.17 Go [2026-07-17 23:00:31] step 510400/1000000 | loss=2.2563 vocab=0.9879 token=1.3237 | 6.11 steps/s | VRAM=27.92 Go [2026-07-17 23:00:39] step 510450/1000000 | loss=2.2838 vocab=0.8916 token=1.3407 | 6.19 steps/s | VRAM=28.61 Go [2026-07-17 23:00:47] step 510500/1000000 | loss=2.2629 vocab=1.1728 token=1.2674 | 6.12 steps/s | VRAM=28.07 Go [2026-07-17 23:00:55] step 510550/1000000 | loss=2.2523 vocab=0.8589 token=1.4003 | 6.12 steps/s | VRAM=29.19 Go [2026-07-17 23:01:04] step 510600/1000000 | loss=2.2475 vocab=1.0753 token=1.4878 | 6.13 steps/s | VRAM=25.76 Go [2026-07-17 23:01:12] step 510650/1000000 | loss=2.2524 vocab=1.9442 token=1.6367 | 6.15 steps/s | VRAM=31.53 Go [2026-07-17 23:01:20] step 510700/1000000 | loss=2.1695 vocab=1.2429 token=1.4810 | 6.12 steps/s | VRAM=28.74 Go [2026-07-17 23:01:28] step 510750/1000000 | loss=2.2068 vocab=0.6836 token=1.2428 | 6.14 steps/s | VRAM=29.24 Go [2026-07-17 23:01:36] step 510800/1000000 | loss=2.2229 vocab=0.8083 token=1.4993 | 6.15 steps/s | VRAM=24.87 Go [2026-07-17 23:01:44] step 510850/1000000 | loss=2.1829 vocab=1.0817 token=1.5781 | 6.16 steps/s | VRAM=31.83 Go [2026-07-17 23:01:52] step 510900/1000000 | loss=2.2598 vocab=0.8429 token=1.1363 | 6.13 steps/s | VRAM=25.96 Go [2026-07-17 23:02:01] step 510950/1000000 | loss=2.3063 vocab=0.7864 token=1.7042 | 6.12 steps/s | VRAM=27.07 Go [2026-07-17 23:02:09] step 511000/1000000 | loss=2.2158 vocab=1.0755 token=1.3772 | 6.15 steps/s | VRAM=31.70 Go [2026-07-17 23:02:17] step 511050/1000000 | loss=2.2008 vocab=0.8537 token=1.2466 | 6.16 steps/s | VRAM=25.30 Go [2026-07-17 23:02:25] step 511100/1000000 | loss=2.3412 vocab=1.3024 token=1.4429 | 6.15 steps/s | VRAM=28.15 Go [2026-07-17 23:02:33] step 511150/1000000 | loss=2.2661 vocab=0.9427 token=1.4667 | 6.12 steps/s | VRAM=29.87 Go [2026-07-17 23:02:41] step 511200/1000000 | loss=2.3726 vocab=1.0017 token=1.7516 | 6.15 steps/s | VRAM=31.67 Go [2026-07-17 23:02:49] step 511250/1000000 | loss=2.2952 vocab=1.2730 token=1.3730 | 6.15 steps/s | VRAM=28.68 Go [2026-07-17 23:02:58] step 511300/1000000 | loss=2.2532 vocab=0.9613 token=1.2039 | 6.14 steps/s | VRAM=31.90 Go [2026-07-17 23:03:06] step 511350/1000000 | loss=2.2114 vocab=1.4651 token=1.6436 | 6.16 steps/s | VRAM=31.55 Go [2026-07-17 23:03:14] step 511400/1000000 | loss=2.2532 vocab=0.7701 token=1.3579 | 6.07 steps/s | VRAM=29.41 Go [2026-07-17 23:03:22] step 511450/1000000 | loss=2.3072 vocab=0.4310 token=1.4437 | 6.14 steps/s | VRAM=25.03 Go [2026-07-17 23:03:30] step 511500/1000000 | loss=2.2393 vocab=1.0516 token=1.2038 | 6.15 steps/s | VRAM=29.42 Go [2026-07-17 23:03:38] step 511550/1000000 | loss=2.2652 vocab=0.4410 token=0.8022 | 6.13 steps/s | VRAM=27.90 Go [2026-07-17 23:03:47] step 511600/1000000 | loss=2.2746 vocab=0.8046 token=1.3981 | 6.15 steps/s | VRAM=25.87 Go [2026-07-17 23:03:55] step 511650/1000000 | loss=2.2397 vocab=0.6436 token=1.3362 | 6.15 steps/s | VRAM=28.57 Go [2026-07-17 23:04:03] step 511700/1000000 | loss=2.2135 vocab=0.6169 token=1.5250 | 6.15 steps/s | VRAM=28.29 Go [2026-07-17 23:04:11] step 511750/1000000 | loss=2.2836 vocab=0.5457 token=1.1030 | 6.12 steps/s | VRAM=28.29 Go [2026-07-17 23:04:19] step 511800/1000000 | loss=2.1970 vocab=1.1105 token=1.2568 | 6.14 steps/s | VRAM=25.75 Go [2026-07-17 23:04:27] step 511850/1000000 | loss=2.2690 vocab=0.9003 token=1.3091 | 6.18 steps/s | VRAM=31.96 Go [2026-07-17 23:04:35] step 511900/1000000 | loss=2.2430 vocab=0.9825 token=1.4910 | 6.16 steps/s | VRAM=32.02 Go [2026-07-17 23:04:43] step 511950/1000000 | loss=2.2570 vocab=0.8166 token=1.4489 | 6.17 steps/s | VRAM=32.81 Go [2026-07-17 23:04:52] step 512000/1000000 | loss=2.2934 vocab=0.7046 token=1.3160 | 6.16 steps/s | VRAM=32.38 Go [2026-07-17 23:05:00] step 512050/1000000 | loss=2.2203 vocab=0.9286 token=1.2128 | 6.14 steps/s | VRAM=28.27 Go [2026-07-17 23:05:08] step 512100/1000000 | loss=2.2328 vocab=1.0137 token=1.1385 | 6.16 steps/s | VRAM=32.63 Go [2026-07-17 23:05:16] step 512150/1000000 | loss=2.2190 vocab=1.0759 token=1.7804 | 6.13 steps/s | VRAM=28.10 Go [2026-07-17 23:05:24] step 512200/1000000 | loss=2.2242 vocab=1.1760 token=1.1722 | 6.16 steps/s | VRAM=25.46 Go [2026-07-17 23:05:32] step 512250/1000000 | loss=2.2616 vocab=0.9054 token=1.3026 | 6.16 steps/s | VRAM=31.33 Go [2026-07-17 23:05:40] step 512300/1000000 | loss=2.3046 vocab=0.7051 token=1.3035 | 6.15 steps/s | VRAM=29.46 Go [2026-07-17 23:05:48] step 512350/1000000 | loss=2.2321 vocab=0.4134 token=1.3449 | 6.14 steps/s | VRAM=31.98 Go [2026-07-17 23:05:57] step 512400/1000000 | loss=2.2256 vocab=0.9804 token=1.4631 | 6.17 steps/s | VRAM=29.02 Go [2026-07-17 23:06:05] step 512450/1000000 | loss=2.1725 vocab=0.8642 token=1.2005 | 6.18 steps/s | VRAM=29.24 Go [2026-07-17 23:06:13] step 512500/1000000 | loss=2.2648 vocab=1.1513 token=1.4533 | 6.16 steps/s | VRAM=27.04 Go [2026-07-17 23:06:21] step 512550/1000000 | loss=2.1801 vocab=0.7431 token=1.5757 | 6.13 steps/s | VRAM=26.94 Go [2026-07-17 23:06:29] step 512600/1000000 | loss=2.2433 vocab=0.9847 token=1.5841 | 6.15 steps/s | VRAM=28.84 Go [2026-07-17 23:06:37] step 512650/1000000 | loss=2.2713 vocab=0.9632 token=1.4605 | 6.14 steps/s | VRAM=29.01 Go [2026-07-17 23:06:45] step 512700/1000000 | loss=2.1784 vocab=0.7768 token=1.1603 | 6.15 steps/s | VRAM=29.13 Go [2026-07-17 23:06:53] step 512750/1000000 | loss=2.2389 vocab=1.2750 token=1.4761 | 6.15 steps/s | VRAM=29.46 Go [2026-07-17 23:07:02] step 512800/1000000 | loss=2.2586 vocab=0.7992 token=1.1974 | 6.14 steps/s | VRAM=25.20 Go [2026-07-17 23:07:10] step 512850/1000000 | loss=2.2595 vocab=1.1199 token=1.1623 | 6.20 steps/s | VRAM=28.82 Go [2026-07-17 23:07:18] step 512900/1000000 | loss=2.3074 vocab=1.2144 token=1.7657 | 6.17 steps/s | VRAM=32.89 Go [2026-07-17 23:07:26] step 512950/1000000 | loss=2.2426 vocab=1.0683 token=1.2700 | 6.15 steps/s | VRAM=32.98 Go [2026-07-17 23:07:34] step 513000/1000000 | loss=2.1836 vocab=1.1024 token=1.8109 | 6.16 steps/s | VRAM=28.30 Go [2026-07-17 23:07:42] step 513050/1000000 | loss=2.2439 vocab=0.8071 token=1.2331 | 6.18 steps/s | VRAM=32.63 Go [2026-07-17 23:07:50] step 513100/1000000 | loss=2.2257 vocab=0.7568 token=1.5023 | 6.16 steps/s | VRAM=24.99 Go [2026-07-17 23:07:58] step 513150/1000000 | loss=2.2367 vocab=1.1222 token=1.0583 | 6.16 steps/s | VRAM=31.40 Go [2026-07-17 23:08:07] step 513200/1000000 | loss=2.3009 vocab=1.3653 token=0.8682 | 6.13 steps/s | VRAM=29.18 Go [2026-07-17 23:08:15] step 513250/1000000 | loss=2.0803 vocab=0.7506 token=1.2971 | 6.14 steps/s | VRAM=29.35 Go [2026-07-17 23:08:23] step 513300/1000000 | loss=2.2754 vocab=1.0214 token=1.7687 | 6.11 steps/s | VRAM=29.89 Go [2026-07-17 23:08:31] step 513350/1000000 | loss=2.2798 vocab=1.3013 token=1.2812 | 6.15 steps/s | VRAM=32.08 Go [2026-07-17 23:08:39] step 513400/1000000 | loss=2.2241 vocab=0.5702 token=1.3476 | 6.13 steps/s | VRAM=32.51 Go [2026-07-17 23:08:47] step 513450/1000000 | loss=2.3555 vocab=0.6788 token=1.3989 | 6.14 steps/s | VRAM=32.65 Go [2026-07-17 23:08:55] step 513500/1000000 | loss=2.2173 vocab=1.0773 token=1.2146 | 6.14 steps/s | VRAM=30.81 Go [2026-07-17 23:09:04] step 513550/1000000 | loss=2.3610 vocab=0.9704 token=1.2875 | 6.15 steps/s | VRAM=25.33 Go [2026-07-17 23:09:12] step 513600/1000000 | loss=2.2880 vocab=1.2416 token=1.7857 | 6.13 steps/s | VRAM=32.35 Go [2026-07-17 23:09:20] step 513650/1000000 | loss=2.2540 vocab=1.1017 token=1.3269 | 6.11 steps/s | VRAM=29.45 Go [2026-07-17 23:09:28] step 513700/1000000 | loss=2.3554 vocab=1.1847 token=1.6353 | 6.08 steps/s | VRAM=25.24 Go [2026-07-17 23:09:36] step 513750/1000000 | loss=2.3455 vocab=1.1088 token=1.1148 | 6.17 steps/s | VRAM=30.50 Go [2026-07-17 23:09:44] step 513800/1000000 | loss=2.1867 vocab=0.6581 token=1.2404 | 6.12 steps/s | VRAM=29.57 Go [2026-07-17 23:09:53] step 513850/1000000 | loss=2.2434 vocab=0.8524 token=1.1042 | 6.16 steps/s | VRAM=25.65 Go [2026-07-17 23:10:01] step 513900/1000000 | loss=2.2099 vocab=0.4008 token=1.1609 | 6.16 steps/s | VRAM=27.86 Go [2026-07-17 23:10:09] step 513950/1000000 | loss=2.1897 vocab=0.5539 token=1.1110 | 6.13 steps/s | VRAM=28.80 Go [2026-07-17 23:10:17] step 514000/1000000 | loss=2.3000 vocab=0.7828 token=1.3754 | 6.12 steps/s | VRAM=32.82 Go [2026-07-17 23:10:25] step 514050/1000000 | loss=2.2527 vocab=0.9471 token=1.3143 | 6.14 steps/s | VRAM=27.96 Go [2026-07-17 23:10:33] step 514100/1000000 | loss=2.1641 vocab=0.9148 token=1.4633 | 6.14 steps/s | VRAM=31.71 Go [2026-07-17 23:10:41] step 514150/1000000 | loss=2.2895 vocab=0.6850 token=1.2309 | 6.15 steps/s | VRAM=29.24 Go [2026-07-17 23:10:50] step 514200/1000000 | loss=2.2268 vocab=0.6031 token=0.7690 | 6.14 steps/s | VRAM=28.58 Go [2026-07-17 23:10:58] step 514250/1000000 | loss=2.2530 vocab=0.6444 token=1.0939 | 6.15 steps/s | VRAM=25.24 Go [2026-07-17 23:11:06] step 514300/1000000 | loss=2.1831 vocab=0.8315 token=1.3088 | 6.15 steps/s | VRAM=24.98 Go [2026-07-17 23:11:14] step 514350/1000000 | loss=2.2467 vocab=1.4113 token=1.4127 | 6.12 steps/s | VRAM=28.54 Go [2026-07-17 23:11:22] step 514400/1000000 | loss=2.2300 vocab=0.7704 token=1.1485 | 6.13 steps/s | VRAM=28.49 Go [2026-07-17 23:11:30] step 514450/1000000 | loss=2.2533 vocab=0.9798 token=1.3551 | 6.12 steps/s | VRAM=28.72 Go [2026-07-17 23:11:38] step 514500/1000000 | loss=2.1566 vocab=0.7728 token=1.4036 | 6.11 steps/s | VRAM=27.38 Go [2026-07-17 23:11:47] step 514550/1000000 | loss=2.2587 vocab=1.8070 token=1.5861 | 6.16 steps/s | VRAM=29.26 Go [2026-07-17 23:11:55] step 514600/1000000 | loss=2.2998 vocab=0.5892 token=1.4182 | 6.15 steps/s | VRAM=32.04 Go [2026-07-17 23:12:03] step 514650/1000000 | loss=2.2196 vocab=0.8894 token=1.3011 | 6.14 steps/s | VRAM=32.68 Go [2026-07-17 23:12:11] step 514700/1000000 | loss=2.2943 vocab=0.8994 token=1.2288 | 6.11 steps/s | VRAM=29.98 Go [2026-07-17 23:12:19] step 514750/1000000 | loss=2.2096 vocab=0.8459 token=0.9526 | 6.16 steps/s | VRAM=25.55 Go [2026-07-17 23:12:27] step 514800/1000000 | loss=2.2245 vocab=0.7148 token=1.2979 | 6.15 steps/s | VRAM=32.00 Go [2026-07-17 23:12:35] step 514850/1000000 | loss=2.2494 vocab=0.8554 token=1.3622 | 6.14 steps/s | VRAM=29.31 Go [2026-07-17 23:12:44] step 514900/1000000 | loss=2.2615 vocab=1.0335 token=1.3724 | 6.13 steps/s | VRAM=28.50 Go [2026-07-17 23:12:52] step 514950/1000000 | loss=2.2395 vocab=0.7806 token=1.3016 | 6.13 steps/s | VRAM=29.31 Go [2026-07-17 23:13:00] step 515000/1000000 | loss=2.2693 vocab=0.9743 token=1.1939 | 6.15 steps/s | VRAM=26.10 Go [2026-07-17 23:13:08] step 515050/1000000 | loss=2.2911 vocab=0.8029 token=1.1001 | 6.16 steps/s | VRAM=28.32 Go [2026-07-17 23:13:16] step 515100/1000000 | loss=2.1849 vocab=1.3341 token=1.4609 | 6.16 steps/s | VRAM=31.63 Go [2026-07-17 23:13:24] step 515150/1000000 | loss=2.2129 vocab=1.3966 token=1.2780 | 6.16 steps/s | VRAM=28.75 Go [2026-07-17 23:13:32] step 515200/1000000 | loss=2.2228 vocab=0.9902 token=1.3810 | 6.14 steps/s | VRAM=32.00 Go [2026-07-17 23:13:41] step 515250/1000000 | loss=2.3165 vocab=1.0337 token=1.4189 | 6.12 steps/s | VRAM=28.71 Go [2026-07-17 23:13:49] step 515300/1000000 | loss=2.2870 vocab=0.9925 token=1.3441 | 6.16 steps/s | VRAM=24.82 Go [2026-07-17 23:13:57] step 515350/1000000 | loss=2.2230 vocab=0.9889 token=1.7415 | 6.14 steps/s | VRAM=29.68 Go [2026-07-17 23:14:05] step 515400/1000000 | loss=2.2582 vocab=1.1846 token=1.4914 | 6.16 steps/s | VRAM=29.05 Go [2026-07-17 23:14:13] step 515450/1000000 | loss=2.2713 vocab=1.4790 token=1.6274 | 6.16 steps/s | VRAM=29.13 Go [2026-07-17 23:14:21] step 515500/1000000 | loss=2.2989 vocab=1.2521 token=1.3698 | 6.15 steps/s | VRAM=29.45 Go [2026-07-17 23:14:29] step 515550/1000000 | loss=2.2087 vocab=1.2662 token=1.2143 | 6.15 steps/s | VRAM=32.04 Go [2026-07-17 23:14:37] step 515600/1000000 | loss=2.1805 vocab=0.8021 token=0.9759 | 6.13 steps/s | VRAM=27.98 Go [2026-07-17 23:14:46] step 515650/1000000 | loss=2.3089 vocab=1.1844 token=1.0702 | 6.15 steps/s | VRAM=32.24 Go [2026-07-17 23:14:54] step 515700/1000000 | loss=2.1940 vocab=0.7508 token=1.2570 | 6.13 steps/s | VRAM=26.07 Go [2026-07-17 23:15:02] step 515750/1000000 | loss=2.2573 vocab=0.8070 token=1.8349 | 6.16 steps/s | VRAM=31.41 Go [2026-07-17 23:15:10] step 515800/1000000 | loss=2.2406 vocab=0.8383 token=1.2436 | 6.14 steps/s | VRAM=28.67 Go [2026-07-17 23:15:18] step 515850/1000000 | loss=2.1957 vocab=0.9755 token=1.2456 | 6.10 steps/s | VRAM=30.05 Go [2026-07-17 23:15:26] step 515900/1000000 | loss=2.2122 vocab=0.9957 token=1.6110 | 6.16 steps/s | VRAM=31.01 Go [2026-07-17 23:15:34] step 515950/1000000 | loss=2.3298 vocab=1.0870 token=1.4011 | 6.13 steps/s | VRAM=32.10 Go [2026-07-17 23:15:43] step 516000/1000000 | loss=2.2236 vocab=0.7514 token=1.6275 | 6.12 steps/s | VRAM=27.06 Go [2026-07-17 23:15:51] step 516050/1000000 | loss=2.3864 vocab=1.2992 token=1.5223 | 6.14 steps/s | VRAM=31.19 Go [2026-07-17 23:15:59] step 516100/1000000 | loss=2.1670 vocab=0.7082 token=1.2071 | 6.10 steps/s | VRAM=25.93 Go [2026-07-17 23:16:07] step 516150/1000000 | loss=2.2261 vocab=1.2996 token=1.1690 | 6.13 steps/s | VRAM=26.57 Go [2026-07-17 23:16:15] step 516200/1000000 | loss=2.1958 vocab=0.9430 token=1.2705 | 6.11 steps/s | VRAM=29.69 Go [2026-07-17 23:16:23] step 516250/1000000 | loss=2.1597 vocab=0.9973 token=1.4213 | 6.10 steps/s | VRAM=32.56 Go [2026-07-17 23:16:32] step 516300/1000000 | loss=2.2574 vocab=0.5644 token=1.0079 | 6.11 steps/s | VRAM=27.95 Go [2026-07-17 23:16:40] step 516350/1000000 | loss=2.3189 vocab=1.0808 token=1.4130 | 6.10 steps/s | VRAM=26.00 Go [2026-07-17 23:16:48] step 516400/1000000 | loss=2.2777 vocab=0.9611 token=1.6033 | 6.15 steps/s | VRAM=28.14 Go [2026-07-17 23:16:56] step 516450/1000000 | loss=2.2653 vocab=0.8656 token=1.4434 | 6.15 steps/s | VRAM=28.75 Go [2026-07-17 23:17:04] step 516500/1000000 | loss=2.3062 vocab=1.2714 token=1.4427 | 6.19 steps/s | VRAM=28.59 Go [2026-07-17 23:17:12] step 516550/1000000 | loss=2.2457 vocab=0.9236 token=1.0544 | 6.15 steps/s | VRAM=31.15 Go [2026-07-17 23:17:20] step 516600/1000000 | loss=2.2115 vocab=1.0992 token=1.4942 | 6.15 steps/s | VRAM=28.61 Go [2026-07-17 23:17:29] step 516650/1000000 | loss=2.3003 vocab=0.8285 token=1.2091 | 6.14 steps/s | VRAM=29.77 Go [2026-07-17 23:17:37] step 516700/1000000 | loss=2.3471 vocab=0.5960 token=1.2665 | 6.14 steps/s | VRAM=28.83 Go [2026-07-17 23:17:45] step 516750/1000000 | loss=2.1971 vocab=0.9703 token=1.2744 | 6.14 steps/s | VRAM=28.37 Go [2026-07-17 23:17:53] step 516800/1000000 | loss=2.2778 vocab=0.9417 token=1.5635 | 6.12 steps/s | VRAM=29.25 Go [2026-07-17 23:18:01] step 516850/1000000 | loss=2.2721 vocab=0.8385 token=1.1971 | 6.15 steps/s | VRAM=32.54 Go [2026-07-17 23:18:09] step 516900/1000000 | loss=2.1466 vocab=0.9678 token=1.2784 | 6.16 steps/s | VRAM=26.22 Go [2026-07-17 23:18:17] step 516950/1000000 | loss=2.3057 vocab=0.8271 token=1.4125 | 6.16 steps/s | VRAM=25.51 Go [2026-07-17 23:18:26] step 517000/1000000 | loss=2.3816 vocab=0.9743 token=1.3487 | 6.14 steps/s | VRAM=26.29 Go [2026-07-17 23:18:34] step 517050/1000000 | loss=2.3114 vocab=0.6957 token=1.3705 | 6.14 steps/s | VRAM=28.83 Go [2026-07-17 23:18:42] step 517100/1000000 | loss=2.2833 vocab=0.8576 token=1.0746 | 6.12 steps/s | VRAM=29.14 Go [2026-07-17 23:18:50] step 517150/1000000 | loss=2.2810 vocab=0.9894 token=1.0893 | 6.15 steps/s | VRAM=32.37 Go [2026-07-17 23:18:58] step 517200/1000000 | loss=2.2884 vocab=1.7108 token=1.6428 | 6.16 steps/s | VRAM=26.85 Go [2026-07-17 23:19:06] step 517250/1000000 | loss=2.3110 vocab=0.8713 token=1.4031 | 6.16 steps/s | VRAM=31.12 Go [2026-07-17 23:19:14] step 517300/1000000 | loss=2.2242 vocab=0.4716 token=1.0286 | 6.12 steps/s | VRAM=29.41 Go [2026-07-17 23:19:23] step 517350/1000000 | loss=2.2211 vocab=0.5683 token=1.1900 | 6.17 steps/s | VRAM=27.62 Go [2026-07-17 23:19:31] step 517400/1000000 | loss=2.2001 vocab=0.9979 token=1.6880 | 6.14 steps/s | VRAM=25.00 Go [2026-07-17 23:19:39] step 517450/1000000 | loss=2.2058 vocab=1.0536 token=1.5825 | 6.15 steps/s | VRAM=28.41 Go [2026-07-17 23:19:47] step 517500/1000000 | loss=2.2621 vocab=1.0084 token=1.3912 | 6.15 steps/s | VRAM=26.20 Go [2026-07-17 23:19:55] step 517550/1000000 | loss=2.2826 vocab=0.7050 token=1.2067 | 6.13 steps/s | VRAM=27.96 Go [2026-07-17 23:20:03] step 517600/1000000 | loss=2.2273 vocab=0.7234 token=0.9772 | 6.14 steps/s | VRAM=26.40 Go [2026-07-17 23:20:11] step 517650/1000000 | loss=2.3118 vocab=1.0705 token=1.1300 | 6.14 steps/s | VRAM=31.57 Go [2026-07-17 23:20:20] step 517700/1000000 | loss=2.2828 vocab=0.5861 token=1.1841 | 6.14 steps/s | VRAM=25.73 Go [2026-07-17 23:20:28] step 517750/1000000 | loss=2.2309 vocab=0.8620 token=1.2525 | 6.14 steps/s | VRAM=26.42 Go [2026-07-17 23:20:36] step 517800/1000000 | loss=2.4102 vocab=0.5271 token=1.0636 | 6.14 steps/s | VRAM=26.03 Go [2026-07-17 23:20:44] step 517850/1000000 | loss=2.2376 vocab=1.6997 token=1.6690 | 6.13 steps/s | VRAM=28.44 Go [2026-07-17 23:20:52] step 517900/1000000 | loss=2.2621 vocab=0.7245 token=1.0028 | 6.12 steps/s | VRAM=26.74 Go [2026-07-17 23:21:00] step 517950/1000000 | loss=2.3129 vocab=0.6983 token=1.3245 | 6.17 steps/s | VRAM=27.35 Go [2026-07-17 23:21:08] step 518000/1000000 | loss=2.2853 vocab=1.3888 token=1.4217 | 6.14 steps/s | VRAM=32.33 Go [2026-07-17 23:21:16] step 518050/1000000 | loss=2.2306 vocab=0.3194 token=0.9381 | 6.16 steps/s | VRAM=32.04 Go [2026-07-17 23:21:25] step 518100/1000000 | loss=2.2829 vocab=0.7369 token=1.7380 | 6.14 steps/s | VRAM=24.91 Go [2026-07-17 23:21:33] step 518150/1000000 | loss=2.2783 vocab=0.7858 token=1.3172 | 6.13 steps/s | VRAM=24.90 Go [2026-07-17 23:21:41] step 518200/1000000 | loss=2.2705 vocab=1.3666 token=1.7822 | 6.13 steps/s | VRAM=28.15 Go [2026-07-17 23:21:49] step 518250/1000000 | loss=2.2782 vocab=0.8477 token=0.9753 | 6.11 steps/s | VRAM=27.05 Go [2026-07-17 23:21:57] step 518300/1000000 | loss=2.2251 vocab=0.5855 token=0.8738 | 6.15 steps/s | VRAM=28.77 Go [2026-07-17 23:22:05] step 518350/1000000 | loss=2.1906 vocab=1.0039 token=1.3694 | 6.14 steps/s | VRAM=28.82 Go [2026-07-17 23:22:14] step 518400/1000000 | loss=2.1987 vocab=1.4651 token=1.6055 | 6.10 steps/s | VRAM=27.01 Go [2026-07-17 23:22:22] step 518450/1000000 | loss=2.2489 vocab=1.1870 token=1.5729 | 6.14 steps/s | VRAM=28.00 Go [2026-07-17 23:22:30] step 518500/1000000 | loss=2.3279 vocab=0.8521 token=1.4568 | 6.14 steps/s | VRAM=32.20 Go [2026-07-17 23:22:38] step 518550/1000000 | loss=2.3222 vocab=0.4255 token=0.8165 | 6.11 steps/s | VRAM=29.82 Go [2026-07-17 23:22:46] step 518600/1000000 | loss=2.2368 vocab=0.4369 token=0.9156 | 6.16 steps/s | VRAM=28.40 Go [2026-07-17 23:22:54] step 518650/1000000 | loss=2.1952 vocab=0.9952 token=1.5410 | 6.14 steps/s | VRAM=28.50 Go [2026-07-17 23:23:02] step 518700/1000000 | loss=2.1783 vocab=0.6610 token=1.3211 | 6.15 steps/s | VRAM=26.78 Go [2026-07-17 23:23:11] step 518750/1000000 | loss=2.2854 vocab=1.0530 token=1.1428 | 6.13 steps/s | VRAM=28.61 Go [2026-07-17 23:23:19] step 518800/1000000 | loss=2.1140 vocab=0.7921 token=1.3103 | 6.14 steps/s | VRAM=32.63 Go [2026-07-17 23:23:27] step 518850/1000000 | loss=2.2654 vocab=1.5149 token=1.8730 | 6.12 steps/s | VRAM=27.92 Go [2026-07-17 23:23:35] step 518900/1000000 | loss=2.2759 vocab=0.8054 token=1.4886 | 6.13 steps/s | VRAM=28.66 Go [2026-07-17 23:23:43] step 518950/1000000 | loss=2.2967 vocab=0.6869 token=1.3781 | 6.14 steps/s | VRAM=25.60 Go [2026-07-17 23:23:51] step 519000/1000000 | loss=2.1361 vocab=0.7795 token=1.1517 | 6.14 steps/s | VRAM=32.85 Go [2026-07-17 23:24:00] step 519050/1000000 | loss=2.2333 vocab=0.6368 token=1.0163 | 6.13 steps/s | VRAM=28.47 Go [2026-07-17 23:24:08] step 519100/1000000 | loss=2.3432 vocab=0.9766 token=1.3417 | 6.14 steps/s | VRAM=26.04 Go [2026-07-17 23:24:16] step 519150/1000000 | loss=2.2888 vocab=0.9044 token=1.3272 | 6.17 steps/s | VRAM=31.86 Go [2026-07-17 23:24:24] step 519200/1000000 | loss=2.2609 vocab=0.5716 token=1.2268 | 6.12 steps/s | VRAM=28.61 Go [2026-07-17 23:24:32] step 519250/1000000 | loss=2.2736 vocab=0.6268 token=1.1216 | 6.09 steps/s | VRAM=28.77 Go [2026-07-17 23:24:40] step 519300/1000000 | loss=2.2230 vocab=0.9297 token=1.4135 | 6.16 steps/s | VRAM=31.51 Go [2026-07-17 23:24:48] step 519350/1000000 | loss=2.1792 vocab=0.7529 token=1.0423 | 6.12 steps/s | VRAM=25.74 Go [2026-07-17 23:24:57] step 519400/1000000 | loss=2.1919 vocab=0.7513 token=1.5826 | 6.15 steps/s | VRAM=29.14 Go [2026-07-17 23:25:05] step 519450/1000000 | loss=2.2790 vocab=0.8879 token=1.2739 | 6.14 steps/s | VRAM=26.13 Go [2026-07-17 23:25:13] step 519500/1000000 | loss=2.2599 vocab=1.1711 token=1.5772 | 6.14 steps/s | VRAM=24.93 Go [2026-07-17 23:25:21] step 519550/1000000 | loss=2.2898 vocab=1.5785 token=1.0224 | 6.10 steps/s | VRAM=23.65 Go [2026-07-17 23:25:29] step 519600/1000000 | loss=2.1999 vocab=0.6719 token=1.4980 | 6.15 steps/s | VRAM=29.10 Go [2026-07-17 23:25:37] step 519650/1000000 | loss=2.3218 vocab=0.5888 token=1.2182 | 6.15 steps/s | VRAM=32.89 Go [2026-07-17 23:25:45] step 519700/1000000 | loss=2.2030 vocab=0.4912 token=1.1309 | 6.17 steps/s | VRAM=26.07 Go [2026-07-17 23:25:54] step 519750/1000000 | loss=2.2095 vocab=0.7816 token=1.1341 | 6.16 steps/s | VRAM=32.60 Go [2026-07-17 23:26:02] step 519800/1000000 | loss=2.2595 vocab=0.6447 token=1.2487 | 6.13 steps/s | VRAM=31.67 Go [2026-07-17 23:26:10] step 519850/1000000 | loss=2.2655 vocab=0.9242 token=1.2540 | 6.18 steps/s | VRAM=32.30 Go [2026-07-17 23:26:18] step 519900/1000000 | loss=2.1374 vocab=0.9116 token=1.0875 | 6.11 steps/s | VRAM=32.38 Go [2026-07-17 23:26:26] step 519950/1000000 | loss=2.1323 vocab=0.3635 token=0.7674 | 6.15 steps/s | VRAM=28.67 Go [2026-07-17 23:26:34] step 520000/1000000 | loss=2.2069 vocab=0.9817 token=1.6305 | 6.14 steps/s | VRAM=28.71 Go [2026-07-17 23:26:42] step 520050/1000000 | loss=2.3093 vocab=0.6691 token=1.0899 | 6.11 steps/s | VRAM=24.41 Go [2026-07-17 23:26:51] step 520100/1000000 | loss=2.2628 vocab=0.6319 token=1.4635 | 6.16 steps/s | VRAM=32.47 Go [2026-07-17 23:26:59] step 520150/1000000 | loss=2.2148 vocab=0.9994 token=1.3832 | 6.13 steps/s | VRAM=28.29 Go [2026-07-17 23:27:07] step 520200/1000000 | loss=2.3443 vocab=1.3583 token=1.5024 | 6.16 steps/s | VRAM=29.94 Go [2026-07-17 23:27:15] step 520250/1000000 | loss=2.2261 vocab=0.5902 token=1.2008 | 6.11 steps/s | VRAM=27.34 Go [2026-07-17 23:27:23] step 520300/1000000 | loss=2.3253 vocab=1.2514 token=1.5567 | 6.14 steps/s | VRAM=31.20 Go [2026-07-17 23:27:31] step 520350/1000000 | loss=2.2909 vocab=0.5707 token=1.4527 | 6.11 steps/s | VRAM=27.13 Go [2026-07-17 23:27:39] step 520400/1000000 | loss=2.1593 vocab=1.3522 token=1.0861 | 6.15 steps/s | VRAM=27.76 Go [2026-07-17 23:27:48] step 520450/1000000 | loss=2.2930 vocab=1.3074 token=1.7465 | 6.13 steps/s | VRAM=32.84 Go [2026-07-17 23:27:56] step 520500/1000000 | loss=2.2332 vocab=0.6217 token=1.1236 | 6.13 steps/s | VRAM=27.78 Go [2026-07-17 23:28:04] step 520550/1000000 | loss=2.1694 vocab=1.2942 token=1.2597 | 6.12 steps/s | VRAM=28.96 Go [2026-07-17 23:28:12] step 520600/1000000 | loss=2.2544 vocab=0.9360 token=1.4051 | 6.13 steps/s | VRAM=25.07 Go [2026-07-17 23:28:20] step 520650/1000000 | loss=2.2110 vocab=0.5736 token=0.9333 | 6.14 steps/s | VRAM=30.51 Go [2026-07-17 23:28:28] step 520700/1000000 | loss=2.2604 vocab=0.6949 token=1.3186 | 6.14 steps/s | VRAM=31.66 Go [2026-07-17 23:28:37] step 520750/1000000 | loss=2.1043 vocab=0.8579 token=1.1565 | 6.11 steps/s | VRAM=26.65 Go [2026-07-17 23:28:45] step 520800/1000000 | loss=2.0669 vocab=0.7128 token=1.7983 | 6.15 steps/s | VRAM=29.43 Go [2026-07-17 23:28:53] step 520850/1000000 | loss=2.2172 vocab=0.5468 token=1.1858 | 6.14 steps/s | VRAM=29.43 Go [2026-07-17 23:29:01] step 520900/1000000 | loss=2.1449 vocab=0.7573 token=1.3111 | 6.12 steps/s | VRAM=27.66 Go [2026-07-17 23:29:09] step 520950/1000000 | loss=2.2610 vocab=0.7236 token=0.7840 | 6.12 steps/s | VRAM=24.21 Go [2026-07-17 23:29:17] step 521000/1000000 | loss=2.1884 vocab=0.8520 token=1.2028 | 6.18 steps/s | VRAM=29.30 Go [2026-07-17 23:29:25] step 521050/1000000 | loss=2.2458 vocab=1.0308 token=1.5713 | 6.10 steps/s | VRAM=25.31 Go [2026-07-17 23:29:34] step 521100/1000000 | loss=2.1692 vocab=0.9326 token=1.5783 | 6.15 steps/s | VRAM=28.40 Go [2026-07-17 23:29:42] step 521150/1000000 | loss=2.2651 vocab=0.6369 token=0.8372 | 6.16 steps/s | VRAM=28.06 Go [2026-07-17 23:29:50] step 521200/1000000 | loss=2.2173 vocab=0.8161 token=1.3166 | 6.15 steps/s | VRAM=26.64 Go [2026-07-17 23:29:58] step 521250/1000000 | loss=2.1712 vocab=1.0754 token=1.4239 | 6.13 steps/s | VRAM=28.04 Go [2026-07-17 23:30:06] step 521300/1000000 | loss=2.1593 vocab=0.7532 token=1.1201 | 6.14 steps/s | VRAM=29.43 Go [2026-07-17 23:30:14] step 521350/1000000 | loss=2.2238 vocab=0.4972 token=1.1162 | 6.10 steps/s | VRAM=25.58 Go [2026-07-17 23:30:23] step 521400/1000000 | loss=2.2490 vocab=1.0776 token=1.1474 | 6.12 steps/s | VRAM=23.41 Go [2026-07-17 23:30:31] step 521450/1000000 | loss=2.2181 vocab=1.2141 token=1.4207 | 6.16 steps/s | VRAM=24.71 Go [2026-07-17 23:30:39] step 521500/1000000 | loss=2.2840 vocab=0.6662 token=1.0419 | 6.13 steps/s | VRAM=32.49 Go [2026-07-17 23:30:47] step 521550/1000000 | loss=2.3212 vocab=0.4960 token=1.3428 | 6.17 steps/s | VRAM=25.60 Go [2026-07-17 23:30:55] step 521600/1000000 | loss=2.2421 vocab=0.7501 token=1.4798 | 6.12 steps/s | VRAM=28.67 Go [2026-07-17 23:31:03] step 521650/1000000 | loss=2.1888 vocab=0.8295 token=1.3048 | 6.13 steps/s | VRAM=32.06 Go [2026-07-17 23:31:11] step 521700/1000000 | loss=2.2920 vocab=1.1036 token=1.0064 | 6.10 steps/s | VRAM=30.11 Go [2026-07-17 23:31:20] step 521750/1000000 | loss=2.1418 vocab=0.9641 token=1.3815 | 6.13 steps/s | VRAM=28.81 Go [2026-07-17 23:31:28] step 521800/1000000 | loss=2.1116 vocab=0.9072 token=1.1169 | 6.12 steps/s | VRAM=28.26 Go [2026-07-17 23:31:36] step 521850/1000000 | loss=2.2086 vocab=1.3840 token=1.5630 | 6.16 steps/s | VRAM=30.53 Go [2026-07-17 23:31:44] step 521900/1000000 | loss=2.3431 vocab=1.1953 token=1.4437 | 6.15 steps/s | VRAM=28.12 Go [2026-07-17 23:31:52] step 521950/1000000 | loss=2.2965 vocab=1.2898 token=1.2395 | 6.11 steps/s | VRAM=30.20 Go [2026-07-17 23:32:00] step 522000/1000000 | loss=2.2159 vocab=0.8774 token=1.5190 | 6.15 steps/s | VRAM=26.54 Go [2026-07-17 23:32:08] step 522050/1000000 | loss=2.1811 vocab=0.5342 token=1.0738 | 6.15 steps/s | VRAM=32.99 Go [2026-07-17 23:32:17] step 522100/1000000 | loss=2.2439 vocab=0.9667 token=1.5205 | 6.14 steps/s | VRAM=27.99 Go [2026-07-17 23:32:25] step 522150/1000000 | loss=2.1373 vocab=0.7578 token=0.8237 | 6.15 steps/s | VRAM=28.13 Go [2026-07-17 23:32:33] step 522200/1000000 | loss=2.2262 vocab=0.4420 token=1.0841 | 6.12 steps/s | VRAM=26.20 Go [2026-07-17 23:32:41] step 522250/1000000 | loss=2.1560 vocab=0.6450 token=1.1542 | 6.14 steps/s | VRAM=30.18 Go [2026-07-17 23:32:49] step 522300/1000000 | loss=2.2133 vocab=0.6526 token=1.0973 | 6.14 steps/s | VRAM=28.24 Go [2026-07-17 23:32:57] step 522350/1000000 | loss=2.3209 vocab=0.7116 token=1.1081 | 6.08 steps/s | VRAM=32.98 Go [2026-07-17 23:33:06] step 522400/1000000 | loss=2.2252 vocab=0.6726 token=1.3301 | 6.16 steps/s | VRAM=24.39 Go [2026-07-17 23:33:14] step 522450/1000000 | loss=2.1690 vocab=1.0438 token=1.3714 | 6.14 steps/s | VRAM=26.16 Go [2026-07-17 23:33:22] step 522500/1000000 | loss=2.2461 vocab=0.8313 token=1.6162 | 6.15 steps/s | VRAM=28.92 Go [2026-07-17 23:33:30] step 522550/1000000 | loss=2.2629 vocab=0.7317 token=1.2726 | 6.11 steps/s | VRAM=26.41 Go [2026-07-17 23:33:38] step 522600/1000000 | loss=2.2268 vocab=0.5060 token=1.1461 | 6.14 steps/s | VRAM=30.11 Go [2026-07-17 23:33:46] step 522650/1000000 | loss=2.2257 vocab=0.6692 token=1.2025 | 6.15 steps/s | VRAM=31.13 Go [2026-07-17 23:33:54] step 522700/1000000 | loss=2.3416 vocab=0.9519 token=1.6236 | 6.14 steps/s | VRAM=28.05 Go [2026-07-17 23:34:03] step 522750/1000000 | loss=2.1735 vocab=0.8373 token=1.3676 | 6.13 steps/s | VRAM=29.32 Go [2026-07-17 23:34:11] step 522800/1000000 | loss=2.1926 vocab=0.8133 token=1.0890 | 6.13 steps/s | VRAM=28.46 Go [2026-07-17 23:34:19] step 522850/1000000 | loss=2.1832 vocab=0.5332 token=1.1849 | 6.12 steps/s | VRAM=32.59 Go [2026-07-17 23:34:27] step 522900/1000000 | loss=2.3624 vocab=1.2938 token=1.5945 | 6.12 steps/s | VRAM=26.98 Go [2026-07-17 23:34:35] step 522950/1000000 | loss=2.2486 vocab=1.0899 token=1.4845 | 6.15 steps/s | VRAM=25.15 Go [2026-07-17 23:34:43] step 523000/1000000 | loss=2.2142 vocab=0.9780 token=0.8287 | 6.15 steps/s | VRAM=28.57 Go [2026-07-17 23:34:52] step 523050/1000000 | loss=2.1547 vocab=0.8807 token=1.3505 | 6.11 steps/s | VRAM=26.29 Go [2026-07-17 23:35:00] step 523100/1000000 | loss=2.1822 vocab=0.9755 token=1.5422 | 6.11 steps/s | VRAM=29.53 Go [2026-07-17 23:35:08] step 523150/1000000 | loss=2.2441 vocab=1.1620 token=1.4422 | 6.11 steps/s | VRAM=29.40 Go [2026-07-17 23:35:16] step 523200/1000000 | loss=2.2464 vocab=0.6286 token=1.1815 | 6.15 steps/s | VRAM=32.28 Go [2026-07-17 23:35:24] step 523250/1000000 | loss=2.2221 vocab=0.9063 token=1.6799 | 6.12 steps/s | VRAM=26.37 Go [2026-07-17 23:35:32] step 523300/1000000 | loss=2.1914 vocab=1.0645 token=1.3971 | 6.17 steps/s | VRAM=31.20 Go [2026-07-17 23:35:40] step 523350/1000000 | loss=2.3540 vocab=0.6735 token=0.9267 | 6.11 steps/s | VRAM=29.06 Go [2026-07-17 23:35:49] step 523400/1000000 | loss=2.2875 vocab=0.8691 token=1.8637 | 6.16 steps/s | VRAM=30.49 Go [2026-07-17 23:35:57] step 523450/1000000 | loss=2.2733 vocab=0.9345 token=1.6832 | 6.15 steps/s | VRAM=31.69 Go [2026-07-17 23:36:05] step 523500/1000000 | loss=2.4229 vocab=1.1485 token=1.6555 | 6.15 steps/s | VRAM=25.80 Go [2026-07-17 23:36:13] step 523550/1000000 | loss=2.2001 vocab=1.0724 token=1.6176 | 6.12 steps/s | VRAM=29.23 Go [2026-07-17 23:36:21] step 523600/1000000 | loss=2.2810 vocab=0.4716 token=1.1492 | 6.15 steps/s | VRAM=26.06 Go [2026-07-17 23:36:29] step 523650/1000000 | loss=2.1400 vocab=0.7149 token=1.2606 | 6.16 steps/s | VRAM=32.83 Go [2026-07-17 23:36:37] step 523700/1000000 | loss=2.2095 vocab=1.1618 token=1.3149 | 6.08 steps/s | VRAM=30.73 Go [2026-07-17 23:36:46] step 523750/1000000 | loss=2.1471 vocab=0.7004 token=1.4291 | 6.17 steps/s | VRAM=31.47 Go [2026-07-17 23:36:54] step 523800/1000000 | loss=2.2215 vocab=0.9865 token=1.2322 | 6.14 steps/s | VRAM=27.09 Go [2026-07-17 23:37:02] step 523850/1000000 | loss=2.1887 vocab=0.5610 token=1.1279 | 6.14 steps/s | VRAM=32.67 Go [2026-07-17 23:37:10] step 523900/1000000 | loss=2.2245 vocab=0.3827 token=0.7204 | 6.14 steps/s | VRAM=31.41 Go [2026-07-17 23:37:18] step 523950/1000000 | loss=2.2771 vocab=0.9651 token=1.3227 | 6.12 steps/s | VRAM=32.52 Go [2026-07-17 23:37:26] step 524000/1000000 | loss=2.1822 vocab=0.8907 token=1.2543 | 6.13 steps/s | VRAM=32.83 Go [2026-07-17 23:37:34] step 524050/1000000 | loss=2.3554 vocab=1.2161 token=1.2108 | 6.13 steps/s | VRAM=25.92 Go [2026-07-17 23:37:43] step 524100/1000000 | loss=2.2858 vocab=0.7116 token=1.5950 | 6.16 steps/s | VRAM=24.53 Go [2026-07-17 23:37:51] step 524150/1000000 | loss=2.2638 vocab=0.9149 token=1.0987 | 6.18 steps/s | VRAM=32.21 Go [2026-07-17 23:37:59] step 524200/1000000 | loss=2.2923 vocab=0.9575 token=0.9782 | 6.13 steps/s | VRAM=29.36 Go [2026-07-17 23:38:07] step 524250/1000000 | loss=2.2847 vocab=0.9827 token=1.4682 | 6.13 steps/s | VRAM=28.66 Go [2026-07-17 23:38:15] step 524300/1000000 | loss=2.3274 vocab=1.3392 token=1.5560 | 6.15 steps/s | VRAM=32.67 Go [2026-07-17 23:38:23] step 524350/1000000 | loss=2.2127 vocab=0.6831 token=1.7778 | 6.14 steps/s | VRAM=29.78 Go [2026-07-17 23:38:31] step 524400/1000000 | loss=2.2069 vocab=1.1715 token=1.3335 | 6.14 steps/s | VRAM=29.75 Go [2026-07-17 23:38:40] step 524450/1000000 | loss=2.2411 vocab=1.0663 token=1.4381 | 6.17 steps/s | VRAM=31.86 Go [2026-07-17 23:38:48] step 524500/1000000 | loss=2.1764 vocab=0.7283 token=1.0812 | 6.12 steps/s | VRAM=30.85 Go [2026-07-17 23:38:56] step 524550/1000000 | loss=2.1865 vocab=1.6157 token=1.3073 | 6.11 steps/s | VRAM=26.29 Go [2026-07-17 23:39:04] step 524600/1000000 | loss=2.3208 vocab=0.8404 token=1.4947 | 6.13 steps/s | VRAM=25.46 Go [2026-07-17 23:39:12] step 524650/1000000 | loss=2.3222 vocab=1.0137 token=1.6936 | 6.15 steps/s | VRAM=32.14 Go [2026-07-17 23:39:20] step 524700/1000000 | loss=2.2437 vocab=0.6637 token=1.3930 | 6.13 steps/s | VRAM=29.42 Go [2026-07-17 23:39:29] step 524750/1000000 | loss=2.3278 vocab=0.7709 token=1.5754 | 6.11 steps/s | VRAM=26.57 Go [2026-07-17 23:39:37] step 524800/1000000 | loss=2.2691 vocab=0.5783 token=1.0448 | 6.16 steps/s | VRAM=32.66 Go [2026-07-17 23:39:45] step 524850/1000000 | loss=2.2004 vocab=0.7630 token=1.0747 | 6.15 steps/s | VRAM=25.76 Go [2026-07-17 23:39:53] step 524900/1000000 | loss=2.2984 vocab=1.0131 token=1.4004 | 6.16 steps/s | VRAM=26.03 Go [2026-07-17 23:40:01] step 524950/1000000 | loss=2.2728 vocab=0.7962 token=1.3068 | 6.12 steps/s | VRAM=29.88 Go [2026-07-17 23:40:09] step 525000/1000000 | loss=2.2672 vocab=0.8509 token=1.2270 | 6.11 steps/s | VRAM=26.67 Go [2026-07-17 23:40:17] step 525050/1000000 | loss=2.2367 vocab=0.9700 token=1.3255 | 6.15 steps/s | VRAM=27.87 Go [2026-07-17 23:40:26] step 525100/1000000 | loss=2.1169 vocab=0.8591 token=1.6542 | 6.15 steps/s | VRAM=29.15 Go [2026-07-17 23:40:34] step 525150/1000000 | loss=2.3765 vocab=0.7085 token=1.3631 | 6.18 steps/s | VRAM=28.42 Go [2026-07-17 23:40:42] step 525200/1000000 | loss=2.2071 vocab=0.5585 token=1.2105 | 6.17 steps/s | VRAM=31.19 Go [2026-07-17 23:40:50] step 525250/1000000 | loss=2.3465 vocab=1.1046 token=1.7276 | 6.12 steps/s | VRAM=28.95 Go [2026-07-17 23:40:58] step 525300/1000000 | loss=2.3107 vocab=0.8513 token=1.4783 | 6.13 steps/s | VRAM=25.43 Go [2026-07-17 23:41:06] step 525350/1000000 | loss=2.2499 vocab=0.6503 token=1.4952 | 6.16 steps/s | VRAM=25.65 Go [2026-07-17 23:41:14] step 525400/1000000 | loss=2.2683 vocab=0.9940 token=1.3486 | 6.17 steps/s | VRAM=28.76 Go [2026-07-17 23:41:22] step 525450/1000000 | loss=2.2638 vocab=0.8192 token=1.7368 | 6.14 steps/s | VRAM=30.86 Go [2026-07-17 23:41:31] step 525500/1000000 | loss=2.2230 vocab=1.0124 token=1.5052 | 6.13 steps/s | VRAM=27.81 Go [2026-07-17 23:41:39] step 525550/1000000 | loss=2.2142 vocab=0.3774 token=1.0644 | 6.17 steps/s | VRAM=31.26 Go [2026-07-17 23:41:47] step 525600/1000000 | loss=2.2148 vocab=1.1942 token=1.5115 | 6.14 steps/s | VRAM=24.75 Go [2026-07-17 23:41:55] step 525650/1000000 | loss=2.2730 vocab=0.8045 token=1.2935 | 6.14 steps/s | VRAM=31.90 Go [2026-07-17 23:42:03] step 525700/1000000 | loss=2.3336 vocab=0.8464 token=1.3073 | 6.10 steps/s | VRAM=24.53 Go [2026-07-17 23:42:11] step 525750/1000000 | loss=2.1763 vocab=0.8608 token=1.2421 | 6.13 steps/s | VRAM=29.46 Go [2026-07-17 23:42:19] step 525800/1000000 | loss=2.3165 vocab=0.9910 token=1.2671 | 6.10 steps/s | VRAM=29.76 Go [2026-07-17 23:42:28] step 525850/1000000 | loss=2.0619 vocab=0.5078 token=0.8682 | 6.11 steps/s | VRAM=28.35 Go [2026-07-17 23:42:36] step 525900/1000000 | loss=2.1840 vocab=0.8109 token=1.1987 | 6.10 steps/s | VRAM=26.65 Go [2026-07-17 23:42:44] step 525950/1000000 | loss=2.2977 vocab=1.0824 token=1.5467 | 6.12 steps/s | VRAM=32.35 Go [2026-07-17 23:42:52] step 526000/1000000 | loss=2.2248 vocab=0.7945 token=1.2124 | 6.08 steps/s | VRAM=29.24 Go [2026-07-17 23:43:00] step 526050/1000000 | loss=2.1408 vocab=0.8003 token=1.4607 | 6.13 steps/s | VRAM=28.35 Go [2026-07-17 23:43:09] step 526100/1000000 | loss=2.2742 vocab=0.8541 token=1.3639 | 6.14 steps/s | VRAM=28.43 Go [2026-07-17 23:43:17] step 526150/1000000 | loss=2.1959 vocab=0.8195 token=1.2854 | 6.15 steps/s | VRAM=32.23 Go [2026-07-17 23:43:25] step 526200/1000000 | loss=2.3218 vocab=0.6834 token=1.4431 | 6.14 steps/s | VRAM=29.39 Go [2026-07-17 23:43:33] step 526250/1000000 | loss=2.2234 vocab=1.0064 token=1.3122 | 6.11 steps/s | VRAM=29.65 Go [2026-07-17 23:43:41] step 526300/1000000 | loss=2.3048 vocab=1.0904 token=1.3742 | 6.15 steps/s | VRAM=28.63 Go [2026-07-17 23:43:49] step 526350/1000000 | loss=2.2170 vocab=1.0473 token=1.4892 | 6.15 steps/s | VRAM=28.74 Go [2026-07-17 23:43:57] step 526400/1000000 | loss=2.2265 vocab=0.9618 token=1.5209 | 6.16 steps/s | VRAM=30.65 Go [2026-07-17 23:44:06] step 526450/1000000 | loss=2.3185 vocab=0.8213 token=1.5334 | 6.13 steps/s | VRAM=29.25 Go [2026-07-17 23:44:14] step 526500/1000000 | loss=2.3121 vocab=0.5748 token=1.5302 | 6.16 steps/s | VRAM=28.57 Go [2026-07-17 23:44:22] step 526550/1000000 | loss=2.2558 vocab=0.9963 token=1.3907 | 6.15 steps/s | VRAM=31.34 Go [2026-07-17 23:44:30] step 526600/1000000 | loss=2.1958 vocab=1.0185 token=1.3204 | 6.15 steps/s | VRAM=31.56 Go [2026-07-17 23:44:38] step 526650/1000000 | loss=2.2112 vocab=1.0982 token=1.3350 | 6.13 steps/s | VRAM=29.06 Go [2026-07-17 23:44:46] step 526700/1000000 | loss=2.2664 vocab=1.3964 token=1.3394 | 6.16 steps/s | VRAM=28.26 Go [2026-07-17 23:44:54] step 526750/1000000 | loss=2.3008 vocab=0.7688 token=1.3769 | 6.14 steps/s | VRAM=32.48 Go [2026-07-17 23:45:03] step 526800/1000000 | loss=2.1780 vocab=1.4593 token=1.3993 | 6.12 steps/s | VRAM=31.52 Go [2026-07-17 23:45:11] step 526850/1000000 | loss=2.2462 vocab=1.1462 token=1.4931 | 6.14 steps/s | VRAM=25.03 Go [2026-07-17 23:45:19] step 526900/1000000 | loss=2.1972 vocab=0.9923 token=1.0521 | 6.12 steps/s | VRAM=31.94 Go [2026-07-17 23:45:27] step 526950/1000000 | loss=2.2685 vocab=1.1251 token=1.8119 | 6.08 steps/s | VRAM=28.54 Go [2026-07-17 23:45:35] step 527000/1000000 | loss=2.2434 vocab=0.7200 token=1.6705 | 6.13 steps/s | VRAM=32.38 Go [2026-07-17 23:45:43] step 527050/1000000 | loss=2.2655 vocab=1.0257 token=1.4431 | 6.11 steps/s | VRAM=28.45 Go [2026-07-17 23:45:52] step 527100/1000000 | loss=2.2096 vocab=0.6642 token=0.9452 | 6.12 steps/s | VRAM=29.22 Go [2026-07-17 23:46:00] step 527150/1000000 | loss=2.2429 vocab=0.9725 token=1.0598 | 6.13 steps/s | VRAM=27.99 Go [2026-07-17 23:46:08] step 527200/1000000 | loss=2.1135 vocab=0.5093 token=1.6157 | 6.11 steps/s | VRAM=28.21 Go [2026-07-17 23:46:16] step 527250/1000000 | loss=2.2296 vocab=0.5399 token=1.2375 | 6.10 steps/s | VRAM=29.23 Go [2026-07-17 23:46:24] step 527300/1000000 | loss=2.2918 vocab=1.0483 token=1.5250 | 6.06 steps/s | VRAM=28.38 Go [2026-07-17 23:46:33] step 527350/1000000 | loss=2.2396 vocab=0.9922 token=1.7210 | 6.11 steps/s | VRAM=24.69 Go [2026-07-17 23:46:41] step 527400/1000000 | loss=2.2074 vocab=0.6701 token=1.1075 | 6.11 steps/s | VRAM=31.96 Go [2026-07-17 23:46:49] step 527450/1000000 | loss=2.3061 vocab=1.2531 token=1.3949 | 6.10 steps/s | VRAM=24.85 Go [2026-07-17 23:46:57] step 527500/1000000 | loss=2.2094 vocab=0.8693 token=1.3548 | 6.11 steps/s | VRAM=32.21 Go [2026-07-17 23:47:05] step 527550/1000000 | loss=2.3168 vocab=1.3810 token=1.8719 | 6.12 steps/s | VRAM=28.24 Go [2026-07-17 23:47:13] step 527600/1000000 | loss=2.2341 vocab=1.1649 token=1.2958 | 6.11 steps/s | VRAM=27.99 Go [2026-07-17 23:47:22] step 527650/1000000 | loss=2.2301 vocab=1.1362 token=1.2318 | 6.12 steps/s | VRAM=31.87 Go [2026-07-17 23:47:30] step 527700/1000000 | loss=2.2631 vocab=1.1620 token=1.3832 | 6.13 steps/s | VRAM=28.91 Go [2026-07-17 23:47:38] step 527750/1000000 | loss=2.1805 vocab=1.0037 token=1.1649 | 6.12 steps/s | VRAM=30.27 Go [2026-07-17 23:47:46] step 527800/1000000 | loss=2.2947 vocab=0.7366 token=1.2588 | 6.14 steps/s | VRAM=29.48 Go [2026-07-17 23:47:54] step 527850/1000000 | loss=2.2725 vocab=0.9977 token=1.6351 | 6.16 steps/s | VRAM=31.65 Go [2026-07-17 23:48:02] step 527900/1000000 | loss=2.2301 vocab=1.2177 token=1.2781 | 6.13 steps/s | VRAM=26.14 Go [2026-07-17 23:48:11] step 527950/1000000 | loss=2.2905 vocab=0.9498 token=1.4375 | 6.15 steps/s | VRAM=32.67 Go [2026-07-17 23:48:19] step 528000/1000000 | loss=2.2463 vocab=1.4057 token=1.6222 | 6.13 steps/s | VRAM=29.09 Go [2026-07-17 23:48:27] step 528050/1000000 | loss=2.2812 vocab=1.2350 token=0.9214 | 6.15 steps/s | VRAM=31.37 Go [2026-07-17 23:48:35] step 528100/1000000 | loss=2.2267 vocab=1.0092 token=1.1776 | 6.12 steps/s | VRAM=28.36 Go [2026-07-17 23:48:43] step 528150/1000000 | loss=2.2651 vocab=0.6333 token=1.2795 | 6.12 steps/s | VRAM=29.30 Go [2026-07-17 23:48:51] step 528200/1000000 | loss=2.2699 vocab=0.3199 token=0.9029 | 6.13 steps/s | VRAM=25.64 Go [2026-07-17 23:48:59] step 528250/1000000 | loss=2.2074 vocab=0.9554 token=1.4735 | 6.13 steps/s | VRAM=26.32 Go [2026-07-17 23:49:08] step 528300/1000000 | loss=2.2516 vocab=0.9268 token=1.4189 | 6.13 steps/s | VRAM=29.48 Go [2026-07-17 23:49:16] step 528350/1000000 | loss=2.2731 vocab=0.9774 token=1.2858 | 6.14 steps/s | VRAM=32.18 Go [2026-07-17 23:49:24] step 528400/1000000 | loss=2.3117 vocab=0.9733 token=1.3863 | 6.13 steps/s | VRAM=26.41 Go [2026-07-17 23:49:32] step 528450/1000000 | loss=2.1724 vocab=0.9325 token=1.4384 | 6.11 steps/s | VRAM=32.08 Go [2026-07-17 23:49:40] step 528500/1000000 | loss=2.1955 vocab=1.1599 token=1.2541 | 6.12 steps/s | VRAM=26.52 Go [2026-07-17 23:49:48] step 528550/1000000 | loss=2.2487 vocab=0.7912 token=0.9870 | 6.17 steps/s | VRAM=31.73 Go [2026-07-17 23:49:57] step 528600/1000000 | loss=2.3284 vocab=0.6202 token=1.1271 | 6.12 steps/s | VRAM=32.31 Go [2026-07-17 23:50:05] step 528650/1000000 | loss=2.2873 vocab=0.6203 token=1.3925 | 6.14 steps/s | VRAM=28.87 Go [2026-07-17 23:50:13] step 528700/1000000 | loss=2.2790 vocab=0.7905 token=1.6295 | 6.10 steps/s | VRAM=23.46 Go [2026-07-17 23:50:21] step 528750/1000000 | loss=2.1883 vocab=0.8187 token=1.5583 | 6.14 steps/s | VRAM=28.46 Go [2026-07-17 23:50:29] step 528800/1000000 | loss=2.1864 vocab=1.1336 token=1.0716 | 6.11 steps/s | VRAM=25.89 Go [2026-07-17 23:50:37] step 528850/1000000 | loss=2.2247 vocab=0.8969 token=1.1352 | 6.14 steps/s | VRAM=29.39 Go [2026-07-17 23:50:45] step 528900/1000000 | loss=2.2829 vocab=0.8292 token=1.3031 | 6.14 steps/s | VRAM=28.23 Go [2026-07-17 23:50:54] step 528950/1000000 | loss=2.2896 vocab=1.0586 token=1.5122 | 6.14 steps/s | VRAM=28.62 Go [2026-07-17 23:51:02] step 529000/1000000 | loss=2.2271 vocab=1.0382 token=1.1994 | 6.11 steps/s | VRAM=26.02 Go [2026-07-17 23:51:10] step 529050/1000000 | loss=2.2995 vocab=1.4673 token=1.8715 | 6.14 steps/s | VRAM=28.96 Go [2026-07-17 23:51:18] step 529100/1000000 | loss=2.2347 vocab=0.6034 token=1.3268 | 6.13 steps/s | VRAM=28.41 Go [2026-07-17 23:51:26] step 529150/1000000 | loss=2.2306 vocab=1.0383 token=1.6156 | 6.12 steps/s | VRAM=32.68 Go [2026-07-17 23:51:34] step 529200/1000000 | loss=2.2552 vocab=0.8331 token=1.0258 | 6.11 steps/s | VRAM=28.61 Go [2026-07-17 23:51:43] step 529250/1000000 | loss=2.2982 vocab=0.8886 token=1.5213 | 6.12 steps/s | VRAM=29.09 Go [2026-07-17 23:51:51] step 529300/1000000 | loss=2.1745 vocab=1.0166 token=1.5317 | 6.11 steps/s | VRAM=28.95 Go [2026-07-17 23:51:59] step 529350/1000000 | loss=2.3401 vocab=1.1114 token=1.3908 | 6.12 steps/s | VRAM=28.82 Go [2026-07-17 23:52:07] step 529400/1000000 | loss=2.3240 vocab=0.7274 token=1.4199 | 6.14 steps/s | VRAM=29.34 Go [2026-07-17 23:52:15] step 529450/1000000 | loss=2.2149 vocab=0.8898 token=1.6139 | 6.13 steps/s | VRAM=31.68 Go [2026-07-17 23:52:23] step 529500/1000000 | loss=2.2626 vocab=1.1801 token=1.0946 | 6.14 steps/s | VRAM=28.37 Go [2026-07-17 23:52:32] step 529550/1000000 | loss=2.2094 vocab=1.0231 token=1.0343 | 6.13 steps/s | VRAM=26.26 Go [2026-07-17 23:52:40] step 529600/1000000 | loss=2.2311 vocab=0.4889 token=1.2762 | 6.15 steps/s | VRAM=28.30 Go [2026-07-17 23:52:48] step 529650/1000000 | loss=2.1985 vocab=0.7940 token=1.1024 | 6.15 steps/s | VRAM=28.67 Go [2026-07-17 23:52:56] step 529700/1000000 | loss=2.2161 vocab=0.5465 token=1.3405 | 6.13 steps/s | VRAM=28.83 Go [2026-07-17 23:53:04] step 529750/1000000 | loss=2.3536 vocab=0.8080 token=1.3450 | 6.15 steps/s | VRAM=31.57 Go [2026-07-17 23:53:12] step 529800/1000000 | loss=2.2370 vocab=0.7804 token=1.8447 | 6.13 steps/s | VRAM=29.54 Go [2026-07-17 23:53:20] step 529850/1000000 | loss=2.1724 vocab=0.6067 token=1.1719 | 6.14 steps/s | VRAM=28.66 Go [2026-07-17 23:53:29] step 529900/1000000 | loss=2.1875 vocab=0.7011 token=1.4620 | 6.13 steps/s | VRAM=28.61 Go [2026-07-17 23:53:37] step 529950/1000000 | loss=2.3112 vocab=0.8141 token=1.1042 | 6.16 steps/s | VRAM=31.86 Go [2026-07-17 23:53:45] step 530000/1000000 | loss=2.1158 vocab=1.0708 token=1.5167 | 6.15 steps/s | VRAM=29.13 Go [2026-07-17 23:53:53] step 530050/1000000 | loss=2.1221 vocab=0.3962 token=0.8996 | 6.14 steps/s | VRAM=28.08 Go [2026-07-17 23:54:01] step 530100/1000000 | loss=2.2218 vocab=0.8930 token=1.1911 | 6.17 steps/s | VRAM=32.00 Go [2026-07-17 23:54:09] step 530150/1000000 | loss=2.3065 vocab=0.9587 token=1.7271 | 6.14 steps/s | VRAM=28.03 Go [2026-07-17 23:54:17] step 530200/1000000 | loss=2.2942 vocab=0.8605 token=1.2543 | 6.16 steps/s | VRAM=32.44 Go [2026-07-17 23:54:26] step 530250/1000000 | loss=2.1710 vocab=1.4603 token=1.2350 | 6.11 steps/s | VRAM=28.50 Go [2026-07-17 23:54:34] step 530300/1000000 | loss=2.2239 vocab=1.2644 token=1.6083 | 6.14 steps/s | VRAM=28.84 Go [2026-07-17 23:54:42] step 530350/1000000 | loss=2.2324 vocab=0.9044 token=1.0728 | 6.17 steps/s | VRAM=28.48 Go [2026-07-17 23:54:50] step 530400/1000000 | loss=2.1341 vocab=0.8383 token=1.1215 | 6.16 steps/s | VRAM=31.54 Go [2026-07-17 23:54:58] step 530450/1000000 | loss=2.2156 vocab=0.5164 token=1.4582 | 6.14 steps/s | VRAM=29.62 Go [2026-07-17 23:55:06] step 530500/1000000 | loss=2.2728 vocab=0.9307 token=1.1467 | 6.14 steps/s | VRAM=28.75 Go [2026-07-17 23:55:14] step 530550/1000000 | loss=2.2242 vocab=1.4518 token=1.2333 | 6.15 steps/s | VRAM=31.11 Go [2026-07-17 23:55:22] step 530600/1000000 | loss=2.1266 vocab=0.7806 token=1.3096 | 6.14 steps/s | VRAM=31.65 Go [2026-07-17 23:55:31] step 530650/1000000 | loss=2.2246 vocab=0.8313 token=1.3182 | 6.13 steps/s | VRAM=28.46 Go [2026-07-17 23:55:39] step 530700/1000000 | loss=2.2657 vocab=0.7304 token=1.0327 | 6.08 steps/s | VRAM=27.56 Go [2026-07-17 23:55:47] step 530750/1000000 | loss=2.1920 vocab=1.0683 token=1.5828 | 6.14 steps/s | VRAM=24.33 Go [2026-07-17 23:55:55] step 530800/1000000 | loss=2.1344 vocab=0.9922 token=1.0532 | 6.14 steps/s | VRAM=29.09 Go [2026-07-17 23:56:03] step 530850/1000000 | loss=2.2163 vocab=0.8710 token=1.5315 | 6.13 steps/s | VRAM=25.77 Go [2026-07-17 23:56:11] step 530900/1000000 | loss=2.2740 vocab=0.8914 token=1.1689 | 6.13 steps/s | VRAM=32.10 Go [2026-07-17 23:56:20] step 530950/1000000 | loss=2.2210 vocab=1.1169 token=1.1367 | 6.12 steps/s | VRAM=32.84 Go [2026-07-17 23:56:28] step 531000/1000000 | loss=2.4080 vocab=1.0488 token=1.0912 | 6.12 steps/s | VRAM=26.62 Go [2026-07-17 23:56:36] step 531050/1000000 | loss=2.2161 vocab=1.4630 token=1.7528 | 6.15 steps/s | VRAM=29.00 Go [2026-07-17 23:56:44] step 531100/1000000 | loss=2.1567 vocab=0.7195 token=1.2456 | 6.16 steps/s | VRAM=29.40 Go [2026-07-17 23:56:52] step 531150/1000000 | loss=2.2124 vocab=0.8139 token=1.3742 | 6.17 steps/s | VRAM=24.82 Go [2026-07-17 23:57:00] step 531200/1000000 | loss=2.2196 vocab=1.2225 token=1.1674 | 6.14 steps/s | VRAM=29.59 Go [2026-07-17 23:57:08] step 531250/1000000 | loss=2.1863 vocab=1.2930 token=1.6090 | 6.14 steps/s | VRAM=28.97 Go [2026-07-17 23:57:17] step 531300/1000000 | loss=2.2592 vocab=0.8922 token=1.3601 | 6.14 steps/s | VRAM=26.69 Go [2026-07-17 23:57:25] step 531350/1000000 | loss=2.2036 vocab=0.8109 token=1.3037 | 6.16 steps/s | VRAM=26.15 Go [2026-07-17 23:57:33] step 531400/1000000 | loss=2.2379 vocab=0.9064 token=1.4063 | 6.14 steps/s | VRAM=28.09 Go [2026-07-17 23:57:41] step 531450/1000000 | loss=2.2395 vocab=0.6248 token=1.2879 | 6.13 steps/s | VRAM=26.68 Go [2026-07-17 23:57:49] step 531500/1000000 | loss=2.2722 vocab=1.2359 token=1.4906 | 6.15 steps/s | VRAM=28.70 Go [2026-07-17 23:57:57] step 531550/1000000 | loss=2.1858 vocab=0.5194 token=1.3717 | 6.14 steps/s | VRAM=32.13 Go [2026-07-17 23:58:05] step 531600/1000000 | loss=2.2003 vocab=1.3183 token=1.4505 | 6.16 steps/s | VRAM=29.22 Go [2026-07-17 23:58:13] step 531650/1000000 | loss=2.2144 vocab=0.5782 token=1.4649 | 6.17 steps/s | VRAM=28.46 Go [2026-07-17 23:58:22] step 531700/1000000 | loss=2.2997 vocab=0.7162 token=0.6935 | 6.15 steps/s | VRAM=31.15 Go [2026-07-17 23:58:30] step 531750/1000000 | loss=2.1631 vocab=0.5455 token=1.0349 | 6.13 steps/s | VRAM=25.88 Go [2026-07-17 23:58:38] step 531800/1000000 | loss=2.2732 vocab=1.0421 token=1.3654 | 6.14 steps/s | VRAM=26.47 Go [2026-07-17 23:58:46] step 531850/1000000 | loss=2.3081 vocab=0.9315 token=1.2895 | 6.15 steps/s | VRAM=25.95 Go [2026-07-17 23:58:54] step 531900/1000000 | loss=2.2539 vocab=1.2101 token=1.6207 | 6.14 steps/s | VRAM=28.46 Go [2026-07-17 23:59:02] step 531950/1000000 | loss=2.2674 vocab=0.8720 token=1.4002 | 6.12 steps/s | VRAM=29.20 Go [2026-07-17 23:59:10] step 532000/1000000 | loss=2.3130 vocab=0.8496 token=1.4537 | 6.14 steps/s | VRAM=25.95 Go [2026-07-17 23:59:19] step 532050/1000000 | loss=2.3104 vocab=0.7738 token=1.2695 | 6.15 steps/s | VRAM=31.79 Go [2026-07-17 23:59:27] step 532100/1000000 | loss=2.2652 vocab=0.7458 token=1.1600 | 6.16 steps/s | VRAM=32.09 Go [2026-07-17 23:59:35] step 532150/1000000 | loss=2.1792 vocab=0.6729 token=1.0156 | 6.15 steps/s | VRAM=29.74 Go [2026-07-17 23:59:43] step 532200/1000000 | loss=2.1961 vocab=0.6274 token=1.5042 | 6.17 steps/s | VRAM=32.03 Go [2026-07-17 23:59:51] step 532250/1000000 | loss=2.2186 vocab=0.6246 token=1.0986 | 6.14 steps/s | VRAM=26.16 Go [2026-07-17 23:59:59] step 532300/1000000 | loss=2.2852 vocab=0.5994 token=1.3769 | 6.15 steps/s | VRAM=25.56 Go [2026-07-18 00:00:07] step 532350/1000000 | loss=2.2995 vocab=1.2393 token=1.4221 | 6.14 steps/s | VRAM=28.50 Go [2026-07-18 00:00:16] step 532400/1000000 | loss=2.2681 vocab=1.1168 token=1.4782 | 6.15 steps/s | VRAM=31.53 Go [2026-07-18 00:00:24] step 532450/1000000 | loss=2.1443 vocab=0.3940 token=0.9232 | 6.16 steps/s | VRAM=26.08 Go [2026-07-18 00:00:32] step 532500/1000000 | loss=2.2658 vocab=1.2601 token=1.6632 | 6.16 steps/s | VRAM=28.65 Go [2026-07-18 00:00:40] step 532550/1000000 | loss=2.2102 vocab=1.3127 token=1.2974 | 6.16 steps/s | VRAM=26.19 Go [2026-07-18 00:00:48] step 532600/1000000 | loss=2.3065 vocab=0.6874 token=1.3933 | 6.14 steps/s | VRAM=33.02 Go [2026-07-18 00:00:56] step 532650/1000000 | loss=2.1669 vocab=1.4148 token=1.1745 | 6.14 steps/s | VRAM=31.66 Go [2026-07-18 00:01:04] step 532700/1000000 | loss=2.2761 vocab=0.5262 token=1.2287 | 6.13 steps/s | VRAM=26.24 Go [2026-07-18 00:01:12] step 532750/1000000 | loss=2.2116 vocab=0.6208 token=1.2418 | 6.14 steps/s | VRAM=25.99 Go [2026-07-18 00:01:21] step 532800/1000000 | loss=2.2275 vocab=1.2908 token=2.0155 | 6.15 steps/s | VRAM=26.50 Go [2026-07-18 00:01:29] step 532850/1000000 | loss=2.3157 vocab=0.8462 token=1.3628 | 6.15 steps/s | VRAM=29.13 Go [2026-07-18 00:01:37] step 532900/1000000 | loss=2.2158 vocab=0.6406 token=1.1401 | 6.15 steps/s | VRAM=28.43 Go [2026-07-18 00:01:45] step 532950/1000000 | loss=2.1722 vocab=0.8616 token=1.1264 | 6.13 steps/s | VRAM=28.37 Go [2026-07-18 00:01:53] step 533000/1000000 | loss=2.2295 vocab=0.7955 token=1.0159 | 6.13 steps/s | VRAM=26.19 Go [2026-07-18 00:02:01] step 533050/1000000 | loss=2.3174 vocab=1.1966 token=1.5510 | 6.15 steps/s | VRAM=25.72 Go [2026-07-18 00:02:09] step 533100/1000000 | loss=2.1891 vocab=0.8837 token=1.6977 | 6.15 steps/s | VRAM=26.32 Go [2026-07-18 00:02:18] step 533150/1000000 | loss=2.2841 vocab=0.5547 token=1.2618 | 6.14 steps/s | VRAM=28.79 Go [2026-07-18 00:02:26] step 533200/1000000 | loss=2.2510 vocab=0.6928 token=1.1847 | 6.15 steps/s | VRAM=32.33 Go [2026-07-18 00:02:34] step 533250/1000000 | loss=2.2498 vocab=0.8328 token=1.6203 | 6.11 steps/s | VRAM=28.67 Go [2026-07-18 00:02:42] step 533300/1000000 | loss=2.2533 vocab=1.0462 token=1.2236 | 6.14 steps/s | VRAM=31.90 Go [2026-07-18 00:02:50] step 533350/1000000 | loss=2.1940 vocab=0.6600 token=1.2261 | 6.13 steps/s | VRAM=29.72 Go [2026-07-18 00:02:58] step 533400/1000000 | loss=2.2139 vocab=0.9158 token=0.9981 | 6.15 steps/s | VRAM=31.50 Go [2026-07-18 00:03:06] step 533450/1000000 | loss=2.2578 vocab=0.9257 token=1.2821 | 6.16 steps/s | VRAM=28.51 Go [2026-07-18 00:03:15] step 533500/1000000 | loss=2.2493 vocab=0.6032 token=1.1760 | 6.13 steps/s | VRAM=28.49 Go [2026-07-18 00:03:23] step 533550/1000000 | loss=2.2390 vocab=1.1936 token=1.5493 | 6.17 steps/s | VRAM=28.09 Go [2026-07-18 00:03:31] step 533600/1000000 | loss=2.2845 vocab=0.7955 token=0.7889 | 6.13 steps/s | VRAM=29.14 Go [2026-07-18 00:03:39] step 533650/1000000 | loss=2.2454 vocab=0.8329 token=1.3809 | 6.14 steps/s | VRAM=28.29 Go [2026-07-18 00:03:47] step 533700/1000000 | loss=2.2516 vocab=1.3281 token=1.1737 | 6.12 steps/s | VRAM=29.42 Go [2026-07-18 00:03:55] step 533750/1000000 | loss=2.2086 vocab=0.8290 token=1.1362 | 6.17 steps/s | VRAM=32.39 Go [2026-07-18 00:04:03] step 533800/1000000 | loss=2.1356 vocab=0.5113 token=1.2741 | 6.12 steps/s | VRAM=28.99 Go [2026-07-18 00:04:12] step 533850/1000000 | loss=2.2706 vocab=0.6980 token=1.3364 | 6.17 steps/s | VRAM=29.02 Go [2026-07-18 00:04:20] step 533900/1000000 | loss=2.2432 vocab=0.5655 token=0.8819 | 6.15 steps/s | VRAM=28.05 Go [2026-07-18 00:04:28] step 533950/1000000 | loss=2.3429 vocab=1.1128 token=1.6995 | 6.14 steps/s | VRAM=24.75 Go [2026-07-18 00:04:36] step 534000/1000000 | loss=2.2369 vocab=0.8333 token=1.4210 | 6.13 steps/s | VRAM=25.59 Go [2026-07-18 00:04:44] step 534050/1000000 | loss=2.1816 vocab=0.6355 token=1.2362 | 6.12 steps/s | VRAM=28.98 Go [2026-07-18 00:04:52] step 534100/1000000 | loss=2.2362 vocab=0.5843 token=1.1679 | 6.12 steps/s | VRAM=29.83 Go [2026-07-18 00:05:00] step 534150/1000000 | loss=2.2105 vocab=1.5666 token=1.4515 | 6.10 steps/s | VRAM=26.83 Go [2026-07-18 00:05:09] step 534200/1000000 | loss=2.1910 vocab=0.6896 token=0.9986 | 6.16 steps/s | VRAM=27.75 Go [2026-07-18 00:05:17] step 534250/1000000 | loss=2.1201 vocab=0.9084 token=1.0470 | 6.16 steps/s | VRAM=28.99 Go [2026-07-18 00:05:25] step 534300/1000000 | loss=2.3073 vocab=0.7375 token=1.4817 | 6.16 steps/s | VRAM=27.92 Go [2026-07-18 00:05:33] step 534350/1000000 | loss=2.1542 vocab=0.8401 token=0.9506 | 6.12 steps/s | VRAM=26.72 Go [2026-07-18 00:05:41] step 534400/1000000 | loss=2.3038 vocab=0.8433 token=1.3368 | 6.14 steps/s | VRAM=26.24 Go [2026-07-18 00:05:49] step 534450/1000000 | loss=2.2259 vocab=0.5572 token=1.0201 | 6.18 steps/s | VRAM=24.82 Go [2026-07-18 00:05:57] step 534500/1000000 | loss=2.2827 vocab=0.6921 token=1.3416 | 6.11 steps/s | VRAM=28.55 Go [2026-07-18 00:06:06] step 534550/1000000 | loss=2.3549 vocab=0.6045 token=1.1426 | 6.14 steps/s | VRAM=29.02 Go [2026-07-18 00:06:14] step 534600/1000000 | loss=2.2507 vocab=1.3452 token=1.3897 | 6.16 steps/s | VRAM=30.70 Go [2026-07-18 00:06:22] step 534650/1000000 | loss=2.2751 vocab=0.8540 token=1.3739 | 6.14 steps/s | VRAM=29.02 Go [2026-07-18 00:06:30] step 534700/1000000 | loss=2.2743 vocab=1.5035 token=1.5424 | 6.10 steps/s | VRAM=30.09 Go [2026-07-18 00:06:38] step 534750/1000000 | loss=2.1589 vocab=0.9489 token=1.2946 | 6.15 steps/s | VRAM=29.12 Go [2026-07-18 00:06:46] step 534800/1000000 | loss=2.2395 vocab=1.0695 token=1.0330 | 6.15 steps/s | VRAM=31.90 Go [2026-07-18 00:06:54] step 534850/1000000 | loss=2.1567 vocab=0.6860 token=1.2366 | 6.13 steps/s | VRAM=28.21 Go [2026-07-18 00:07:03] step 534900/1000000 | loss=2.1714 vocab=0.7233 token=1.0059 | 6.14 steps/s | VRAM=26.47 Go [2026-07-18 00:07:11] step 534950/1000000 | loss=2.3680 vocab=0.7625 token=1.1907 | 6.11 steps/s | VRAM=29.76 Go [2026-07-18 00:07:19] step 535000/1000000 | loss=2.1652 vocab=1.1777 token=1.4585 | 6.13 steps/s | VRAM=29.48 Go [2026-07-18 00:07:27] step 535050/1000000 | loss=2.1761 vocab=0.4502 token=1.3554 | 6.13 steps/s | VRAM=25.90 Go [2026-07-18 00:07:35] step 535100/1000000 | loss=2.2394 vocab=0.5609 token=0.9734 | 6.13 steps/s | VRAM=27.00 Go [2026-07-18 00:07:43] step 535150/1000000 | loss=2.2077 vocab=1.3413 token=1.6056 | 6.12 steps/s | VRAM=25.59 Go [2026-07-18 00:07:52] step 535200/1000000 | loss=2.2326 vocab=0.6869 token=1.4124 | 6.14 steps/s | VRAM=23.43 Go [2026-07-18 00:08:00] step 535250/1000000 | loss=2.3107 vocab=0.6511 token=1.0044 | 6.13 steps/s | VRAM=31.77 Go [2026-07-18 00:08:08] step 535300/1000000 | loss=2.2318 vocab=0.8983 token=1.4296 | 6.16 steps/s | VRAM=31.98 Go [2026-07-18 00:08:16] step 535350/1000000 | loss=2.2441 vocab=0.6457 token=1.1721 | 6.13 steps/s | VRAM=28.64 Go [2026-07-18 00:08:24] step 535400/1000000 | loss=2.1428 vocab=0.9591 token=1.0566 | 6.16 steps/s | VRAM=28.58 Go [2026-07-18 00:08:32] step 535450/1000000 | loss=2.2146 vocab=0.6011 token=1.3710 | 6.14 steps/s | VRAM=28.94 Go [2026-07-18 00:08:40] step 535500/1000000 | loss=2.1695 vocab=0.9605 token=1.2264 | 6.13 steps/s | VRAM=27.84 Go [2026-07-18 00:08:49] step 535550/1000000 | loss=2.2145 vocab=0.7756 token=1.2363 | 6.17 steps/s | VRAM=29.22 Go [2026-07-18 00:08:57] step 535600/1000000 | loss=2.1939 vocab=0.6617 token=1.3832 | 6.13 steps/s | VRAM=26.01 Go [2026-07-18 00:09:05] step 535650/1000000 | loss=2.1940 vocab=0.7494 token=1.1992 | 6.16 steps/s | VRAM=32.69 Go [2026-07-18 00:09:13] step 535700/1000000 | loss=2.2215 vocab=0.9829 token=1.4116 | 6.14 steps/s | VRAM=25.52 Go [2026-07-18 00:09:21] step 535750/1000000 | loss=2.2727 vocab=0.7986 token=1.1891 | 6.17 steps/s | VRAM=28.19 Go [2026-07-18 00:09:29] step 535800/1000000 | loss=2.3353 vocab=0.9004 token=1.0515 | 6.17 steps/s | VRAM=32.11 Go [2026-07-18 00:09:37] step 535850/1000000 | loss=2.2045 vocab=1.1275 token=1.3953 | 6.15 steps/s | VRAM=29.74 Go [2026-07-18 00:09:45] step 535900/1000000 | loss=2.2368 vocab=1.2646 token=1.6978 | 6.17 steps/s | VRAM=31.38 Go [2026-07-18 00:09:53] step 535950/1000000 | loss=2.1088 vocab=0.7689 token=1.1136 | 6.15 steps/s | VRAM=32.16 Go [2026-07-18 00:10:02] step 536000/1000000 | loss=2.2813 vocab=0.7162 token=1.2518 | 6.18 steps/s | VRAM=28.97 Go [2026-07-18 00:10:10] step 536050/1000000 | loss=2.2513 vocab=1.1024 token=1.3358 | 6.18 steps/s | VRAM=27.99 Go [2026-07-18 00:10:18] step 536100/1000000 | loss=2.2174 vocab=0.7840 token=1.6504 | 6.16 steps/s | VRAM=28.58 Go [2026-07-18 00:10:26] step 536150/1000000 | loss=2.1609 vocab=0.7713 token=1.2479 | 6.13 steps/s | VRAM=25.98 Go [2026-07-18 00:10:34] step 536200/1000000 | loss=2.1139 vocab=0.7350 token=1.0642 | 6.15 steps/s | VRAM=26.10 Go [2026-07-18 00:10:42] step 536250/1000000 | loss=2.2140 vocab=0.9105 token=1.6216 | 6.17 steps/s | VRAM=32.71 Go [2026-07-18 00:10:50] step 536300/1000000 | loss=2.2294 vocab=1.6431 token=1.6911 | 6.16 steps/s | VRAM=24.87 Go [2026-07-18 00:10:58] step 536350/1000000 | loss=2.2800 vocab=0.6663 token=0.9975 | 6.16 steps/s | VRAM=32.79 Go [2026-07-18 00:11:07] step 536400/1000000 | loss=2.1596 vocab=0.8622 token=1.3987 | 6.16 steps/s | VRAM=25.55 Go [2026-07-18 00:11:15] step 536450/1000000 | loss=2.3057 vocab=1.3277 token=1.6209 | 6.16 steps/s | VRAM=31.47 Go [2026-07-18 00:11:23] step 536500/1000000 | loss=2.1441 vocab=0.7594 token=1.3987 | 6.13 steps/s | VRAM=31.72 Go [2026-07-18 00:11:31] step 536550/1000000 | loss=2.3511 vocab=0.9980 token=1.4487 | 6.14 steps/s | VRAM=28.48 Go [2026-07-18 00:11:39] step 536600/1000000 | loss=2.2289 vocab=0.6525 token=1.3733 | 6.14 steps/s | VRAM=25.57 Go [2026-07-18 00:11:47] step 536650/1000000 | loss=2.2901 vocab=0.6756 token=1.3202 | 6.15 steps/s | VRAM=32.54 Go [2026-07-18 00:11:55] step 536700/1000000 | loss=2.1848 vocab=0.8617 token=1.0422 | 6.14 steps/s | VRAM=30.13 Go [2026-07-18 00:12:03] step 536750/1000000 | loss=2.3477 vocab=1.3596 token=1.4859 | 6.17 steps/s | VRAM=28.36 Go [2026-07-18 00:12:12] step 536800/1000000 | loss=2.1606 vocab=0.7873 token=1.5836 | 6.16 steps/s | VRAM=28.71 Go [2026-07-18 00:12:20] step 536850/1000000 | loss=2.2682 vocab=0.8822 token=1.4318 | 6.16 steps/s | VRAM=28.72 Go [2026-07-18 00:12:28] step 536900/1000000 | loss=2.2204 vocab=1.5211 token=1.6175 | 6.16 steps/s | VRAM=27.39 Go [2026-07-18 00:12:36] step 536950/1000000 | loss=2.2704 vocab=2.6670 token=1.4828 | 6.16 steps/s | VRAM=27.53 Go [2026-07-18 00:12:44] step 537000/1000000 | loss=2.2710 vocab=0.9010 token=1.3919 | 6.15 steps/s | VRAM=25.52 Go [2026-07-18 00:12:52] step 537050/1000000 | loss=2.2258 vocab=0.8268 token=1.2972 | 6.15 steps/s | VRAM=29.30 Go [2026-07-18 00:13:00] step 537100/1000000 | loss=2.0591 vocab=0.5823 token=1.3085 | 6.16 steps/s | VRAM=28.65 Go [2026-07-18 00:13:08] step 537150/1000000 | loss=2.1909 vocab=0.8456 token=0.8951 | 6.12 steps/s | VRAM=29.59 Go [2026-07-18 00:13:17] step 537200/1000000 | loss=2.3099 vocab=1.1632 token=1.8207 | 6.13 steps/s | VRAM=25.97 Go [2026-07-18 00:13:25] step 537250/1000000 | loss=2.1523 vocab=0.4931 token=1.4412 | 6.14 steps/s | VRAM=25.27 Go [2026-07-18 00:13:33] step 537300/1000000 | loss=2.3430 vocab=0.6747 token=0.8956 | 6.15 steps/s | VRAM=28.96 Go [2026-07-18 00:13:41] step 537350/1000000 | loss=2.2023 vocab=0.9341 token=1.6418 | 6.13 steps/s | VRAM=30.87 Go [2026-07-18 00:13:49] step 537400/1000000 | loss=2.1663 vocab=0.7510 token=1.3798 | 6.16 steps/s | VRAM=25.27 Go [2026-07-18 00:13:57] step 537450/1000000 | loss=2.3003 vocab=0.8898 token=1.3625 | 6.15 steps/s | VRAM=31.92 Go [2026-07-18 00:14:05] step 537500/1000000 | loss=2.2105 vocab=0.7020 token=0.9150 | 6.13 steps/s | VRAM=26.59 Go [2026-07-18 00:14:14] step 537550/1000000 | loss=2.2193 vocab=0.9746 token=0.9502 | 6.16 steps/s | VRAM=31.52 Go [2026-07-18 00:14:22] step 537600/1000000 | loss=2.1960 vocab=0.6824 token=1.3770 | 6.13 steps/s | VRAM=29.45 Go [2026-07-18 00:14:30] step 537650/1000000 | loss=2.2580 vocab=1.0640 token=1.1952 | 6.15 steps/s | VRAM=28.92 Go [2026-07-18 00:14:38] step 537700/1000000 | loss=2.2827 vocab=1.0729 token=0.9648 | 6.15 steps/s | VRAM=28.87 Go [2026-07-18 00:14:46] step 537750/1000000 | loss=2.2073 vocab=0.9722 token=1.7784 | 6.13 steps/s | VRAM=29.10 Go [2026-07-18 00:14:54] step 537800/1000000 | loss=2.2231 vocab=0.7989 token=1.0613 | 6.11 steps/s | VRAM=26.24 Go [2026-07-18 00:15:02] step 537850/1000000 | loss=2.1928 vocab=0.7114 token=1.1707 | 6.16 steps/s | VRAM=28.51 Go [2026-07-18 00:15:11] step 537900/1000000 | loss=2.2732 vocab=0.5845 token=1.5540 | 6.16 steps/s | VRAM=28.76 Go [2026-07-18 00:15:19] step 537950/1000000 | loss=2.2944 vocab=1.4738 token=1.5792 | 6.12 steps/s | VRAM=30.22 Go [2026-07-18 00:15:27] step 538000/1000000 | loss=2.2413 vocab=0.7994 token=1.3430 | 6.17 steps/s | VRAM=25.00 Go [2026-07-18 00:15:35] step 538050/1000000 | loss=2.1578 vocab=0.8009 token=1.2826 | 6.17 steps/s | VRAM=28.59 Go [2026-07-18 00:15:43] step 538100/1000000 | loss=2.1483 vocab=0.9966 token=1.3818 | 6.16 steps/s | VRAM=31.10 Go [2026-07-18 00:15:51] step 538150/1000000 | loss=2.1344 vocab=1.6493 token=1.7868 | 6.14 steps/s | VRAM=26.61 Go [2026-07-18 00:15:59] step 538200/1000000 | loss=2.1173 vocab=0.7689 token=0.9152 | 6.15 steps/s | VRAM=28.59 Go [2026-07-18 00:16:07] step 538250/1000000 | loss=2.4086 vocab=0.9532 token=1.6383 | 6.14 steps/s | VRAM=27.79 Go [2026-07-18 00:16:16] step 538300/1000000 | loss=2.3225 vocab=0.6065 token=1.1325 | 6.14 steps/s | VRAM=32.20 Go [2026-07-18 00:16:24] step 538350/1000000 | loss=2.2589 vocab=0.7123 token=1.4236 | 6.11 steps/s | VRAM=29.01 Go [2026-07-18 00:16:32] step 538400/1000000 | loss=2.1560 vocab=1.0162 token=1.1018 | 6.12 steps/s | VRAM=26.90 Go [2026-07-18 00:16:40] step 538450/1000000 | loss=2.2749 vocab=0.4999 token=0.9322 | 6.19 steps/s | VRAM=31.52 Go [2026-07-18 00:16:48] step 538500/1000000 | loss=2.1972 vocab=0.9822 token=1.9314 | 6.17 steps/s | VRAM=28.53 Go [2026-07-18 00:16:56] step 538550/1000000 | loss=2.3045 vocab=0.5477 token=1.3167 | 6.17 steps/s | VRAM=27.84 Go [2026-07-18 00:17:04] step 538600/1000000 | loss=2.1770 vocab=1.0205 token=1.6860 | 6.14 steps/s | VRAM=27.62 Go [2026-07-18 00:17:13] step 538650/1000000 | loss=2.3322 vocab=0.7755 token=1.2899 | 6.17 steps/s | VRAM=32.20 Go [2026-07-18 00:17:21] step 538700/1000000 | loss=2.1056 vocab=0.4348 token=1.3005 | 6.14 steps/s | VRAM=25.81 Go [2026-07-18 00:17:29] step 538750/1000000 | loss=2.3021 vocab=1.1517 token=1.3906 | 6.18 steps/s | VRAM=31.61 Go [2026-07-18 00:17:37] step 538800/1000000 | loss=2.3083 vocab=1.2713 token=1.5009 | 6.15 steps/s | VRAM=32.95 Go [2026-07-18 00:17:45] step 538850/1000000 | loss=2.2303 vocab=1.0795 token=1.1830 | 6.13 steps/s | VRAM=32.44 Go [2026-07-18 00:17:53] step 538900/1000000 | loss=2.2338 vocab=0.5816 token=1.3691 | 6.14 steps/s | VRAM=25.87 Go [2026-07-18 00:18:01] step 538950/1000000 | loss=2.2595 vocab=1.0444 token=1.2380 | 6.15 steps/s | VRAM=31.25 Go [2026-07-18 00:18:09] step 539000/1000000 | loss=2.2142 vocab=0.4967 token=1.1378 | 6.15 steps/s | VRAM=32.61 Go [2026-07-18 00:18:18] step 539050/1000000 | loss=2.2224 vocab=0.6462 token=0.9240 | 6.14 steps/s | VRAM=32.75 Go [2026-07-18 00:18:26] step 539100/1000000 | loss=2.3010 vocab=0.9584 token=1.2353 | 6.13 steps/s | VRAM=29.37 Go [2026-07-18 00:18:34] step 539150/1000000 | loss=2.2024 vocab=0.8408 token=1.3734 | 6.15 steps/s | VRAM=28.28 Go [2026-07-18 00:18:42] step 539200/1000000 | loss=2.1107 vocab=0.7450 token=1.3979 | 6.14 steps/s | VRAM=28.90 Go [2026-07-18 00:18:50] step 539250/1000000 | loss=2.1191 vocab=0.7013 token=1.2235 | 6.13 steps/s | VRAM=29.47 Go [2026-07-18 00:18:58] step 539300/1000000 | loss=2.3144 vocab=2.0069 token=1.7296 | 6.18 steps/s | VRAM=24.37 Go [2026-07-18 00:19:06] step 539350/1000000 | loss=2.2947 vocab=0.9436 token=1.0959 | 6.15 steps/s | VRAM=25.72 Go [2026-07-18 00:19:15] step 539400/1000000 | loss=2.2772 vocab=0.5735 token=1.4506 | 6.14 steps/s | VRAM=29.10 Go [2026-07-18 00:19:23] step 539450/1000000 | loss=2.3166 vocab=1.1073 token=1.1503 | 6.16 steps/s | VRAM=32.35 Go [2026-07-18 00:19:31] step 539500/1000000 | loss=2.2513 vocab=0.9521 token=1.2896 | 6.13 steps/s | VRAM=24.96 Go [2026-07-18 00:19:39] step 539550/1000000 | loss=2.2759 vocab=0.8129 token=1.3461 | 6.13 steps/s | VRAM=28.53 Go [2026-07-18 00:19:47] step 539600/1000000 | loss=2.3135 vocab=1.1009 token=1.2657 | 6.16 steps/s | VRAM=25.27 Go [2026-07-18 00:19:55] step 539650/1000000 | loss=2.1915 vocab=0.7391 token=1.0853 | 6.16 steps/s | VRAM=31.15 Go [2026-07-18 00:20:03] step 539700/1000000 | loss=2.1990 vocab=0.9750 token=1.3765 | 6.14 steps/s | VRAM=32.38 Go [2026-07-18 00:20:12] step 539750/1000000 | loss=2.2050 vocab=0.7007 token=1.8279 | 6.11 steps/s | VRAM=26.28 Go [2026-07-18 00:20:20] step 539800/1000000 | loss=2.2759 vocab=0.8231 token=0.9796 | 6.15 steps/s | VRAM=31.78 Go [2026-07-18 00:20:28] step 539850/1000000 | loss=2.3488 vocab=0.9339 token=1.5450 | 6.12 steps/s | VRAM=32.12 Go [2026-07-18 00:20:36] step 539900/1000000 | loss=2.2049 vocab=0.7846 token=1.3064 | 6.11 steps/s | VRAM=29.13 Go [2026-07-18 00:20:44] step 539950/1000000 | loss=2.3005 vocab=1.3079 token=1.5231 | 6.16 steps/s | VRAM=32.36 Go [2026-07-18 00:20:52] step 540000/1000000 | loss=2.2251 vocab=0.6520 token=1.4440 | 6.11 steps/s | VRAM=32.28 Go [2026-07-18 00:21:00] step 540050/1000000 | loss=2.1981 vocab=1.0934 token=1.6370 | 6.14 steps/s | VRAM=31.86 Go [2026-07-18 00:21:09] step 540100/1000000 | loss=2.2021 vocab=0.9608 token=1.5309 | 6.11 steps/s | VRAM=25.89 Go [2026-07-18 00:21:17] step 540150/1000000 | loss=2.2540 vocab=1.1753 token=1.2484 | 6.14 steps/s | VRAM=28.82 Go [2026-07-18 00:21:25] step 540200/1000000 | loss=2.1464 vocab=0.9999 token=1.3399 | 6.12 steps/s | VRAM=32.87 Go [2026-07-18 00:21:33] step 540250/1000000 | loss=2.3855 vocab=1.1492 token=1.7747 | 6.12 steps/s | VRAM=32.20 Go [2026-07-18 00:21:41] step 540300/1000000 | loss=2.2786 vocab=0.8356 token=1.4465 | 6.13 steps/s | VRAM=25.70 Go [2026-07-18 00:21:49] step 540350/1000000 | loss=2.2898 vocab=1.0027 token=1.2474 | 6.15 steps/s | VRAM=29.10 Go [2026-07-18 00:21:58] step 540400/1000000 | loss=2.1979 vocab=0.9780 token=1.5068 | 6.13 steps/s | VRAM=32.24 Go [2026-07-18 00:22:06] step 540450/1000000 | loss=2.1271 vocab=0.9174 token=1.7407 | 6.13 steps/s | VRAM=26.20 Go [2026-07-18 00:22:14] step 540500/1000000 | loss=2.3571 vocab=1.1496 token=1.3060 | 6.08 steps/s | VRAM=28.31 Go [2026-07-18 00:22:22] step 540550/1000000 | loss=2.3653 vocab=0.7654 token=1.4659 | 6.14 steps/s | VRAM=28.24 Go [2026-07-18 00:22:30] step 540600/1000000 | loss=2.1740 vocab=0.7187 token=1.2868 | 6.14 steps/s | VRAM=28.86 Go [2026-07-18 00:22:38] step 540650/1000000 | loss=2.2823 vocab=0.8020 token=1.4721 | 6.14 steps/s | VRAM=28.36 Go [2026-07-18 00:22:47] step 540700/1000000 | loss=2.2702 vocab=0.5933 token=1.3280 | 6.14 steps/s | VRAM=27.18 Go [2026-07-18 00:22:55] step 540750/1000000 | loss=2.2536 vocab=0.5742 token=1.2059 | 6.12 steps/s | VRAM=29.18 Go [2026-07-18 00:23:03] step 540800/1000000 | loss=2.1223 vocab=0.8118 token=1.2614 | 6.16 steps/s | VRAM=28.28 Go [2026-07-18 00:23:11] step 540850/1000000 | loss=2.2460 vocab=1.1934 token=1.6523 | 6.11 steps/s | VRAM=28.28 Go [2026-07-18 00:23:19] step 540900/1000000 | loss=2.3197 vocab=1.4626 token=1.4726 | 6.14 steps/s | VRAM=32.13 Go [2026-07-18 00:23:27] step 540950/1000000 | loss=2.3897 vocab=1.2552 token=1.4870 | 6.11 steps/s | VRAM=26.13 Go [2026-07-18 00:23:36] step 541000/1000000 | loss=2.2709 vocab=1.1022 token=1.2286 | 6.12 steps/s | VRAM=26.55 Go [2026-07-18 00:23:44] step 541050/1000000 | loss=2.2985 vocab=1.1528 token=1.2556 | 6.12 steps/s | VRAM=29.14 Go [2026-07-18 00:23:52] step 541100/1000000 | loss=2.1559 vocab=1.0260 token=1.5680 | 6.11 steps/s | VRAM=29.58 Go [2026-07-18 00:24:00] step 541150/1000000 | loss=2.2954 vocab=0.9092 token=1.3636 | 6.13 steps/s | VRAM=28.82 Go [2026-07-18 00:24:08] step 541200/1000000 | loss=2.2269 vocab=0.5351 token=1.1418 | 6.12 steps/s | VRAM=32.84 Go [2026-07-18 00:24:16] step 541250/1000000 | loss=2.3223 vocab=0.9589 token=1.3595 | 6.14 steps/s | VRAM=26.19 Go [2026-07-18 00:24:25] step 541300/1000000 | loss=2.2109 vocab=0.8573 token=1.2346 | 6.16 steps/s | VRAM=25.96 Go [2026-07-18 00:24:33] step 541350/1000000 | loss=2.1808 vocab=0.6012 token=0.9434 | 6.17 steps/s | VRAM=28.49 Go [2026-07-18 00:24:41] step 541400/1000000 | loss=2.1816 vocab=1.0266 token=1.2967 | 6.14 steps/s | VRAM=28.65 Go [2026-07-18 00:24:49] step 541450/1000000 | loss=2.2428 vocab=0.9623 token=1.3704 | 6.13 steps/s | VRAM=26.82 Go [2026-07-18 00:24:57] step 541500/1000000 | loss=2.2195 vocab=0.9198 token=1.5468 | 6.11 steps/s | VRAM=28.97 Go [2026-07-18 00:25:05] step 541550/1000000 | loss=2.2539 vocab=1.0007 token=1.2799 | 6.11 steps/s | VRAM=28.16 Go [2026-07-18 00:25:13] step 541600/1000000 | loss=2.3135 vocab=1.0289 token=1.8149 | 6.15 steps/s | VRAM=33.02 Go [2026-07-18 00:25:22] step 541650/1000000 | loss=2.3058 vocab=0.9629 token=1.2148 | 6.16 steps/s | VRAM=28.82 Go [2026-07-18 00:25:30] step 541700/1000000 | loss=2.1124 vocab=1.3823 token=1.4000 | 6.13 steps/s | VRAM=26.28 Go [2026-07-18 00:25:38] step 541750/1000000 | loss=2.2940 vocab=0.7525 token=1.4671 | 6.11 steps/s | VRAM=29.12 Go [2026-07-18 00:25:46] step 541800/1000000 | loss=2.2488 vocab=1.0224 token=1.2342 | 6.13 steps/s | VRAM=28.50 Go [2026-07-18 00:25:54] step 541850/1000000 | loss=2.2303 vocab=0.7620 token=1.4588 | 6.12 steps/s | VRAM=31.97 Go [2026-07-18 00:26:02] step 541900/1000000 | loss=2.2263 vocab=0.8941 token=1.1991 | 6.14 steps/s | VRAM=24.46 Go [2026-07-18 00:26:10] step 541950/1000000 | loss=2.2763 vocab=1.4492 token=1.5260 | 6.15 steps/s | VRAM=26.06 Go [2026-07-18 00:26:19] step 542000/1000000 | loss=2.2039 vocab=0.9486 token=1.5989 | 6.14 steps/s | VRAM=30.63 Go [2026-07-18 00:26:27] step 542050/1000000 | loss=2.2005 vocab=1.3887 token=2.1181 | 6.13 steps/s | VRAM=32.38 Go [2026-07-18 00:26:35] step 542100/1000000 | loss=2.2713 vocab=0.9430 token=1.3328 | 6.15 steps/s | VRAM=32.77 Go [2026-07-18 00:26:43] step 542150/1000000 | loss=2.2469 vocab=1.5262 token=1.3841 | 6.13 steps/s | VRAM=25.64 Go [2026-07-18 00:26:51] step 542200/1000000 | loss=2.1979 vocab=0.9193 token=1.3626 | 6.15 steps/s | VRAM=28.40 Go [2026-07-18 00:26:59] step 542250/1000000 | loss=2.1453 vocab=0.9488 token=1.1918 | 6.13 steps/s | VRAM=32.12 Go [2026-07-18 00:27:08] step 542300/1000000 | loss=2.2797 vocab=0.8688 token=1.1585 | 6.12 steps/s | VRAM=23.45 Go [2026-07-18 00:27:16] step 542350/1000000 | loss=2.2651 vocab=0.6228 token=1.0887 | 6.15 steps/s | VRAM=28.13 Go [2026-07-18 00:27:24] step 542400/1000000 | loss=2.2202 vocab=0.8161 token=1.3740 | 6.12 steps/s | VRAM=29.66 Go [2026-07-18 00:27:32] step 542450/1000000 | loss=2.1165 vocab=1.2843 token=1.2203 | 6.13 steps/s | VRAM=25.12 Go [2026-07-18 00:27:40] step 542500/1000000 | loss=2.2594 vocab=1.1834 token=1.4484 | 6.15 steps/s | VRAM=25.47 Go [2026-07-18 00:27:48] step 542550/1000000 | loss=2.3545 vocab=1.0911 token=1.7181 | 6.12 steps/s | VRAM=28.66 Go [2026-07-18 00:27:56] step 542600/1000000 | loss=2.2717 vocab=1.0598 token=0.9809 | 6.11 steps/s | VRAM=29.36 Go [2026-07-18 00:28:05] step 542650/1000000 | loss=2.2604 vocab=0.7496 token=0.9584 | 6.15 steps/s | VRAM=32.49 Go [2026-07-18 00:28:13] step 542700/1000000 | loss=2.3162 vocab=1.1565 token=1.1469 | 6.15 steps/s | VRAM=32.54 Go [2026-07-18 00:28:21] step 542750/1000000 | loss=2.2753 vocab=0.6534 token=1.3224 | 6.12 steps/s | VRAM=30.02 Go [2026-07-18 00:28:29] step 542800/1000000 | loss=2.3638 vocab=1.0757 token=1.2140 | 6.15 steps/s | VRAM=29.05 Go [2026-07-18 00:28:37] step 542850/1000000 | loss=2.1826 vocab=0.7853 token=1.1307 | 6.13 steps/s | VRAM=28.75 Go [2026-07-18 00:28:45] step 542900/1000000 | loss=2.2601 vocab=0.4904 token=1.0814 | 6.17 steps/s | VRAM=31.15 Go [2026-07-18 00:28:53] step 542950/1000000 | loss=2.1989 vocab=0.5845 token=1.0279 | 6.16 steps/s | VRAM=26.39 Go [2026-07-18 00:29:02] step 543000/1000000 | loss=2.2349 vocab=0.6479 token=1.1304 | 6.13 steps/s | VRAM=25.77 Go [2026-07-18 00:29:10] step 543050/1000000 | loss=2.3162 vocab=0.9502 token=1.4759 | 6.15 steps/s | VRAM=29.18 Go [2026-07-18 00:29:18] step 543100/1000000 | loss=2.2242 vocab=0.7655 token=1.0536 | 6.17 steps/s | VRAM=31.15 Go [2026-07-18 00:29:26] step 543150/1000000 | loss=2.2107 vocab=0.5205 token=1.1458 | 6.13 steps/s | VRAM=29.09 Go [2026-07-18 00:29:34] step 543200/1000000 | loss=2.2317 vocab=1.3859 token=1.6348 | 6.16 steps/s | VRAM=31.77 Go [2026-07-18 00:29:42] step 543250/1000000 | loss=2.2448 vocab=0.8870 token=1.4399 | 6.12 steps/s | VRAM=28.78 Go [2026-07-18 00:29:50] step 543300/1000000 | loss=2.2728 vocab=0.6484 token=1.0555 | 6.14 steps/s | VRAM=28.44 Go [2026-07-18 00:29:59] step 543350/1000000 | loss=2.2572 vocab=0.5673 token=0.8433 | 6.12 steps/s | VRAM=29.44 Go [2026-07-18 00:30:07] step 543400/1000000 | loss=2.2455 vocab=0.4887 token=1.2174 | 6.15 steps/s | VRAM=29.91 Go [2026-07-18 00:30:15] step 543450/1000000 | loss=2.2119 vocab=0.6432 token=1.2215 | 6.14 steps/s | VRAM=32.69 Go [2026-07-18 00:30:23] step 543500/1000000 | loss=2.1903 vocab=0.9120 token=1.1353 | 6.15 steps/s | VRAM=28.23 Go [2026-07-18 00:30:31] step 543550/1000000 | loss=2.1627 vocab=0.8443 token=1.2543 | 6.14 steps/s | VRAM=32.36 Go [2026-07-18 00:30:39] step 543600/1000000 | loss=2.2086 vocab=1.6457 token=1.3983 | 6.18 steps/s | VRAM=30.74 Go [2026-07-18 00:30:47] step 543650/1000000 | loss=2.1220 vocab=0.6288 token=1.3196 | 6.15 steps/s | VRAM=28.78 Go [2026-07-18 00:30:55] step 543700/1000000 | loss=2.1737 vocab=0.6763 token=1.3162 | 6.14 steps/s | VRAM=28.49 Go [2026-07-18 00:31:04] step 543750/1000000 | loss=2.2634 vocab=0.7536 token=1.1152 | 6.11 steps/s | VRAM=26.71 Go [2026-07-18 00:31:12] step 543800/1000000 | loss=2.3328 vocab=0.8455 token=1.2813 | 6.13 steps/s | VRAM=28.61 Go [2026-07-18 00:31:20] step 543850/1000000 | loss=2.1827 vocab=0.8193 token=1.3743 | 6.11 steps/s | VRAM=25.99 Go [2026-07-18 00:31:28] step 543900/1000000 | loss=2.2073 vocab=1.0491 token=1.4613 | 6.15 steps/s | VRAM=31.56 Go [2026-07-18 00:31:36] step 543950/1000000 | loss=2.2118 vocab=0.5530 token=1.1986 | 6.15 steps/s | VRAM=28.15 Go [2026-07-18 00:31:44] step 544000/1000000 | loss=2.2643 vocab=0.9519 token=1.6872 | 6.13 steps/s | VRAM=28.99 Go [2026-07-18 00:31:53] step 544050/1000000 | loss=2.1780 vocab=0.7888 token=1.3135 | 6.15 steps/s | VRAM=28.83 Go [2026-07-18 00:32:01] step 544100/1000000 | loss=2.2335 vocab=1.1133 token=1.7435 | 6.13 steps/s | VRAM=31.59 Go [2026-07-18 00:32:09] step 544150/1000000 | loss=2.3388 vocab=0.6983 token=1.3779 | 6.14 steps/s | VRAM=33.02 Go [2026-07-18 00:32:17] step 544200/1000000 | loss=2.1505 vocab=1.0708 token=1.2966 | 6.13 steps/s | VRAM=27.76 Go [2026-07-18 00:32:25] step 544250/1000000 | loss=2.1916 vocab=0.9159 token=1.3562 | 6.14 steps/s | VRAM=32.61 Go [2026-07-18 00:32:33] step 544300/1000000 | loss=2.2134 vocab=0.7028 token=1.2542 | 6.13 steps/s | VRAM=29.06 Go [2026-07-18 00:32:41] step 544350/1000000 | loss=2.3473 vocab=0.9430 token=1.2215 | 6.15 steps/s | VRAM=29.04 Go [2026-07-18 00:32:49] step 544400/1000000 | loss=2.2859 vocab=1.1022 token=1.6211 | 6.17 steps/s | VRAM=28.08 Go [2026-07-18 00:32:58] step 544450/1000000 | loss=2.2878 vocab=0.8700 token=1.7471 | 6.14 steps/s | VRAM=32.69 Go [2026-07-18 00:33:06] step 544500/1000000 | loss=2.2448 vocab=0.8463 token=1.2150 | 6.15 steps/s | VRAM=28.77 Go [2026-07-18 00:33:14] step 544550/1000000 | loss=2.2501 vocab=1.0520 token=1.7620 | 6.13 steps/s | VRAM=27.94 Go [2026-07-18 00:33:22] step 544600/1000000 | loss=2.2705 vocab=0.9148 token=1.4364 | 6.15 steps/s | VRAM=25.56 Go [2026-07-18 00:33:30] step 544650/1000000 | loss=2.2950 vocab=0.6922 token=1.3551 | 6.14 steps/s | VRAM=29.41 Go [2026-07-18 00:33:38] step 544700/1000000 | loss=2.3045 vocab=1.0085 token=1.3975 | 6.13 steps/s | VRAM=28.32 Go [2026-07-18 00:33:47] step 544750/1000000 | loss=2.2536 vocab=0.9803 token=1.5366 | 6.14 steps/s | VRAM=30.88 Go [2026-07-18 00:33:55] step 544800/1000000 | loss=2.1460 vocab=0.6078 token=1.3877 | 6.12 steps/s | VRAM=28.47 Go [2026-07-18 00:34:03] step 544850/1000000 | loss=2.2529 vocab=0.5821 token=0.9011 | 6.12 steps/s | VRAM=30.61 Go [2026-07-18 00:34:11] step 544900/1000000 | loss=2.2090 vocab=1.1007 token=1.3883 | 6.10 steps/s | VRAM=26.82 Go [2026-07-18 00:34:19] step 544950/1000000 | loss=2.1456 vocab=1.0436 token=1.2788 | 6.16 steps/s | VRAM=32.08 Go [2026-07-18 00:34:27] step 545000/1000000 | loss=2.2960 vocab=1.0336 token=1.4672 | 6.15 steps/s | VRAM=28.41 Go [2026-07-18 00:34:35] step 545050/1000000 | loss=2.2568 vocab=0.7994 token=1.0522 | 6.12 steps/s | VRAM=26.80 Go [2026-07-18 00:34:44] step 545100/1000000 | loss=2.2027 vocab=1.3363 token=1.3820 | 6.14 steps/s | VRAM=29.47 Go [2026-07-18 00:34:52] step 545150/1000000 | loss=2.2590 vocab=0.8705 token=1.3409 | 6.15 steps/s | VRAM=31.52 Go [2026-07-18 00:35:00] step 545200/1000000 | loss=2.2776 vocab=0.6427 token=1.4112 | 6.16 steps/s | VRAM=31.63 Go [2026-07-18 00:35:08] step 545250/1000000 | loss=2.2470 vocab=0.6899 token=1.3903 | 6.13 steps/s | VRAM=29.19 Go [2026-07-18 00:35:16] step 545300/1000000 | loss=2.3164 vocab=0.9947 token=1.3088 | 6.14 steps/s | VRAM=32.10 Go [2026-07-18 00:35:24] step 545350/1000000 | loss=2.3793 vocab=0.8765 token=1.6590 | 6.14 steps/s | VRAM=26.20 Go [2026-07-18 00:35:32] step 545400/1000000 | loss=2.2459 vocab=0.5476 token=0.9799 | 6.14 steps/s | VRAM=28.51 Go [2026-07-18 00:35:41] step 545450/1000000 | loss=2.2727 vocab=0.7624 token=1.3411 | 6.11 steps/s | VRAM=28.79 Go [2026-07-18 00:35:49] step 545500/1000000 | loss=2.3123 vocab=0.9924 token=1.5191 | 6.12 steps/s | VRAM=26.41 Go [2026-07-18 00:35:57] step 545550/1000000 | loss=2.2254 vocab=2.1257 token=1.6417 | 6.11 steps/s | VRAM=26.16 Go [2026-07-18 00:36:05] step 545600/1000000 | loss=2.2569 vocab=1.1060 token=1.2189 | 6.17 steps/s | VRAM=28.60 Go [2026-07-18 00:36:13] step 545650/1000000 | loss=2.2135 vocab=0.8503 token=1.3510 | 6.14 steps/s | VRAM=28.75 Go [2026-07-18 00:36:21] step 545700/1000000 | loss=2.2277 vocab=0.9826 token=1.2863 | 6.13 steps/s | VRAM=28.62 Go [2026-07-18 00:36:30] step 545750/1000000 | loss=2.2188 vocab=0.8887 token=1.1262 | 6.12 steps/s | VRAM=26.46 Go [2026-07-18 00:36:38] step 545800/1000000 | loss=2.2672 vocab=0.9374 token=1.1631 | 6.15 steps/s | VRAM=27.62 Go [2026-07-18 00:36:46] step 545850/1000000 | loss=2.1364 vocab=0.9230 token=1.6546 | 6.13 steps/s | VRAM=29.56 Go [2026-07-18 00:36:54] step 545900/1000000 | loss=2.2899 vocab=1.6188 token=1.2057 | 6.14 steps/s | VRAM=26.72 Go [2026-07-18 00:37:02] step 545950/1000000 | loss=2.1764 vocab=0.3752 token=0.9173 | 6.15 steps/s | VRAM=27.27 Go [2026-07-18 00:37:10] step 546000/1000000 | loss=2.2677 vocab=0.8377 token=1.3330 | 6.15 steps/s | VRAM=29.17 Go [2026-07-18 00:37:18] step 546050/1000000 | loss=2.2898 vocab=0.5944 token=1.6021 | 6.17 steps/s | VRAM=25.34 Go [2026-07-18 00:37:26] step 546100/1000000 | loss=2.2602 vocab=0.8297 token=1.4378 | 6.14 steps/s | VRAM=28.89 Go [2026-07-18 00:37:35] step 546150/1000000 | loss=2.1499 vocab=0.8014 token=1.2211 | 6.17 steps/s | VRAM=31.98 Go [2026-07-18 00:37:43] step 546200/1000000 | loss=2.1960 vocab=0.9723 token=1.0307 | 6.14 steps/s | VRAM=29.57 Go [2026-07-18 00:37:51] step 546250/1000000 | loss=2.1940 vocab=0.7068 token=1.1894 | 6.12 steps/s | VRAM=25.06 Go [2026-07-18 00:37:59] step 546300/1000000 | loss=2.2746 vocab=0.8410 token=1.4899 | 6.16 steps/s | VRAM=28.09 Go [2026-07-18 00:38:07] step 546350/1000000 | loss=2.2856 vocab=0.8816 token=1.1019 | 6.14 steps/s | VRAM=25.16 Go [2026-07-18 00:38:15] step 546400/1000000 | loss=2.2684 vocab=1.3510 token=1.3106 | 6.12 steps/s | VRAM=28.82 Go [2026-07-18 00:38:23] step 546450/1000000 | loss=2.1636 vocab=0.7026 token=1.3593 | 6.15 steps/s | VRAM=31.00 Go [2026-07-18 00:38:32] step 546500/1000000 | loss=2.3324 vocab=1.3580 token=1.8971 | 6.14 steps/s | VRAM=32.34 Go [2026-07-18 00:38:40] step 546550/1000000 | loss=2.1488 vocab=1.1577 token=1.2837 | 6.12 steps/s | VRAM=25.93 Go [2026-07-18 00:38:48] step 546600/1000000 | loss=2.2530 vocab=0.4952 token=1.2610 | 6.16 steps/s | VRAM=32.21 Go [2026-07-18 00:38:56] step 546650/1000000 | loss=2.2637 vocab=0.6145 token=1.2422 | 6.14 steps/s | VRAM=29.30 Go [2026-07-18 00:39:04] step 546700/1000000 | loss=2.2317 vocab=0.8739 token=1.3733 | 6.12 steps/s | VRAM=28.88 Go [2026-07-18 00:39:12] step 546750/1000000 | loss=2.3576 vocab=1.4464 token=1.1052 | 6.12 steps/s | VRAM=31.51 Go [2026-07-18 00:39:21] step 546800/1000000 | loss=2.3344 vocab=1.1602 token=1.6961 | 6.13 steps/s | VRAM=28.44 Go [2026-07-18 00:39:29] step 546850/1000000 | loss=2.1806 vocab=1.2249 token=1.5412 | 6.16 steps/s | VRAM=25.18 Go [2026-07-18 00:39:37] step 546900/1000000 | loss=2.2752 vocab=0.5031 token=1.3299 | 6.14 steps/s | VRAM=29.12 Go [2026-07-18 00:39:45] step 546950/1000000 | loss=2.3386 vocab=1.2675 token=1.2636 | 6.11 steps/s | VRAM=30.26 Go [2026-07-18 00:39:53] step 547000/1000000 | loss=2.2853 vocab=0.8582 token=1.5792 | 6.15 steps/s | VRAM=28.14 Go [2026-07-18 00:40:01] step 547050/1000000 | loss=2.3252 vocab=1.4815 token=1.4440 | 6.14 steps/s | VRAM=29.65 Go [2026-07-18 00:40:09] step 547100/1000000 | loss=2.2760 vocab=0.3485 token=1.0842 | 6.14 steps/s | VRAM=32.19 Go [2026-07-18 00:40:18] step 547150/1000000 | loss=2.1253 vocab=1.2591 token=1.4902 | 6.13 steps/s | VRAM=32.71 Go [2026-07-18 00:40:26] step 547200/1000000 | loss=2.1842 vocab=0.6018 token=1.4064 | 6.14 steps/s | VRAM=30.85 Go [2026-07-18 00:40:34] step 547250/1000000 | loss=2.2276 vocab=0.9973 token=1.4345 | 6.13 steps/s | VRAM=29.60 Go [2026-07-18 00:40:42] step 547300/1000000 | loss=2.3561 vocab=1.1313 token=1.6748 | 6.14 steps/s | VRAM=25.64 Go [2026-07-18 00:40:50] step 547350/1000000 | loss=2.2572 vocab=1.1776 token=1.2755 | 6.16 steps/s | VRAM=29.38 Go [2026-07-18 00:40:58] step 547400/1000000 | loss=2.2115 vocab=0.6337 token=1.1425 | 6.13 steps/s | VRAM=29.89 Go [2026-07-18 00:41:06] step 547450/1000000 | loss=2.1721 vocab=0.7739 token=1.2966 | 6.16 steps/s | VRAM=31.89 Go [2026-07-18 00:41:15] step 547500/1000000 | loss=2.1026 vocab=0.9383 token=1.3773 | 6.13 steps/s | VRAM=28.21 Go [2026-07-18 00:41:23] step 547550/1000000 | loss=2.2663 vocab=1.3635 token=1.6591 | 6.15 steps/s | VRAM=25.56 Go [2026-07-18 00:41:31] step 547600/1000000 | loss=2.2005 vocab=1.2022 token=1.5123 | 6.15 steps/s | VRAM=31.56 Go [2026-07-18 00:41:39] step 547650/1000000 | loss=2.2318 vocab=1.0224 token=1.1337 | 6.13 steps/s | VRAM=32.57 Go [2026-07-18 00:41:47] step 547700/1000000 | loss=2.3138 vocab=0.8545 token=1.5220 | 6.13 steps/s | VRAM=25.71 Go [2026-07-18 00:41:55] step 547750/1000000 | loss=2.2812 vocab=0.6540 token=1.3184 | 6.13 steps/s | VRAM=28.67 Go [2026-07-18 00:42:03] step 547800/1000000 | loss=2.1042 vocab=0.7727 token=1.3458 | 6.15 steps/s | VRAM=31.29 Go [2026-07-18 00:42:11] step 547850/1000000 | loss=2.1885 vocab=0.6683 token=0.9811 | 6.16 steps/s | VRAM=31.14 Go [2026-07-18 00:42:20] step 547900/1000000 | loss=2.1519 vocab=1.0607 token=1.6878 | 6.13 steps/s | VRAM=31.73 Go [2026-07-18 00:42:28] step 547950/1000000 | loss=2.2625 vocab=1.0307 token=0.9420 | 6.13 steps/s | VRAM=29.74 Go [2026-07-18 00:42:36] step 548000/1000000 | loss=2.2651 vocab=0.7995 token=1.4753 | 6.10 steps/s | VRAM=28.96 Go [2026-07-18 00:42:44] step 548050/1000000 | loss=2.2262 vocab=0.6587 token=1.2450 | 6.10 steps/s | VRAM=28.50 Go [2026-07-18 00:42:52] step 548100/1000000 | loss=2.3155 vocab=0.8590 token=1.2805 | 6.13 steps/s | VRAM=28.36 Go [2026-07-18 00:43:01] step 548150/1000000 | loss=2.2362 vocab=0.8149 token=1.2863 | 6.14 steps/s | VRAM=32.63 Go [2026-07-18 00:43:09] step 548200/1000000 | loss=2.2794 vocab=0.5125 token=0.8790 | 6.09 steps/s | VRAM=29.48 Go [2026-07-18 00:43:17] step 548250/1000000 | loss=2.2343 vocab=0.5297 token=1.3054 | 6.13 steps/s | VRAM=29.95 Go [2026-07-18 00:43:25] step 548300/1000000 | loss=2.1752 vocab=1.1491 token=1.3959 | 6.15 steps/s | VRAM=32.61 Go [2026-07-18 00:43:33] step 548350/1000000 | loss=2.2291 vocab=0.6217 token=1.3688 | 6.10 steps/s | VRAM=30.02 Go [2026-07-18 00:43:41] step 548400/1000000 | loss=2.2875 vocab=0.7131 token=0.9910 | 6.15 steps/s | VRAM=28.49 Go [2026-07-18 00:43:49] step 548450/1000000 | loss=2.1392 vocab=0.3947 token=1.2049 | 6.15 steps/s | VRAM=30.62 Go [2026-07-18 00:43:58] step 548500/1000000 | loss=2.2527 vocab=0.6411 token=1.5004 | 6.15 steps/s | VRAM=31.47 Go [2026-07-18 00:44:06] step 548550/1000000 | loss=2.1558 vocab=0.8780 token=1.1762 | 6.16 steps/s | VRAM=31.54 Go [2026-07-18 00:44:14] step 548600/1000000 | loss=2.2849 vocab=0.7048 token=1.1836 | 6.12 steps/s | VRAM=29.20 Go [2026-07-18 00:44:22] step 548650/1000000 | loss=2.2426 vocab=0.6824 token=1.3636 | 6.14 steps/s | VRAM=29.41 Go [2026-07-18 00:44:30] step 548700/1000000 | loss=2.2927 vocab=1.4808 token=1.4585 | 6.15 steps/s | VRAM=28.95 Go [2026-07-18 00:44:38] step 548750/1000000 | loss=2.1881 vocab=0.3989 token=1.1803 | 6.10 steps/s | VRAM=24.86 Go [2026-07-18 00:44:46] step 548800/1000000 | loss=2.2098 vocab=0.8100 token=1.3069 | 6.16 steps/s | VRAM=27.55 Go [2026-07-18 00:44:55] step 548850/1000000 | loss=2.2261 vocab=1.0598 token=1.5451 | 6.12 steps/s | VRAM=25.75 Go [2026-07-18 00:45:03] step 548900/1000000 | loss=2.1655 vocab=0.9041 token=1.0118 | 6.13 steps/s | VRAM=24.95 Go [2026-07-18 00:45:11] step 548950/1000000 | loss=2.2616 vocab=1.1459 token=1.3060 | 6.15 steps/s | VRAM=31.03 Go [2026-07-18 00:45:19] step 549000/1000000 | loss=2.2916 vocab=1.3037 token=1.2332 | 6.13 steps/s | VRAM=29.70 Go [2026-07-18 00:45:27] step 549050/1000000 | loss=2.2913 vocab=0.5857 token=0.9278 | 6.13 steps/s | VRAM=25.94 Go [2026-07-18 00:45:35] step 549100/1000000 | loss=2.2016 vocab=0.6379 token=1.1933 | 6.16 steps/s | VRAM=31.49 Go [2026-07-18 00:45:43] step 549150/1000000 | loss=2.2175 vocab=0.4877 token=1.4230 | 6.15 steps/s | VRAM=31.99 Go [2026-07-18 00:45:52] step 549200/1000000 | loss=2.2804 vocab=0.6445 token=0.9766 | 6.12 steps/s | VRAM=28.02 Go [2026-07-18 00:46:00] step 549250/1000000 | loss=2.2030 vocab=0.6772 token=1.1228 | 6.17 steps/s | VRAM=29.07 Go [2026-07-18 00:46:08] step 549300/1000000 | loss=2.2491 vocab=1.0988 token=1.2215 | 6.12 steps/s | VRAM=25.86 Go [2026-07-18 00:46:16] step 549350/1000000 | loss=2.2850 vocab=0.9418 token=1.2149 | 6.13 steps/s | VRAM=24.32 Go [2026-07-18 00:46:24] step 549400/1000000 | loss=2.1838 vocab=0.7729 token=1.4206 | 6.12 steps/s | VRAM=27.16 Go [2026-07-18 00:46:32] step 549450/1000000 | loss=2.2050 vocab=1.0713 token=1.2040 | 6.17 steps/s | VRAM=28.93 Go [2026-07-18 00:46:40] step 549500/1000000 | loss=2.1400 vocab=1.4274 token=1.5247 | 6.15 steps/s | VRAM=29.46 Go [2026-07-18 00:46:49] step 549550/1000000 | loss=2.1424 vocab=0.7479 token=1.3528 | 6.13 steps/s | VRAM=25.42 Go [2026-07-18 00:46:57] step 549600/1000000 | loss=2.2875 vocab=0.9308 token=1.2807 | 6.16 steps/s | VRAM=24.73 Go [2026-07-18 00:47:05] step 549650/1000000 | loss=2.2424 vocab=0.7690 token=1.1041 | 6.13 steps/s | VRAM=25.23 Go [2026-07-18 00:47:13] step 549700/1000000 | loss=2.1378 vocab=0.6996 token=1.2267 | 6.13 steps/s | VRAM=31.95 Go [2026-07-18 00:47:21] step 549750/1000000 | loss=2.1692 vocab=1.0678 token=1.1541 | 6.17 steps/s | VRAM=24.97 Go [2026-07-18 00:47:29] step 549800/1000000 | loss=2.2216 vocab=0.7338 token=1.3999 | 6.15 steps/s | VRAM=25.00 Go [2026-07-18 00:47:37] step 549850/1000000 | loss=2.2955 vocab=1.2475 token=1.5464 | 6.15 steps/s | VRAM=31.88 Go [2026-07-18 00:47:46] step 549900/1000000 | loss=2.2444 vocab=0.6556 token=1.2384 | 6.11 steps/s | VRAM=28.07 Go [2026-07-18 00:47:54] step 549950/1000000 | loss=2.1825 vocab=1.2648 token=1.2453 | 6.10 steps/s | VRAM=29.25 Go [2026-07-18 00:48:02] step 550000/1000000 | loss=2.2078 vocab=0.6519 token=1.0052 | 6.10 steps/s | VRAM=26.80 Go [push] step 550000 OK (best-of-2, loss_token=1.0052) [2026-07-18 00:49:27] step 550050/1000000 | loss=2.1886 vocab=0.9533 token=0.8343 | 0.59 steps/s | VRAM=32.30 Go [2026-07-18 00:49:35] step 550100/1000000 | loss=2.2056 vocab=1.1939 token=1.6891 | 6.19 steps/s | VRAM=28.92 Go [2026-07-18 00:49:43] step 550150/1000000 | loss=2.2960 vocab=0.8956 token=1.5149 | 6.18 steps/s | VRAM=26.36 Go [2026-07-18 00:49:51] step 550200/1000000 | loss=2.3495 vocab=1.2062 token=1.1027 | 6.19 steps/s | VRAM=32.76 Go [2026-07-18 00:50:00] step 550250/1000000 | loss=2.4454 vocab=1.5330 token=1.3935 | 6.15 steps/s | VRAM=24.25 Go [2026-07-18 00:50:08] step 550300/1000000 | loss=2.2741 vocab=0.5897 token=1.0675 | 6.17 steps/s | VRAM=28.43 Go [2026-07-18 00:50:16] step 550350/1000000 | loss=2.2646 vocab=1.0057 token=1.7140 | 6.16 steps/s | VRAM=28.07 Go [2026-07-18 00:50:24] step 550400/1000000 | loss=2.2930 vocab=0.6770 token=1.0771 | 6.14 steps/s | VRAM=25.06 Go [2026-07-18 00:50:32] step 550450/1000000 | loss=2.2209 vocab=0.9223 token=1.3501 | 6.11 steps/s | VRAM=28.83 Go [2026-07-18 00:50:40] step 550500/1000000 | loss=2.2034 vocab=0.9052 token=1.4044 | 6.13 steps/s | VRAM=28.18 Go [2026-07-18 00:50:49] step 550550/1000000 | loss=2.1444 vocab=1.0728 token=1.3899 | 6.11 steps/s | VRAM=28.77 Go [2026-07-18 00:50:57] step 550600/1000000 | loss=2.1273 vocab=0.8420 token=1.6309 | 6.12 steps/s | VRAM=29.01 Go [2026-07-18 00:51:05] step 550650/1000000 | loss=2.1844 vocab=1.2972 token=1.1486 | 6.09 steps/s | VRAM=26.52 Go [2026-07-18 00:51:13] step 550700/1000000 | loss=2.1500 vocab=0.8142 token=1.3455 | 6.13 steps/s | VRAM=28.84 Go [2026-07-18 00:51:21] step 550750/1000000 | loss=2.3023 vocab=1.0541 token=1.3337 | 6.10 steps/s | VRAM=27.82 Go [2026-07-18 00:51:29] step 550800/1000000 | loss=2.2541 vocab=0.6664 token=1.3750 | 6.14 steps/s | VRAM=25.27 Go [2026-07-18 00:51:38] step 550850/1000000 | loss=2.2514 vocab=0.7364 token=1.1899 | 6.12 steps/s | VRAM=31.78 Go [2026-07-18 00:51:46] step 550900/1000000 | loss=2.1435 vocab=0.7942 token=0.9054 | 6.10 steps/s | VRAM=26.25 Go [2026-07-18 00:51:54] step 550950/1000000 | loss=2.1874 vocab=0.5483 token=1.2020 | 6.11 steps/s | VRAM=26.84 Go [2026-07-18 00:52:02] step 551000/1000000 | loss=2.2719 vocab=1.1375 token=1.3212 | 6.11 steps/s | VRAM=27.83 Go [2026-07-18 00:52:10] step 551050/1000000 | loss=2.2419 vocab=0.8475 token=1.3902 | 6.10 steps/s | VRAM=32.99 Go [2026-07-18 00:52:19] step 551100/1000000 | loss=2.2492 vocab=1.0190 token=1.3333 | 6.10 steps/s | VRAM=28.94 Go [2026-07-18 00:52:27] step 551150/1000000 | loss=2.2666 vocab=1.0640 token=1.5019 | 6.14 steps/s | VRAM=24.18 Go [2026-07-18 00:52:35] step 551200/1000000 | loss=2.2048 vocab=0.9738 token=1.2601 | 6.13 steps/s | VRAM=32.41 Go [2026-07-18 00:52:43] step 551250/1000000 | loss=2.2487 vocab=0.7531 token=1.0385 | 6.10 steps/s | VRAM=28.26 Go [2026-07-18 00:52:51] step 551300/1000000 | loss=2.2006 vocab=0.9970 token=1.2353 | 6.12 steps/s | VRAM=28.22 Go [2026-07-18 00:52:59] step 551350/1000000 | loss=2.2441 vocab=0.7502 token=1.5226 | 6.10 steps/s | VRAM=29.24 Go [2026-07-18 00:53:08] step 551400/1000000 | loss=2.1918 vocab=0.8575 token=1.0453 | 6.14 steps/s | VRAM=28.61 Go [2026-07-18 00:53:16] step 551450/1000000 | loss=2.3115 vocab=0.8997 token=1.5108 | 6.10 steps/s | VRAM=26.03 Go [2026-07-18 00:53:24] step 551500/1000000 | loss=2.2469 vocab=1.5204 token=1.2674 | 6.08 steps/s | VRAM=28.57 Go [2026-07-18 00:53:32] step 551550/1000000 | loss=2.3123 vocab=0.7222 token=1.4309 | 6.10 steps/s | VRAM=26.20 Go [2026-07-18 00:53:40] step 551600/1000000 | loss=2.2243 vocab=0.7379 token=1.5136 | 6.10 steps/s | VRAM=25.87 Go [2026-07-18 00:53:48] step 551650/1000000 | loss=2.2595 vocab=1.0341 token=1.0153 | 6.13 steps/s | VRAM=28.23 Go [2026-07-18 00:53:57] step 551700/1000000 | loss=2.3106 vocab=1.3823 token=1.7877 | 6.11 steps/s | VRAM=27.27 Go [2026-07-18 00:54:05] step 551750/1000000 | loss=2.3223 vocab=0.8511 token=1.3327 | 6.11 steps/s | VRAM=30.42 Go [2026-07-18 00:54:13] step 551800/1000000 | loss=2.2722 vocab=0.7474 token=1.1521 | 6.11 steps/s | VRAM=28.70 Go [2026-07-18 00:54:21] step 551850/1000000 | loss=2.3136 vocab=1.1744 token=1.8059 | 6.10 steps/s | VRAM=28.35 Go [2026-07-18 00:54:29] step 551900/1000000 | loss=2.2679 vocab=1.1703 token=1.5320 | 6.11 steps/s | VRAM=28.88 Go [2026-07-18 00:54:38] step 551950/1000000 | loss=2.3504 vocab=0.8296 token=1.7456 | 6.08 steps/s | VRAM=30.34 Go [2026-07-18 00:54:46] step 552000/1000000 | loss=2.1496 vocab=0.6477 token=0.7818 | 6.14 steps/s | VRAM=28.72 Go [2026-07-18 00:54:54] step 552050/1000000 | loss=2.3053 vocab=1.0584 token=1.5597 | 6.17 steps/s | VRAM=31.18 Go [2026-07-18 00:55:02] step 552100/1000000 | loss=2.2086 vocab=0.9546 token=1.4453 | 6.17 steps/s | VRAM=31.25 Go [2026-07-18 00:55:10] step 552150/1000000 | loss=2.3050 vocab=0.6787 token=1.3705 | 6.14 steps/s | VRAM=25.59 Go [2026-07-18 00:55:18] step 552200/1000000 | loss=2.3037 vocab=0.9045 token=1.1081 | 6.14 steps/s | VRAM=31.35 Go [2026-07-18 00:55:26] step 552250/1000000 | loss=2.2346 vocab=0.4795 token=1.0411 | 6.12 steps/s | VRAM=29.05 Go [2026-07-18 00:55:35] step 552300/1000000 | loss=2.2706 vocab=1.4020 token=1.6557 | 6.13 steps/s | VRAM=31.40 Go [2026-07-18 00:55:43] step 552350/1000000 | loss=2.2262 vocab=0.7517 token=1.6645 | 6.18 steps/s | VRAM=28.67 Go [2026-07-18 00:55:51] step 552400/1000000 | loss=2.2795 vocab=0.7044 token=0.9669 | 6.15 steps/s | VRAM=31.75 Go [2026-07-18 00:55:59] step 552450/1000000 | loss=2.2706 vocab=0.8193 token=1.3274 | 6.15 steps/s | VRAM=31.85 Go [2026-07-18 00:56:07] step 552500/1000000 | loss=2.1890 vocab=0.4790 token=1.1275 | 6.11 steps/s | VRAM=29.37 Go [2026-07-18 00:56:15] step 552550/1000000 | loss=2.2400 vocab=0.6029 token=1.1194 | 6.16 steps/s | VRAM=28.56 Go [2026-07-18 00:56:23] step 552600/1000000 | loss=2.2658 vocab=0.7124 token=1.1173 | 6.14 steps/s | VRAM=28.04 Go [2026-07-18 00:56:32] step 552650/1000000 | loss=2.2008 vocab=0.8698 token=1.1279 | 6.12 steps/s | VRAM=28.05 Go [2026-07-18 00:56:40] step 552700/1000000 | loss=2.2204 vocab=0.9237 token=1.2364 | 6.15 steps/s | VRAM=31.34 Go [2026-07-18 00:56:48] step 552750/1000000 | loss=2.2877 vocab=0.7250 token=0.9828 | 6.12 steps/s | VRAM=32.64 Go [2026-07-18 00:56:56] step 552800/1000000 | loss=2.2572 vocab=1.6760 token=1.4509 | 6.14 steps/s | VRAM=29.02 Go [2026-07-18 00:57:04] step 552850/1000000 | loss=2.3688 vocab=1.2297 token=1.4437 | 6.14 steps/s | VRAM=27.80 Go [2026-07-18 00:57:12] step 552900/1000000 | loss=2.1961 vocab=0.9667 token=1.0368 | 6.14 steps/s | VRAM=28.24 Go [2026-07-18 00:57:20] step 552950/1000000 | loss=2.1588 vocab=0.5516 token=0.9249 | 6.12 steps/s | VRAM=29.32 Go [2026-07-18 00:57:29] step 553000/1000000 | loss=2.1807 vocab=0.8009 token=1.2642 | 6.10 steps/s | VRAM=29.10 Go [2026-07-18 00:57:37] step 553050/1000000 | loss=2.2229 vocab=0.9859 token=1.1658 | 6.09 steps/s | VRAM=26.51 Go [2026-07-18 00:57:45] step 553100/1000000 | loss=2.1307 vocab=1.4081 token=1.2555 | 6.17 steps/s | VRAM=32.01 Go [2026-07-18 00:57:53] step 553150/1000000 | loss=2.2257 vocab=2.3992 token=1.7242 | 6.12 steps/s | VRAM=26.64 Go [2026-07-18 00:58:01] step 553200/1000000 | loss=2.2775 vocab=0.9153 token=1.2596 | 6.18 steps/s | VRAM=28.49 Go [2026-07-18 00:58:09] step 553250/1000000 | loss=2.1848 vocab=0.9185 token=1.2604 | 6.15 steps/s | VRAM=32.02 Go [2026-07-18 00:58:18] step 553300/1000000 | loss=2.2961 vocab=0.7976 token=1.0811 | 6.12 steps/s | VRAM=32.98 Go [2026-07-18 00:58:26] step 553350/1000000 | loss=2.2050 vocab=0.7330 token=1.4144 | 6.11 steps/s | VRAM=26.32 Go [2026-07-18 00:58:34] step 553400/1000000 | loss=2.2137 vocab=1.0743 token=1.2352 | 6.16 steps/s | VRAM=29.04 Go [2026-07-18 00:58:42] step 553450/1000000 | loss=2.2415 vocab=0.9215 token=1.2209 | 6.13 steps/s | VRAM=29.73 Go [2026-07-18 00:58:50] step 553500/1000000 | loss=2.1965 vocab=0.9618 token=1.6145 | 6.15 steps/s | VRAM=28.23 Go [2026-07-18 00:58:58] step 553550/1000000 | loss=2.2634 vocab=1.1326 token=1.7259 | 6.16 steps/s | VRAM=30.96 Go [2026-07-18 00:59:06] step 553600/1000000 | loss=2.3170 vocab=1.1902 token=1.6601 | 6.10 steps/s | VRAM=29.93 Go [2026-07-18 00:59:15] step 553650/1000000 | loss=2.0777 vocab=0.9903 token=1.5339 | 6.15 steps/s | VRAM=32.31 Go [2026-07-18 00:59:23] step 553700/1000000 | loss=2.2120 vocab=0.5276 token=1.2643 | 6.13 steps/s | VRAM=29.00 Go [2026-07-18 00:59:31] step 553750/1000000 | loss=2.2563 vocab=0.5849 token=1.1120 | 6.16 steps/s | VRAM=28.72 Go [2026-07-18 00:59:39] step 553800/1000000 | loss=2.2261 vocab=1.2205 token=1.5137 | 6.12 steps/s | VRAM=28.28 Go [2026-07-18 00:59:47] step 553850/1000000 | loss=2.1604 vocab=0.8860 token=0.9225 | 6.08 steps/s | VRAM=26.42 Go [2026-07-18 00:59:55] step 553900/1000000 | loss=2.2730 vocab=0.9259 token=1.8146 | 6.13 steps/s | VRAM=32.42 Go [2026-07-18 01:00:04] step 553950/1000000 | loss=2.1756 vocab=0.9885 token=1.5130 | 6.14 steps/s | VRAM=28.35 Go [2026-07-18 01:00:12] step 554000/1000000 | loss=2.1931 vocab=1.4500 token=1.5166 | 6.11 steps/s | VRAM=29.44 Go [2026-07-18 01:00:20] step 554050/1000000 | loss=2.2503 vocab=0.8714 token=1.3082 | 6.14 steps/s | VRAM=33.01 Go [2026-07-18 01:00:28] step 554100/1000000 | loss=2.3015 vocab=1.0930 token=1.5499 | 6.15 steps/s | VRAM=29.21 Go [2026-07-18 01:00:36] step 554150/1000000 | loss=2.2042 vocab=1.1500 token=1.4727 | 6.11 steps/s | VRAM=26.11 Go [2026-07-18 01:00:44] step 554200/1000000 | loss=2.1772 vocab=1.0647 token=1.5223 | 6.17 steps/s | VRAM=32.86 Go [2026-07-18 01:00:52] step 554250/1000000 | loss=2.1624 vocab=0.9174 token=1.1586 | 6.14 steps/s | VRAM=31.91 Go [2026-07-18 01:01:01] step 554300/1000000 | loss=2.2283 vocab=1.1515 token=1.3598 | 6.10 steps/s | VRAM=28.45 Go [2026-07-18 01:01:09] step 554350/1000000 | loss=2.2546 vocab=1.1137 token=1.5872 | 6.14 steps/s | VRAM=27.90 Go [2026-07-18 01:01:17] step 554400/1000000 | loss=2.2077 vocab=0.7046 token=1.4950 | 6.13 steps/s | VRAM=28.58 Go [2026-07-18 01:01:25] step 554450/1000000 | loss=2.2521 vocab=0.6721 token=1.4015 | 6.12 steps/s | VRAM=26.54 Go [2026-07-18 01:01:33] step 554500/1000000 | loss=2.1961 vocab=0.9998 token=1.5548 | 6.13 steps/s | VRAM=31.67 Go [2026-07-18 01:01:41] step 554550/1000000 | loss=2.2154 vocab=0.8431 token=0.9103 | 6.16 steps/s | VRAM=31.69 Go [2026-07-18 01:01:50] step 554600/1000000 | loss=2.2525 vocab=0.6726 token=1.3415 | 6.14 steps/s | VRAM=23.80 Go [2026-07-18 01:01:58] step 554650/1000000 | loss=2.1356 vocab=0.9285 token=1.4519 | 6.16 steps/s | VRAM=29.56 Go [2026-07-18 01:02:06] step 554700/1000000 | loss=2.2287 vocab=0.8111 token=1.5195 | 6.12 steps/s | VRAM=26.58 Go [2026-07-18 01:02:14] step 554750/1000000 | loss=2.2145 vocab=0.7102 token=1.0906 | 6.15 steps/s | VRAM=31.61 Go [2026-07-18 01:02:22] step 554800/1000000 | loss=2.2809 vocab=0.8904 token=1.4091 | 6.15 steps/s | VRAM=31.76 Go [2026-07-18 01:02:30] step 554850/1000000 | loss=2.3217 vocab=0.6792 token=1.1482 | 6.17 steps/s | VRAM=31.65 Go [2026-07-18 01:02:38] step 554900/1000000 | loss=2.2603 vocab=0.9756 token=1.3470 | 6.13 steps/s | VRAM=28.45 Go [2026-07-18 01:02:47] step 554950/1000000 | loss=2.3336 vocab=0.7351 token=1.0558 | 6.12 steps/s | VRAM=29.09 Go [2026-07-18 01:02:55] step 555000/1000000 | loss=2.2508 vocab=0.8943 token=1.1500 | 6.15 steps/s | VRAM=32.50 Go [2026-07-18 01:03:03] step 555050/1000000 | loss=2.1460 vocab=0.9322 token=1.4771 | 6.15 steps/s | VRAM=31.91 Go [2026-07-18 01:03:11] step 555100/1000000 | loss=2.3357 vocab=1.2436 token=1.3787 | 6.16 steps/s | VRAM=24.64 Go [2026-07-18 01:03:19] step 555150/1000000 | loss=2.3156 vocab=1.1991 token=1.6911 | 6.13 steps/s | VRAM=28.56 Go [2026-07-18 01:03:27] step 555200/1000000 | loss=2.2714 vocab=0.8590 token=1.7418 | 6.14 steps/s | VRAM=28.80 Go [2026-07-18 01:03:35] step 555250/1000000 | loss=2.0914 vocab=0.6212 token=1.3638 | 6.11 steps/s | VRAM=28.91 Go [2026-07-18 01:03:44] step 555300/1000000 | loss=2.2824 vocab=0.9685 token=1.2525 | 6.15 steps/s | VRAM=32.01 Go [2026-07-18 01:03:52] step 555350/1000000 | loss=2.1496 vocab=0.6469 token=1.1792 | 6.13 steps/s | VRAM=28.32 Go [2026-07-18 01:04:00] step 555400/1000000 | loss=2.2422 vocab=1.0902 token=1.4691 | 6.15 steps/s | VRAM=32.52 Go [2026-07-18 01:04:08] step 555450/1000000 | loss=2.2934 vocab=1.3860 token=1.4508 | 6.12 steps/s | VRAM=29.31 Go [2026-07-18 01:04:16] step 555500/1000000 | loss=2.2204 vocab=1.0391 token=1.7001 | 6.12 steps/s | VRAM=26.12 Go [2026-07-18 01:04:24] step 555550/1000000 | loss=2.3823 vocab=0.9749 token=1.5963 | 6.14 steps/s | VRAM=25.68 Go [2026-07-18 01:04:32] step 555600/1000000 | loss=2.1849 vocab=0.6961 token=1.0411 | 6.15 steps/s | VRAM=32.64 Go [2026-07-18 01:04:41] step 555650/1000000 | loss=2.2819 vocab=0.9619 token=1.3679 | 6.14 steps/s | VRAM=31.08 Go [2026-07-18 01:04:49] step 555700/1000000 | loss=2.1588 vocab=1.1630 token=0.9854 | 6.13 steps/s | VRAM=28.64 Go [2026-07-18 01:04:57] step 555750/1000000 | loss=2.2007 vocab=0.9994 token=1.4797 | 6.14 steps/s | VRAM=29.25 Go [2026-07-18 01:05:05] step 555800/1000000 | loss=2.2648 vocab=0.7307 token=1.0028 | 6.12 steps/s | VRAM=29.12 Go [2026-07-18 01:05:13] step 555850/1000000 | loss=2.2679 vocab=0.7934 token=1.2483 | 6.14 steps/s | VRAM=28.51 Go [2026-07-18 01:05:21] step 555900/1000000 | loss=2.3376 vocab=1.4864 token=1.3339 | 6.13 steps/s | VRAM=28.76 Go [2026-07-18 01:05:29] step 555950/1000000 | loss=2.2747 vocab=0.8568 token=1.5688 | 6.17 steps/s | VRAM=32.38 Go [2026-07-18 01:05:38] step 556000/1000000 | loss=2.3071 vocab=0.9100 token=1.4753 | 6.12 steps/s | VRAM=27.08 Go [2026-07-18 01:05:46] step 556050/1000000 | loss=2.2448 vocab=0.7574 token=1.3920 | 6.14 steps/s | VRAM=27.81 Go [2026-07-18 01:05:54] step 556100/1000000 | loss=2.3949 vocab=0.5308 token=1.3729 | 6.15 steps/s | VRAM=28.74 Go [2026-07-18 01:06:02] step 556150/1000000 | loss=2.2477 vocab=0.8876 token=1.3822 | 6.13 steps/s | VRAM=24.55 Go [2026-07-18 01:06:10] step 556200/1000000 | loss=2.3091 vocab=1.1425 token=1.4275 | 6.13 steps/s | VRAM=29.59 Go [2026-07-18 01:06:18] step 556250/1000000 | loss=2.2333 vocab=0.8987 token=1.4706 | 6.15 steps/s | VRAM=28.08 Go [2026-07-18 01:06:26] step 556300/1000000 | loss=2.2323 vocab=0.7292 token=1.3568 | 6.15 steps/s | VRAM=28.76 Go [2026-07-18 01:06:35] step 556350/1000000 | loss=2.1104 vocab=0.8487 token=1.2505 | 6.13 steps/s | VRAM=32.76 Go [2026-07-18 01:06:43] step 556400/1000000 | loss=2.2289 vocab=1.0072 token=1.1322 | 6.14 steps/s | VRAM=28.12 Go [2026-07-18 01:06:51] step 556450/1000000 | loss=2.3090 vocab=0.7947 token=1.2667 | 6.16 steps/s | VRAM=28.87 Go [2026-07-18 01:06:59] step 556500/1000000 | loss=2.2973 vocab=0.9097 token=1.3506 | 6.13 steps/s | VRAM=28.86 Go [2026-07-18 01:07:07] step 556550/1000000 | loss=2.1938 vocab=1.1883 token=1.4103 | 6.16 steps/s | VRAM=27.84 Go [2026-07-18 01:07:15] step 556600/1000000 | loss=2.3804 vocab=0.7198 token=1.2952 | 6.13 steps/s | VRAM=32.53 Go [2026-07-18 01:07:23] step 556650/1000000 | loss=2.2481 vocab=0.6441 token=0.9701 | 6.15 steps/s | VRAM=29.08 Go [2026-07-18 01:07:32] step 556700/1000000 | loss=2.2623 vocab=1.2086 token=1.6362 | 6.14 steps/s | VRAM=30.15 Go [2026-07-18 01:07:40] step 556750/1000000 | loss=2.2122 vocab=0.8054 token=1.4678 | 6.14 steps/s | VRAM=31.92 Go [2026-07-18 01:07:48] step 556800/1000000 | loss=2.3232 vocab=1.2693 token=1.4561 | 6.17 steps/s | VRAM=32.06 Go [2026-07-18 01:07:56] step 556850/1000000 | loss=2.1749 vocab=0.8798 token=1.3354 | 6.11 steps/s | VRAM=32.53 Go [2026-07-18 01:08:04] step 556900/1000000 | loss=2.1655 vocab=0.3724 token=1.0309 | 6.14 steps/s | VRAM=25.22 Go [2026-07-18 01:08:12] step 556950/1000000 | loss=2.2421 vocab=0.8413 token=1.0122 | 6.14 steps/s | VRAM=28.18 Go [2026-07-18 01:08:20] step 557000/1000000 | loss=2.2955 vocab=0.6873 token=1.2033 | 6.14 steps/s | VRAM=29.01 Go [2026-07-18 01:08:29] step 557050/1000000 | loss=2.1889 vocab=0.6802 token=1.5589 | 6.15 steps/s | VRAM=28.84 Go [2026-07-18 01:08:37] step 557100/1000000 | loss=2.2963 vocab=0.6233 token=1.2700 | 6.14 steps/s | VRAM=29.11 Go [2026-07-18 01:08:45] step 557150/1000000 | loss=2.2560 vocab=0.8912 token=1.0349 | 6.13 steps/s | VRAM=28.88 Go [2026-07-18 01:08:53] step 557200/1000000 | loss=2.3776 vocab=0.7058 token=1.1366 | 6.15 steps/s | VRAM=31.42 Go [2026-07-18 01:09:01] step 557250/1000000 | loss=2.1897 vocab=1.3345 token=1.2965 | 6.16 steps/s | VRAM=28.63 Go [2026-07-18 01:09:09] step 557300/1000000 | loss=2.1706 vocab=0.7470 token=1.1149 | 6.15 steps/s | VRAM=31.50 Go [2026-07-18 01:09:17] step 557350/1000000 | loss=2.2340 vocab=0.5414 token=1.4103 | 6.11 steps/s | VRAM=28.95 Go [2026-07-18 01:09:26] step 557400/1000000 | loss=2.2402 vocab=1.4914 token=1.2206 | 6.12 steps/s | VRAM=29.66 Go [2026-07-18 01:09:34] step 557450/1000000 | loss=2.2964 vocab=0.6903 token=1.0368 | 6.13 steps/s | VRAM=28.68 Go [2026-07-18 01:09:42] step 557500/1000000 | loss=2.2415 vocab=1.1388 token=1.6129 | 6.14 steps/s | VRAM=26.39 Go [2026-07-18 01:09:50] step 557550/1000000 | loss=2.2332 vocab=0.9462 token=1.2682 | 6.15 steps/s | VRAM=27.92 Go [2026-07-18 01:09:58] step 557600/1000000 | loss=2.3045 vocab=0.9936 token=1.4198 | 6.16 steps/s | VRAM=27.76 Go [2026-07-18 01:10:06] step 557650/1000000 | loss=2.2482 vocab=0.9203 token=1.2846 | 6.13 steps/s | VRAM=24.66 Go [2026-07-18 01:10:14] step 557700/1000000 | loss=2.2745 vocab=0.7424 token=1.1916 | 6.15 steps/s | VRAM=28.46 Go [2026-07-18 01:10:23] step 557750/1000000 | loss=2.1372 vocab=0.9379 token=1.4111 | 6.14 steps/s | VRAM=25.88 Go [2026-07-18 01:10:31] step 557800/1000000 | loss=2.2413 vocab=1.4232 token=1.2638 | 6.16 steps/s | VRAM=32.45 Go [2026-07-18 01:10:39] step 557850/1000000 | loss=2.2929 vocab=0.9178 token=1.2673 | 6.08 steps/s | VRAM=26.00 Go [2026-07-18 01:10:47] step 557900/1000000 | loss=2.2870 vocab=1.0645 token=1.4768 | 6.11 steps/s | VRAM=28.64 Go [2026-07-18 01:10:55] step 557950/1000000 | loss=2.3043 vocab=0.6919 token=0.9477 | 6.12 steps/s | VRAM=29.57 Go [2026-07-18 01:11:04] step 558000/1000000 | loss=2.2270 vocab=0.9700 token=1.3726 | 6.11 steps/s | VRAM=28.36 Go [2026-07-18 01:11:12] step 558050/1000000 | loss=2.3370 vocab=0.6298 token=1.1512 | 6.14 steps/s | VRAM=26.11 Go [2026-07-18 01:11:20] step 558100/1000000 | loss=2.3582 vocab=1.1431 token=1.6677 | 6.14 steps/s | VRAM=29.43 Go [2026-07-18 01:11:28] step 558150/1000000 | loss=2.2633 vocab=1.1118 token=1.2901 | 6.13 steps/s | VRAM=29.35 Go [2026-07-18 01:11:36] step 558200/1000000 | loss=2.3057 vocab=1.0443 token=1.3895 | 6.14 steps/s | VRAM=28.41 Go [2026-07-18 01:11:44] step 558250/1000000 | loss=2.2788 vocab=0.9984 token=1.5676 | 6.11 steps/s | VRAM=28.83 Go [2026-07-18 01:11:52] step 558300/1000000 | loss=2.2260 vocab=0.9207 token=1.1479 | 6.11 steps/s | VRAM=29.22 Go [2026-07-18 01:12:01] step 558350/1000000 | loss=2.0923 vocab=0.2608 token=0.9072 | 6.14 steps/s | VRAM=28.36 Go [2026-07-18 01:12:09] step 558400/1000000 | loss=2.2694 vocab=0.8371 token=1.4550 | 6.17 steps/s | VRAM=31.45 Go [2026-07-18 01:12:17] step 558450/1000000 | loss=2.1743 vocab=0.8241 token=1.4121 | 6.13 steps/s | VRAM=26.48 Go [2026-07-18 01:12:25] step 558500/1000000 | loss=2.3286 vocab=0.7228 token=1.1607 | 6.14 steps/s | VRAM=28.70 Go [2026-07-18 01:12:33] step 558550/1000000 | loss=2.3253 vocab=1.0049 token=1.6496 | 6.11 steps/s | VRAM=31.37 Go [2026-07-18 01:12:41] step 558600/1000000 | loss=2.2867 vocab=1.5722 token=1.8739 | 6.13 steps/s | VRAM=28.25 Go [2026-07-18 01:12:50] step 558650/1000000 | loss=2.2154 vocab=0.9561 token=1.4439 | 6.13 steps/s | VRAM=29.92 Go [2026-07-18 01:12:58] step 558700/1000000 | loss=2.2560 vocab=1.2886 token=1.3737 | 6.14 steps/s | VRAM=28.74 Go [2026-07-18 01:13:06] step 558750/1000000 | loss=2.3076 vocab=0.6932 token=1.5189 | 6.14 steps/s | VRAM=25.63 Go [2026-07-18 01:13:14] step 558800/1000000 | loss=2.3024 vocab=0.3355 token=0.9516 | 6.14 steps/s | VRAM=28.51 Go [2026-07-18 01:13:22] step 558850/1000000 | loss=2.3368 vocab=0.9195 token=1.1687 | 6.10 steps/s | VRAM=23.81 Go [2026-07-18 01:13:30] step 558900/1000000 | loss=2.2402 vocab=0.7997 token=1.4679 | 6.12 steps/s | VRAM=25.46 Go [2026-07-18 01:13:38] step 558950/1000000 | loss=2.2676 vocab=1.2223 token=1.2894 | 6.14 steps/s | VRAM=29.46 Go [2026-07-18 01:13:47] step 559000/1000000 | loss=2.0523 vocab=0.5717 token=1.0923 | 6.13 steps/s | VRAM=32.03 Go [2026-07-18 01:13:55] step 559050/1000000 | loss=2.1994 vocab=0.6455 token=1.1117 | 6.12 steps/s | VRAM=25.87 Go [2026-07-18 01:14:03] step 559100/1000000 | loss=2.1569 vocab=1.1926 token=1.5867 | 6.14 steps/s | VRAM=28.63 Go [2026-07-18 01:14:11] step 559150/1000000 | loss=2.3298 vocab=0.3772 token=1.0673 | 6.14 steps/s | VRAM=28.60 Go [2026-07-18 01:14:19] step 559200/1000000 | loss=2.2000 vocab=1.4826 token=1.4520 | 6.09 steps/s | VRAM=29.15 Go [2026-07-18 01:14:27] step 559250/1000000 | loss=2.2532 vocab=0.8961 token=1.5536 | 6.16 steps/s | VRAM=31.80 Go [2026-07-18 01:14:36] step 559300/1000000 | loss=2.3003 vocab=1.0667 token=1.4127 | 6.12 steps/s | VRAM=28.72 Go [2026-07-18 01:14:44] step 559350/1000000 | loss=2.2300 vocab=0.8960 token=1.3978 | 6.16 steps/s | VRAM=28.06 Go [2026-07-18 01:14:52] step 559400/1000000 | loss=2.2087 vocab=0.9234 token=1.3574 | 6.13 steps/s | VRAM=28.12 Go [2026-07-18 01:15:00] step 559450/1000000 | loss=2.2127 vocab=0.7484 token=1.2125 | 6.12 steps/s | VRAM=29.30 Go [2026-07-18 01:15:08] step 559500/1000000 | loss=2.1961 vocab=1.3563 token=1.4589 | 6.11 steps/s | VRAM=28.47 Go [2026-07-18 01:15:16] step 559550/1000000 | loss=2.3087 vocab=1.0134 token=1.5866 | 6.14 steps/s | VRAM=31.46 Go [2026-07-18 01:15:24] step 559600/1000000 | loss=2.2315 vocab=0.7954 token=1.1050 | 6.14 steps/s | VRAM=29.23 Go [2026-07-18 01:15:33] step 559650/1000000 | loss=2.2471 vocab=0.5231 token=1.0927 | 6.12 steps/s | VRAM=28.84 Go [2026-07-18 01:15:41] step 559700/1000000 | loss=2.1718 vocab=1.3836 token=1.4275 | 6.13 steps/s | VRAM=32.02 Go [2026-07-18 01:15:49] step 559750/1000000 | loss=2.3627 vocab=0.9542 token=1.8654 | 6.11 steps/s | VRAM=27.78 Go [2026-07-18 01:15:57] step 559800/1000000 | loss=2.3716 vocab=0.8989 token=1.4547 | 6.16 steps/s | VRAM=26.08 Go [2026-07-18 01:16:05] step 559850/1000000 | loss=2.2336 vocab=1.0509 token=1.6232 | 6.15 steps/s | VRAM=32.34 Go [2026-07-18 01:16:13] step 559900/1000000 | loss=2.3050 vocab=0.7216 token=1.3946 | 6.12 steps/s | VRAM=26.42 Go [2026-07-18 01:16:22] step 559950/1000000 | loss=2.1172 vocab=0.7208 token=1.4914 | 6.16 steps/s | VRAM=32.66 Go [2026-07-18 01:16:30] step 560000/1000000 | loss=2.1145 vocab=0.5604 token=1.3317 | 6.10 steps/s | VRAM=30.72 Go [2026-07-18 01:16:38] step 560050/1000000 | loss=2.1490 vocab=0.8294 token=1.4634 | 6.15 steps/s | VRAM=29.57 Go [2026-07-18 01:16:46] step 560100/1000000 | loss=2.3717 vocab=0.8763 token=1.1424 | 6.14 steps/s | VRAM=28.37 Go [2026-07-18 01:16:54] step 560150/1000000 | loss=2.1775 vocab=0.9951 token=1.4358 | 6.10 steps/s | VRAM=27.02 Go [2026-07-18 01:17:02] step 560200/1000000 | loss=2.2870 vocab=0.7375 token=1.6223 | 6.15 steps/s | VRAM=28.53 Go [2026-07-18 01:17:10] step 560250/1000000 | loss=2.1613 vocab=0.5667 token=1.4324 | 6.15 steps/s | VRAM=29.03 Go [2026-07-18 01:17:19] step 560300/1000000 | loss=2.2510 vocab=0.9045 token=1.3125 | 6.12 steps/s | VRAM=26.01 Go [2026-07-18 01:17:27] step 560350/1000000 | loss=2.2954 vocab=0.9685 token=1.3587 | 6.14 steps/s | VRAM=31.11 Go [2026-07-18 01:17:35] step 560400/1000000 | loss=2.2180 vocab=1.0759 token=1.2713 | 6.13 steps/s | VRAM=28.79 Go [2026-07-18 01:17:43] step 560450/1000000 | loss=2.0832 vocab=0.6626 token=1.3041 | 6.14 steps/s | VRAM=32.01 Go [2026-07-18 01:17:51] step 560500/1000000 | loss=2.2788 vocab=0.4993 token=1.0726 | 6.12 steps/s | VRAM=29.08 Go [2026-07-18 01:17:59] step 560550/1000000 | loss=2.2514 vocab=1.1847 token=1.3617 | 6.15 steps/s | VRAM=31.31 Go [2026-07-18 01:18:08] step 560600/1000000 | loss=2.2614 vocab=1.0032 token=1.4190 | 6.13 steps/s | VRAM=28.91 Go [2026-07-18 01:18:16] step 560650/1000000 | loss=2.1109 vocab=0.9398 token=1.2476 | 6.14 steps/s | VRAM=29.03 Go [2026-07-18 01:18:24] step 560700/1000000 | loss=2.2234 vocab=0.5609 token=1.0791 | 6.15 steps/s | VRAM=32.27 Go [2026-07-18 01:18:32] step 560750/1000000 | loss=2.1333 vocab=0.8434 token=1.5042 | 6.13 steps/s | VRAM=26.04 Go [2026-07-18 01:18:40] step 560800/1000000 | loss=2.1994 vocab=0.7036 token=1.3920 | 6.13 steps/s | VRAM=28.68 Go [2026-07-18 01:18:48] step 560850/1000000 | loss=2.2575 vocab=0.4289 token=0.8741 | 6.15 steps/s | VRAM=28.57 Go [2026-07-18 01:18:56] step 560900/1000000 | loss=2.2390 vocab=0.5881 token=0.9583 | 6.13 steps/s | VRAM=32.63 Go [2026-07-18 01:19:05] step 560950/1000000 | loss=2.2500 vocab=1.2495 token=1.0781 | 6.11 steps/s | VRAM=29.08 Go [2026-07-18 01:19:13] step 561000/1000000 | loss=2.2066 vocab=0.6955 token=1.3714 | 6.10 steps/s | VRAM=25.25 Go [2026-07-18 01:19:21] step 561050/1000000 | loss=2.2555 vocab=1.2185 token=1.3612 | 6.10 steps/s | VRAM=26.11 Go [2026-07-18 01:19:29] step 561100/1000000 | loss=2.1541 vocab=0.9378 token=1.7615 | 6.12 steps/s | VRAM=28.55 Go [2026-07-18 01:19:37] step 561150/1000000 | loss=2.2327 vocab=0.8184 token=1.1124 | 6.10 steps/s | VRAM=26.67 Go [2026-07-18 01:19:45] step 561200/1000000 | loss=2.1373 vocab=1.0009 token=1.2964 | 6.13 steps/s | VRAM=32.95 Go [2026-07-18 01:19:54] step 561250/1000000 | loss=2.2114 vocab=0.7558 token=1.2719 | 6.13 steps/s | VRAM=31.07 Go [2026-07-18 01:20:02] step 561300/1000000 | loss=2.1834 vocab=0.7048 token=1.2851 | 6.12 steps/s | VRAM=25.87 Go [2026-07-18 01:20:10] step 561350/1000000 | loss=2.2138 vocab=0.7494 token=1.3571 | 6.13 steps/s | VRAM=29.01 Go [2026-07-18 01:20:18] step 561400/1000000 | loss=2.1762 vocab=0.7188 token=1.1438 | 6.14 steps/s | VRAM=28.57 Go [2026-07-18 01:20:26] step 561450/1000000 | loss=2.2378 vocab=0.9533 token=1.1307 | 6.14 steps/s | VRAM=29.73 Go [2026-07-18 01:20:34] step 561500/1000000 | loss=2.1795 vocab=1.0429 token=1.5126 | 6.13 steps/s | VRAM=25.69 Go [2026-07-18 01:20:43] step 561550/1000000 | loss=2.2961 vocab=0.8359 token=1.2051 | 6.14 steps/s | VRAM=28.78 Go [2026-07-18 01:20:51] step 561600/1000000 | loss=2.3428 vocab=1.1096 token=1.0062 | 6.14 steps/s | VRAM=31.96 Go [2026-07-18 01:20:59] step 561650/1000000 | loss=2.2009 vocab=0.5778 token=1.4703 | 6.09 steps/s | VRAM=29.32 Go [2026-07-18 01:21:07] step 561700/1000000 | loss=2.2379 vocab=0.8290 token=1.2789 | 6.11 steps/s | VRAM=29.09 Go [2026-07-18 01:21:15] step 561750/1000000 | loss=2.2008 vocab=0.8517 token=1.5665 | 6.15 steps/s | VRAM=31.66 Go [2026-07-18 01:21:23] step 561800/1000000 | loss=2.3182 vocab=1.2171 token=1.6249 | 6.10 steps/s | VRAM=26.40 Go [2026-07-18 01:21:32] step 561850/1000000 | loss=2.3234 vocab=0.8783 token=1.1766 | 6.12 steps/s | VRAM=29.22 Go [2026-07-18 01:21:40] step 561900/1000000 | loss=2.3063 vocab=1.0279 token=1.6279 | 6.14 steps/s | VRAM=24.79 Go [2026-07-18 01:21:48] step 561950/1000000 | loss=2.2995 vocab=0.8427 token=1.3046 | 6.10 steps/s | VRAM=27.70 Go [2026-07-18 01:21:56] step 562000/1000000 | loss=2.2152 vocab=0.8091 token=1.1983 | 6.10 steps/s | VRAM=29.85 Go [2026-07-18 01:22:04] step 562050/1000000 | loss=2.2056 vocab=0.8425 token=1.6795 | 6.13 steps/s | VRAM=25.25 Go [2026-07-18 01:22:12] step 562100/1000000 | loss=2.2573 vocab=1.7951 token=1.6007 | 6.14 steps/s | VRAM=29.26 Go [2026-07-18 01:22:21] step 562150/1000000 | loss=2.3412 vocab=1.2544 token=1.5709 | 6.08 steps/s | VRAM=29.06 Go [2026-07-18 01:22:29] step 562200/1000000 | loss=2.2846 vocab=0.9337 token=1.6097 | 6.14 steps/s | VRAM=29.34 Go [2026-07-18 01:22:37] step 562250/1000000 | loss=2.1182 vocab=0.7428 token=1.1081 | 6.14 steps/s | VRAM=28.06 Go [2026-07-18 01:22:45] step 562300/1000000 | loss=2.1711 vocab=0.5319 token=1.2028 | 6.14 steps/s | VRAM=28.55 Go [2026-07-18 01:22:53] step 562350/1000000 | loss=2.1850 vocab=1.0768 token=1.1586 | 6.13 steps/s | VRAM=28.10 Go [2026-07-18 01:23:01] step 562400/1000000 | loss=2.4019 vocab=0.7791 token=1.5396 | 6.13 steps/s | VRAM=31.78 Go [2026-07-18 01:23:10] step 562450/1000000 | loss=2.3144 vocab=1.0590 token=1.8439 | 6.13 steps/s | VRAM=28.81 Go [2026-07-18 01:23:18] step 562500/1000000 | loss=2.2743 vocab=1.1862 token=0.9963 | 6.10 steps/s | VRAM=28.95 Go [2026-07-18 01:23:26] step 562550/1000000 | loss=2.2176 vocab=0.6589 token=1.5201 | 6.11 steps/s | VRAM=29.23 Go [2026-07-18 01:23:34] step 562600/1000000 | loss=2.3189 vocab=1.2244 token=1.5964 | 6.14 steps/s | VRAM=31.54 Go [2026-07-18 01:23:42] step 562650/1000000 | loss=2.2393 vocab=1.1472 token=1.4205 | 6.15 steps/s | VRAM=31.88 Go [2026-07-18 01:23:50] step 562700/1000000 | loss=2.2146 vocab=0.8089 token=0.9886 | 6.12 steps/s | VRAM=26.15 Go [2026-07-18 01:23:59] step 562750/1000000 | loss=2.3123 vocab=1.1687 token=1.7008 | 6.11 steps/s | VRAM=25.94 Go [2026-07-18 01:24:07] step 562800/1000000 | loss=2.2302 vocab=0.8059 token=1.3185 | 6.10 steps/s | VRAM=26.20 Go [2026-07-18 01:24:15] step 562850/1000000 | loss=2.2192 vocab=0.8975 token=1.3943 | 6.13 steps/s | VRAM=26.11 Go [2026-07-18 01:24:23] step 562900/1000000 | loss=2.2532 vocab=0.7288 token=1.1704 | 6.12 steps/s | VRAM=28.83 Go [2026-07-18 01:24:31] step 562950/1000000 | loss=2.2393 vocab=0.5768 token=1.2383 | 6.11 steps/s | VRAM=28.31 Go [2026-07-18 01:24:39] step 563000/1000000 | loss=2.1664 vocab=0.9896 token=1.5499 | 6.16 steps/s | VRAM=32.99 Go [2026-07-18 01:24:48] step 563050/1000000 | loss=2.2784 vocab=0.6011 token=1.2919 | 6.08 steps/s | VRAM=28.41 Go [2026-07-18 01:24:56] step 563100/1000000 | loss=2.2398 vocab=1.4279 token=1.7039 | 6.13 steps/s | VRAM=28.44 Go [2026-07-18 01:25:04] step 563150/1000000 | loss=2.3320 vocab=1.0463 token=1.4242 | 6.09 steps/s | VRAM=28.53 Go [2026-07-18 01:25:12] step 563200/1000000 | loss=2.0964 vocab=0.6176 token=1.5340 | 6.15 steps/s | VRAM=28.16 Go [2026-07-18 01:25:20] step 563250/1000000 | loss=2.2680 vocab=0.8322 token=1.3585 | 6.11 steps/s | VRAM=27.01 Go [2026-07-18 01:25:28] step 563300/1000000 | loss=2.2805 vocab=1.5697 token=1.9167 | 6.14 steps/s | VRAM=32.68 Go [2026-07-18 01:25:37] step 563350/1000000 | loss=2.2432 vocab=0.6769 token=1.0482 | 6.15 steps/s | VRAM=31.10 Go [2026-07-18 01:25:45] step 563400/1000000 | loss=2.1405 vocab=1.4490 token=1.6716 | 6.13 steps/s | VRAM=29.42 Go [2026-07-18 01:25:53] step 563450/1000000 | loss=2.2872 vocab=0.4303 token=0.7668 | 6.13 steps/s | VRAM=28.32 Go [2026-07-18 01:26:01] step 563500/1000000 | loss=2.2458 vocab=0.7518 token=1.0725 | 6.15 steps/s | VRAM=32.28 Go [2026-07-18 01:26:09] step 563550/1000000 | loss=2.2357 vocab=0.4768 token=0.9294 | 6.12 steps/s | VRAM=28.13 Go [2026-07-18 01:26:17] step 563600/1000000 | loss=2.2561 vocab=0.4512 token=0.8733 | 6.14 steps/s | VRAM=27.86 Go [2026-07-18 01:26:25] step 563650/1000000 | loss=2.3333 vocab=1.0538 token=1.5978 | 6.13 steps/s | VRAM=28.99 Go [2026-07-18 01:26:34] step 563700/1000000 | loss=2.2110 vocab=0.8240 token=1.2337 | 6.14 steps/s | VRAM=31.17 Go [2026-07-18 01:26:42] step 563750/1000000 | loss=2.1618 vocab=0.2486 token=0.8841 | 6.12 steps/s | VRAM=26.54 Go [2026-07-18 01:26:50] step 563800/1000000 | loss=2.2873 vocab=0.9686 token=1.4682 | 6.14 steps/s | VRAM=29.34 Go [2026-07-18 01:26:58] step 563850/1000000 | loss=2.2463 vocab=1.1434 token=1.4953 | 6.13 steps/s | VRAM=28.37 Go [2026-07-18 01:27:06] step 563900/1000000 | loss=2.1698 vocab=0.5079 token=1.1205 | 6.12 steps/s | VRAM=32.97 Go [2026-07-18 01:27:14] step 563950/1000000 | loss=2.1712 vocab=1.1035 token=1.4738 | 6.14 steps/s | VRAM=28.92 Go [2026-07-18 01:27:23] step 564000/1000000 | loss=2.1842 vocab=0.8181 token=1.4138 | 6.13 steps/s | VRAM=26.29 Go [2026-07-18 01:27:31] step 564050/1000000 | loss=2.2081 vocab=0.7190 token=1.5208 | 6.14 steps/s | VRAM=31.78 Go [2026-07-18 01:27:39] step 564100/1000000 | loss=2.2616 vocab=0.7702 token=1.3998 | 6.14 steps/s | VRAM=26.03 Go [2026-07-18 01:27:47] step 564150/1000000 | loss=2.1975 vocab=0.9516 token=1.0849 | 6.14 steps/s | VRAM=28.93 Go [2026-07-18 01:27:55] step 564200/1000000 | loss=2.2372 vocab=1.1827 token=1.2455 | 6.13 steps/s | VRAM=29.45 Go [2026-07-18 01:28:03] step 564250/1000000 | loss=2.1861 vocab=0.9821 token=1.2663 | 6.14 steps/s | VRAM=28.31 Go [2026-07-18 01:28:11] step 564300/1000000 | loss=2.1551 vocab=1.2498 token=1.0407 | 6.16 steps/s | VRAM=25.72 Go [2026-07-18 01:28:20] step 564350/1000000 | loss=2.2028 vocab=0.9493 token=1.0917 | 6.17 steps/s | VRAM=30.56 Go [2026-07-18 01:28:28] step 564400/1000000 | loss=2.1591 vocab=1.1969 token=1.5751 | 6.17 steps/s | VRAM=32.34 Go [2026-07-18 01:28:36] step 564450/1000000 | loss=2.2917 vocab=1.4652 token=1.6557 | 6.15 steps/s | VRAM=29.48 Go [2026-07-18 01:28:44] step 564500/1000000 | loss=2.2061 vocab=1.2309 token=1.5456 | 6.15 steps/s | VRAM=26.39 Go [2026-07-18 01:28:52] step 564550/1000000 | loss=2.2901 vocab=1.0218 token=1.3825 | 6.14 steps/s | VRAM=28.48 Go [2026-07-18 01:29:00] step 564600/1000000 | loss=2.2362 vocab=1.0430 token=1.4041 | 6.16 steps/s | VRAM=32.58 Go [2026-07-18 01:29:08] step 564650/1000000 | loss=2.2577 vocab=0.6864 token=1.3510 | 6.16 steps/s | VRAM=32.38 Go [2026-07-18 01:29:16] step 564700/1000000 | loss=2.2626 vocab=0.9329 token=1.3472 | 6.16 steps/s | VRAM=28.31 Go [2026-07-18 01:29:24] step 564750/1000000 | loss=2.3152 vocab=1.5681 token=1.3806 | 6.17 steps/s | VRAM=31.76 Go [2026-07-18 01:29:33] step 564800/1000000 | loss=2.1769 vocab=0.5778 token=1.3353 | 6.16 steps/s | VRAM=25.07 Go [2026-07-18 01:29:41] step 564850/1000000 | loss=2.2213 vocab=0.8332 token=1.5962 | 6.16 steps/s | VRAM=31.37 Go [2026-07-18 01:29:49] step 564900/1000000 | loss=2.2637 vocab=1.3351 token=1.4902 | 6.15 steps/s | VRAM=32.22 Go [2026-07-18 01:29:57] step 564950/1000000 | loss=2.2567 vocab=0.9830 token=1.3229 | 6.16 steps/s | VRAM=24.55 Go [2026-07-18 01:30:05] step 565000/1000000 | loss=2.2577 vocab=0.8917 token=0.9789 | 6.15 steps/s | VRAM=32.78 Go [2026-07-18 01:30:13] step 565050/1000000 | loss=2.2194 vocab=1.1329 token=1.5050 | 6.11 steps/s | VRAM=28.79 Go [2026-07-18 01:30:21] step 565100/1000000 | loss=2.2145 vocab=1.1246 token=1.3330 | 6.14 steps/s | VRAM=28.83 Go [2026-07-18 01:30:30] step 565150/1000000 | loss=2.2587 vocab=0.8847 token=1.3046 | 6.17 steps/s | VRAM=32.58 Go [2026-07-18 01:30:38] step 565200/1000000 | loss=2.1712 vocab=0.9640 token=1.3715 | 6.13 steps/s | VRAM=25.91 Go [2026-07-18 01:30:46] step 565250/1000000 | loss=2.2497 vocab=0.7348 token=0.9760 | 6.13 steps/s | VRAM=28.05 Go [2026-07-18 01:30:54] step 565300/1000000 | loss=2.2330 vocab=0.8397 token=1.5255 | 6.16 steps/s | VRAM=28.26 Go [2026-07-18 01:31:02] step 565350/1000000 | loss=2.2433 vocab=0.6020 token=1.5176 | 6.16 steps/s | VRAM=28.83 Go [2026-07-18 01:31:10] step 565400/1000000 | loss=2.3387 vocab=1.4216 token=1.4374 | 6.08 steps/s | VRAM=23.87 Go [2026-07-18 01:31:18] step 565450/1000000 | loss=2.1407 vocab=0.4570 token=0.8628 | 6.14 steps/s | VRAM=27.66 Go [2026-07-18 01:31:27] step 565500/1000000 | loss=2.1577 vocab=1.0395 token=1.3591 | 6.14 steps/s | VRAM=30.65 Go [2026-07-18 01:31:35] step 565550/1000000 | loss=2.2444 vocab=0.7834 token=1.2494 | 6.11 steps/s | VRAM=26.41 Go [2026-07-18 01:31:43] step 565600/1000000 | loss=2.3469 vocab=1.1087 token=1.3312 | 6.14 steps/s | VRAM=28.32 Go [2026-07-18 01:31:51] step 565650/1000000 | loss=2.2274 vocab=1.0189 token=1.8287 | 6.12 steps/s | VRAM=31.73 Go [2026-07-18 01:31:59] step 565700/1000000 | loss=2.2012 vocab=0.7030 token=1.5310 | 6.13 steps/s | VRAM=31.71 Go [2026-07-18 01:32:07] step 565750/1000000 | loss=2.3368 vocab=1.1372 token=1.5445 | 6.14 steps/s | VRAM=25.96 Go [2026-07-18 01:32:16] step 565800/1000000 | loss=2.1780 vocab=1.1250 token=1.2568 | 6.13 steps/s | VRAM=30.15 Go [2026-07-18 01:32:24] step 565850/1000000 | loss=2.2012 vocab=0.5237 token=1.2156 | 6.16 steps/s | VRAM=28.51 Go [2026-07-18 01:32:32] step 565900/1000000 | loss=2.1851 vocab=0.6577 token=1.2721 | 6.12 steps/s | VRAM=29.41 Go [2026-07-18 01:32:40] step 565950/1000000 | loss=2.1850 vocab=0.3868 token=1.3729 | 6.15 steps/s | VRAM=29.06 Go [2026-07-18 01:32:48] step 566000/1000000 | loss=2.1242 vocab=1.3112 token=1.6611 | 6.08 steps/s | VRAM=28.72 Go [2026-07-18 01:32:56] step 566050/1000000 | loss=2.1605 vocab=0.8785 token=1.0575 | 6.13 steps/s | VRAM=32.37 Go [2026-07-18 01:33:04] step 566100/1000000 | loss=2.1889 vocab=0.8989 token=1.4552 | 6.14 steps/s | VRAM=29.02 Go [2026-07-18 01:33:13] step 566150/1000000 | loss=2.3072 vocab=0.6557 token=1.2581 | 6.13 steps/s | VRAM=28.46 Go [2026-07-18 01:33:21] step 566200/1000000 | loss=2.2109 vocab=0.9109 token=1.6209 | 6.15 steps/s | VRAM=28.52 Go [2026-07-18 01:33:29] step 566250/1000000 | loss=2.3168 vocab=0.7642 token=1.1759 | 6.14 steps/s | VRAM=28.86 Go [2026-07-18 01:33:37] step 566300/1000000 | loss=2.1956 vocab=1.1134 token=1.2660 | 6.12 steps/s | VRAM=28.77 Go [2026-07-18 01:33:45] step 566350/1000000 | loss=2.2200 vocab=1.0341 token=1.6674 | 6.13 steps/s | VRAM=29.50 Go [2026-07-18 01:33:53] step 566400/1000000 | loss=2.1925 vocab=0.4723 token=0.9611 | 6.19 steps/s | VRAM=27.96 Go [2026-07-18 01:34:01] step 566450/1000000 | loss=2.2184 vocab=0.9945 token=1.1629 | 6.14 steps/s | VRAM=32.90 Go [2026-07-18 01:34:10] step 566500/1000000 | loss=2.0781 vocab=0.5933 token=1.3515 | 6.12 steps/s | VRAM=32.73 Go [2026-07-18 01:34:18] step 566550/1000000 | loss=2.3641 vocab=0.7549 token=0.9563 | 6.14 steps/s | VRAM=27.93 Go [2026-07-18 01:34:26] step 566600/1000000 | loss=2.4131 vocab=1.1874 token=1.6840 | 6.16 steps/s | VRAM=28.89 Go [2026-07-18 01:34:34] step 566650/1000000 | loss=2.1863 vocab=0.6570 token=1.1236 | 6.14 steps/s | VRAM=31.10 Go [2026-07-18 01:34:42] step 566700/1000000 | loss=2.2828 vocab=0.7208 token=1.2406 | 6.13 steps/s | VRAM=29.27 Go [2026-07-18 01:34:50] step 566750/1000000 | loss=2.2469 vocab=1.6212 token=1.6293 | 6.16 steps/s | VRAM=31.85 Go [2026-07-18 01:34:58] step 566800/1000000 | loss=2.2667 vocab=1.0489 token=1.3307 | 6.12 steps/s | VRAM=28.53 Go [2026-07-18 01:35:07] step 566850/1000000 | loss=2.2657 vocab=0.8426 token=0.9284 | 6.16 steps/s | VRAM=31.38 Go [2026-07-18 01:35:15] step 566900/1000000 | loss=2.3215 vocab=1.5887 token=1.7582 | 6.15 steps/s | VRAM=31.86 Go [2026-07-18 01:35:23] step 566950/1000000 | loss=2.1481 vocab=1.3246 token=1.7820 | 6.14 steps/s | VRAM=28.78 Go [2026-07-18 01:35:31] step 567000/1000000 | loss=2.2624 vocab=0.8177 token=1.2716 | 6.13 steps/s | VRAM=26.49 Go [2026-07-18 01:35:39] step 567050/1000000 | loss=2.3817 vocab=1.4688 token=1.3259 | 6.12 steps/s | VRAM=29.24 Go [2026-07-18 01:35:47] step 567100/1000000 | loss=2.2432 vocab=0.7436 token=0.9974 | 6.14 steps/s | VRAM=31.67 Go [2026-07-18 01:35:55] step 567150/1000000 | loss=2.3322 vocab=0.5985 token=0.9627 | 6.14 steps/s | VRAM=32.05 Go [2026-07-18 01:36:04] step 567200/1000000 | loss=2.3029 vocab=0.7108 token=1.3685 | 6.14 steps/s | VRAM=30.95 Go [2026-07-18 01:36:12] step 567250/1000000 | loss=2.2330 vocab=1.4889 token=1.4758 | 6.11 steps/s | VRAM=27.17 Go [2026-07-18 01:36:20] step 567300/1000000 | loss=2.1279 vocab=0.5190 token=1.0996 | 6.13 steps/s | VRAM=28.32 Go [2026-07-18 01:36:28] step 567350/1000000 | loss=2.3127 vocab=0.7318 token=1.2576 | 6.14 steps/s | VRAM=32.28 Go [2026-07-18 01:36:36] step 567400/1000000 | loss=2.2276 vocab=0.6051 token=1.2569 | 6.11 steps/s | VRAM=31.58 Go [2026-07-18 01:36:44] step 567450/1000000 | loss=2.1599 vocab=0.6643 token=1.2021 | 6.13 steps/s | VRAM=31.97 Go [2026-07-18 01:36:53] step 567500/1000000 | loss=2.3087 vocab=0.7254 token=1.4771 | 6.10 steps/s | VRAM=26.33 Go [2026-07-18 01:37:01] step 567550/1000000 | loss=2.2989 vocab=0.6575 token=1.3107 | 6.12 steps/s | VRAM=31.94 Go [2026-07-18 01:37:09] step 567600/1000000 | loss=2.3526 vocab=0.4535 token=0.9167 | 6.13 steps/s | VRAM=25.57 Go [2026-07-18 01:37:17] step 567650/1000000 | loss=2.3170 vocab=1.5945 token=1.9638 | 6.17 steps/s | VRAM=30.94 Go [2026-07-18 01:37:25] step 567700/1000000 | loss=2.1570 vocab=0.9827 token=1.3174 | 6.11 steps/s | VRAM=24.88 Go [2026-07-18 01:37:33] step 567750/1000000 | loss=2.1760 vocab=0.8569 token=1.3922 | 6.13 steps/s | VRAM=25.93 Go [2026-07-18 01:37:42] step 567800/1000000 | loss=2.3226 vocab=1.0048 token=1.3825 | 6.15 steps/s | VRAM=28.87 Go [2026-07-18 01:37:50] step 567850/1000000 | loss=2.2341 vocab=0.7508 token=1.5710 | 6.16 steps/s | VRAM=28.80 Go [2026-07-18 01:37:58] step 567900/1000000 | loss=2.1557 vocab=0.7082 token=1.4272 | 6.14 steps/s | VRAM=31.91 Go [2026-07-18 01:38:06] step 567950/1000000 | loss=2.2040 vocab=0.9144 token=0.8710 | 6.13 steps/s | VRAM=32.42 Go [2026-07-18 01:38:14] step 568000/1000000 | loss=2.2692 vocab=0.8459 token=1.8782 | 6.13 steps/s | VRAM=28.72 Go [2026-07-18 01:38:22] step 568050/1000000 | loss=2.2165 vocab=0.9567 token=1.3319 | 6.15 steps/s | VRAM=31.75 Go [2026-07-18 01:38:30] step 568100/1000000 | loss=2.3486 vocab=1.2800 token=1.6749 | 6.12 steps/s | VRAM=27.39 Go [2026-07-18 01:38:39] step 568150/1000000 | loss=2.3619 vocab=1.5284 token=1.6004 | 6.11 steps/s | VRAM=28.18 Go [2026-07-18 01:38:47] step 568200/1000000 | loss=2.2506 vocab=1.3932 token=1.5666 | 6.12 steps/s | VRAM=23.36 Go [2026-07-18 01:38:55] step 568250/1000000 | loss=2.2023 vocab=0.6669 token=0.9943 | 6.16 steps/s | VRAM=31.42 Go [2026-07-18 01:39:03] step 568300/1000000 | loss=2.3148 vocab=0.7879 token=1.3332 | 6.11 steps/s | VRAM=26.76 Go [2026-07-18 01:39:11] step 568350/1000000 | loss=2.1843 vocab=1.1767 token=1.6433 | 6.12 steps/s | VRAM=30.16 Go [2026-07-18 01:39:19] step 568400/1000000 | loss=2.1780 vocab=1.0021 token=1.2924 | 6.12 steps/s | VRAM=29.70 Go [2026-07-18 01:39:27] step 568450/1000000 | loss=2.1956 vocab=1.0402 token=1.6023 | 6.17 steps/s | VRAM=29.12 Go [2026-07-18 01:39:36] step 568500/1000000 | loss=2.1587 vocab=0.8839 token=1.1131 | 6.12 steps/s | VRAM=31.88 Go [2026-07-18 01:39:44] step 568550/1000000 | loss=2.2470 vocab=0.9759 token=1.2147 | 6.10 steps/s | VRAM=28.48 Go [2026-07-18 01:39:52] step 568600/1000000 | loss=2.2352 vocab=0.6709 token=1.3141 | 6.14 steps/s | VRAM=31.39 Go [2026-07-18 01:40:00] step 568650/1000000 | loss=2.2248 vocab=0.8259 token=1.0326 | 6.14 steps/s | VRAM=32.84 Go [2026-07-18 01:40:08] step 568700/1000000 | loss=2.2080 vocab=0.7939 token=0.9315 | 6.15 steps/s | VRAM=28.04 Go [2026-07-18 01:40:16] step 568750/1000000 | loss=2.2234 vocab=0.7794 token=1.3175 | 6.16 steps/s | VRAM=32.48 Go [2026-07-18 01:40:24] step 568800/1000000 | loss=2.1821 vocab=0.5759 token=1.0303 | 6.16 steps/s | VRAM=28.32 Go [2026-07-18 01:40:33] step 568850/1000000 | loss=2.2094 vocab=0.7305 token=1.1137 | 6.15 steps/s | VRAM=32.88 Go [2026-07-18 01:40:41] step 568900/1000000 | loss=2.0978 vocab=0.6444 token=1.4595 | 6.14 steps/s | VRAM=28.59 Go [2026-07-18 01:40:49] step 568950/1000000 | loss=2.2363 vocab=0.9553 token=1.4206 | 6.13 steps/s | VRAM=28.75 Go [2026-07-18 01:40:57] step 569000/1000000 | loss=2.2829 vocab=1.3310 token=1.6350 | 6.12 steps/s | VRAM=32.51 Go [2026-07-18 01:41:05] step 569050/1000000 | loss=2.1804 vocab=0.8118 token=1.5218 | 6.15 steps/s | VRAM=27.88 Go [2026-07-18 01:41:13] step 569100/1000000 | loss=2.3325 vocab=1.3165 token=1.2854 | 6.14 steps/s | VRAM=31.51 Go [2026-07-18 01:41:21] step 569150/1000000 | loss=2.2440 vocab=1.0946 token=1.6963 | 6.12 steps/s | VRAM=29.36 Go [2026-07-18 01:41:30] step 569200/1000000 | loss=2.2503 vocab=1.5855 token=1.3596 | 6.15 steps/s | VRAM=28.98 Go [2026-07-18 01:41:38] step 569250/1000000 | loss=2.2418 vocab=0.6762 token=1.2346 | 6.12 steps/s | VRAM=32.75 Go [2026-07-18 01:41:46] step 569300/1000000 | loss=2.2676 vocab=0.6245 token=1.0841 | 6.12 steps/s | VRAM=27.99 Go [2026-07-18 01:41:54] step 569350/1000000 | loss=2.2024 vocab=1.3974 token=1.6089 | 6.15 steps/s | VRAM=28.48 Go [2026-07-18 01:42:02] step 569400/1000000 | loss=2.1887 vocab=0.4616 token=0.9562 | 6.13 steps/s | VRAM=25.71 Go [2026-07-18 01:42:10] step 569450/1000000 | loss=2.2324 vocab=1.0191 token=1.1344 | 6.15 steps/s | VRAM=32.49 Go [2026-07-18 01:42:18] step 569500/1000000 | loss=2.2113 vocab=1.2864 token=1.5462 | 6.17 steps/s | VRAM=31.81 Go [2026-07-18 01:42:27] step 569550/1000000 | loss=2.2823 vocab=1.6239 token=1.5728 | 6.13 steps/s | VRAM=28.86 Go [2026-07-18 01:42:35] step 569600/1000000 | loss=2.2042 vocab=1.5184 token=1.4034 | 6.14 steps/s | VRAM=26.60 Go [2026-07-18 01:42:43] step 569650/1000000 | loss=2.2745 vocab=0.8028 token=1.3184 | 6.13 steps/s | VRAM=28.55 Go [2026-07-18 01:42:51] step 569700/1000000 | loss=2.2414 vocab=1.3003 token=1.3619 | 6.16 steps/s | VRAM=32.99 Go [2026-07-18 01:42:59] step 569750/1000000 | loss=2.3030 vocab=1.0847 token=1.4666 | 6.13 steps/s | VRAM=27.91 Go [2026-07-18 01:43:07] step 569800/1000000 | loss=2.2722 vocab=1.2201 token=1.8931 | 6.13 steps/s | VRAM=32.65 Go [2026-07-18 01:43:16] step 569850/1000000 | loss=2.2317 vocab=0.8825 token=1.2860 | 6.11 steps/s | VRAM=29.00 Go [2026-07-18 01:43:24] step 569900/1000000 | loss=2.0835 vocab=0.4408 token=1.0571 | 6.11 steps/s | VRAM=26.74 Go [2026-07-18 01:43:32] step 569950/1000000 | loss=2.2586 vocab=0.7644 token=1.4196 | 6.14 steps/s | VRAM=28.91 Go [2026-07-18 01:43:40] step 570000/1000000 | loss=2.2028 vocab=0.9863 token=1.2857 | 6.12 steps/s | VRAM=28.61 Go [2026-07-18 01:43:48] step 570050/1000000 | loss=2.2224 vocab=0.8819 token=1.0045 | 6.14 steps/s | VRAM=28.85 Go [2026-07-18 01:43:56] step 570100/1000000 | loss=2.2947 vocab=0.6819 token=1.1776 | 6.13 steps/s | VRAM=28.19 Go [2026-07-18 01:44:05] step 570150/1000000 | loss=2.2201 vocab=0.7681 token=1.3849 | 6.13 steps/s | VRAM=32.36 Go [2026-07-18 01:44:13] step 570200/1000000 | loss=2.1573 vocab=1.0090 token=1.4476 | 6.11 steps/s | VRAM=26.12 Go [2026-07-18 01:44:21] step 570250/1000000 | loss=2.2194 vocab=0.8637 token=1.6574 | 6.14 steps/s | VRAM=26.70 Go [2026-07-18 01:44:29] step 570300/1000000 | loss=2.3426 vocab=0.8745 token=1.2835 | 6.14 steps/s | VRAM=28.50 Go [2026-07-18 01:44:37] step 570350/1000000 | loss=2.2911 vocab=1.0855 token=1.2656 | 6.16 steps/s | VRAM=25.22 Go [2026-07-18 01:44:45] step 570400/1000000 | loss=2.2143 vocab=0.9292 token=1.1735 | 6.14 steps/s | VRAM=26.53 Go [2026-07-18 01:44:53] step 570450/1000000 | loss=2.2254 vocab=0.8982 token=1.1554 | 6.10 steps/s | VRAM=28.93 Go [2026-07-18 01:45:02] step 570500/1000000 | loss=2.1703 vocab=0.5714 token=0.9768 | 6.13 steps/s | VRAM=31.75 Go [2026-07-18 01:45:10] step 570550/1000000 | loss=2.2595 vocab=1.2144 token=1.5918 | 6.10 steps/s | VRAM=29.40 Go [2026-07-18 01:45:18] step 570600/1000000 | loss=2.1762 vocab=0.7520 token=1.1000 | 6.13 steps/s | VRAM=32.52 Go [2026-07-18 01:45:26] step 570650/1000000 | loss=2.1995 vocab=0.5858 token=1.0819 | 6.14 steps/s | VRAM=28.41 Go [2026-07-18 01:45:34] step 570700/1000000 | loss=2.2893 vocab=0.6021 token=1.2017 | 6.13 steps/s | VRAM=29.25 Go [2026-07-18 01:45:42] step 570750/1000000 | loss=2.2415 vocab=0.8442 token=1.1128 | 6.08 steps/s | VRAM=26.25 Go [2026-07-18 01:45:51] step 570800/1000000 | loss=2.2441 vocab=0.9848 token=1.3371 | 6.16 steps/s | VRAM=32.26 Go [2026-07-18 01:45:59] step 570850/1000000 | loss=2.1451 vocab=0.7634 token=1.0909 | 6.11 steps/s | VRAM=28.76 Go [2026-07-18 01:46:07] step 570900/1000000 | loss=2.1599 vocab=0.5296 token=0.9066 | 6.14 steps/s | VRAM=28.27 Go [2026-07-18 01:46:15] step 570950/1000000 | loss=2.1083 vocab=0.7284 token=1.1493 | 6.13 steps/s | VRAM=26.33 Go [2026-07-18 01:46:23] step 571000/1000000 | loss=2.1623 vocab=1.0504 token=1.2012 | 6.14 steps/s | VRAM=28.51 Go [2026-07-18 01:46:31] step 571050/1000000 | loss=2.2332 vocab=1.2313 token=1.3245 | 6.13 steps/s | VRAM=28.94 Go [2026-07-18 01:46:40] step 571100/1000000 | loss=2.1154 vocab=0.6774 token=1.4583 | 6.08 steps/s | VRAM=26.42 Go [2026-07-18 01:46:48] step 571150/1000000 | loss=2.3172 vocab=1.3509 token=1.6756 | 6.11 steps/s | VRAM=24.45 Go [2026-07-18 01:46:56] step 571200/1000000 | loss=2.2470 vocab=0.7047 token=1.4856 | 6.13 steps/s | VRAM=29.60 Go [2026-07-18 01:47:04] step 571250/1000000 | loss=2.3334 vocab=1.0396 token=1.7263 | 6.12 steps/s | VRAM=28.36 Go [2026-07-18 01:47:12] step 571300/1000000 | loss=2.1864 vocab=1.2754 token=1.4744 | 6.13 steps/s | VRAM=29.47 Go [2026-07-18 01:47:20] step 571350/1000000 | loss=2.2119 vocab=1.1660 token=1.4693 | 6.09 steps/s | VRAM=29.00 Go [2026-07-18 01:47:29] step 571400/1000000 | loss=2.1200 vocab=0.4673 token=0.9961 | 6.11 steps/s | VRAM=25.04 Go [2026-07-18 01:47:37] step 571450/1000000 | loss=2.1545 vocab=1.2113 token=1.1689 | 6.14 steps/s | VRAM=29.38 Go [2026-07-18 01:47:45] step 571500/1000000 | loss=2.2900 vocab=0.7154 token=1.5258 | 6.13 steps/s | VRAM=31.24 Go [2026-07-18 01:47:53] step 571550/1000000 | loss=2.2126 vocab=0.5178 token=1.0920 | 6.12 steps/s | VRAM=29.89 Go [2026-07-18 01:48:01] step 571600/1000000 | loss=2.2469 vocab=0.6324 token=1.0925 | 6.13 steps/s | VRAM=25.41 Go [2026-07-18 01:48:09] step 571650/1000000 | loss=2.1462 vocab=1.1101 token=1.6201 | 6.11 steps/s | VRAM=28.83 Go [2026-07-18 01:48:18] step 571700/1000000 | loss=2.1835 vocab=0.9264 token=1.4070 | 6.13 steps/s | VRAM=32.48 Go [2026-07-18 01:48:26] step 571750/1000000 | loss=2.2647 vocab=0.9916 token=1.3855 | 6.15 steps/s | VRAM=31.90 Go [2026-07-18 01:48:34] step 571800/1000000 | loss=2.1745 vocab=0.5981 token=1.1233 | 6.15 steps/s | VRAM=25.70 Go [2026-07-18 01:48:42] step 571850/1000000 | loss=2.3750 vocab=0.8414 token=0.9943 | 6.14 steps/s | VRAM=31.99 Go [2026-07-18 01:48:50] step 571900/1000000 | loss=2.2468 vocab=0.7628 token=1.7108 | 6.13 steps/s | VRAM=28.13 Go [2026-07-18 01:48:58] step 571950/1000000 | loss=2.2609 vocab=0.8159 token=1.5707 | 6.15 steps/s | VRAM=26.07 Go [2026-07-18 01:49:06] step 572000/1000000 | loss=2.2485 vocab=0.7983 token=1.3257 | 6.12 steps/s | VRAM=31.87 Go [2026-07-18 01:49:15] step 572050/1000000 | loss=2.1433 vocab=1.2200 token=1.5239 | 6.09 steps/s | VRAM=26.18 Go [2026-07-18 01:49:23] step 572100/1000000 | loss=2.3095 vocab=1.3452 token=1.9928 | 6.14 steps/s | VRAM=25.44 Go [2026-07-18 01:49:31] step 572150/1000000 | loss=2.1934 vocab=1.0233 token=1.0039 | 6.13 steps/s | VRAM=28.21 Go [2026-07-18 01:49:39] step 572200/1000000 | loss=2.2939 vocab=1.3218 token=1.7746 | 6.15 steps/s | VRAM=27.33 Go [2026-07-18 01:49:47] step 572250/1000000 | loss=2.1712 vocab=0.5514 token=1.0877 | 6.12 steps/s | VRAM=27.67 Go [2026-07-18 01:49:55] step 572300/1000000 | loss=2.2520 vocab=0.8190 token=1.0631 | 6.14 steps/s | VRAM=25.97 Go [2026-07-18 01:50:04] step 572350/1000000 | loss=2.2196 vocab=0.8587 token=1.1728 | 6.11 steps/s | VRAM=23.26 Go [2026-07-18 01:50:12] step 572400/1000000 | loss=2.3037 vocab=0.4752 token=1.0237 | 6.13 steps/s | VRAM=28.38 Go [2026-07-18 01:50:20] step 572450/1000000 | loss=2.1042 vocab=0.5704 token=1.2434 | 6.10 steps/s | VRAM=30.11 Go [2026-07-18 01:50:28] step 572500/1000000 | loss=2.3632 vocab=0.8945 token=1.2520 | 6.18 steps/s | VRAM=30.50 Go [2026-07-18 01:50:36] step 572550/1000000 | loss=2.2059 vocab=0.6637 token=1.0163 | 6.13 steps/s | VRAM=29.41 Go [2026-07-18 01:50:44] step 572600/1000000 | loss=2.1715 vocab=0.6509 token=1.3137 | 6.14 steps/s | VRAM=32.25 Go [2026-07-18 01:50:52] step 572650/1000000 | loss=2.3192 vocab=0.6998 token=1.4898 | 6.18 steps/s | VRAM=31.97 Go [2026-07-18 01:51:01] step 572700/1000000 | loss=2.2856 vocab=1.0199 token=1.4535 | 6.15 steps/s | VRAM=32.29 Go [2026-07-18 01:51:09] step 572750/1000000 | loss=2.1239 vocab=0.9007 token=1.1866 | 6.15 steps/s | VRAM=29.43 Go [2026-07-18 01:51:17] step 572800/1000000 | loss=2.2168 vocab=1.1135 token=1.3870 | 6.15 steps/s | VRAM=29.41 Go [2026-07-18 01:51:25] step 572850/1000000 | loss=2.1527 vocab=0.7218 token=1.0460 | 6.16 steps/s | VRAM=32.34 Go [2026-07-18 01:51:33] step 572900/1000000 | loss=2.2239 vocab=0.9166 token=1.2238 | 6.13 steps/s | VRAM=29.88 Go [2026-07-18 01:51:41] step 572950/1000000 | loss=2.2706 vocab=1.3191 token=1.2594 | 6.16 steps/s | VRAM=28.10 Go [2026-07-18 01:51:49] step 573000/1000000 | loss=2.1512 vocab=0.8023 token=1.1249 | 6.11 steps/s | VRAM=29.33 Go [2026-07-18 01:51:58] step 573050/1000000 | loss=2.1724 vocab=0.9977 token=1.3640 | 6.11 steps/s | VRAM=29.41 Go [2026-07-18 01:52:06] step 573100/1000000 | loss=2.1957 vocab=1.1572 token=1.2794 | 6.13 steps/s | VRAM=28.32 Go [2026-07-18 01:52:14] step 573150/1000000 | loss=2.1656 vocab=1.0544 token=1.5440 | 6.12 steps/s | VRAM=24.32 Go [2026-07-18 01:52:22] step 573200/1000000 | loss=2.2522 vocab=1.3986 token=1.3440 | 6.17 steps/s | VRAM=31.45 Go [2026-07-18 01:52:30] step 573250/1000000 | loss=2.2774 vocab=1.0589 token=1.2842 | 6.13 steps/s | VRAM=31.81 Go [2026-07-18 01:52:38] step 573300/1000000 | loss=2.2082 vocab=0.7008 token=1.1122 | 6.11 steps/s | VRAM=29.50 Go [2026-07-18 01:52:47] step 573350/1000000 | loss=2.2870 vocab=0.7346 token=1.4909 | 6.15 steps/s | VRAM=25.08 Go [2026-07-18 01:52:55] step 573400/1000000 | loss=2.2442 vocab=0.8524 token=1.4807 | 6.14 steps/s | VRAM=32.55 Go [2026-07-18 01:53:03] step 573450/1000000 | loss=2.3323 vocab=0.8173 token=1.2573 | 6.12 steps/s | VRAM=24.89 Go [2026-07-18 01:53:11] step 573500/1000000 | loss=2.2319 vocab=1.1384 token=1.2904 | 6.16 steps/s | VRAM=25.44 Go [2026-07-18 01:53:19] step 573550/1000000 | loss=2.1759 vocab=0.8526 token=1.4544 | 6.15 steps/s | VRAM=27.84 Go [2026-07-18 01:53:27] step 573600/1000000 | loss=2.2742 vocab=0.7524 token=1.3312 | 6.13 steps/s | VRAM=28.64 Go [2026-07-18 01:53:35] step 573650/1000000 | loss=2.2586 vocab=0.6940 token=1.1507 | 6.15 steps/s | VRAM=28.21 Go [2026-07-18 01:53:44] step 573700/1000000 | loss=2.2116 vocab=0.3992 token=0.9941 | 6.11 steps/s | VRAM=28.68 Go [2026-07-18 01:53:52] step 573750/1000000 | loss=2.1583 vocab=0.8753 token=1.2264 | 6.15 steps/s | VRAM=28.72 Go [2026-07-18 01:54:00] step 573800/1000000 | loss=2.3075 vocab=0.8568 token=1.3320 | 6.12 steps/s | VRAM=28.41 Go [2026-07-18 01:54:08] step 573850/1000000 | loss=2.3465 vocab=0.8221 token=1.1162 | 6.15 steps/s | VRAM=31.69 Go [2026-07-18 01:54:16] step 573900/1000000 | loss=2.3264 vocab=1.0015 token=1.5716 | 6.15 steps/s | VRAM=25.23 Go [2026-07-18 01:54:24] step 573950/1000000 | loss=2.1759 vocab=0.8641 token=1.1801 | 6.16 steps/s | VRAM=28.27 Go [2026-07-18 01:54:32] step 574000/1000000 | loss=2.2037 vocab=1.0755 token=1.4637 | 6.11 steps/s | VRAM=29.34 Go [2026-07-18 01:54:41] step 574050/1000000 | loss=2.2723 vocab=0.5888 token=0.9317 | 6.14 steps/s | VRAM=29.09 Go [2026-07-18 01:54:49] step 574100/1000000 | loss=2.3213 vocab=1.2271 token=1.2586 | 6.14 steps/s | VRAM=28.84 Go [2026-07-18 01:54:57] step 574150/1000000 | loss=2.2446 vocab=0.8139 token=1.2328 | 6.15 steps/s | VRAM=28.28 Go [2026-07-18 01:55:05] step 574200/1000000 | loss=2.2882 vocab=0.6797 token=1.6739 | 6.15 steps/s | VRAM=28.81 Go [2026-07-18 01:55:13] step 574250/1000000 | loss=2.2325 vocab=0.7629 token=1.4736 | 6.13 steps/s | VRAM=32.67 Go [2026-07-18 01:55:21] step 574300/1000000 | loss=2.3161 vocab=1.9886 token=1.7837 | 6.12 steps/s | VRAM=32.60 Go [2026-07-18 01:55:29] step 574350/1000000 | loss=2.1718 vocab=0.8410 token=1.4300 | 6.13 steps/s | VRAM=29.60 Go [2026-07-18 01:55:38] step 574400/1000000 | loss=2.2124 vocab=0.7858 token=1.2562 | 6.17 steps/s | VRAM=31.25 Go [2026-07-18 01:55:46] step 574450/1000000 | loss=2.1683 vocab=0.6291 token=0.9895 | 6.16 steps/s | VRAM=28.94 Go [2026-07-18 01:55:54] step 574500/1000000 | loss=2.2302 vocab=0.9486 token=1.1956 | 6.12 steps/s | VRAM=25.86 Go [2026-07-18 01:56:02] step 574550/1000000 | loss=2.2496 vocab=1.0145 token=1.3458 | 6.15 steps/s | VRAM=31.24 Go [2026-07-18 01:56:10] step 574600/1000000 | loss=2.2623 vocab=1.4688 token=1.3720 | 6.13 steps/s | VRAM=28.29 Go [2026-07-18 01:56:18] step 574650/1000000 | loss=2.3958 vocab=0.7096 token=1.1573 | 6.11 steps/s | VRAM=26.19 Go [2026-07-18 01:56:26] step 574700/1000000 | loss=2.2305 vocab=0.9422 token=1.3414 | 6.10 steps/s | VRAM=26.90 Go [2026-07-18 01:56:35] step 574750/1000000 | loss=2.1281 vocab=0.9285 token=1.4589 | 6.13 steps/s | VRAM=28.40 Go [2026-07-18 01:56:43] step 574800/1000000 | loss=2.2184 vocab=1.3061 token=1.6709 | 6.15 steps/s | VRAM=28.27 Go [2026-07-18 01:56:51] step 574850/1000000 | loss=2.1882 vocab=0.7440 token=1.5359 | 6.14 steps/s | VRAM=32.80 Go [2026-07-18 01:56:59] step 574900/1000000 | loss=2.2191 vocab=0.8498 token=1.5112 | 6.12 steps/s | VRAM=29.22 Go [2026-07-18 01:57:07] step 574950/1000000 | loss=2.1941 vocab=0.8403 token=1.2227 | 6.13 steps/s | VRAM=31.73 Go [2026-07-18 01:57:15] step 575000/1000000 | loss=2.2932 vocab=1.0131 token=1.5015 | 6.12 steps/s | VRAM=28.99 Go [2026-07-18 01:57:23] step 575050/1000000 | loss=2.1542 vocab=0.4434 token=0.7009 | 6.17 steps/s | VRAM=30.91 Go [2026-07-18 01:57:32] step 575100/1000000 | loss=2.3265 vocab=0.4773 token=1.1330 | 6.09 steps/s | VRAM=25.89 Go [2026-07-18 01:57:40] step 575150/1000000 | loss=2.2978 vocab=1.0387 token=1.6730 | 6.14 steps/s | VRAM=32.10 Go [2026-07-18 01:57:48] step 575200/1000000 | loss=2.1508 vocab=0.7992 token=1.4993 | 6.15 steps/s | VRAM=31.82 Go [2026-07-18 01:57:56] step 575250/1000000 | loss=2.2241 vocab=1.1888 token=0.8938 | 6.15 steps/s | VRAM=28.32 Go [2026-07-18 01:58:04] step 575300/1000000 | loss=2.2320 vocab=0.5285 token=1.2107 | 6.15 steps/s | VRAM=28.00 Go [2026-07-18 01:58:12] step 575350/1000000 | loss=2.2624 vocab=1.1346 token=1.4275 | 6.13 steps/s | VRAM=26.53 Go [2026-07-18 01:58:21] step 575400/1000000 | loss=2.2897 vocab=0.8543 token=1.4210 | 6.15 steps/s | VRAM=30.76 Go [2026-07-18 01:58:29] step 575450/1000000 | loss=2.2151 vocab=0.9763 token=1.3510 | 6.11 steps/s | VRAM=28.35 Go [2026-07-18 01:58:37] step 575500/1000000 | loss=2.1926 vocab=1.0098 token=1.5639 | 6.12 steps/s | VRAM=26.37 Go [2026-07-18 01:58:45] step 575550/1000000 | loss=2.2186 vocab=1.2439 token=1.3302 | 6.14 steps/s | VRAM=29.49 Go [2026-07-18 01:58:53] step 575600/1000000 | loss=2.2202 vocab=0.8555 token=1.0146 | 6.14 steps/s | VRAM=32.48 Go [2026-07-18 01:59:01] step 575650/1000000 | loss=2.0508 vocab=1.2855 token=1.5094 | 6.11 steps/s | VRAM=28.80 Go [2026-07-18 01:59:10] step 575700/1000000 | loss=2.2305 vocab=1.0459 token=1.6121 | 6.10 steps/s | VRAM=26.88 Go [2026-07-18 01:59:18] step 575750/1000000 | loss=2.1965 vocab=0.9526 token=1.4440 | 6.11 steps/s | VRAM=28.49 Go [2026-07-18 01:59:26] step 575800/1000000 | loss=2.1940 vocab=0.9923 token=1.5492 | 6.15 steps/s | VRAM=31.99 Go [2026-07-18 01:59:34] step 575850/1000000 | loss=2.1353 vocab=0.8586 token=1.3441 | 6.13 steps/s | VRAM=28.15 Go [2026-07-18 01:59:42] step 575900/1000000 | loss=2.2112 vocab=0.9593 token=1.4462 | 6.12 steps/s | VRAM=28.81 Go [2026-07-18 01:59:50] step 575950/1000000 | loss=2.1881 vocab=0.6499 token=1.3232 | 6.12 steps/s | VRAM=28.54 Go [2026-07-18 01:59:59] step 576000/1000000 | loss=2.1388 vocab=0.6919 token=1.2744 | 6.11 steps/s | VRAM=29.41 Go [2026-07-18 02:00:07] step 576050/1000000 | loss=2.2075 vocab=0.8410 token=1.2610 | 6.15 steps/s | VRAM=27.89 Go [2026-07-18 02:00:15] step 576100/1000000 | loss=2.3207 vocab=1.1718 token=1.3624 | 6.10 steps/s | VRAM=26.14 Go [2026-07-18 02:00:23] step 576150/1000000 | loss=2.2652 vocab=1.0681 token=1.5163 | 6.14 steps/s | VRAM=24.62 Go [2026-07-18 02:00:31] step 576200/1000000 | loss=2.1550 vocab=1.4960 token=1.8304 | 6.14 steps/s | VRAM=31.93 Go [2026-07-18 02:00:39] step 576250/1000000 | loss=2.2317 vocab=0.5876 token=1.3509 | 6.10 steps/s | VRAM=29.37 Go [2026-07-18 02:00:47] step 576300/1000000 | loss=2.1484 vocab=0.9207 token=1.2033 | 6.14 steps/s | VRAM=28.41 Go [2026-07-18 02:00:56] step 576350/1000000 | loss=2.1841 vocab=0.7130 token=1.5610 | 6.14 steps/s | VRAM=28.25 Go [2026-07-18 02:01:04] step 576400/1000000 | loss=2.2983 vocab=0.8799 token=1.1976 | 6.14 steps/s | VRAM=25.63 Go [2026-07-18 02:01:12] step 576450/1000000 | loss=2.2660 vocab=0.8742 token=1.2654 | 6.17 steps/s | VRAM=32.13 Go [2026-07-18 02:01:20] step 576500/1000000 | loss=2.4282 vocab=1.3636 token=1.6181 | 6.14 steps/s | VRAM=31.64 Go [2026-07-18 02:01:28] step 576550/1000000 | loss=2.1996 vocab=0.6789 token=1.2405 | 6.13 steps/s | VRAM=28.65 Go [2026-07-18 02:01:36] step 576600/1000000 | loss=2.2061 vocab=0.5891 token=1.0452 | 6.14 steps/s | VRAM=28.41 Go [2026-07-18 02:01:44] step 576650/1000000 | loss=2.1932 vocab=1.1231 token=0.9959 | 6.14 steps/s | VRAM=28.92 Go [2026-07-18 02:01:53] step 576700/1000000 | loss=2.3701 vocab=1.4954 token=1.5617 | 6.14 steps/s | VRAM=32.47 Go [2026-07-18 02:02:01] step 576750/1000000 | loss=2.0756 vocab=1.0259 token=1.0123 | 6.13 steps/s | VRAM=25.99 Go [2026-07-18 02:02:09] step 576800/1000000 | loss=2.1313 vocab=0.8519 token=1.0769 | 6.15 steps/s | VRAM=28.88 Go [2026-07-18 02:02:17] step 576850/1000000 | loss=2.1559 vocab=0.8508 token=1.1967 | 6.16 steps/s | VRAM=32.85 Go [2026-07-18 02:02:25] step 576900/1000000 | loss=2.1652 vocab=0.6854 token=0.7879 | 6.13 steps/s | VRAM=28.31 Go [2026-07-18 02:02:33] step 576950/1000000 | loss=2.1897 vocab=0.6245 token=0.8847 | 6.14 steps/s | VRAM=29.07 Go [2026-07-18 02:02:41] step 577000/1000000 | loss=2.1772 vocab=0.7253 token=1.2795 | 6.15 steps/s | VRAM=32.23 Go [2026-07-18 02:02:50] step 577050/1000000 | loss=2.2153 vocab=0.9797 token=1.2205 | 6.13 steps/s | VRAM=29.10 Go [2026-07-18 02:02:58] step 577100/1000000 | loss=2.1575 vocab=0.6848 token=1.2329 | 6.16 steps/s | VRAM=27.87 Go [2026-07-18 02:03:06] step 577150/1000000 | loss=2.1540 vocab=0.9525 token=1.6806 | 6.15 steps/s | VRAM=31.49 Go [2026-07-18 02:03:14] step 577200/1000000 | loss=2.2110 vocab=0.7724 token=0.9244 | 6.12 steps/s | VRAM=30.05 Go [2026-07-18 02:03:22] step 577250/1000000 | loss=2.2536 vocab=0.7323 token=1.4128 | 6.13 steps/s | VRAM=29.27 Go [2026-07-18 02:03:30] step 577300/1000000 | loss=2.3214 vocab=0.8468 token=1.4599 | 6.12 steps/s | VRAM=25.64 Go [2026-07-18 02:03:38] step 577350/1000000 | loss=2.2694 vocab=0.8019 token=1.4578 | 6.12 steps/s | VRAM=29.75 Go [2026-07-18 02:03:47] step 577400/1000000 | loss=2.1779 vocab=1.2705 token=1.3158 | 6.11 steps/s | VRAM=28.36 Go [2026-07-18 02:03:55] step 577450/1000000 | loss=2.2232 vocab=0.9427 token=1.1759 | 6.13 steps/s | VRAM=28.38 Go [2026-07-18 02:04:03] step 577500/1000000 | loss=2.2443 vocab=0.8686 token=0.9322 | 6.10 steps/s | VRAM=28.22 Go [2026-07-18 02:04:11] step 577550/1000000 | loss=2.2335 vocab=0.8511 token=1.3865 | 6.12 steps/s | VRAM=28.03 Go [2026-07-18 02:04:19] step 577600/1000000 | loss=2.2361 vocab=0.9847 token=1.4133 | 6.11 steps/s | VRAM=28.72 Go [2026-07-18 02:04:28] step 577650/1000000 | loss=2.2791 vocab=1.2305 token=1.4301 | 6.11 steps/s | VRAM=28.25 Go [2026-07-18 02:04:36] step 577700/1000000 | loss=2.3424 vocab=1.0250 token=1.4508 | 6.11 steps/s | VRAM=26.49 Go [2026-07-18 02:04:44] step 577750/1000000 | loss=2.2400 vocab=0.7620 token=1.3024 | 6.12 steps/s | VRAM=25.49 Go [2026-07-18 02:04:52] step 577800/1000000 | loss=2.2028 vocab=1.2830 token=1.5172 | 6.15 steps/s | VRAM=31.58 Go [2026-07-18 02:05:00] step 577850/1000000 | loss=2.2095 vocab=0.6476 token=1.3497 | 6.14 steps/s | VRAM=28.89 Go [2026-07-18 02:05:08] step 577900/1000000 | loss=2.1514 vocab=0.8434 token=1.5852 | 6.12 steps/s | VRAM=29.72 Go [2026-07-18 02:05:16] step 577950/1000000 | loss=2.2696 vocab=1.1259 token=1.8817 | 6.13 steps/s | VRAM=28.12 Go [2026-07-18 02:05:25] step 578000/1000000 | loss=2.2582 vocab=0.8851 token=1.2939 | 6.09 steps/s | VRAM=27.00 Go [2026-07-18 02:05:33] step 578050/1000000 | loss=2.2361 vocab=1.0470 token=1.5980 | 6.12 steps/s | VRAM=28.86 Go [2026-07-18 02:05:41] step 578100/1000000 | loss=2.2219 vocab=0.5710 token=1.2201 | 6.12 steps/s | VRAM=28.95 Go [2026-07-18 02:05:49] step 578150/1000000 | loss=2.1995 vocab=1.0390 token=1.9088 | 6.14 steps/s | VRAM=28.81 Go [2026-07-18 02:05:57] step 578200/1000000 | loss=2.2882 vocab=0.6912 token=1.1297 | 6.16 steps/s | VRAM=29.34 Go [2026-07-18 02:06:05] step 578250/1000000 | loss=2.2496 vocab=0.9130 token=0.9549 | 6.13 steps/s | VRAM=31.86 Go [2026-07-18 02:06:14] step 578300/1000000 | loss=2.1992 vocab=1.2196 token=1.6716 | 6.14 steps/s | VRAM=32.28 Go [2026-07-18 02:06:22] step 578350/1000000 | loss=2.2713 vocab=1.0597 token=1.4899 | 6.12 steps/s | VRAM=26.46 Go [2026-07-18 02:06:30] step 578400/1000000 | loss=2.0830 vocab=0.8413 token=1.0473 | 6.12 steps/s | VRAM=29.22 Go [2026-07-18 02:06:38] step 578450/1000000 | loss=2.1932 vocab=0.8917 token=1.5257 | 6.12 steps/s | VRAM=28.66 Go [2026-07-18 02:06:46] step 578500/1000000 | loss=2.1714 vocab=0.9658 token=1.0842 | 6.14 steps/s | VRAM=32.25 Go [2026-07-18 02:06:54] step 578550/1000000 | loss=2.2318 vocab=0.6565 token=1.2132 | 6.14 steps/s | VRAM=29.53 Go [2026-07-18 02:07:03] step 578600/1000000 | loss=2.2880 vocab=0.6300 token=1.1336 | 6.13 steps/s | VRAM=31.20 Go [2026-07-18 02:07:11] step 578650/1000000 | loss=2.1602 vocab=0.7291 token=1.2268 | 6.14 steps/s | VRAM=32.28 Go [2026-07-18 02:07:19] step 578700/1000000 | loss=2.2447 vocab=0.6803 token=1.3277 | 6.14 steps/s | VRAM=28.14 Go [2026-07-18 02:07:27] step 578750/1000000 | loss=2.2091 vocab=0.8671 token=1.4878 | 6.13 steps/s | VRAM=28.62 Go [2026-07-18 02:07:35] step 578800/1000000 | loss=2.1848 vocab=1.2971 token=1.5700 | 6.14 steps/s | VRAM=29.06 Go [2026-07-18 02:07:43] step 578850/1000000 | loss=2.2157 vocab=0.6191 token=1.1912 | 6.13 steps/s | VRAM=32.73 Go [2026-07-18 02:07:51] step 578900/1000000 | loss=2.2782 vocab=0.5955 token=1.6826 | 6.11 steps/s | VRAM=29.59 Go [2026-07-18 02:08:00] step 578950/1000000 | loss=2.2325 vocab=0.9516 token=1.5217 | 6.12 steps/s | VRAM=28.47 Go [2026-07-18 02:08:08] step 579000/1000000 | loss=2.2177 vocab=1.0496 token=1.3280 | 6.11 steps/s | VRAM=25.76 Go [2026-07-18 02:08:16] step 579050/1000000 | loss=2.1984 vocab=0.7785 token=1.0581 | 6.14 steps/s | VRAM=32.96 Go [2026-07-18 02:08:24] step 579100/1000000 | loss=2.2912 vocab=1.0144 token=1.2059 | 6.11 steps/s | VRAM=29.38 Go [2026-07-18 02:08:32] step 579150/1000000 | loss=2.2380 vocab=0.7773 token=1.1669 | 6.12 steps/s | VRAM=32.80 Go [2026-07-18 02:08:40] step 579200/1000000 | loss=2.2952 vocab=0.7152 token=1.1912 | 6.14 steps/s | VRAM=32.67 Go [2026-07-18 02:08:49] step 579250/1000000 | loss=2.1821 vocab=0.6876 token=1.0409 | 6.15 steps/s | VRAM=31.26 Go [2026-07-18 02:08:57] step 579300/1000000 | loss=2.2221 vocab=0.7869 token=0.7948 | 6.13 steps/s | VRAM=28.67 Go [2026-07-18 02:09:05] step 579350/1000000 | loss=2.1055 vocab=1.6728 token=1.7466 | 6.11 steps/s | VRAM=26.22 Go [2026-07-18 02:09:13] step 579400/1000000 | loss=2.2556 vocab=0.6465 token=1.0229 | 6.14 steps/s | VRAM=29.22 Go [2026-07-18 02:09:21] step 579450/1000000 | loss=2.2412 vocab=0.9860 token=1.2801 | 6.13 steps/s | VRAM=32.54 Go [2026-07-18 02:09:29] step 579500/1000000 | loss=2.2862 vocab=0.8019 token=1.0384 | 6.10 steps/s | VRAM=28.63 Go [2026-07-18 02:09:38] step 579550/1000000 | loss=2.2160 vocab=1.2460 token=1.5030 | 6.09 steps/s | VRAM=26.56 Go [2026-07-18 02:09:46] step 579600/1000000 | loss=2.1865 vocab=0.6092 token=1.5425 | 6.12 steps/s | VRAM=32.58 Go [2026-07-18 02:09:54] step 579650/1000000 | loss=2.2151 vocab=1.6593 token=1.7064 | 6.11 steps/s | VRAM=29.03 Go [2026-07-18 02:10:02] step 579700/1000000 | loss=2.2329 vocab=1.0996 token=1.6218 | 6.13 steps/s | VRAM=28.68 Go [2026-07-18 02:10:10] step 579750/1000000 | loss=2.2833 vocab=0.3987 token=0.7755 | 6.12 steps/s | VRAM=28.55 Go [2026-07-18 02:10:18] step 579800/1000000 | loss=2.0952 vocab=1.1669 token=1.8932 | 6.15 steps/s | VRAM=31.38 Go [2026-07-18 02:10:27] step 579850/1000000 | loss=2.1779 vocab=0.9197 token=1.0277 | 6.14 steps/s | VRAM=28.86 Go [2026-07-18 02:10:35] step 579900/1000000 | loss=2.2466 vocab=1.2126 token=1.6669 | 6.11 steps/s | VRAM=29.99 Go [2026-07-18 02:10:43] step 579950/1000000 | loss=2.0909 vocab=1.1511 token=1.5907 | 6.14 steps/s | VRAM=31.61 Go [2026-07-18 02:10:51] step 580000/1000000 | loss=2.1777 vocab=1.4601 token=1.5348 | 6.11 steps/s | VRAM=28.39 Go [2026-07-18 02:10:59] step 580050/1000000 | loss=2.2408 vocab=1.2516 token=1.2949 | 6.16 steps/s | VRAM=32.85 Go [2026-07-18 02:11:07] step 580100/1000000 | loss=2.2350 vocab=1.1343 token=0.9258 | 6.11 steps/s | VRAM=31.96 Go [2026-07-18 02:11:16] step 580150/1000000 | loss=2.1736 vocab=0.9515 token=1.2922 | 6.14 steps/s | VRAM=32.12 Go [2026-07-18 02:11:24] step 580200/1000000 | loss=2.3541 vocab=0.8822 token=1.2263 | 6.12 steps/s | VRAM=32.72 Go [2026-07-18 02:11:32] step 580250/1000000 | loss=2.1408 vocab=0.8138 token=0.9616 | 6.13 steps/s | VRAM=25.40 Go [2026-07-18 02:11:40] step 580300/1000000 | loss=2.1828 vocab=1.0017 token=1.2468 | 6.13 steps/s | VRAM=26.89 Go [2026-07-18 02:11:48] step 580350/1000000 | loss=2.3623 vocab=1.4317 token=1.2257 | 6.11 steps/s | VRAM=26.34 Go [2026-07-18 02:11:56] step 580400/1000000 | loss=2.3191 vocab=0.9321 token=1.4327 | 6.13 steps/s | VRAM=28.43 Go [2026-07-18 02:12:05] step 580450/1000000 | loss=2.2994 vocab=2.0174 token=1.3126 | 6.12 steps/s | VRAM=29.38 Go [2026-07-18 02:12:13] step 580500/1000000 | loss=2.2944 vocab=1.3498 token=1.5011 | 6.12 steps/s | VRAM=29.23 Go [2026-07-18 02:12:21] step 580550/1000000 | loss=2.2568 vocab=1.4825 token=1.2159 | 6.15 steps/s | VRAM=32.23 Go [2026-07-18 02:12:29] step 580600/1000000 | loss=2.2462 vocab=1.2524 token=1.5397 | 6.10 steps/s | VRAM=29.98 Go [2026-07-18 02:12:37] step 580650/1000000 | loss=2.1791 vocab=0.5286 token=1.3022 | 6.13 steps/s | VRAM=29.29 Go [2026-07-18 02:12:45] step 580700/1000000 | loss=2.1618 vocab=0.4299 token=1.1711 | 6.09 steps/s | VRAM=24.79 Go [2026-07-18 02:12:54] step 580750/1000000 | loss=2.2474 vocab=0.9553 token=1.5529 | 6.13 steps/s | VRAM=28.62 Go [2026-07-18 02:13:02] step 580800/1000000 | loss=2.2551 vocab=1.1340 token=1.2572 | 6.12 steps/s | VRAM=32.95 Go [2026-07-18 02:13:10] step 580850/1000000 | loss=2.2491 vocab=1.2317 token=1.5162 | 6.13 steps/s | VRAM=28.05 Go [2026-07-18 02:13:18] step 580900/1000000 | loss=2.1877 vocab=0.7740 token=1.0394 | 6.13 steps/s | VRAM=31.58 Go [2026-07-18 02:13:26] step 580950/1000000 | loss=2.1879 vocab=0.8772 token=1.1923 | 6.15 steps/s | VRAM=31.37 Go [2026-07-18 02:13:34] step 581000/1000000 | loss=2.1673 vocab=0.6606 token=0.9672 | 6.12 steps/s | VRAM=32.89 Go [2026-07-18 02:13:43] step 581050/1000000 | loss=2.1919 vocab=0.6082 token=1.1121 | 6.09 steps/s | VRAM=31.07 Go [2026-07-18 02:13:51] step 581100/1000000 | loss=2.1778 vocab=1.0303 token=1.4328 | 6.14 steps/s | VRAM=28.36 Go [2026-07-18 02:13:59] step 581150/1000000 | loss=2.2906 vocab=1.4010 token=1.6197 | 6.13 steps/s | VRAM=32.42 Go [2026-07-18 02:14:07] step 581200/1000000 | loss=2.2863 vocab=1.5759 token=1.6998 | 6.11 steps/s | VRAM=24.91 Go [2026-07-18 02:14:15] step 581250/1000000 | loss=2.3243 vocab=0.9084 token=1.8961 | 6.14 steps/s | VRAM=25.14 Go [2026-07-18 02:14:23] step 581300/1000000 | loss=2.2757 vocab=1.1276 token=1.0596 | 6.13 steps/s | VRAM=25.84 Go [2026-07-18 02:14:32] step 581350/1000000 | loss=2.2059 vocab=0.6957 token=1.3362 | 6.12 steps/s | VRAM=29.19 Go [2026-07-18 02:14:40] step 581400/1000000 | loss=2.2018 vocab=0.8022 token=1.5818 | 6.13 steps/s | VRAM=29.28 Go [2026-07-18 02:14:48] step 581450/1000000 | loss=2.2650 vocab=1.0124 token=1.4466 | 6.11 steps/s | VRAM=32.36 Go [2026-07-18 02:14:56] step 581500/1000000 | loss=2.3286 vocab=0.8354 token=1.8582 | 6.12 steps/s | VRAM=25.25 Go [2026-07-18 02:15:04] step 581550/1000000 | loss=2.2131 vocab=1.4418 token=1.7927 | 6.13 steps/s | VRAM=29.10 Go [2026-07-18 02:15:12] step 581600/1000000 | loss=2.2617 vocab=0.9905 token=1.3790 | 6.11 steps/s | VRAM=28.93 Go [2026-07-18 02:15:21] step 581650/1000000 | loss=2.1638 vocab=1.8602 token=1.8493 | 6.13 steps/s | VRAM=24.49 Go [2026-07-18 02:15:29] step 581700/1000000 | loss=2.2046 vocab=0.6532 token=1.3691 | 6.13 steps/s | VRAM=23.61 Go [2026-07-18 02:15:37] step 581750/1000000 | loss=2.2156 vocab=1.6283 token=1.5468 | 6.15 steps/s | VRAM=29.13 Go [2026-07-18 02:15:45] step 581800/1000000 | loss=2.2273 vocab=0.6710 token=1.5542 | 6.15 steps/s | VRAM=25.43 Go [2026-07-18 02:15:53] step 581850/1000000 | loss=2.2531 vocab=0.9214 token=1.6613 | 6.11 steps/s | VRAM=29.50 Go [2026-07-18 02:16:01] step 581900/1000000 | loss=2.2555 vocab=0.9282 token=1.2479 | 6.10 steps/s | VRAM=27.72 Go [2026-07-18 02:16:10] step 581950/1000000 | loss=2.2540 vocab=0.8499 token=1.6178 | 6.11 steps/s | VRAM=32.88 Go [2026-07-18 02:16:18] step 582000/1000000 | loss=2.1909 vocab=0.7157 token=1.2513 | 6.15 steps/s | VRAM=28.66 Go [2026-07-18 02:16:26] step 582050/1000000 | loss=2.3901 vocab=1.1539 token=1.4912 | 6.14 steps/s | VRAM=26.09 Go [2026-07-18 02:16:34] step 582100/1000000 | loss=2.1719 vocab=0.8218 token=1.2925 | 6.15 steps/s | VRAM=32.27 Go [2026-07-18 02:16:42] step 582150/1000000 | loss=2.1966 vocab=0.6377 token=1.4092 | 6.11 steps/s | VRAM=30.25 Go [2026-07-18 02:16:50] step 582200/1000000 | loss=2.1401 vocab=0.9928 token=1.0665 | 6.12 steps/s | VRAM=28.23 Go [2026-07-18 02:16:58] step 582250/1000000 | loss=2.2055 vocab=0.7812 token=1.1102 | 6.09 steps/s | VRAM=32.92 Go [2026-07-18 02:17:07] step 582300/1000000 | loss=2.4007 vocab=0.8240 token=1.0708 | 6.14 steps/s | VRAM=31.44 Go [2026-07-18 02:17:15] step 582350/1000000 | loss=2.2144 vocab=0.8704 token=1.4084 | 6.10 steps/s | VRAM=27.23 Go [2026-07-18 02:17:23] step 582400/1000000 | loss=2.2455 vocab=1.0603 token=1.3628 | 6.12 steps/s | VRAM=28.65 Go [2026-07-18 02:17:31] step 582450/1000000 | loss=2.2860 vocab=2.0708 token=1.5792 | 6.11 steps/s | VRAM=28.77 Go [2026-07-18 02:17:39] step 582500/1000000 | loss=2.2483 vocab=0.7889 token=1.0746 | 6.11 steps/s | VRAM=29.14 Go [2026-07-18 02:17:48] step 582550/1000000 | loss=2.2238 vocab=0.8939 token=1.3620 | 6.12 steps/s | VRAM=32.47 Go [2026-07-18 02:17:56] step 582600/1000000 | loss=2.2451 vocab=1.0205 token=1.3241 | 6.13 steps/s | VRAM=28.39 Go [2026-07-18 02:18:04] step 582650/1000000 | loss=2.1942 vocab=0.7937 token=1.2615 | 6.13 steps/s | VRAM=26.13 Go [2026-07-18 02:18:12] step 582700/1000000 | loss=2.2480 vocab=0.4490 token=1.1198 | 6.10 steps/s | VRAM=26.67 Go [2026-07-18 02:18:20] step 582750/1000000 | loss=2.2748 vocab=0.7496 token=1.2130 | 6.14 steps/s | VRAM=32.66 Go [2026-07-18 02:18:28] step 582800/1000000 | loss=2.2404 vocab=1.2262 token=1.6435 | 6.10 steps/s | VRAM=26.50 Go [2026-07-18 02:18:37] step 582850/1000000 | loss=2.1764 vocab=0.9198 token=1.6840 | 6.14 steps/s | VRAM=28.16 Go [2026-07-18 02:18:45] step 582900/1000000 | loss=2.2959 vocab=0.3646 token=1.1134 | 6.13 steps/s | VRAM=32.67 Go [2026-07-18 02:18:53] step 582950/1000000 | loss=2.1727 vocab=0.6023 token=1.2579 | 6.09 steps/s | VRAM=27.10 Go [2026-07-18 02:19:01] step 583000/1000000 | loss=2.2721 vocab=0.9537 token=1.5967 | 6.14 steps/s | VRAM=28.25 Go [2026-07-18 02:19:09] step 583050/1000000 | loss=2.2441 vocab=0.3341 token=0.8739 | 6.11 steps/s | VRAM=26.52 Go [2026-07-18 02:19:17] step 583100/1000000 | loss=2.1998 vocab=0.8891 token=1.2822 | 6.14 steps/s | VRAM=30.22 Go [2026-07-18 02:19:26] step 583150/1000000 | loss=2.2308 vocab=1.0982 token=1.2928 | 6.14 steps/s | VRAM=28.61 Go [2026-07-18 02:19:34] step 583200/1000000 | loss=2.2950 vocab=1.0734 token=1.5521 | 6.12 steps/s | VRAM=26.34 Go [2026-07-18 02:19:42] step 583250/1000000 | loss=2.1073 vocab=0.9439 token=1.1168 | 6.13 steps/s | VRAM=28.30 Go [2026-07-18 02:19:50] step 583300/1000000 | loss=2.2110 vocab=0.6046 token=1.2989 | 6.15 steps/s | VRAM=31.28 Go [2026-07-18 02:19:58] step 583350/1000000 | loss=2.3138 vocab=1.1780 token=1.3526 | 6.14 steps/s | VRAM=29.11 Go [2026-07-18 02:20:06] step 583400/1000000 | loss=2.3103 vocab=0.6757 token=0.8260 | 6.13 steps/s | VRAM=28.14 Go [2026-07-18 02:20:14] step 583450/1000000 | loss=2.1349 vocab=0.8281 token=1.8326 | 6.15 steps/s | VRAM=29.09 Go [2026-07-18 02:20:23] step 583500/1000000 | loss=2.3013 vocab=0.9434 token=1.4485 | 6.11 steps/s | VRAM=28.14 Go [2026-07-18 02:20:31] step 583550/1000000 | loss=2.1994 vocab=0.6018 token=1.4087 | 6.15 steps/s | VRAM=27.91 Go [2026-07-18 02:20:39] step 583600/1000000 | loss=2.2391 vocab=1.5736 token=1.3182 | 6.12 steps/s | VRAM=31.39 Go [2026-07-18 02:20:47] step 583650/1000000 | loss=2.1684 vocab=1.0083 token=0.9697 | 6.13 steps/s | VRAM=25.20 Go [2026-07-18 02:20:55] step 583700/1000000 | loss=2.3547 vocab=0.9442 token=1.7027 | 6.09 steps/s | VRAM=27.41 Go [2026-07-18 02:21:03] step 583750/1000000 | loss=2.2083 vocab=0.7119 token=1.0818 | 6.15 steps/s | VRAM=31.55 Go [2026-07-18 02:21:12] step 583800/1000000 | loss=2.1885 vocab=0.8166 token=1.1190 | 6.14 steps/s | VRAM=32.86 Go [2026-07-18 02:21:20] step 583850/1000000 | loss=2.2787 vocab=0.9765 token=1.3782 | 6.14 steps/s | VRAM=32.60 Go [2026-07-18 02:21:28] step 583900/1000000 | loss=2.2387 vocab=0.8628 token=1.5296 | 6.13 steps/s | VRAM=28.18 Go [2026-07-18 02:21:36] step 583950/1000000 | loss=2.1138 vocab=0.6288 token=0.7483 | 6.13 steps/s | VRAM=29.05 Go [2026-07-18 02:21:44] step 584000/1000000 | loss=2.2089 vocab=1.1441 token=1.5425 | 6.11 steps/s | VRAM=28.90 Go [2026-07-18 02:21:52] step 584050/1000000 | loss=2.1466 vocab=1.3107 token=1.4844 | 6.10 steps/s | VRAM=26.30 Go [2026-07-18 02:22:01] step 584100/1000000 | loss=2.2196 vocab=0.7863 token=1.2625 | 6.14 steps/s | VRAM=28.61 Go [2026-07-18 02:22:09] step 584150/1000000 | loss=2.1720 vocab=0.9497 token=1.0738 | 6.12 steps/s | VRAM=29.25 Go [2026-07-18 02:22:17] step 584200/1000000 | loss=2.2169 vocab=1.5061 token=1.4238 | 6.14 steps/s | VRAM=32.97 Go [2026-07-18 02:22:25] step 584250/1000000 | loss=2.1212 vocab=0.9181 token=1.3800 | 6.15 steps/s | VRAM=25.85 Go [2026-07-18 02:22:33] step 584300/1000000 | loss=2.1443 vocab=1.0537 token=1.0913 | 6.13 steps/s | VRAM=28.31 Go [2026-07-18 02:22:41] step 584350/1000000 | loss=2.2129 vocab=0.9297 token=1.6631 | 6.17 steps/s | VRAM=24.50 Go [2026-07-18 02:22:49] step 584400/1000000 | loss=2.2188 vocab=0.2470 token=0.9330 | 6.16 steps/s | VRAM=28.97 Go [2026-07-18 02:22:57] step 584450/1000000 | loss=2.2580 vocab=0.5945 token=1.0229 | 6.17 steps/s | VRAM=31.39 Go [2026-07-18 02:23:06] step 584500/1000000 | loss=2.2365 vocab=0.5405 token=1.4134 | 6.16 steps/s | VRAM=31.69 Go [2026-07-18 02:23:14] step 584550/1000000 | loss=2.0780 vocab=1.2372 token=1.7148 | 6.14 steps/s | VRAM=29.19 Go [2026-07-18 02:23:22] step 584600/1000000 | loss=2.3312 vocab=1.4990 token=1.7478 | 6.14 steps/s | VRAM=31.87 Go [2026-07-18 02:23:30] step 584650/1000000 | loss=2.2271 vocab=1.0998 token=1.4101 | 6.11 steps/s | VRAM=25.72 Go [2026-07-18 02:23:38] step 584700/1000000 | loss=2.2910 vocab=1.0374 token=1.5570 | 6.17 steps/s | VRAM=31.79 Go [2026-07-18 02:23:46] step 584750/1000000 | loss=2.2331 vocab=1.1122 token=1.6677 | 6.14 steps/s | VRAM=32.21 Go [2026-07-18 02:23:54] step 584800/1000000 | loss=2.1039 vocab=0.5938 token=1.2101 | 6.13 steps/s | VRAM=29.06 Go [2026-07-18 02:24:03] step 584850/1000000 | loss=2.3049 vocab=1.2069 token=1.4012 | 6.15 steps/s | VRAM=29.69 Go [2026-07-18 02:24:11] step 584900/1000000 | loss=2.1874 vocab=0.7769 token=1.5872 | 6.09 steps/s | VRAM=29.15 Go [2026-07-18 02:24:19] step 584950/1000000 | loss=2.1931 vocab=0.8395 token=1.2595 | 6.14 steps/s | VRAM=28.68 Go [2026-07-18 02:24:27] step 585000/1000000 | loss=2.2282 vocab=0.7237 token=1.2098 | 6.10 steps/s | VRAM=29.15 Go [2026-07-18 02:24:35] step 585050/1000000 | loss=2.2459 vocab=0.7193 token=1.4094 | 6.13 steps/s | VRAM=28.59 Go [2026-07-18 02:24:43] step 585100/1000000 | loss=2.2228 vocab=0.6798 token=0.9020 | 6.10 steps/s | VRAM=28.17 Go [2026-07-18 02:24:52] step 585150/1000000 | loss=2.1587 vocab=0.9955 token=1.2038 | 6.09 steps/s | VRAM=29.58 Go [2026-07-18 02:25:00] step 585200/1000000 | loss=2.1667 vocab=0.8730 token=1.3307 | 6.13 steps/s | VRAM=25.79 Go [2026-07-18 02:25:08] step 585250/1000000 | loss=2.1660 vocab=0.6239 token=1.1174 | 6.14 steps/s | VRAM=28.70 Go [2026-07-18 02:25:16] step 585300/1000000 | loss=2.1911 vocab=0.7319 token=1.3578 | 6.15 steps/s | VRAM=29.61 Go [2026-07-18 02:25:24] step 585350/1000000 | loss=2.1436 vocab=1.5394 token=1.5671 | 6.13 steps/s | VRAM=31.89 Go [2026-07-18 02:25:32] step 585400/1000000 | loss=2.3027 vocab=0.7473 token=1.3093 | 6.11 steps/s | VRAM=29.23 Go [2026-07-18 02:25:41] step 585450/1000000 | loss=2.1499 vocab=0.8608 token=1.0524 | 6.13 steps/s | VRAM=25.72 Go [2026-07-18 02:25:49] step 585500/1000000 | loss=2.1838 vocab=1.3290 token=1.4578 | 6.09 steps/s | VRAM=30.30 Go [2026-07-18 02:25:57] step 585550/1000000 | loss=2.2170 vocab=0.8533 token=1.4235 | 6.12 steps/s | VRAM=29.22 Go [2026-07-18 02:26:05] step 585600/1000000 | loss=2.2907 vocab=1.3797 token=1.2969 | 6.12 steps/s | VRAM=24.61 Go [2026-07-18 02:26:13] step 585650/1000000 | loss=2.2403 vocab=0.9384 token=1.2476 | 6.11 steps/s | VRAM=26.88 Go [2026-07-18 02:26:22] step 585700/1000000 | loss=2.2131 vocab=0.6427 token=0.9607 | 6.11 steps/s | VRAM=28.99 Go [2026-07-18 02:26:30] step 585750/1000000 | loss=2.2819 vocab=1.2489 token=1.5433 | 6.11 steps/s | VRAM=29.07 Go [2026-07-18 02:26:38] step 585800/1000000 | loss=2.2769 vocab=0.5836 token=0.8260 | 6.11 steps/s | VRAM=28.03 Go [2026-07-18 02:26:46] step 585850/1000000 | loss=2.1707 vocab=0.9077 token=1.0835 | 6.11 steps/s | VRAM=29.02 Go [2026-07-18 02:26:54] step 585900/1000000 | loss=2.1947 vocab=0.5229 token=1.1256 | 6.11 steps/s | VRAM=28.32 Go [2026-07-18 02:27:02] step 585950/1000000 | loss=2.2005 vocab=0.9513 token=1.3906 | 6.08 steps/s | VRAM=29.24 Go [2026-07-18 02:27:11] step 586000/1000000 | loss=2.2151 vocab=1.1527 token=0.9852 | 6.12 steps/s | VRAM=31.86 Go [2026-07-18 02:27:19] step 586050/1000000 | loss=2.1920 vocab=0.6461 token=0.9885 | 6.11 steps/s | VRAM=29.31 Go [2026-07-18 02:27:27] step 586100/1000000 | loss=2.1365 vocab=0.8034 token=1.2356 | 6.12 steps/s | VRAM=29.18 Go [2026-07-18 02:27:35] step 586150/1000000 | loss=2.2473 vocab=1.4724 token=1.5731 | 6.12 steps/s | VRAM=27.18 Go [2026-07-18 02:27:43] step 586200/1000000 | loss=2.3467 vocab=1.0286 token=1.6768 | 6.13 steps/s | VRAM=28.34 Go [2026-07-18 02:27:51] step 586250/1000000 | loss=2.2813 vocab=0.6039 token=1.0831 | 6.15 steps/s | VRAM=28.41 Go [2026-07-18 02:28:00] step 586300/1000000 | loss=2.3381 vocab=1.2146 token=1.3658 | 6.14 steps/s | VRAM=32.58 Go [2026-07-18 02:28:08] step 586350/1000000 | loss=2.2270 vocab=1.0410 token=1.3552 | 6.14 steps/s | VRAM=28.79 Go [2026-07-18 02:28:16] step 586400/1000000 | loss=2.3249 vocab=1.1983 token=1.4194 | 6.11 steps/s | VRAM=28.76 Go [2026-07-18 02:28:24] step 586450/1000000 | loss=2.1961 vocab=0.5001 token=0.9075 | 6.13 steps/s | VRAM=27.87 Go [2026-07-18 02:28:32] step 586500/1000000 | loss=2.2242 vocab=1.0864 token=1.0139 | 6.13 steps/s | VRAM=27.29 Go [2026-07-18 02:28:40] step 586550/1000000 | loss=2.3011 vocab=0.8025 token=1.5558 | 6.14 steps/s | VRAM=24.93 Go [2026-07-18 02:28:49] step 586600/1000000 | loss=2.1091 vocab=0.7028 token=1.2060 | 6.13 steps/s | VRAM=30.06 Go [2026-07-18 02:28:57] step 586650/1000000 | loss=2.1322 vocab=0.5383 token=1.0925 | 6.13 steps/s | VRAM=28.59 Go [2026-07-18 02:29:05] step 586700/1000000 | loss=2.2410 vocab=0.6412 token=1.0921 | 6.10 steps/s | VRAM=26.11 Go [2026-07-18 02:29:13] step 586750/1000000 | loss=2.2195 vocab=0.6030 token=1.2061 | 6.13 steps/s | VRAM=24.49 Go [2026-07-18 02:29:21] step 586800/1000000 | loss=2.2345 vocab=0.7720 token=1.1518 | 6.14 steps/s | VRAM=28.27 Go [2026-07-18 02:29:29] step 586850/1000000 | loss=2.1668 vocab=0.6327 token=0.9146 | 6.11 steps/s | VRAM=26.18 Go [2026-07-18 02:29:38] step 586900/1000000 | loss=2.3539 vocab=0.7998 token=1.5128 | 6.14 steps/s | VRAM=29.13 Go [2026-07-18 02:29:46] step 586950/1000000 | loss=2.1619 vocab=0.9229 token=1.3888 | 6.11 steps/s | VRAM=25.24 Go [2026-07-18 02:29:54] step 587000/1000000 | loss=2.1459 vocab=0.7206 token=1.0871 | 6.16 steps/s | VRAM=31.58 Go [2026-07-18 02:30:02] step 587050/1000000 | loss=2.2613 vocab=0.7648 token=1.5243 | 6.11 steps/s | VRAM=29.33 Go [2026-07-18 02:30:10] step 587100/1000000 | loss=2.2587 vocab=0.8085 token=1.2000 | 6.12 steps/s | VRAM=32.59 Go [2026-07-18 02:30:18] step 587150/1000000 | loss=2.1462 vocab=0.6072 token=1.2051 | 6.11 steps/s | VRAM=28.38 Go [2026-07-18 02:30:27] step 587200/1000000 | loss=2.1223 vocab=0.5864 token=1.1421 | 6.13 steps/s | VRAM=31.50 Go [2026-07-18 02:30:35] step 587250/1000000 | loss=2.2144 vocab=0.9999 token=1.5920 | 6.10 steps/s | VRAM=32.20 Go [2026-07-18 02:30:43] step 587300/1000000 | loss=2.2008 vocab=1.0312 token=1.3609 | 6.13 steps/s | VRAM=32.39 Go [2026-07-18 02:30:51] step 587350/1000000 | loss=2.1865 vocab=0.7016 token=1.4169 | 6.14 steps/s | VRAM=24.66 Go [2026-07-18 02:30:59] step 587400/1000000 | loss=2.2494 vocab=0.9918 token=1.3204 | 6.12 steps/s | VRAM=28.33 Go [2026-07-18 02:31:07] step 587450/1000000 | loss=2.3263 vocab=1.3479 token=1.6924 | 6.12 steps/s | VRAM=28.48 Go [2026-07-18 02:31:16] step 587500/1000000 | loss=2.2884 vocab=0.6334 token=1.6545 | 6.14 steps/s | VRAM=28.87 Go [2026-07-18 02:31:24] step 587550/1000000 | loss=2.2636 vocab=0.5774 token=1.2085 | 6.11 steps/s | VRAM=32.85 Go [2026-07-18 02:31:32] step 587600/1000000 | loss=2.1353 vocab=0.7007 token=1.0893 | 6.12 steps/s | VRAM=31.81 Go [2026-07-18 02:31:40] step 587650/1000000 | loss=2.1644 vocab=0.8847 token=1.1138 | 6.12 steps/s | VRAM=32.20 Go [2026-07-18 02:31:48] step 587700/1000000 | loss=2.0836 vocab=0.8550 token=1.6136 | 6.09 steps/s | VRAM=32.90 Go [2026-07-18 02:31:56] step 587750/1000000 | loss=2.3029 vocab=0.7797 token=1.7817 | 6.10 steps/s | VRAM=26.75 Go [2026-07-18 02:32:05] step 587800/1000000 | loss=2.2728 vocab=1.2926 token=1.4992 | 6.13 steps/s | VRAM=24.99 Go [2026-07-18 02:32:13] step 587850/1000000 | loss=2.2915 vocab=0.9478 token=1.4312 | 6.13 steps/s | VRAM=31.63 Go [2026-07-18 02:32:21] step 587900/1000000 | loss=2.2823 vocab=1.1863 token=1.2070 | 6.09 steps/s | VRAM=25.82 Go [2026-07-18 02:32:29] step 587950/1000000 | loss=2.2046 vocab=0.9400 token=1.4061 | 6.13 steps/s | VRAM=25.25 Go [2026-07-18 02:32:37] step 588000/1000000 | loss=2.2106 vocab=1.3045 token=1.4895 | 6.09 steps/s | VRAM=28.87 Go [2026-07-18 02:32:46] step 588050/1000000 | loss=2.1834 vocab=0.7792 token=1.3268 | 6.10 steps/s | VRAM=26.43 Go [2026-07-18 02:32:54] step 588100/1000000 | loss=2.2857 vocab=0.9191 token=1.8654 | 6.11 steps/s | VRAM=25.04 Go [2026-07-18 02:33:02] step 588150/1000000 | loss=2.2101 vocab=1.2737 token=1.3538 | 6.12 steps/s | VRAM=28.29 Go [2026-07-18 02:33:10] step 588200/1000000 | loss=2.2455 vocab=0.4223 token=1.2922 | 6.11 steps/s | VRAM=29.52 Go [2026-07-18 02:33:18] step 588250/1000000 | loss=2.2548 vocab=0.8284 token=1.5129 | 6.10 steps/s | VRAM=26.89 Go [2026-07-18 02:33:26] step 588300/1000000 | loss=2.2175 vocab=0.9161 token=1.5299 | 6.14 steps/s | VRAM=31.79 Go [2026-07-18 02:33:35] step 588350/1000000 | loss=2.1631 vocab=0.5701 token=0.9895 | 6.10 steps/s | VRAM=28.80 Go [2026-07-18 02:33:43] step 588400/1000000 | loss=2.2195 vocab=1.0908 token=1.5329 | 6.13 steps/s | VRAM=28.01 Go [2026-07-18 02:33:51] step 588450/1000000 | loss=2.3156 vocab=0.8046 token=0.9746 | 6.14 steps/s | VRAM=31.44 Go [2026-07-18 02:33:59] step 588500/1000000 | loss=2.2381 vocab=1.2834 token=1.3541 | 6.09 steps/s | VRAM=23.89 Go [2026-07-18 02:34:07] step 588550/1000000 | loss=2.3303 vocab=0.5687 token=1.1321 | 6.15 steps/s | VRAM=29.15 Go [2026-07-18 02:34:15] step 588600/1000000 | loss=2.2420 vocab=1.3194 token=1.5951 | 6.09 steps/s | VRAM=29.87 Go [2026-07-18 02:34:24] step 588650/1000000 | loss=2.2310 vocab=0.6382 token=1.4829 | 6.13 steps/s | VRAM=28.43 Go [2026-07-18 02:34:32] step 588700/1000000 | loss=2.1326 vocab=0.9681 token=1.0277 | 6.14 steps/s | VRAM=25.68 Go [2026-07-18 02:34:40] step 588750/1000000 | loss=2.2063 vocab=0.6517 token=1.1875 | 6.11 steps/s | VRAM=29.66 Go [2026-07-18 02:34:48] step 588800/1000000 | loss=2.1657 vocab=0.9905 token=1.6264 | 6.08 steps/s | VRAM=28.81 Go [2026-07-18 02:34:56] step 588850/1000000 | loss=2.1425 vocab=0.9193 token=1.3321 | 6.07 steps/s | VRAM=26.75 Go [2026-07-18 02:35:05] step 588900/1000000 | loss=2.1264 vocab=0.5561 token=1.2957 | 6.11 steps/s | VRAM=29.11 Go [2026-07-18 02:35:13] step 588950/1000000 | loss=2.2316 vocab=0.4911 token=1.0612 | 6.10 steps/s | VRAM=28.29 Go [2026-07-18 02:35:21] step 589000/1000000 | loss=2.1766 vocab=0.7072 token=1.3438 | 6.10 steps/s | VRAM=26.63 Go [2026-07-18 02:35:29] step 589050/1000000 | loss=2.1309 vocab=0.9257 token=1.4967 | 6.08 steps/s | VRAM=31.51 Go [2026-07-18 02:35:37] step 589100/1000000 | loss=2.2687 vocab=0.8207 token=1.4150 | 6.13 steps/s | VRAM=31.55 Go [2026-07-18 02:35:46] step 589150/1000000 | loss=2.1869 vocab=0.8305 token=0.9971 | 6.12 steps/s | VRAM=31.52 Go [2026-07-18 02:35:54] step 589200/1000000 | loss=2.1314 vocab=0.9206 token=1.4715 | 6.11 steps/s | VRAM=32.85 Go [2026-07-18 02:36:02] step 589250/1000000 | loss=2.3082 vocab=1.2975 token=1.2803 | 6.12 steps/s | VRAM=22.82 Go [2026-07-18 02:36:10] step 589300/1000000 | loss=2.2244 vocab=1.0380 token=1.4358 | 6.11 steps/s | VRAM=30.80 Go [2026-07-18 02:36:18] step 589350/1000000 | loss=2.1305 vocab=0.7361 token=1.3325 | 6.09 steps/s | VRAM=27.05 Go [2026-07-18 02:36:26] step 589400/1000000 | loss=2.3055 vocab=0.8647 token=1.1453 | 6.13 steps/s | VRAM=31.44 Go [2026-07-18 02:36:35] step 589450/1000000 | loss=2.1792 vocab=0.6999 token=1.1130 | 6.11 steps/s | VRAM=29.20 Go [2026-07-18 02:36:43] step 589500/1000000 | loss=2.1616 vocab=0.5330 token=1.2491 | 6.10 steps/s | VRAM=28.32 Go [2026-07-18 02:36:51] step 589550/1000000 | loss=2.2575 vocab=1.3430 token=1.7028 | 6.12 steps/s | VRAM=26.84 Go [2026-07-18 02:36:59] step 589600/1000000 | loss=2.3357 vocab=1.5214 token=1.5286 | 6.12 steps/s | VRAM=27.98 Go [2026-07-18 02:37:07] step 589650/1000000 | loss=2.3266 vocab=0.8877 token=1.7007 | 6.12 steps/s | VRAM=30.01 Go [2026-07-18 02:37:15] step 589700/1000000 | loss=2.1906 vocab=0.7418 token=1.0780 | 6.13 steps/s | VRAM=28.07 Go [2026-07-18 02:37:24] step 589750/1000000 | loss=2.1187 vocab=0.7025 token=0.9716 | 6.12 steps/s | VRAM=28.05 Go [2026-07-18 02:37:32] step 589800/1000000 | loss=2.2971 vocab=1.3187 token=1.5690 | 6.16 steps/s | VRAM=28.08 Go [2026-07-18 02:37:40] step 589850/1000000 | loss=2.2163 vocab=0.9059 token=1.0140 | 6.12 steps/s | VRAM=29.31 Go [2026-07-18 02:37:48] step 589900/1000000 | loss=2.2444 vocab=0.8580 token=1.0860 | 6.13 steps/s | VRAM=28.23 Go [2026-07-18 02:37:56] step 589950/1000000 | loss=2.2301 vocab=1.1269 token=1.7489 | 6.14 steps/s | VRAM=27.64 Go [2026-07-18 02:38:04] step 590000/1000000 | loss=2.2372 vocab=1.2343 token=1.2222 | 6.14 steps/s | VRAM=28.69 Go [2026-07-18 02:38:13] step 590050/1000000 | loss=2.2315 vocab=0.7465 token=1.6384 | 6.15 steps/s | VRAM=32.01 Go [2026-07-18 02:38:21] step 590100/1000000 | loss=2.1592 vocab=0.5150 token=0.8638 | 6.10 steps/s | VRAM=27.71 Go [2026-07-18 02:38:29] step 590150/1000000 | loss=2.1570 vocab=1.4170 token=1.2913 | 6.12 steps/s | VRAM=29.16 Go [2026-07-18 02:38:37] step 590200/1000000 | loss=2.1678 vocab=1.1090 token=1.3717 | 6.09 steps/s | VRAM=26.61 Go [2026-07-18 02:38:45] step 590250/1000000 | loss=2.2183 vocab=1.0320 token=1.2886 | 6.14 steps/s | VRAM=32.65 Go [2026-07-18 02:38:53] step 590300/1000000 | loss=2.1537 vocab=1.0658 token=1.2955 | 6.11 steps/s | VRAM=29.02 Go [2026-07-18 02:39:02] step 590350/1000000 | loss=2.1275 vocab=0.8998 token=1.3792 | 6.14 steps/s | VRAM=26.17 Go [2026-07-18 02:39:10] step 590400/1000000 | loss=2.2939 vocab=1.4313 token=1.3937 | 6.14 steps/s | VRAM=26.68 Go [2026-07-18 02:39:18] step 590450/1000000 | loss=2.2108 vocab=0.4615 token=0.8888 | 6.15 steps/s | VRAM=28.50 Go [2026-07-18 02:39:26] step 590500/1000000 | loss=2.2041 vocab=0.7104 token=1.4625 | 6.12 steps/s | VRAM=28.84 Go [2026-07-18 02:39:34] step 590550/1000000 | loss=2.2466 vocab=0.6771 token=1.4527 | 6.10 steps/s | VRAM=26.78 Go [2026-07-18 02:39:42] step 590600/1000000 | loss=2.2514 vocab=0.9107 token=1.7450 | 6.14 steps/s | VRAM=28.89 Go [2026-07-18 02:39:50] step 590650/1000000 | loss=2.1845 vocab=0.8667 token=1.1790 | 6.14 steps/s | VRAM=28.23 Go [2026-07-18 02:39:59] step 590700/1000000 | loss=2.2008 vocab=0.8038 token=1.1335 | 6.14 steps/s | VRAM=31.93 Go [2026-07-18 02:40:07] step 590750/1000000 | loss=2.2145 vocab=0.5092 token=0.9375 | 6.16 steps/s | VRAM=32.25 Go [2026-07-18 02:40:15] step 590800/1000000 | loss=2.1959 vocab=0.9473 token=1.3991 | 6.13 steps/s | VRAM=28.64 Go [2026-07-18 02:40:23] step 590850/1000000 | loss=2.2066 vocab=0.7589 token=1.3474 | 6.11 steps/s | VRAM=29.86 Go [2026-07-18 02:40:31] step 590900/1000000 | loss=2.2965 vocab=1.2040 token=1.7365 | 6.12 steps/s | VRAM=26.16 Go [2026-07-18 02:40:39] step 590950/1000000 | loss=2.1542 vocab=0.7339 token=1.2957 | 6.13 steps/s | VRAM=32.27 Go [2026-07-18 02:40:48] step 591000/1000000 | loss=2.1843 vocab=0.8582 token=1.1507 | 6.12 steps/s | VRAM=25.44 Go [2026-07-18 02:40:56] step 591050/1000000 | loss=2.3181 vocab=0.7967 token=1.5301 | 6.13 steps/s | VRAM=30.91 Go [2026-07-18 02:41:04] step 591100/1000000 | loss=2.1382 vocab=1.0918 token=1.1989 | 6.13 steps/s | VRAM=28.33 Go [2026-07-18 02:41:12] step 591150/1000000 | loss=2.2732 vocab=1.1886 token=1.2253 | 6.10 steps/s | VRAM=29.31 Go [2026-07-18 02:41:20] step 591200/1000000 | loss=2.2197 vocab=1.0540 token=1.5857 | 6.14 steps/s | VRAM=32.00 Go [2026-07-18 02:41:28] step 591250/1000000 | loss=2.1942 vocab=1.4491 token=1.5460 | 6.08 steps/s | VRAM=29.63 Go [2026-07-18 02:41:37] step 591300/1000000 | loss=2.1930 vocab=0.8038 token=1.3863 | 6.12 steps/s | VRAM=27.40 Go [2026-07-18 02:41:45] step 591350/1000000 | loss=2.2304 vocab=0.8415 token=1.1900 | 6.12 steps/s | VRAM=30.95 Go [2026-07-18 02:41:53] step 591400/1000000 | loss=2.1820 vocab=0.9020 token=1.6368 | 6.13 steps/s | VRAM=26.65 Go [2026-07-18 02:42:01] step 591450/1000000 | loss=2.2375 vocab=1.1649 token=1.3041 | 6.16 steps/s | VRAM=24.37 Go [2026-07-18 02:42:09] step 591500/1000000 | loss=2.2067 vocab=1.1423 token=1.8398 | 6.11 steps/s | VRAM=26.58 Go [2026-07-18 02:42:17] step 591550/1000000 | loss=2.2822 vocab=0.7359 token=1.2833 | 6.14 steps/s | VRAM=32.30 Go [2026-07-18 02:42:26] step 591600/1000000 | loss=2.1706 vocab=1.2318 token=1.3582 | 6.10 steps/s | VRAM=28.49 Go [2026-07-18 02:42:34] step 591650/1000000 | loss=2.1765 vocab=1.0466 token=1.1702 | 6.12 steps/s | VRAM=26.22 Go [2026-07-18 02:42:42] step 591700/1000000 | loss=2.1846 vocab=0.7777 token=1.0813 | 6.11 steps/s | VRAM=26.54 Go [2026-07-18 02:42:50] step 591750/1000000 | loss=2.3094 vocab=1.2517 token=1.3993 | 6.12 steps/s | VRAM=28.73 Go [2026-07-18 02:42:58] step 591800/1000000 | loss=2.2464 vocab=1.1935 token=1.4358 | 6.11 steps/s | VRAM=26.70 Go [2026-07-18 02:43:07] step 591850/1000000 | loss=2.2848 vocab=0.9346 token=1.0598 | 6.10 steps/s | VRAM=29.20 Go [2026-07-18 02:43:15] step 591900/1000000 | loss=2.2258 vocab=1.1849 token=1.5897 | 6.11 steps/s | VRAM=27.64 Go [2026-07-18 02:43:23] step 591950/1000000 | loss=2.2354 vocab=1.1974 token=1.5489 | 6.12 steps/s | VRAM=30.61 Go [2026-07-18 02:43:31] step 592000/1000000 | loss=2.2462 vocab=1.2753 token=1.4621 | 6.13 steps/s | VRAM=31.81 Go [2026-07-18 02:43:39] step 592050/1000000 | loss=2.2527 vocab=0.9070 token=1.4126 | 6.08 steps/s | VRAM=27.94 Go [2026-07-18 02:43:47] step 592100/1000000 | loss=2.2110 vocab=0.6317 token=1.3455 | 6.13 steps/s | VRAM=32.90 Go [2026-07-18 02:43:56] step 592150/1000000 | loss=2.3086 vocab=0.6218 token=0.8966 | 6.09 steps/s | VRAM=28.78 Go [2026-07-18 02:44:04] step 592200/1000000 | loss=2.3012 vocab=1.1953 token=1.8431 | 6.13 steps/s | VRAM=25.92 Go [2026-07-18 02:44:12] step 592250/1000000 | loss=2.1844 vocab=0.9023 token=1.3962 | 6.09 steps/s | VRAM=28.61 Go [2026-07-18 02:44:20] step 592300/1000000 | loss=2.1731 vocab=1.0811 token=1.4346 | 6.07 steps/s | VRAM=28.09 Go [2026-07-18 02:44:28] step 592350/1000000 | loss=2.3190 vocab=0.6757 token=0.9847 | 6.07 steps/s | VRAM=28.70 Go [2026-07-18 02:44:37] step 592400/1000000 | loss=2.2006 vocab=0.5764 token=1.1105 | 6.06 steps/s | VRAM=25.64 Go [2026-07-18 02:44:45] step 592450/1000000 | loss=2.2785 vocab=0.7411 token=1.2828 | 6.09 steps/s | VRAM=24.36 Go [2026-07-18 02:44:53] step 592500/1000000 | loss=2.2196 vocab=0.6779 token=1.0523 | 6.08 steps/s | VRAM=26.68 Go [2026-07-18 02:45:01] step 592550/1000000 | loss=2.3505 vocab=1.1630 token=1.1250 | 6.06 steps/s | VRAM=29.11 Go [2026-07-18 02:45:10] step 592600/1000000 | loss=2.2275 vocab=0.9972 token=1.3660 | 6.08 steps/s | VRAM=29.24 Go [2026-07-18 02:45:18] step 592650/1000000 | loss=2.1054 vocab=1.2147 token=1.0239 | 6.06 steps/s | VRAM=30.02 Go [2026-07-18 02:45:26] step 592700/1000000 | loss=2.2129 vocab=1.7739 token=1.2875 | 6.08 steps/s | VRAM=28.06 Go [2026-07-18 02:45:34] step 592750/1000000 | loss=2.2109 vocab=1.0556 token=1.3084 | 6.10 steps/s | VRAM=32.13 Go [2026-07-18 02:45:43] step 592800/1000000 | loss=2.1548 vocab=0.7691 token=1.1879 | 6.06 steps/s | VRAM=25.81 Go [2026-07-18 02:45:51] step 592850/1000000 | loss=2.2380 vocab=0.4438 token=1.6108 | 6.08 steps/s | VRAM=31.46 Go [2026-07-18 02:45:59] step 592900/1000000 | loss=2.1840 vocab=0.8417 token=1.4803 | 6.10 steps/s | VRAM=31.40 Go [2026-07-18 02:46:07] step 592950/1000000 | loss=2.1415 vocab=1.1777 token=1.3704 | 6.08 steps/s | VRAM=31.76 Go [2026-07-18 02:46:15] step 593000/1000000 | loss=2.1308 vocab=0.9463 token=1.0805 | 6.03 steps/s | VRAM=26.16 Go [2026-07-18 02:46:24] step 593050/1000000 | loss=2.1828 vocab=0.6628 token=1.0860 | 6.06 steps/s | VRAM=26.38 Go [2026-07-18 02:46:32] step 593100/1000000 | loss=2.2625 vocab=0.5954 token=1.5600 | 6.10 steps/s | VRAM=32.30 Go [2026-07-18 02:46:40] step 593150/1000000 | loss=2.2398 vocab=0.6107 token=1.5991 | 6.08 steps/s | VRAM=32.09 Go [2026-07-18 02:46:48] step 593200/1000000 | loss=2.1252 vocab=1.0139 token=1.3071 | 6.08 steps/s | VRAM=30.10 Go [2026-07-18 02:46:57] step 593250/1000000 | loss=2.1866 vocab=0.7297 token=0.8779 | 6.05 steps/s | VRAM=29.53 Go [2026-07-18 02:47:05] step 593300/1000000 | loss=2.2308 vocab=1.2235 token=1.4995 | 6.07 steps/s | VRAM=32.54 Go [2026-07-18 02:47:13] step 593350/1000000 | loss=2.2090 vocab=0.9595 token=1.3805 | 6.04 steps/s | VRAM=26.32 Go [2026-07-18 02:47:21] step 593400/1000000 | loss=2.2358 vocab=0.7363 token=0.9722 | 6.08 steps/s | VRAM=28.58 Go [2026-07-18 02:47:30] step 593450/1000000 | loss=2.2140 vocab=0.7214 token=1.3694 | 6.09 steps/s | VRAM=28.56 Go [2026-07-18 02:47:38] step 593500/1000000 | loss=2.1812 vocab=1.2412 token=1.2527 | 6.07 steps/s | VRAM=28.56 Go [2026-07-18 02:47:46] step 593550/1000000 | loss=2.0782 vocab=0.7358 token=0.8711 | 6.07 steps/s | VRAM=25.48 Go [2026-07-18 02:47:54] step 593600/1000000 | loss=2.2822 vocab=1.1457 token=1.8872 | 6.04 steps/s | VRAM=32.09 Go [2026-07-18 02:48:03] step 593650/1000000 | loss=2.1678 vocab=0.6629 token=1.1136 | 6.08 steps/s | VRAM=33.00 Go [2026-07-18 02:48:11] step 593700/1000000 | loss=2.2890 vocab=0.9776 token=1.4744 | 6.09 steps/s | VRAM=29.27 Go [2026-07-18 02:48:19] step 593750/1000000 | loss=2.2796 vocab=1.0407 token=1.3228 | 6.06 steps/s | VRAM=28.50 Go [2026-07-18 02:48:27] step 593800/1000000 | loss=2.1964 vocab=0.7420 token=1.3551 | 6.04 steps/s | VRAM=29.30 Go [2026-07-18 02:48:36] step 593850/1000000 | loss=2.2675 vocab=0.7292 token=1.4717 | 6.08 steps/s | VRAM=24.93 Go [2026-07-18 02:48:44] step 593900/1000000 | loss=2.2632 vocab=0.7557 token=1.2074 | 6.06 steps/s | VRAM=26.30 Go [2026-07-18 02:48:52] step 593950/1000000 | loss=2.1675 vocab=1.0230 token=1.3084 | 6.08 steps/s | VRAM=29.22 Go [2026-07-18 02:49:00] step 594000/1000000 | loss=2.2461 vocab=1.2044 token=1.5123 | 6.05 steps/s | VRAM=28.03 Go [2026-07-18 02:49:09] step 594050/1000000 | loss=2.1575 vocab=0.7331 token=1.2054 | 6.07 steps/s | VRAM=28.33 Go [2026-07-18 02:49:17] step 594100/1000000 | loss=2.2211 vocab=1.0150 token=1.3239 | 6.07 steps/s | VRAM=26.94 Go [2026-07-18 02:49:25] step 594150/1000000 | loss=2.2088 vocab=0.5218 token=0.9368 | 6.09 steps/s | VRAM=31.35 Go [2026-07-18 02:49:33] step 594200/1000000 | loss=2.1549 vocab=0.4483 token=1.0409 | 6.08 steps/s | VRAM=28.50 Go [2026-07-18 02:49:41] step 594250/1000000 | loss=2.2953 vocab=0.6866 token=1.2042 | 6.13 steps/s | VRAM=24.48 Go [2026-07-18 02:49:49] step 594300/1000000 | loss=2.1371 vocab=0.9323 token=1.2318 | 6.14 steps/s | VRAM=24.26 Go [2026-07-18 02:49:58] step 594350/1000000 | loss=2.2253 vocab=0.7069 token=1.3620 | 6.14 steps/s | VRAM=32.77 Go [2026-07-18 02:50:06] step 594400/1000000 | loss=2.2328 vocab=1.3875 token=1.5701 | 6.10 steps/s | VRAM=29.08 Go [2026-07-18 02:50:14] step 594450/1000000 | loss=2.1658 vocab=1.0351 token=1.1296 | 6.10 steps/s | VRAM=29.15 Go [2026-07-18 02:50:22] step 594500/1000000 | loss=2.2299 vocab=0.7776 token=1.6508 | 6.09 steps/s | VRAM=23.10 Go [2026-07-18 02:50:30] step 594550/1000000 | loss=2.1986 vocab=0.6778 token=1.8907 | 6.12 steps/s | VRAM=31.38 Go [2026-07-18 02:50:39] step 594600/1000000 | loss=2.3284 vocab=0.8781 token=1.4266 | 6.13 steps/s | VRAM=28.68 Go [2026-07-18 02:50:47] step 594650/1000000 | loss=2.1785 vocab=0.6800 token=1.3008 | 6.16 steps/s | VRAM=32.46 Go [2026-07-18 02:50:55] step 594700/1000000 | loss=2.2711 vocab=1.2313 token=1.2057 | 6.10 steps/s | VRAM=31.68 Go [2026-07-18 02:51:03] step 594750/1000000 | loss=2.1790 vocab=1.1740 token=1.1156 | 6.12 steps/s | VRAM=29.88 Go [2026-07-18 02:51:11] step 594800/1000000 | loss=2.2916 vocab=0.7557 token=1.2493 | 6.11 steps/s | VRAM=31.45 Go [2026-07-18 02:51:19] step 594850/1000000 | loss=2.1626 vocab=1.1915 token=1.6272 | 6.15 steps/s | VRAM=24.55 Go [2026-07-18 02:51:28] step 594900/1000000 | loss=2.1422 vocab=0.7093 token=1.0706 | 6.10 steps/s | VRAM=32.71 Go [2026-07-18 02:51:36] step 594950/1000000 | loss=2.1434 vocab=1.3104 token=1.2091 | 6.14 steps/s | VRAM=31.49 Go [2026-07-18 02:51:44] step 595000/1000000 | loss=2.1375 vocab=1.0187 token=1.0180 | 6.11 steps/s | VRAM=25.70 Go [2026-07-18 02:51:52] step 595050/1000000 | loss=2.2170 vocab=0.7395 token=1.3943 | 6.11 steps/s | VRAM=27.07 Go [2026-07-18 02:52:00] step 595100/1000000 | loss=2.2388 vocab=0.5715 token=1.2627 | 6.17 steps/s | VRAM=30.16 Go [2026-07-18 02:52:08] step 595150/1000000 | loss=2.2488 vocab=1.0970 token=1.3553 | 6.12 steps/s | VRAM=25.79 Go [2026-07-18 02:52:16] step 595200/1000000 | loss=2.1937 vocab=0.7607 token=1.1971 | 6.11 steps/s | VRAM=28.38 Go [2026-07-18 02:52:25] step 595250/1000000 | loss=2.1755 vocab=0.9636 token=1.2734 | 6.08 steps/s | VRAM=26.73 Go [2026-07-18 02:52:33] step 595300/1000000 | loss=2.1941 vocab=0.4657 token=1.2265 | 6.11 steps/s | VRAM=28.11 Go [2026-07-18 02:52:41] step 595350/1000000 | loss=2.2589 vocab=0.7344 token=1.5061 | 6.14 steps/s | VRAM=32.02 Go [2026-07-18 02:52:49] step 595400/1000000 | loss=2.2291 vocab=0.8180 token=1.0244 | 6.14 steps/s | VRAM=31.26 Go [2026-07-18 02:52:57] step 595450/1000000 | loss=2.2180 vocab=0.8218 token=1.5444 | 6.10 steps/s | VRAM=26.68 Go [2026-07-18 02:53:06] step 595500/1000000 | loss=2.2388 vocab=0.7186 token=1.1576 | 6.12 steps/s | VRAM=32.22 Go [2026-07-18 02:53:14] step 595550/1000000 | loss=2.2502 vocab=1.2498 token=1.6174 | 6.11 steps/s | VRAM=31.14 Go [2026-07-18 02:53:22] step 595600/1000000 | loss=2.2313 vocab=0.8701 token=1.1938 | 6.10 steps/s | VRAM=28.49 Go [2026-07-18 02:53:30] step 595650/1000000 | loss=2.2570 vocab=1.2758 token=1.4812 | 6.09 steps/s | VRAM=26.54 Go [2026-07-18 02:53:38] step 595700/1000000 | loss=2.2065 vocab=0.9284 token=1.3872 | 6.08 steps/s | VRAM=23.82 Go [2026-07-18 02:53:47] step 595750/1000000 | loss=2.1492 vocab=0.4614 token=1.1449 | 6.13 steps/s | VRAM=27.80 Go [2026-07-18 02:53:55] step 595800/1000000 | loss=2.3240 vocab=1.0361 token=1.2485 | 6.15 steps/s | VRAM=25.13 Go [2026-07-18 02:54:03] step 595850/1000000 | loss=2.3253 vocab=0.8236 token=1.2498 | 6.14 steps/s | VRAM=28.70 Go [2026-07-18 02:54:11] step 595900/1000000 | loss=2.1102 vocab=0.9620 token=1.1661 | 6.11 steps/s | VRAM=24.26 Go [2026-07-18 02:54:19] step 595950/1000000 | loss=2.3200 vocab=0.4612 token=1.1960 | 6.11 steps/s | VRAM=28.54 Go [2026-07-18 02:54:27] step 596000/1000000 | loss=2.1947 vocab=0.6586 token=1.5574 | 6.11 steps/s | VRAM=26.56 Go [2026-07-18 02:54:36] step 596050/1000000 | loss=2.0639 vocab=1.0354 token=1.6086 | 6.11 steps/s | VRAM=26.87 Go [2026-07-18 02:54:44] step 596100/1000000 | loss=2.2340 vocab=1.0706 token=1.4032 | 6.13 steps/s | VRAM=32.10 Go [2026-07-18 02:54:52] step 596150/1000000 | loss=2.2485 vocab=0.8520 token=1.4763 | 6.11 steps/s | VRAM=29.37 Go [2026-07-18 02:55:00] step 596200/1000000 | loss=2.1076 vocab=0.7245 token=1.4972 | 6.15 steps/s | VRAM=25.69 Go [2026-07-18 02:55:08] step 596250/1000000 | loss=2.3070 vocab=1.1551 token=1.5870 | 6.13 steps/s | VRAM=32.35 Go [2026-07-18 02:55:16] step 596300/1000000 | loss=2.2307 vocab=1.0735 token=1.5395 | 6.14 steps/s | VRAM=24.51 Go [2026-07-18 02:55:24] step 596350/1000000 | loss=2.2794 vocab=1.1857 token=1.0774 | 6.14 steps/s | VRAM=30.61 Go [2026-07-18 02:55:33] step 596400/1000000 | loss=2.2289 vocab=0.5805 token=1.1050 | 6.14 steps/s | VRAM=29.02 Go [2026-07-18 02:55:41] step 596450/1000000 | loss=2.1794 vocab=0.9337 token=1.4282 | 6.16 steps/s | VRAM=27.10 Go [2026-07-18 02:55:49] step 596500/1000000 | loss=2.2277 vocab=0.7220 token=1.5056 | 6.14 steps/s | VRAM=32.94 Go [2026-07-18 02:55:57] step 596550/1000000 | loss=2.2305 vocab=0.2612 token=0.5394 | 6.15 steps/s | VRAM=24.87 Go [2026-07-18 02:56:05] step 596600/1000000 | loss=2.1978 vocab=0.8654 token=1.3073 | 6.09 steps/s | VRAM=26.31 Go [2026-07-18 02:56:13] step 596650/1000000 | loss=2.2274 vocab=1.1445 token=1.5369 | 6.09 steps/s | VRAM=29.48 Go [2026-07-18 02:56:22] step 596700/1000000 | loss=2.2764 vocab=1.1701 token=1.4338 | 6.11 steps/s | VRAM=29.83 Go [2026-07-18 02:56:30] step 596750/1000000 | loss=2.2680 vocab=0.8507 token=1.4427 | 6.12 steps/s | VRAM=32.26 Go [2026-07-18 02:56:38] step 596800/1000000 | loss=2.2439 vocab=0.7899 token=1.2463 | 6.13 steps/s | VRAM=28.29 Go [2026-07-18 02:56:46] step 596850/1000000 | loss=2.2263 vocab=0.7741 token=1.0655 | 6.12 steps/s | VRAM=28.10 Go [2026-07-18 02:56:54] step 596900/1000000 | loss=2.2230 vocab=1.2737 token=1.4174 | 6.16 steps/s | VRAM=29.90 Go [2026-07-18 02:57:02] step 596950/1000000 | loss=2.1556 vocab=0.5581 token=0.6845 | 6.14 steps/s | VRAM=27.89 Go [2026-07-18 02:57:10] step 597000/1000000 | loss=2.2190 vocab=0.6698 token=1.8741 | 6.14 steps/s | VRAM=28.62 Go [2026-07-18 02:57:19] step 597050/1000000 | loss=2.1873 vocab=0.7660 token=1.3000 | 6.16 steps/s | VRAM=31.93 Go [2026-07-18 02:57:27] step 597100/1000000 | loss=2.1739 vocab=0.7486 token=1.2524 | 6.13 steps/s | VRAM=32.60 Go [2026-07-18 02:57:35] step 597150/1000000 | loss=2.0873 vocab=1.1247 token=1.2517 | 6.11 steps/s | VRAM=28.28 Go [2026-07-18 02:57:43] step 597200/1000000 | loss=2.2321 vocab=1.4035 token=1.6643 | 6.13 steps/s | VRAM=28.79 Go [2026-07-18 02:57:51] step 597250/1000000 | loss=2.2297 vocab=0.9852 token=1.4829 | 6.13 steps/s | VRAM=31.91 Go [2026-07-18 02:57:59] step 597300/1000000 | loss=2.2150 vocab=0.8138 token=1.4608 | 6.12 steps/s | VRAM=26.19 Go [2026-07-18 02:58:08] step 597350/1000000 | loss=2.1677 vocab=0.9542 token=1.1921 | 6.16 steps/s | VRAM=32.65 Go [2026-07-18 02:58:16] step 597400/1000000 | loss=2.1882 vocab=1.0293 token=1.3690 | 6.14 steps/s | VRAM=25.56 Go [2026-07-18 02:58:24] step 597450/1000000 | loss=2.3174 vocab=1.1168 token=1.6393 | 6.10 steps/s | VRAM=25.09 Go [2026-07-18 02:58:32] step 597500/1000000 | loss=2.1927 vocab=0.5897 token=1.2178 | 6.13 steps/s | VRAM=29.45 Go [2026-07-18 02:58:40] step 597550/1000000 | loss=2.2272 vocab=0.5217 token=1.3775 | 6.11 steps/s | VRAM=32.28 Go [2026-07-18 02:58:48] step 597600/1000000 | loss=2.2395 vocab=1.0114 token=1.1766 | 6.14 steps/s | VRAM=31.87 Go [2026-07-18 02:58:57] step 597650/1000000 | loss=2.2143 vocab=1.0639 token=1.6260 | 6.10 steps/s | VRAM=28.39 Go [2026-07-18 02:59:05] step 597700/1000000 | loss=2.2213 vocab=1.0773 token=1.8824 | 6.15 steps/s | VRAM=28.54 Go [2026-07-18 02:59:13] step 597750/1000000 | loss=2.2466 vocab=0.8730 token=0.9575 | 6.13 steps/s | VRAM=32.97 Go [2026-07-18 02:59:21] step 597800/1000000 | loss=2.2434 vocab=1.1142 token=1.0747 | 6.12 steps/s | VRAM=31.72 Go [2026-07-18 02:59:29] step 597850/1000000 | loss=2.1832 vocab=0.8556 token=1.4359 | 6.15 steps/s | VRAM=31.13 Go [2026-07-18 02:59:37] step 597900/1000000 | loss=2.3046 vocab=0.7430 token=1.2242 | 6.16 steps/s | VRAM=28.14 Go [2026-07-18 02:59:45] step 597950/1000000 | loss=2.1437 vocab=0.7605 token=1.5673 | 6.12 steps/s | VRAM=29.24 Go [2026-07-18 02:59:54] step 598000/1000000 | loss=2.1474 vocab=1.1596 token=1.4807 | 6.11 steps/s | VRAM=32.76 Go [2026-07-18 03:00:02] step 598050/1000000 | loss=2.2345 vocab=0.8853 token=1.6272 | 6.12 steps/s | VRAM=26.15 Go [2026-07-18 03:00:10] step 598100/1000000 | loss=2.1948 vocab=0.5983 token=1.0109 | 6.15 steps/s | VRAM=31.92 Go [2026-07-18 03:00:18] step 598150/1000000 | loss=2.2768 vocab=0.7460 token=1.3556 | 6.14 steps/s | VRAM=24.62 Go [2026-07-18 03:00:26] step 598200/1000000 | loss=2.2676 vocab=0.5162 token=1.4932 | 6.14 steps/s | VRAM=32.55 Go [2026-07-18 03:00:34] step 598250/1000000 | loss=2.3261 vocab=0.7629 token=1.3884 | 6.13 steps/s | VRAM=29.28 Go [2026-07-18 03:00:43] step 598300/1000000 | loss=2.2212 vocab=0.4790 token=0.9915 | 6.12 steps/s | VRAM=28.98 Go [2026-07-18 03:00:51] step 598350/1000000 | loss=2.2106 vocab=0.4697 token=0.9906 | 6.12 steps/s | VRAM=29.43 Go [2026-07-18 03:00:59] step 598400/1000000 | loss=2.2530 vocab=0.8868 token=1.4702 | 6.14 steps/s | VRAM=32.49 Go [2026-07-18 03:01:07] step 598450/1000000 | loss=2.2858 vocab=0.3813 token=1.0776 | 6.13 steps/s | VRAM=28.49 Go [2026-07-18 03:01:15] step 598500/1000000 | loss=2.3457 vocab=1.3385 token=1.1750 | 6.09 steps/s | VRAM=26.88 Go [2026-07-18 03:01:23] step 598550/1000000 | loss=2.2574 vocab=0.7062 token=1.2410 | 6.14 steps/s | VRAM=28.80 Go [2026-07-18 03:01:31] step 598600/1000000 | loss=2.1080 vocab=0.6399 token=0.9440 | 6.15 steps/s | VRAM=32.88 Go [2026-07-18 03:01:40] step 598650/1000000 | loss=2.1854 vocab=0.6694 token=1.3631 | 6.12 steps/s | VRAM=29.43 Go [2026-07-18 03:01:48] step 598700/1000000 | loss=2.2821 vocab=1.2957 token=1.5638 | 6.10 steps/s | VRAM=30.84 Go [2026-07-18 03:01:56] step 598750/1000000 | loss=2.2536 vocab=0.9748 token=1.2405 | 6.15 steps/s | VRAM=28.49 Go [2026-07-18 03:02:04] step 598800/1000000 | loss=2.2262 vocab=0.8047 token=1.3749 | 6.16 steps/s | VRAM=28.19 Go [2026-07-18 03:02:12] step 598850/1000000 | loss=2.2578 vocab=0.8202 token=1.1099 | 6.15 steps/s | VRAM=28.74 Go [2026-07-18 03:02:20] step 598900/1000000 | loss=2.2371 vocab=2.0296 token=1.9847 | 6.10 steps/s | VRAM=26.61 Go [2026-07-18 03:02:29] step 598950/1000000 | loss=2.2950 vocab=0.7923 token=1.1862 | 6.15 steps/s | VRAM=31.35 Go [2026-07-18 03:02:37] step 599000/1000000 | loss=2.1461 vocab=0.8769 token=1.3623 | 6.11 steps/s | VRAM=31.60 Go [2026-07-18 03:02:45] step 599050/1000000 | loss=2.2550 vocab=0.7484 token=1.2405 | 6.13 steps/s | VRAM=29.99 Go [2026-07-18 03:02:53] step 599100/1000000 | loss=2.2066 vocab=0.5744 token=1.2709 | 6.12 steps/s | VRAM=29.50 Go [2026-07-18 03:03:01] step 599150/1000000 | loss=2.3087 vocab=0.7972 token=1.3044 | 6.11 steps/s | VRAM=26.05 Go [2026-07-18 03:03:09] step 599200/1000000 | loss=2.2547 vocab=0.7379 token=1.2765 | 6.15 steps/s | VRAM=29.09 Go [2026-07-18 03:03:18] step 599250/1000000 | loss=2.1740 vocab=1.0208 token=1.4982 | 6.16 steps/s | VRAM=32.25 Go [2026-07-18 03:03:26] step 599300/1000000 | loss=2.1492 vocab=0.6606 token=0.9073 | 6.09 steps/s | VRAM=28.92 Go [2026-07-18 03:03:34] step 599350/1000000 | loss=2.3012 vocab=1.2369 token=1.4472 | 6.14 steps/s | VRAM=29.89 Go [2026-07-18 03:03:42] step 599400/1000000 | loss=2.1820 vocab=0.9033 token=1.0788 | 6.11 steps/s | VRAM=29.94 Go [2026-07-18 03:03:50] step 599450/1000000 | loss=2.1346 vocab=0.9525 token=1.4692 | 6.12 steps/s | VRAM=28.98 Go [2026-07-18 03:03:58] step 599500/1000000 | loss=2.2089 vocab=0.4460 token=0.6929 | 6.15 steps/s | VRAM=28.48 Go [2026-07-18 03:04:06] step 599550/1000000 | loss=2.1836 vocab=1.5584 token=1.3594 | 6.16 steps/s | VRAM=31.06 Go [2026-07-18 03:04:15] step 599600/1000000 | loss=2.2721 vocab=0.5614 token=1.2594 | 6.15 steps/s | VRAM=29.42 Go [2026-07-18 03:04:23] step 599650/1000000 | loss=2.2657 vocab=0.8451 token=1.2282 | 6.12 steps/s | VRAM=28.31 Go [2026-07-18 03:04:31] step 599700/1000000 | loss=2.3029 vocab=1.2292 token=1.5126 | 6.16 steps/s | VRAM=24.55 Go [2026-07-18 03:04:39] step 599750/1000000 | loss=2.2822 vocab=1.0750 token=1.0419 | 6.14 steps/s | VRAM=28.44 Go [2026-07-18 03:04:47] step 599800/1000000 | loss=2.2344 vocab=0.5605 token=1.1578 | 6.15 steps/s | VRAM=32.83 Go [2026-07-18 03:04:55] step 599850/1000000 | loss=2.0758 vocab=0.5644 token=1.1209 | 6.15 steps/s | VRAM=31.78 Go [2026-07-18 03:05:03] step 599900/1000000 | loss=2.1851 vocab=0.7422 token=0.7709 | 6.12 steps/s | VRAM=26.20 Go [2026-07-18 03:05:12] step 599950/1000000 | loss=2.2819 vocab=0.6464 token=1.3886 | 6.10 steps/s | VRAM=28.17 Go [2026-07-18 03:05:20] step 600000/1000000 | loss=2.2439 vocab=1.0963 token=1.0528 | 6.14 steps/s | VRAM=32.87 Go [step 600000] loss_token=1.0528 > min recent (1.0052) -> push saute [2026-07-18 03:05:28] step 600050/1000000 | loss=2.1743 vocab=0.7988 token=1.0278 | 6.13 steps/s | VRAM=31.83 Go [2026-07-18 03:05:36] step 600100/1000000 | loss=2.3434 vocab=0.9379 token=1.1606 | 6.11 steps/s | VRAM=26.64 Go [2026-07-18 03:05:44] step 600150/1000000 | loss=2.2071 vocab=0.6856 token=1.1480 | 6.12 steps/s | VRAM=25.27 Go [2026-07-18 03:05:52] step 600200/1000000 | loss=2.3308 vocab=1.1696 token=1.1954 | 6.14 steps/s | VRAM=28.64 Go [2026-07-18 03:06:01] step 600250/1000000 | loss=2.2640 vocab=0.5825 token=1.2001 | 6.14 steps/s | VRAM=28.92 Go [2026-07-18 03:06:09] step 600300/1000000 | loss=2.1468 vocab=0.6973 token=1.0789 | 6.14 steps/s | VRAM=31.64 Go [2026-07-18 03:06:17] step 600350/1000000 | loss=2.2335 vocab=0.7158 token=1.2564 | 6.11 steps/s | VRAM=24.70 Go [2026-07-18 03:06:25] step 600400/1000000 | loss=2.1962 vocab=0.9594 token=1.0470 | 6.12 steps/s | VRAM=26.18 Go [2026-07-18 03:06:33] step 600450/1000000 | loss=2.2331 vocab=0.8173 token=1.2800 | 6.12 steps/s | VRAM=29.02 Go [2026-07-18 03:06:41] step 600500/1000000 | loss=2.1350 vocab=0.6360 token=1.2388 | 6.13 steps/s | VRAM=28.57 Go [2026-07-18 03:06:50] step 600550/1000000 | loss=2.2213 vocab=0.7926 token=1.1152 | 6.15 steps/s | VRAM=28.46 Go [2026-07-18 03:06:58] step 600600/1000000 | loss=2.2358 vocab=0.9685 token=1.1606 | 6.15 steps/s | VRAM=26.17 Go [2026-07-18 03:07:06] step 600650/1000000 | loss=2.2425 vocab=0.7889 token=0.9273 | 6.12 steps/s | VRAM=32.68 Go [2026-07-18 03:07:14] step 600700/1000000 | loss=2.1899 vocab=0.8249 token=1.4386 | 6.15 steps/s | VRAM=32.45 Go [2026-07-18 03:07:22] step 600750/1000000 | loss=2.2182 vocab=1.0110 token=1.4279 | 6.13 steps/s | VRAM=26.40 Go [2026-07-18 03:07:30] step 600800/1000000 | loss=2.2385 vocab=0.7755 token=0.9134 | 6.12 steps/s | VRAM=29.35 Go [2026-07-18 03:07:38] step 600850/1000000 | loss=2.2642 vocab=0.6544 token=1.1463 | 6.14 steps/s | VRAM=28.11 Go [2026-07-18 03:07:47] step 600900/1000000 | loss=2.1842 vocab=0.9695 token=1.7296 | 6.13 steps/s | VRAM=29.42 Go [2026-07-18 03:07:55] step 600950/1000000 | loss=2.1803 vocab=1.4385 token=1.4648 | 6.14 steps/s | VRAM=28.48 Go [2026-07-18 03:08:03] step 601000/1000000 | loss=2.2281 vocab=1.0033 token=1.3831 | 6.12 steps/s | VRAM=29.36 Go [2026-07-18 03:08:11] step 601050/1000000 | loss=2.2826 vocab=1.3648 token=1.1647 | 6.13 steps/s | VRAM=28.46 Go [2026-07-18 03:08:19] step 601100/1000000 | loss=2.2251 vocab=0.9286 token=1.0589 | 6.16 steps/s | VRAM=25.85 Go [2026-07-18 03:08:27] step 601150/1000000 | loss=2.2343 vocab=0.9740 token=1.5259 | 6.14 steps/s | VRAM=27.74 Go [2026-07-18 03:08:35] step 601200/1000000 | loss=2.1454 vocab=0.6820 token=1.2835 | 6.14 steps/s | VRAM=28.19 Go [2026-07-18 03:08:44] step 601250/1000000 | loss=2.2411 vocab=1.0889 token=1.0744 | 6.15 steps/s | VRAM=32.39 Go [2026-07-18 03:08:52] step 601300/1000000 | loss=2.2422 vocab=0.6337 token=1.1577 | 6.11 steps/s | VRAM=28.48 Go [2026-07-18 03:09:00] step 601350/1000000 | loss=2.1655 vocab=0.9473 token=1.3267 | 6.14 steps/s | VRAM=31.95 Go [2026-07-18 03:09:08] step 601400/1000000 | loss=2.2044 vocab=0.9821 token=1.3821 | 6.16 steps/s | VRAM=32.20 Go [2026-07-18 03:09:16] step 601450/1000000 | loss=2.2825 vocab=0.7308 token=1.5907 | 6.14 steps/s | VRAM=27.97 Go [2026-07-18 03:09:24] step 601500/1000000 | loss=2.1773 vocab=1.0052 token=1.4770 | 6.08 steps/s | VRAM=30.39 Go [2026-07-18 03:09:33] step 601550/1000000 | loss=2.2033 vocab=1.6572 token=1.3333 | 6.13 steps/s | VRAM=28.23 Go [2026-07-18 03:09:41] step 601600/1000000 | loss=2.1604 vocab=0.8119 token=0.8549 | 6.12 steps/s | VRAM=29.00 Go [2026-07-18 03:09:49] step 601650/1000000 | loss=2.2235 vocab=0.7590 token=1.2097 | 6.13 steps/s | VRAM=28.64 Go [2026-07-18 03:09:57] step 601700/1000000 | loss=2.3172 vocab=0.8114 token=1.3514 | 6.11 steps/s | VRAM=30.16 Go [2026-07-18 03:10:05] step 601750/1000000 | loss=2.2229 vocab=0.5747 token=1.1674 | 6.16 steps/s | VRAM=25.68 Go [2026-07-18 03:10:13] step 601800/1000000 | loss=2.2363 vocab=0.6408 token=1.3188 | 6.16 steps/s | VRAM=27.52 Go [2026-07-18 03:10:22] step 601850/1000000 | loss=2.3036 vocab=0.6990 token=1.0474 | 6.11 steps/s | VRAM=32.52 Go [2026-07-18 03:10:30] step 601900/1000000 | loss=2.1443 vocab=0.8630 token=1.3852 | 6.11 steps/s | VRAM=29.22 Go [2026-07-18 03:10:38] step 601950/1000000 | loss=2.2770 vocab=0.8262 token=1.1200 | 6.14 steps/s | VRAM=28.75 Go [2026-07-18 03:10:46] step 602000/1000000 | loss=2.1629 vocab=0.7855 token=0.9732 | 6.13 steps/s | VRAM=28.97 Go [2026-07-18 03:10:54] step 602050/1000000 | loss=2.2459 vocab=0.6093 token=1.3083 | 6.15 steps/s | VRAM=32.80 Go [2026-07-18 03:11:02] step 602100/1000000 | loss=2.2286 vocab=1.1335 token=1.6788 | 6.11 steps/s | VRAM=26.46 Go [2026-07-18 03:11:10] step 602150/1000000 | loss=2.1120 vocab=0.8401 token=1.3392 | 6.16 steps/s | VRAM=31.94 Go [2026-07-18 03:11:19] step 602200/1000000 | loss=2.2897 vocab=1.0521 token=1.5604 | 6.12 steps/s | VRAM=28.64 Go [2026-07-18 03:11:27] step 602250/1000000 | loss=2.1944 vocab=0.6817 token=1.3316 | 6.09 steps/s | VRAM=28.86 Go [2026-07-18 03:11:35] step 602300/1000000 | loss=2.1665 vocab=0.9481 token=1.2304 | 6.14 steps/s | VRAM=31.71 Go [2026-07-18 03:11:43] step 602350/1000000 | loss=2.1889 vocab=0.9957 token=1.4075 | 6.11 steps/s | VRAM=27.00 Go [2026-07-18 03:11:51] step 602400/1000000 | loss=2.2472 vocab=0.9770 token=1.5147 | 6.13 steps/s | VRAM=25.66 Go [2026-07-18 03:11:59] step 602450/1000000 | loss=2.3047 vocab=0.3649 token=0.7467 | 6.16 steps/s | VRAM=24.63 Go [2026-07-18 03:12:08] step 602500/1000000 | loss=2.2632 vocab=0.5343 token=1.2717 | 6.14 steps/s | VRAM=32.63 Go [2026-07-18 03:12:16] step 602550/1000000 | loss=2.1960 vocab=0.8815 token=0.8811 | 6.13 steps/s | VRAM=28.20 Go [2026-07-18 03:12:24] step 602600/1000000 | loss=2.2462 vocab=1.2858 token=1.2773 | 6.12 steps/s | VRAM=28.57 Go [2026-07-18 03:12:32] step 602650/1000000 | loss=2.2403 vocab=1.0831 token=0.9458 | 6.13 steps/s | VRAM=31.49 Go [2026-07-18 03:12:40] step 602700/1000000 | loss=2.2118 vocab=0.7816 token=1.4614 | 6.12 steps/s | VRAM=26.61 Go [2026-07-18 03:12:48] step 602750/1000000 | loss=2.1831 vocab=0.7798 token=1.4862 | 6.13 steps/s | VRAM=31.79 Go [2026-07-18 03:12:57] step 602800/1000000 | loss=2.2151 vocab=1.4293 token=1.6893 | 6.12 steps/s | VRAM=29.17 Go [2026-07-18 03:13:05] step 602850/1000000 | loss=2.2792 vocab=0.7670 token=1.3641 | 6.09 steps/s | VRAM=29.22 Go [2026-07-18 03:13:13] step 602900/1000000 | loss=2.1375 vocab=0.9579 token=1.2386 | 6.12 steps/s | VRAM=28.12 Go [2026-07-18 03:13:21] step 602950/1000000 | loss=2.2547 vocab=1.2433 token=1.3924 | 6.15 steps/s | VRAM=27.47 Go [2026-07-18 03:13:29] step 603000/1000000 | loss=2.2149 vocab=1.6136 token=1.6839 | 6.11 steps/s | VRAM=27.81 Go [2026-07-18 03:13:37] step 603050/1000000 | loss=2.1637 vocab=1.0381 token=1.1743 | 6.12 steps/s | VRAM=28.61 Go [2026-07-18 03:13:46] step 603100/1000000 | loss=2.2057 vocab=1.4635 token=1.2023 | 6.10 steps/s | VRAM=28.39 Go [2026-07-18 03:13:54] step 603150/1000000 | loss=2.1869 vocab=0.9835 token=1.1487 | 6.11 steps/s | VRAM=25.76 Go [2026-07-18 03:14:02] step 603200/1000000 | loss=2.2951 vocab=0.9290 token=1.4708 | 6.12 steps/s | VRAM=28.22 Go [2026-07-18 03:14:10] step 603250/1000000 | loss=2.2909 vocab=0.8450 token=1.2657 | 6.14 steps/s | VRAM=32.91 Go [2026-07-18 03:14:18] step 603300/1000000 | loss=2.2384 vocab=1.1177 token=0.9829 | 6.09 steps/s | VRAM=29.17 Go [2026-07-18 03:14:26] step 603350/1000000 | loss=2.1739 vocab=1.1885 token=1.5374 | 6.12 steps/s | VRAM=33.02 Go [2026-07-18 03:14:35] step 603400/1000000 | loss=2.2431 vocab=1.2939 token=1.3899 | 6.12 steps/s | VRAM=26.13 Go [2026-07-18 03:14:43] step 603450/1000000 | loss=2.3072 vocab=0.7810 token=1.4920 | 6.12 steps/s | VRAM=27.57 Go [2026-07-18 03:14:51] step 603500/1000000 | loss=2.2237 vocab=1.4027 token=1.1688 | 6.11 steps/s | VRAM=26.16 Go [2026-07-18 03:14:59] step 603550/1000000 | loss=2.2110 vocab=0.9495 token=1.3268 | 6.11 steps/s | VRAM=29.56 Go [2026-07-18 03:15:07] step 603600/1000000 | loss=2.2166 vocab=0.8794 token=1.3600 | 6.12 steps/s | VRAM=28.24 Go [2026-07-18 03:15:16] step 603650/1000000 | loss=2.1793 vocab=0.7900 token=1.2831 | 6.13 steps/s | VRAM=26.26 Go [2026-07-18 03:15:24] step 603700/1000000 | loss=2.1384 vocab=1.2074 token=1.3661 | 6.10 steps/s | VRAM=26.74 Go [2026-07-18 03:15:32] step 603750/1000000 | loss=2.3110 vocab=0.7290 token=1.5366 | 6.13 steps/s | VRAM=31.22 Go [2026-07-18 03:15:40] step 603800/1000000 | loss=2.2104 vocab=0.4681 token=0.8494 | 6.13 steps/s | VRAM=24.44 Go [2026-07-18 03:15:48] step 603850/1000000 | loss=2.2299 vocab=0.7654 token=1.5444 | 6.13 steps/s | VRAM=25.93 Go [2026-07-18 03:15:56] step 603900/1000000 | loss=2.1274 vocab=0.5750 token=1.2935 | 6.14 steps/s | VRAM=32.87 Go [2026-07-18 03:16:04] step 603950/1000000 | loss=2.2261 vocab=0.8902 token=1.5729 | 6.17 steps/s | VRAM=31.49 Go [2026-07-18 03:16:13] step 604000/1000000 | loss=2.2466 vocab=1.1851 token=1.7488 | 6.15 steps/s | VRAM=29.33 Go [2026-07-18 03:16:21] step 604050/1000000 | loss=2.2901 vocab=0.6528 token=0.9951 | 6.12 steps/s | VRAM=32.52 Go [2026-07-18 03:16:29] step 604100/1000000 | loss=2.2226 vocab=0.8907 token=1.5739 | 6.13 steps/s | VRAM=32.21 Go [2026-07-18 03:16:37] step 604150/1000000 | loss=2.0895 vocab=0.7168 token=1.3707 | 6.14 steps/s | VRAM=32.30 Go [2026-07-18 03:16:45] step 604200/1000000 | loss=2.1821 vocab=1.2812 token=1.4162 | 6.12 steps/s | VRAM=28.10 Go [2026-07-18 03:16:53] step 604250/1000000 | loss=2.2229 vocab=0.7780 token=0.9586 | 6.14 steps/s | VRAM=25.47 Go [2026-07-18 03:17:02] step 604300/1000000 | loss=2.3022 vocab=0.9020 token=0.7981 | 6.14 steps/s | VRAM=31.78 Go [2026-07-18 03:17:10] step 604350/1000000 | loss=2.0671 vocab=0.8109 token=1.0083 | 6.17 steps/s | VRAM=28.52 Go [2026-07-18 03:17:18] step 604400/1000000 | loss=2.1975 vocab=0.5816 token=1.0910 | 6.13 steps/s | VRAM=29.44 Go [2026-07-18 03:17:26] step 604450/1000000 | loss=2.2139 vocab=0.6796 token=1.1690 | 6.16 steps/s | VRAM=28.98 Go [2026-07-18 03:17:34] step 604500/1000000 | loss=2.1816 vocab=0.5688 token=1.2434 | 6.14 steps/s | VRAM=32.35 Go [2026-07-18 03:17:42] step 604550/1000000 | loss=2.1972 vocab=1.2805 token=0.8275 | 6.15 steps/s | VRAM=29.89 Go [2026-07-18 03:17:50] step 604600/1000000 | loss=2.2040 vocab=0.7324 token=0.8231 | 6.14 steps/s | VRAM=25.46 Go [2026-07-18 03:17:58] step 604650/1000000 | loss=2.2169 vocab=0.7781 token=1.2256 | 6.13 steps/s | VRAM=29.51 Go [2026-07-18 03:18:07] step 604700/1000000 | loss=2.1929 vocab=0.8109 token=1.3207 | 6.13 steps/s | VRAM=29.09 Go [2026-07-18 03:18:15] step 604750/1000000 | loss=2.2250 vocab=0.8837 token=1.1317 | 6.14 steps/s | VRAM=29.18 Go [2026-07-18 03:18:23] step 604800/1000000 | loss=2.1788 vocab=0.7573 token=1.3419 | 6.13 steps/s | VRAM=25.30 Go [2026-07-18 03:18:31] step 604850/1000000 | loss=2.1655 vocab=0.6723 token=1.5877 | 6.15 steps/s | VRAM=29.83 Go [2026-07-18 03:18:39] step 604900/1000000 | loss=2.1695 vocab=1.1608 token=1.5506 | 6.15 steps/s | VRAM=29.06 Go [2026-07-18 03:18:47] step 604950/1000000 | loss=2.2599 vocab=0.5656 token=1.2166 | 6.13 steps/s | VRAM=31.11 Go [2026-07-18 03:18:55] step 605000/1000000 | loss=2.2397 vocab=0.8252 token=1.0914 | 6.14 steps/s | VRAM=26.72 Go [2026-07-18 03:19:04] step 605050/1000000 | loss=2.1898 vocab=0.6407 token=0.9499 | 6.15 steps/s | VRAM=27.18 Go [2026-07-18 03:19:12] step 605100/1000000 | loss=2.1577 vocab=0.8159 token=1.1045 | 6.16 steps/s | VRAM=31.24 Go [2026-07-18 03:19:20] step 605150/1000000 | loss=2.2143 vocab=0.6955 token=1.1807 | 6.13 steps/s | VRAM=29.50 Go [2026-07-18 03:19:28] step 605200/1000000 | loss=2.1401 vocab=1.0404 token=1.1505 | 6.15 steps/s | VRAM=31.20 Go [2026-07-18 03:19:36] step 605250/1000000 | loss=2.1630 vocab=0.6492 token=1.0355 | 6.13 steps/s | VRAM=24.83 Go [2026-07-18 03:19:44] step 605300/1000000 | loss=2.3124 vocab=1.3045 token=1.6627 | 6.14 steps/s | VRAM=30.41 Go [2026-07-18 03:19:52] step 605350/1000000 | loss=2.1604 vocab=0.4717 token=1.4602 | 6.14 steps/s | VRAM=29.10 Go [2026-07-18 03:20:01] step 605400/1000000 | loss=2.3125 vocab=1.0856 token=1.0652 | 6.16 steps/s | VRAM=32.02 Go [2026-07-18 03:20:09] step 605450/1000000 | loss=2.2293 vocab=0.6840 token=1.4918 | 6.17 steps/s | VRAM=28.73 Go [2026-07-18 03:20:17] step 605500/1000000 | loss=2.2328 vocab=0.5095 token=1.3271 | 6.12 steps/s | VRAM=28.54 Go [2026-07-18 03:20:25] step 605550/1000000 | loss=2.1908 vocab=0.5483 token=1.3785 | 6.15 steps/s | VRAM=32.44 Go [2026-07-18 03:20:33] step 605600/1000000 | loss=2.2324 vocab=0.9748 token=2.0619 | 6.16 steps/s | VRAM=28.79 Go [2026-07-18 03:20:41] step 605650/1000000 | loss=2.3448 vocab=0.5137 token=1.3878 | 6.14 steps/s | VRAM=28.45 Go [2026-07-18 03:20:49] step 605700/1000000 | loss=2.2579 vocab=0.6326 token=1.1668 | 6.11 steps/s | VRAM=32.26 Go [2026-07-18 03:20:58] step 605750/1000000 | loss=2.2178 vocab=0.9145 token=1.2651 | 6.12 steps/s | VRAM=26.19 Go [2026-07-18 03:21:06] step 605800/1000000 | loss=2.2511 vocab=0.6822 token=1.3939 | 6.14 steps/s | VRAM=25.52 Go [2026-07-18 03:21:14] step 605850/1000000 | loss=2.3424 vocab=0.7401 token=0.7176 | 6.13 steps/s | VRAM=30.14 Go [2026-07-18 03:21:22] step 605900/1000000 | loss=2.2299 vocab=0.4569 token=1.0476 | 6.14 steps/s | VRAM=27.96 Go [2026-07-18 03:21:30] step 605950/1000000 | loss=2.2381 vocab=1.2241 token=1.4061 | 6.11 steps/s | VRAM=25.90 Go [2026-07-18 03:21:38] step 606000/1000000 | loss=2.2030 vocab=0.6085 token=1.3050 | 6.16 steps/s | VRAM=31.38 Go [2026-07-18 03:21:47] step 606050/1000000 | loss=2.1941 vocab=0.9579 token=1.4918 | 6.13 steps/s | VRAM=29.89 Go [2026-07-18 03:21:55] step 606100/1000000 | loss=2.2819 vocab=0.9837 token=1.4381 | 6.14 steps/s | VRAM=25.01 Go [2026-07-18 03:22:03] step 606150/1000000 | loss=2.2426 vocab=0.5773 token=1.1166 | 6.15 steps/s | VRAM=31.80 Go [2026-07-18 03:22:11] step 606200/1000000 | loss=2.2872 vocab=0.7604 token=1.1858 | 6.15 steps/s | VRAM=25.41 Go [2026-07-18 03:22:19] step 606250/1000000 | loss=2.3035 vocab=1.0906 token=1.6194 | 6.15 steps/s | VRAM=28.75 Go [2026-07-18 03:22:27] step 606300/1000000 | loss=2.2390 vocab=0.5565 token=1.3084 | 6.10 steps/s | VRAM=27.23 Go [2026-07-18 03:22:35] step 606350/1000000 | loss=2.1673 vocab=1.2946 token=1.7426 | 6.16 steps/s | VRAM=30.87 Go [2026-07-18 03:22:44] step 606400/1000000 | loss=2.2468 vocab=0.8198 token=1.1591 | 6.12 steps/s | VRAM=32.89 Go [2026-07-18 03:22:52] step 606450/1000000 | loss=2.1677 vocab=0.5455 token=1.2838 | 6.14 steps/s | VRAM=28.91 Go [2026-07-18 03:23:00] step 606500/1000000 | loss=2.2890 vocab=0.9360 token=1.3275 | 6.13 steps/s | VRAM=28.93 Go [2026-07-18 03:23:08] step 606550/1000000 | loss=2.1985 vocab=0.6916 token=1.0275 | 6.14 steps/s | VRAM=29.65 Go [2026-07-18 03:23:16] step 606600/1000000 | loss=2.3041 vocab=0.6721 token=1.3581 | 6.14 steps/s | VRAM=29.21 Go [2026-07-18 03:23:24] step 606650/1000000 | loss=2.3248 vocab=1.6850 token=1.2916 | 6.13 steps/s | VRAM=27.30 Go [2026-07-18 03:23:32] step 606700/1000000 | loss=2.1916 vocab=1.0528 token=1.4140 | 6.13 steps/s | VRAM=32.89 Go [2026-07-18 03:23:41] step 606750/1000000 | loss=2.2423 vocab=0.9148 token=1.6584 | 6.14 steps/s | VRAM=25.16 Go [2026-07-18 03:23:49] step 606800/1000000 | loss=2.2143 vocab=0.7675 token=1.0868 | 6.11 steps/s | VRAM=26.07 Go [2026-07-18 03:23:57] step 606850/1000000 | loss=2.1445 vocab=1.0742 token=1.3788 | 6.15 steps/s | VRAM=27.92 Go [2026-07-18 03:24:05] step 606900/1000000 | loss=2.2510 vocab=0.8435 token=1.1873 | 6.14 steps/s | VRAM=28.90 Go [2026-07-18 03:24:13] step 606950/1000000 | loss=2.3128 vocab=1.5532 token=1.8377 | 6.13 steps/s | VRAM=28.70 Go [2026-07-18 03:24:21] step 607000/1000000 | loss=2.1436 vocab=0.8999 token=1.6914 | 6.12 steps/s | VRAM=28.99 Go [2026-07-18 03:24:29] step 607050/1000000 | loss=2.2826 vocab=1.1290 token=1.0876 | 6.14 steps/s | VRAM=31.44 Go [2026-07-18 03:24:38] step 607100/1000000 | loss=2.2368 vocab=0.5114 token=1.3917 | 6.15 steps/s | VRAM=28.69 Go [2026-07-18 03:24:46] step 607150/1000000 | loss=2.2447 vocab=0.7391 token=1.2997 | 6.09 steps/s | VRAM=29.36 Go [2026-07-18 03:24:54] step 607200/1000000 | loss=2.2570 vocab=0.6606 token=1.1266 | 6.14 steps/s | VRAM=31.57 Go [2026-07-18 03:25:02] step 607250/1000000 | loss=2.2169 vocab=0.9728 token=1.1653 | 6.16 steps/s | VRAM=26.40 Go [2026-07-18 03:25:10] step 607300/1000000 | loss=2.3351 vocab=1.0218 token=1.4402 | 6.13 steps/s | VRAM=32.33 Go [2026-07-18 03:25:18] step 607350/1000000 | loss=2.3909 vocab=1.9252 token=1.4035 | 6.10 steps/s | VRAM=29.32 Go [2026-07-18 03:25:27] step 607400/1000000 | loss=2.1940 vocab=1.4321 token=1.5493 | 6.13 steps/s | VRAM=28.67 Go [2026-07-18 03:25:35] step 607450/1000000 | loss=2.1804 vocab=0.6868 token=1.3494 | 6.14 steps/s | VRAM=29.04 Go [2026-07-18 03:25:43] step 607500/1000000 | loss=2.2669 vocab=0.8283 token=1.1955 | 6.13 steps/s | VRAM=32.09 Go [2026-07-18 03:25:51] step 607550/1000000 | loss=2.1657 vocab=1.4367 token=1.5277 | 6.12 steps/s | VRAM=25.45 Go [2026-07-18 03:25:59] step 607600/1000000 | loss=2.1631 vocab=0.8680 token=1.1223 | 6.13 steps/s | VRAM=27.87 Go [2026-07-18 03:26:07] step 607650/1000000 | loss=2.2457 vocab=0.5486 token=1.4339 | 6.12 steps/s | VRAM=32.39 Go [2026-07-18 03:26:16] step 607700/1000000 | loss=2.1826 vocab=0.8521 token=1.4290 | 6.13 steps/s | VRAM=26.66 Go [2026-07-18 03:26:24] step 607750/1000000 | loss=2.2290 vocab=0.7649 token=1.2335 | 6.12 steps/s | VRAM=26.16 Go [2026-07-18 03:26:32] step 607800/1000000 | loss=2.1988 vocab=1.0099 token=1.6878 | 6.13 steps/s | VRAM=28.54 Go [2026-07-18 03:26:40] step 607850/1000000 | loss=2.2634 vocab=0.8051 token=0.8966 | 6.10 steps/s | VRAM=26.73 Go [2026-07-18 03:26:48] step 607900/1000000 | loss=2.2303 vocab=0.5602 token=1.0682 | 6.15 steps/s | VRAM=27.60 Go [2026-07-18 03:26:56] step 607950/1000000 | loss=2.1985 vocab=0.5362 token=0.9565 | 6.10 steps/s | VRAM=25.79 Go [2026-07-18 03:27:05] step 608000/1000000 | loss=2.3410 vocab=1.4272 token=1.9225 | 6.15 steps/s | VRAM=31.60 Go [2026-07-18 03:27:13] step 608050/1000000 | loss=2.2428 vocab=1.1423 token=1.5126 | 6.13 steps/s | VRAM=27.95 Go [2026-07-18 03:27:21] step 608100/1000000 | loss=2.2461 vocab=0.7025 token=1.0733 | 6.15 steps/s | VRAM=28.73 Go [2026-07-18 03:27:29] step 608150/1000000 | loss=2.1675 vocab=0.5706 token=1.0294 | 6.13 steps/s | VRAM=28.21 Go [2026-07-18 03:27:37] step 608200/1000000 | loss=2.0999 vocab=0.6215 token=1.0631 | 6.09 steps/s | VRAM=27.82 Go [2026-07-18 03:27:45] step 608250/1000000 | loss=2.2415 vocab=0.8887 token=1.2992 | 6.12 steps/s | VRAM=26.20 Go [2026-07-18 03:27:54] step 608300/1000000 | loss=2.2795 vocab=0.7702 token=1.1072 | 6.08 steps/s | VRAM=25.94 Go [2026-07-18 03:28:02] step 608350/1000000 | loss=2.1843 vocab=1.3043 token=1.8888 | 6.14 steps/s | VRAM=24.91 Go [2026-07-18 03:28:10] step 608400/1000000 | loss=2.2506 vocab=0.9339 token=1.4820 | 6.14 steps/s | VRAM=24.80 Go [2026-07-18 03:28:18] step 608450/1000000 | loss=2.3375 vocab=1.9505 token=1.7539 | 6.13 steps/s | VRAM=32.34 Go [2026-07-18 03:28:26] step 608500/1000000 | loss=2.0831 vocab=1.1841 token=1.4181 | 6.13 steps/s | VRAM=27.87 Go [2026-07-18 03:28:34] step 608550/1000000 | loss=2.2079 vocab=0.6387 token=1.3595 | 6.09 steps/s | VRAM=26.92 Go [2026-07-18 03:28:43] step 608600/1000000 | loss=2.2575 vocab=1.3305 token=1.4552 | 6.10 steps/s | VRAM=29.48 Go [2026-07-18 03:28:51] step 608650/1000000 | loss=2.3143 vocab=0.6543 token=1.1093 | 6.11 steps/s | VRAM=28.22 Go [2026-07-18 03:28:59] step 608700/1000000 | loss=2.2475 vocab=1.5743 token=1.8563 | 6.11 steps/s | VRAM=27.20 Go [2026-07-18 03:29:07] step 608750/1000000 | loss=2.1588 vocab=1.1053 token=1.1898 | 6.15 steps/s | VRAM=31.22 Go [2026-07-18 03:29:15] step 608800/1000000 | loss=2.2774 vocab=0.7043 token=1.3441 | 6.13 steps/s | VRAM=28.19 Go [2026-07-18 03:29:23] step 608850/1000000 | loss=2.2190 vocab=0.9647 token=1.0459 | 6.13 steps/s | VRAM=31.69 Go [2026-07-18 03:29:32] step 608900/1000000 | loss=2.2803 vocab=0.9620 token=1.2192 | 6.12 steps/s | VRAM=26.67 Go [2026-07-18 03:29:40] step 608950/1000000 | loss=2.3433 vocab=1.0764 token=1.5030 | 6.16 steps/s | VRAM=31.18 Go [2026-07-18 03:29:48] step 609000/1000000 | loss=2.2671 vocab=0.5916 token=1.1153 | 6.11 steps/s | VRAM=25.87 Go [2026-07-18 03:29:56] step 609050/1000000 | loss=2.1798 vocab=0.9939 token=1.6911 | 6.14 steps/s | VRAM=28.64 Go [2026-07-18 03:30:04] step 609100/1000000 | loss=2.1310 vocab=1.2151 token=1.4318 | 6.15 steps/s | VRAM=25.73 Go [2026-07-18 03:30:12] step 609150/1000000 | loss=2.2466 vocab=0.6694 token=1.2866 | 6.10 steps/s | VRAM=28.74 Go [2026-07-18 03:30:20] step 609200/1000000 | loss=2.2791 vocab=0.8147 token=1.4406 | 6.14 steps/s | VRAM=25.44 Go [2026-07-18 03:30:29] step 609250/1000000 | loss=2.2471 vocab=0.8352 token=1.5056 | 6.14 steps/s | VRAM=32.04 Go [2026-07-18 03:30:37] step 609300/1000000 | loss=2.1408 vocab=0.4697 token=1.2467 | 6.13 steps/s | VRAM=31.92 Go [2026-07-18 03:30:45] step 609350/1000000 | loss=2.2553 vocab=0.6096 token=0.9172 | 6.08 steps/s | VRAM=27.77 Go [2026-07-18 03:30:53] step 609400/1000000 | loss=2.2316 vocab=0.5879 token=0.7724 | 6.13 steps/s | VRAM=28.82 Go [2026-07-18 03:31:01] step 609450/1000000 | loss=2.3475 vocab=0.8824 token=0.9115 | 6.15 steps/s | VRAM=25.76 Go [2026-07-18 03:31:09] step 609500/1000000 | loss=2.1984 vocab=0.4548 token=0.9228 | 6.17 steps/s | VRAM=28.90 Go [2026-07-18 03:31:18] step 609550/1000000 | loss=2.2586 vocab=1.0054 token=1.4142 | 6.15 steps/s | VRAM=24.39 Go [2026-07-18 03:31:26] step 609600/1000000 | loss=2.1450 vocab=1.0017 token=1.4794 | 6.10 steps/s | VRAM=24.94 Go [2026-07-18 03:31:34] step 609650/1000000 | loss=2.2166 vocab=0.8051 token=0.9217 | 6.12 steps/s | VRAM=24.84 Go [2026-07-18 03:31:42] step 609700/1000000 | loss=2.3306 vocab=0.5120 token=1.1239 | 6.11 steps/s | VRAM=29.18 Go [2026-07-18 03:31:50] step 609750/1000000 | loss=2.3854 vocab=0.8907 token=1.2000 | 6.13 steps/s | VRAM=31.01 Go [2026-07-18 03:31:58] step 609800/1000000 | loss=2.1539 vocab=0.6439 token=0.7019 | 6.15 steps/s | VRAM=28.92 Go [2026-07-18 03:32:07] step 609850/1000000 | loss=2.2070 vocab=0.7543 token=0.7539 | 6.11 steps/s | VRAM=28.92 Go [2026-07-18 03:32:15] step 609900/1000000 | loss=2.2149 vocab=0.5686 token=1.2426 | 6.13 steps/s | VRAM=32.21 Go [2026-07-18 03:32:23] step 609950/1000000 | loss=2.2969 vocab=1.6564 token=1.8007 | 6.11 steps/s | VRAM=29.39 Go [2026-07-18 03:32:31] step 610000/1000000 | loss=2.2337 vocab=1.0364 token=1.2851 | 6.09 steps/s | VRAM=23.45 Go [2026-07-18 03:32:39] step 610050/1000000 | loss=2.1622 vocab=1.4597 token=1.4163 | 6.16 steps/s | VRAM=30.77 Go [2026-07-18 03:32:47] step 610100/1000000 | loss=2.2648 vocab=0.9718 token=1.1193 | 6.12 steps/s | VRAM=31.29 Go [2026-07-18 03:32:56] step 610150/1000000 | loss=2.3129 vocab=0.7921 token=1.5054 | 6.13 steps/s | VRAM=27.80 Go [2026-07-18 03:33:04] step 610200/1000000 | loss=2.2695 vocab=1.3178 token=1.4948 | 6.12 steps/s | VRAM=26.64 Go [2026-07-18 03:33:12] step 610250/1000000 | loss=2.3199 vocab=0.5812 token=0.9156 | 6.13 steps/s | VRAM=26.24 Go [2026-07-18 03:33:20] step 610300/1000000 | loss=2.1537 vocab=0.8736 token=1.6992 | 6.13 steps/s | VRAM=28.83 Go [2026-07-18 03:33:28] step 610350/1000000 | loss=2.2195 vocab=1.2750 token=1.5544 | 6.17 steps/s | VRAM=26.11 Go [2026-07-18 03:33:36] step 610400/1000000 | loss=2.3376 vocab=0.9805 token=1.6381 | 6.13 steps/s | VRAM=24.65 Go [2026-07-18 03:33:44] step 610450/1000000 | loss=2.1756 vocab=0.8191 token=1.6124 | 6.14 steps/s | VRAM=28.71 Go [2026-07-18 03:33:53] step 610500/1000000 | loss=2.2084 vocab=0.1487 token=0.9444 | 6.13 steps/s | VRAM=25.81 Go [2026-07-18 03:34:01] step 610550/1000000 | loss=2.2520 vocab=0.5875 token=1.2242 | 6.14 steps/s | VRAM=25.79 Go [2026-07-18 03:34:09] step 610600/1000000 | loss=2.2781 vocab=0.8406 token=1.3873 | 6.15 steps/s | VRAM=28.54 Go [2026-07-18 03:34:17] step 610650/1000000 | loss=2.3125 vocab=0.8325 token=1.2418 | 6.14 steps/s | VRAM=26.34 Go [2026-07-18 03:34:25] step 610700/1000000 | loss=2.2049 vocab=0.5219 token=1.2086 | 6.14 steps/s | VRAM=27.81 Go [2026-07-18 03:34:33] step 610750/1000000 | loss=2.1565 vocab=1.2853 token=1.4542 | 6.11 steps/s | VRAM=26.43 Go [2026-07-18 03:34:41] step 610800/1000000 | loss=2.0460 vocab=0.8845 token=1.3311 | 6.14 steps/s | VRAM=31.44 Go [2026-07-18 03:34:50] step 610850/1000000 | loss=2.3378 vocab=1.3505 token=1.5079 | 6.14 steps/s | VRAM=28.17 Go [2026-07-18 03:34:58] step 610900/1000000 | loss=2.1820 vocab=0.9823 token=1.1801 | 6.15 steps/s | VRAM=28.83 Go [2026-07-18 03:35:06] step 610950/1000000 | loss=2.2262 vocab=2.4236 token=1.1881 | 6.13 steps/s | VRAM=26.33 Go [2026-07-18 03:35:14] step 611000/1000000 | loss=2.1683 vocab=0.9431 token=1.2745 | 6.14 steps/s | VRAM=26.86 Go [2026-07-18 03:35:22] step 611050/1000000 | loss=2.2023 vocab=0.6644 token=1.1311 | 6.15 steps/s | VRAM=29.09 Go [2026-07-18 03:35:30] step 611100/1000000 | loss=2.2085 vocab=0.7152 token=1.5683 | 6.11 steps/s | VRAM=30.21 Go [2026-07-18 03:35:39] step 611150/1000000 | loss=2.2619 vocab=0.8579 token=1.4042 | 6.12 steps/s | VRAM=28.65 Go [2026-07-18 03:35:47] step 611200/1000000 | loss=2.1864 vocab=0.7912 token=1.0860 | 6.14 steps/s | VRAM=26.30 Go [2026-07-18 03:35:55] step 611250/1000000 | loss=2.2653 vocab=0.7704 token=1.1705 | 6.13 steps/s | VRAM=28.75 Go [2026-07-18 03:36:03] step 611300/1000000 | loss=2.1447 vocab=0.9049 token=1.1886 | 6.14 steps/s | VRAM=29.11 Go [2026-07-18 03:36:11] step 611350/1000000 | loss=2.2960 vocab=0.7594 token=1.2180 | 6.12 steps/s | VRAM=30.58 Go [2026-07-18 03:36:19] step 611400/1000000 | loss=2.2423 vocab=0.9346 token=1.3904 | 6.13 steps/s | VRAM=28.25 Go [2026-07-18 03:36:27] step 611450/1000000 | loss=2.1932 vocab=1.2983 token=1.2372 | 6.10 steps/s | VRAM=25.94 Go [2026-07-18 03:36:36] step 611500/1000000 | loss=2.1715 vocab=1.0670 token=1.1800 | 6.12 steps/s | VRAM=32.27 Go [2026-07-18 03:36:44] step 611550/1000000 | loss=2.3026 vocab=1.0210 token=1.3493 | 6.12 steps/s | VRAM=32.11 Go [2026-07-18 03:36:52] step 611600/1000000 | loss=2.2348 vocab=1.0545 token=1.5417 | 6.10 steps/s | VRAM=28.85 Go [2026-07-18 03:37:00] step 611650/1000000 | loss=2.2556 vocab=0.9280 token=0.9702 | 6.10 steps/s | VRAM=26.34 Go [2026-07-18 03:37:08] step 611700/1000000 | loss=2.2040 vocab=0.5583 token=1.4091 | 6.17 steps/s | VRAM=29.15 Go [2026-07-18 03:37:16] step 611750/1000000 | loss=2.1598 vocab=0.6171 token=1.3808 | 6.15 steps/s | VRAM=32.22 Go [2026-07-18 03:37:25] step 611800/1000000 | loss=2.2973 vocab=0.7635 token=1.1809 | 6.09 steps/s | VRAM=29.92 Go [2026-07-18 03:37:33] step 611850/1000000 | loss=2.1062 vocab=1.0795 token=1.2092 | 6.14 steps/s | VRAM=28.98 Go [2026-07-18 03:37:41] step 611900/1000000 | loss=2.1920 vocab=1.1834 token=1.3736 | 6.15 steps/s | VRAM=32.02 Go [2026-07-18 03:37:49] step 611950/1000000 | loss=2.3111 vocab=0.5999 token=1.1711 | 6.12 steps/s | VRAM=29.42 Go [2026-07-18 03:37:57] step 612000/1000000 | loss=2.2947 vocab=0.7982 token=1.1569 | 6.18 steps/s | VRAM=25.58 Go [2026-07-18 03:38:05] step 612050/1000000 | loss=2.1977 vocab=1.6246 token=1.3347 | 6.15 steps/s | VRAM=29.17 Go [2026-07-18 03:38:13] step 612100/1000000 | loss=2.2268 vocab=0.8840 token=1.1067 | 6.16 steps/s | VRAM=31.53 Go [2026-07-18 03:38:22] step 612150/1000000 | loss=2.2120 vocab=0.7051 token=1.1532 | 6.15 steps/s | VRAM=32.72 Go [2026-07-18 03:38:30] step 612200/1000000 | loss=2.2599 vocab=0.9874 token=1.1321 | 6.14 steps/s | VRAM=29.87 Go [2026-07-18 03:38:38] step 612250/1000000 | loss=2.1847 vocab=0.5699 token=1.2630 | 6.12 steps/s | VRAM=27.77 Go [2026-07-18 03:38:46] step 612300/1000000 | loss=2.1143 vocab=0.6772 token=1.2751 | 6.14 steps/s | VRAM=28.96 Go [2026-07-18 03:38:54] step 612350/1000000 | loss=2.2775 vocab=0.8379 token=1.1188 | 6.12 steps/s | VRAM=29.47 Go [2026-07-18 03:39:02] step 612400/1000000 | loss=2.2251 vocab=1.1262 token=1.1742 | 6.16 steps/s | VRAM=32.07 Go [2026-07-18 03:39:10] step 612450/1000000 | loss=2.2022 vocab=1.1689 token=1.1943 | 6.13 steps/s | VRAM=32.88 Go [2026-07-18 03:39:19] step 612500/1000000 | loss=2.2543 vocab=1.0750 token=1.5005 | 6.14 steps/s | VRAM=25.86 Go [2026-07-18 03:39:27] step 612550/1000000 | loss=2.1878 vocab=0.9500 token=1.5875 | 6.15 steps/s | VRAM=28.65 Go [2026-07-18 03:39:35] step 612600/1000000 | loss=2.2295 vocab=0.7582 token=1.2159 | 6.12 steps/s | VRAM=30.38 Go [2026-07-18 03:39:43] step 612650/1000000 | loss=2.1609 vocab=0.7100 token=1.3020 | 6.13 steps/s | VRAM=32.74 Go [2026-07-18 03:39:51] step 612700/1000000 | loss=2.1508 vocab=0.8887 token=1.6129 | 6.14 steps/s | VRAM=31.72 Go [2026-07-18 03:39:59] step 612750/1000000 | loss=2.1577 vocab=1.0832 token=1.4373 | 6.13 steps/s | VRAM=32.93 Go [2026-07-18 03:40:08] step 612800/1000000 | loss=2.3027 vocab=0.9418 token=1.3301 | 6.14 steps/s | VRAM=26.09 Go [2026-07-18 03:40:16] step 612850/1000000 | loss=2.1757 vocab=0.5635 token=1.4008 | 6.15 steps/s | VRAM=28.64 Go [2026-07-18 03:40:24] step 612900/1000000 | loss=2.2402 vocab=0.7961 token=1.2085 | 6.15 steps/s | VRAM=28.24 Go [2026-07-18 03:40:32] step 612950/1000000 | loss=2.2395 vocab=1.4307 token=1.6079 | 6.13 steps/s | VRAM=26.42 Go [2026-07-18 03:40:40] step 613000/1000000 | loss=2.2500 vocab=0.9371 token=1.3356 | 6.11 steps/s | VRAM=28.72 Go [2026-07-18 03:40:48] step 613050/1000000 | loss=2.1721 vocab=1.0301 token=1.2912 | 6.14 steps/s | VRAM=32.40 Go [2026-07-18 03:40:56] step 613100/1000000 | loss=2.2925 vocab=0.8608 token=1.4272 | 6.13 steps/s | VRAM=26.07 Go [2026-07-18 03:41:05] step 613150/1000000 | loss=2.1809 vocab=0.9653 token=1.0214 | 6.10 steps/s | VRAM=26.82 Go [2026-07-18 03:41:13] step 613200/1000000 | loss=2.2034 vocab=1.0039 token=1.2365 | 6.11 steps/s | VRAM=28.95 Go [2026-07-18 03:41:21] step 613250/1000000 | loss=2.1964 vocab=0.6019 token=0.9244 | 6.10 steps/s | VRAM=28.28 Go [2026-07-18 03:41:29] step 613300/1000000 | loss=2.2707 vocab=1.3304 token=1.4603 | 6.13 steps/s | VRAM=32.16 Go [2026-07-18 03:41:37] step 613350/1000000 | loss=2.2716 vocab=0.6551 token=1.6357 | 6.13 steps/s | VRAM=26.38 Go [2026-07-18 03:41:45] step 613400/1000000 | loss=2.2045 vocab=0.9145 token=1.3200 | 6.13 steps/s | VRAM=26.22 Go [2026-07-18 03:41:54] step 613450/1000000 | loss=2.1676 vocab=1.0029 token=1.7596 | 6.11 steps/s | VRAM=28.96 Go [2026-07-18 03:42:02] step 613500/1000000 | loss=2.1668 vocab=0.6811 token=1.3817 | 6.13 steps/s | VRAM=28.42 Go [2026-07-18 03:42:10] step 613550/1000000 | loss=2.2272 vocab=0.6810 token=0.9566 | 6.14 steps/s | VRAM=28.73 Go [2026-07-18 03:42:18] step 613600/1000000 | loss=2.2370 vocab=0.8552 token=1.6538 | 6.12 steps/s | VRAM=24.87 Go [2026-07-18 03:42:26] step 613650/1000000 | loss=2.1592 vocab=1.2164 token=1.4128 | 6.11 steps/s | VRAM=28.20 Go [2026-07-18 03:42:34] step 613700/1000000 | loss=2.1977 vocab=1.2482 token=1.6046 | 6.15 steps/s | VRAM=31.86 Go [2026-07-18 03:42:43] step 613750/1000000 | loss=2.2472 vocab=0.9160 token=1.2810 | 6.11 steps/s | VRAM=29.59 Go [2026-07-18 03:42:51] step 613800/1000000 | loss=2.2664 vocab=0.6398 token=1.4914 | 6.14 steps/s | VRAM=28.12 Go [2026-07-18 03:42:59] step 613850/1000000 | loss=2.2083 vocab=0.7972 token=1.2554 | 6.13 steps/s | VRAM=27.24 Go [2026-07-18 03:43:07] step 613900/1000000 | loss=2.1646 vocab=0.7298 token=1.3331 | 6.12 steps/s | VRAM=31.53 Go [2026-07-18 03:43:15] step 613950/1000000 | loss=2.2521 vocab=1.6033 token=1.0618 | 6.13 steps/s | VRAM=31.12 Go [2026-07-18 03:43:23] step 614000/1000000 | loss=2.3361 vocab=1.1154 token=1.5798 | 6.14 steps/s | VRAM=29.18 Go [2026-07-18 03:43:32] step 614050/1000000 | loss=2.1955 vocab=0.9958 token=1.4310 | 6.12 steps/s | VRAM=24.83 Go [2026-07-18 03:43:40] step 614100/1000000 | loss=2.2665 vocab=0.9949 token=1.4016 | 6.13 steps/s | VRAM=28.88 Go [2026-07-18 03:43:48] step 614150/1000000 | loss=2.2410 vocab=0.7046 token=1.5539 | 6.10 steps/s | VRAM=23.81 Go [2026-07-18 03:43:56] step 614200/1000000 | loss=2.2553 vocab=1.0683 token=1.8814 | 6.14 steps/s | VRAM=28.34 Go [2026-07-18 03:44:04] step 614250/1000000 | loss=2.2034 vocab=1.1193 token=1.2080 | 6.11 steps/s | VRAM=28.23 Go [2026-07-18 03:44:12] step 614300/1000000 | loss=2.3558 vocab=0.6748 token=1.7098 | 6.11 steps/s | VRAM=28.25 Go [2026-07-18 03:44:21] step 614350/1000000 | loss=2.1980 vocab=0.4340 token=1.3246 | 6.11 steps/s | VRAM=27.87 Go [2026-07-18 03:44:29] step 614400/1000000 | loss=2.2164 vocab=0.9229 token=1.8797 | 6.13 steps/s | VRAM=29.99 Go [2026-07-18 03:44:37] step 614450/1000000 | loss=2.2250 vocab=1.1720 token=1.2639 | 6.10 steps/s | VRAM=28.86 Go [2026-07-18 03:44:45] step 614500/1000000 | loss=2.1848 vocab=0.6163 token=1.1062 | 6.12 steps/s | VRAM=24.31 Go [2026-07-18 03:44:53] step 614550/1000000 | loss=2.1896 vocab=0.7975 token=0.8077 | 6.14 steps/s | VRAM=32.43 Go [2026-07-18 03:45:01] step 614600/1000000 | loss=2.3129 vocab=0.5340 token=1.2187 | 6.12 steps/s | VRAM=25.55 Go [2026-07-18 03:45:10] step 614650/1000000 | loss=2.2336 vocab=0.9873 token=1.4265 | 6.11 steps/s | VRAM=29.00 Go [2026-07-18 03:45:18] step 614700/1000000 | loss=2.1969 vocab=1.2496 token=1.4306 | 6.15 steps/s | VRAM=27.81 Go [2026-07-18 03:45:26] step 614750/1000000 | loss=2.1038 vocab=0.7979 token=1.2202 | 6.13 steps/s | VRAM=29.17 Go [2026-07-18 03:45:34] step 614800/1000000 | loss=2.2989 vocab=0.9878 token=1.7228 | 6.14 steps/s | VRAM=26.24 Go [2026-07-18 03:45:42] step 614850/1000000 | loss=2.1749 vocab=0.4881 token=0.9390 | 6.11 steps/s | VRAM=28.77 Go [2026-07-18 03:45:50] step 614900/1000000 | loss=2.2367 vocab=0.8058 token=1.6691 | 6.12 steps/s | VRAM=32.55 Go [2026-07-18 03:45:59] step 614950/1000000 | loss=2.2289 vocab=0.7968 token=1.1699 | 6.13 steps/s | VRAM=31.87 Go [2026-07-18 03:46:07] step 615000/1000000 | loss=2.2955 vocab=0.6945 token=1.0403 | 6.12 steps/s | VRAM=28.65 Go [2026-07-18 03:46:15] step 615050/1000000 | loss=2.2718 vocab=1.5426 token=1.4412 | 6.14 steps/s | VRAM=24.49 Go [2026-07-18 03:46:23] step 615100/1000000 | loss=2.1975 vocab=0.7071 token=1.3573 | 6.17 steps/s | VRAM=30.53 Go [2026-07-18 03:46:31] step 615150/1000000 | loss=2.3098 vocab=0.8417 token=1.5524 | 6.13 steps/s | VRAM=31.01 Go [2026-07-18 03:46:39] step 615200/1000000 | loss=2.1731 vocab=1.0647 token=1.0752 | 6.11 steps/s | VRAM=29.66 Go [2026-07-18 03:46:47] step 615250/1000000 | loss=2.2140 vocab=1.0413 token=1.2392 | 6.13 steps/s | VRAM=28.25 Go [2026-07-18 03:46:56] step 615300/1000000 | loss=2.1224 vocab=0.7098 token=1.0545 | 6.13 steps/s | VRAM=32.41 Go [2026-07-18 03:47:04] step 615350/1000000 | loss=2.1623 vocab=0.9528 token=1.5066 | 6.09 steps/s | VRAM=25.93 Go [2026-07-18 03:47:12] step 615400/1000000 | loss=2.2064 vocab=0.8204 token=1.1976 | 6.16 steps/s | VRAM=31.92 Go [2026-07-18 03:47:20] step 615450/1000000 | loss=2.1688 vocab=0.9037 token=1.3428 | 6.12 steps/s | VRAM=28.61 Go [2026-07-18 03:47:28] step 615500/1000000 | loss=2.3411 vocab=0.5741 token=1.1875 | 6.13 steps/s | VRAM=29.61 Go [2026-07-18 03:47:36] step 615550/1000000 | loss=2.1464 vocab=0.7530 token=1.4074 | 6.09 steps/s | VRAM=29.07 Go [2026-07-18 03:47:45] step 615600/1000000 | loss=2.1775 vocab=0.7894 token=1.0481 | 6.14 steps/s | VRAM=31.30 Go [2026-07-18 03:47:53] step 615650/1000000 | loss=2.3185 vocab=0.7896 token=1.1732 | 6.13 steps/s | VRAM=28.82 Go [2026-07-18 03:48:01] step 615700/1000000 | loss=2.2280 vocab=1.0295 token=1.1266 | 6.12 steps/s | VRAM=25.45 Go [2026-07-18 03:48:09] step 615750/1000000 | loss=2.1270 vocab=0.6532 token=1.4387 | 6.15 steps/s | VRAM=32.20 Go [2026-07-18 03:48:17] step 615800/1000000 | loss=2.2449 vocab=1.7545 token=1.4475 | 6.12 steps/s | VRAM=23.50 Go [2026-07-18 03:48:25] step 615850/1000000 | loss=2.2963 vocab=0.6058 token=1.0459 | 6.13 steps/s | VRAM=28.49 Go [2026-07-18 03:48:34] step 615900/1000000 | loss=2.3002 vocab=1.0487 token=1.3684 | 6.14 steps/s | VRAM=31.55 Go [2026-07-18 03:48:42] step 615950/1000000 | loss=2.1716 vocab=1.0466 token=1.6306 | 6.15 steps/s | VRAM=32.46 Go [2026-07-18 03:48:50] step 616000/1000000 | loss=2.3604 vocab=0.6600 token=1.2379 | 6.14 steps/s | VRAM=24.82 Go [2026-07-18 03:48:58] step 616050/1000000 | loss=2.1854 vocab=0.7538 token=0.9802 | 6.12 steps/s | VRAM=26.58 Go [2026-07-18 03:49:06] step 616100/1000000 | loss=2.2247 vocab=0.8812 token=1.1690 | 6.14 steps/s | VRAM=24.47 Go [2026-07-18 03:49:14] step 616150/1000000 | loss=2.0886 vocab=0.5781 token=1.2077 | 6.09 steps/s | VRAM=30.04 Go [2026-07-18 03:49:23] step 616200/1000000 | loss=2.2211 vocab=1.1918 token=1.1078 | 6.08 steps/s | VRAM=28.91 Go [2026-07-18 03:49:31] step 616250/1000000 | loss=2.2121 vocab=0.6820 token=1.2648 | 6.10 steps/s | VRAM=28.70 Go [2026-07-18 03:49:39] step 616300/1000000 | loss=2.1665 vocab=1.1013 token=1.5042 | 6.10 steps/s | VRAM=26.36 Go [2026-07-18 03:49:47] step 616350/1000000 | loss=2.2362 vocab=0.9778 token=1.8203 | 6.13 steps/s | VRAM=28.75 Go [2026-07-18 03:49:55] step 616400/1000000 | loss=2.3247 vocab=1.0404 token=1.1632 | 6.13 steps/s | VRAM=28.96 Go [2026-07-18 03:50:03] step 616450/1000000 | loss=2.2332 vocab=0.7216 token=1.3486 | 6.13 steps/s | VRAM=32.63 Go [2026-07-18 03:50:12] step 616500/1000000 | loss=2.1979 vocab=0.5759 token=1.4913 | 6.11 steps/s | VRAM=26.47 Go [2026-07-18 03:50:20] step 616550/1000000 | loss=2.1001 vocab=0.7517 token=1.2637 | 6.10 steps/s | VRAM=29.20 Go [2026-07-18 03:50:28] step 616600/1000000 | loss=2.1995 vocab=0.9997 token=1.5649 | 6.14 steps/s | VRAM=28.45 Go [2026-07-18 03:50:36] step 616650/1000000 | loss=2.2624 vocab=1.0542 token=1.4415 | 6.11 steps/s | VRAM=31.46 Go [2026-07-18 03:50:44] step 616700/1000000 | loss=2.2056 vocab=0.9579 token=1.5002 | 6.13 steps/s | VRAM=29.44 Go [2026-07-18 03:50:52] step 616750/1000000 | loss=2.1764 vocab=0.9268 token=1.4047 | 6.12 steps/s | VRAM=32.52 Go [2026-07-18 03:51:01] step 616800/1000000 | loss=2.2793 vocab=1.1229 token=1.3510 | 6.11 steps/s | VRAM=28.00 Go [2026-07-18 03:51:09] step 616850/1000000 | loss=2.1781 vocab=1.5343 token=1.2493 | 6.15 steps/s | VRAM=28.76 Go [2026-07-18 03:51:17] step 616900/1000000 | loss=2.2934 vocab=1.1536 token=1.5787 | 6.11 steps/s | VRAM=31.65 Go [2026-07-18 03:51:25] step 616950/1000000 | loss=2.2638 vocab=1.0238 token=1.0357 | 6.11 steps/s | VRAM=29.33 Go [2026-07-18 03:51:33] step 617000/1000000 | loss=2.1514 vocab=0.7772 token=1.3163 | 6.12 steps/s | VRAM=26.55 Go [2026-07-18 03:51:41] step 617050/1000000 | loss=2.2109 vocab=1.2304 token=1.8769 | 6.13 steps/s | VRAM=32.78 Go [2026-07-18 03:51:50] step 617100/1000000 | loss=2.1356 vocab=0.5817 token=1.2012 | 6.13 steps/s | VRAM=27.88 Go [2026-07-18 03:51:58] step 617150/1000000 | loss=2.3289 vocab=1.1559 token=1.7397 | 6.13 steps/s | VRAM=29.37 Go [2026-07-18 03:52:06] step 617200/1000000 | loss=2.2158 vocab=1.0920 token=1.5282 | 6.12 steps/s | VRAM=29.50 Go [2026-07-18 03:52:14] step 617250/1000000 | loss=2.2620 vocab=0.4529 token=0.9028 | 6.17 steps/s | VRAM=28.84 Go [2026-07-18 03:52:22] step 617300/1000000 | loss=2.3580 vocab=1.0185 token=1.4656 | 6.13 steps/s | VRAM=25.34 Go [2026-07-18 03:52:30] step 617350/1000000 | loss=2.2564 vocab=0.7649 token=1.3610 | 6.14 steps/s | VRAM=26.02 Go [2026-07-18 03:52:39] step 617400/1000000 | loss=2.3094 vocab=1.2155 token=1.7098 | 6.12 steps/s | VRAM=28.55 Go [2026-07-18 03:52:47] step 617450/1000000 | loss=2.1694 vocab=0.9252 token=1.7695 | 6.15 steps/s | VRAM=31.39 Go [2026-07-18 03:52:55] step 617500/1000000 | loss=2.1341 vocab=0.9514 token=1.7589 | 6.12 steps/s | VRAM=29.07 Go [2026-07-18 03:53:03] step 617550/1000000 | loss=2.2257 vocab=1.4325 token=1.5873 | 6.14 steps/s | VRAM=29.11 Go [2026-07-18 03:53:11] step 617600/1000000 | loss=2.2428 vocab=1.1354 token=1.4924 | 6.15 steps/s | VRAM=27.76 Go [2026-07-18 03:53:19] step 617650/1000000 | loss=2.1453 vocab=0.3877 token=1.0917 | 6.13 steps/s | VRAM=28.96 Go [2026-07-18 03:53:27] step 617700/1000000 | loss=2.2102 vocab=0.6043 token=1.2669 | 6.14 steps/s | VRAM=32.77 Go [2026-07-18 03:53:36] step 617750/1000000 | loss=2.1191 vocab=0.5594 token=1.1497 | 6.12 steps/s | VRAM=29.79 Go [2026-07-18 03:53:44] step 617800/1000000 | loss=2.1122 vocab=0.8872 token=1.1471 | 6.15 steps/s | VRAM=28.04 Go [2026-07-18 03:53:52] step 617850/1000000 | loss=2.3506 vocab=2.0924 token=1.5554 | 6.11 steps/s | VRAM=29.31 Go [2026-07-18 03:54:00] step 617900/1000000 | loss=2.1976 vocab=0.5374 token=1.2298 | 6.13 steps/s | VRAM=30.33 Go [2026-07-18 03:54:08] step 617950/1000000 | loss=2.2630 vocab=1.5732 token=1.6860 | 6.16 steps/s | VRAM=32.06 Go [2026-07-18 03:54:16] step 618000/1000000 | loss=2.3194 vocab=0.8631 token=1.4056 | 6.11 steps/s | VRAM=23.88 Go [2026-07-18 03:54:24] step 618050/1000000 | loss=2.1909 vocab=0.8259 token=1.4923 | 6.18 steps/s | VRAM=32.33 Go [2026-07-18 03:54:33] step 618100/1000000 | loss=2.1502 vocab=0.6611 token=1.2111 | 6.16 steps/s | VRAM=32.50 Go [2026-07-18 03:54:41] step 618150/1000000 | loss=2.3061 vocab=0.3076 token=1.3853 | 6.13 steps/s | VRAM=25.98 Go [2026-07-18 03:54:49] step 618200/1000000 | loss=2.3311 vocab=0.9518 token=1.5041 | 6.16 steps/s | VRAM=24.54 Go [2026-07-18 03:54:57] step 618250/1000000 | loss=2.2155 vocab=1.0149 token=1.4752 | 6.14 steps/s | VRAM=27.86 Go [2026-07-18 03:55:05] step 618300/1000000 | loss=2.2465 vocab=1.0897 token=1.6327 | 6.14 steps/s | VRAM=28.80 Go [2026-07-18 03:55:13] step 618350/1000000 | loss=2.2451 vocab=0.5797 token=1.0015 | 6.15 steps/s | VRAM=28.32 Go [2026-07-18 03:55:21] step 618400/1000000 | loss=2.1954 vocab=1.0590 token=1.4209 | 6.11 steps/s | VRAM=31.40 Go [2026-07-18 03:55:30] step 618450/1000000 | loss=2.1635 vocab=0.9129 token=1.2668 | 6.12 steps/s | VRAM=25.82 Go [2026-07-18 03:55:38] step 618500/1000000 | loss=2.2283 vocab=0.9704 token=1.6123 | 6.15 steps/s | VRAM=26.52 Go [2026-07-18 03:55:46] step 618550/1000000 | loss=2.2574 vocab=0.7665 token=1.1605 | 6.13 steps/s | VRAM=28.31 Go [2026-07-18 03:55:54] step 618600/1000000 | loss=2.2591 vocab=0.5815 token=1.2392 | 6.12 steps/s | VRAM=29.08 Go [2026-07-18 03:56:02] step 618650/1000000 | loss=2.2105 vocab=0.8499 token=1.4410 | 6.11 steps/s | VRAM=25.99 Go [2026-07-18 03:56:10] step 618700/1000000 | loss=2.1952 vocab=0.8442 token=1.2317 | 6.12 steps/s | VRAM=32.30 Go [2026-07-18 03:56:19] step 618750/1000000 | loss=2.2451 vocab=0.7243 token=0.9925 | 6.14 steps/s | VRAM=28.83 Go [2026-07-18 03:56:27] step 618800/1000000 | loss=2.2709 vocab=0.5476 token=1.5248 | 6.13 steps/s | VRAM=30.24 Go [2026-07-18 03:56:35] step 618850/1000000 | loss=2.2031 vocab=0.5878 token=1.1748 | 6.14 steps/s | VRAM=25.31 Go [2026-07-18 03:56:43] step 618900/1000000 | loss=2.2750 vocab=1.5171 token=1.3028 | 6.16 steps/s | VRAM=32.28 Go [2026-07-18 03:56:51] step 618950/1000000 | loss=2.1515 vocab=0.5255 token=1.3395 | 6.11 steps/s | VRAM=28.54 Go [2026-07-18 03:56:59] step 619000/1000000 | loss=2.1953 vocab=1.0089 token=1.3067 | 6.16 steps/s | VRAM=28.04 Go [2026-07-18 03:57:07] step 619050/1000000 | loss=2.1809 vocab=0.8100 token=0.9694 | 6.13 steps/s | VRAM=32.05 Go [2026-07-18 03:57:16] step 619100/1000000 | loss=2.2302 vocab=1.0577 token=1.5495 | 6.13 steps/s | VRAM=27.94 Go [2026-07-18 03:57:24] step 619150/1000000 | loss=2.2947 vocab=0.7112 token=1.4026 | 6.14 steps/s | VRAM=28.67 Go [2026-07-18 03:57:32] step 619200/1000000 | loss=2.2041 vocab=0.8647 token=1.2181 | 6.15 steps/s | VRAM=28.02 Go [2026-07-18 03:57:40] step 619250/1000000 | loss=2.2038 vocab=0.5746 token=1.3359 | 6.11 steps/s | VRAM=26.52 Go [2026-07-18 03:57:48] step 619300/1000000 | loss=2.2093 vocab=0.3730 token=1.0011 | 6.13 steps/s | VRAM=28.32 Go [2026-07-18 03:57:56] step 619350/1000000 | loss=2.2145 vocab=0.8172 token=1.2019 | 6.15 steps/s | VRAM=25.99 Go [2026-07-18 03:58:05] step 619400/1000000 | loss=2.2087 vocab=1.1593 token=1.2323 | 6.15 steps/s | VRAM=31.64 Go [2026-07-18 03:58:13] step 619450/1000000 | loss=2.1713 vocab=1.4382 token=1.5955 | 6.15 steps/s | VRAM=32.60 Go [2026-07-18 03:58:21] step 619500/1000000 | loss=2.1971 vocab=0.3897 token=0.9185 | 6.11 steps/s | VRAM=32.29 Go [2026-07-18 03:58:29] step 619550/1000000 | loss=2.2873 vocab=1.0265 token=1.1366 | 6.14 steps/s | VRAM=28.60 Go [2026-07-18 03:58:37] step 619600/1000000 | loss=2.2073 vocab=0.9277 token=1.6236 | 6.12 steps/s | VRAM=26.19 Go [2026-07-18 03:58:45] step 619650/1000000 | loss=2.3002 vocab=1.3573 token=1.6671 | 6.15 steps/s | VRAM=27.96 Go [2026-07-18 03:58:53] step 619700/1000000 | loss=2.2481 vocab=0.9845 token=1.5219 | 6.16 steps/s | VRAM=32.41 Go [2026-07-18 03:59:02] step 619750/1000000 | loss=2.1038 vocab=0.8600 token=1.4578 | 6.16 steps/s | VRAM=28.17 Go [2026-07-18 03:59:10] step 619800/1000000 | loss=2.2178 vocab=0.9072 token=1.1355 | 6.10 steps/s | VRAM=26.81 Go [2026-07-18 03:59:18] step 619850/1000000 | loss=2.3057 vocab=0.8957 token=1.2548 | 6.15 steps/s | VRAM=31.90 Go [2026-07-18 03:59:26] step 619900/1000000 | loss=2.1559 vocab=0.8162 token=1.6813 | 6.14 steps/s | VRAM=26.02 Go [2026-07-18 03:59:34] step 619950/1000000 | loss=2.1787 vocab=0.8400 token=1.1453 | 6.12 steps/s | VRAM=29.75 Go [2026-07-18 03:59:42] step 620000/1000000 | loss=2.1645 vocab=0.9020 token=1.2684 | 6.15 steps/s | VRAM=31.30 Go [2026-07-18 03:59:50] step 620050/1000000 | loss=2.2439 vocab=0.6932 token=1.1541 | 6.15 steps/s | VRAM=28.36 Go [2026-07-18 03:59:59] step 620100/1000000 | loss=2.2365 vocab=0.7341 token=1.1973 | 6.15 steps/s | VRAM=28.99 Go [2026-07-18 04:00:07] step 620150/1000000 | loss=2.1694 vocab=0.6953 token=1.5055 | 6.13 steps/s | VRAM=26.34 Go [2026-07-18 04:00:15] step 620200/1000000 | loss=2.2723 vocab=0.7771 token=0.6920 | 6.12 steps/s | VRAM=25.43 Go [2026-07-18 04:00:23] step 620250/1000000 | loss=2.1791 vocab=1.1490 token=1.1465 | 6.20 steps/s | VRAM=32.80 Go [2026-07-18 04:00:31] step 620300/1000000 | loss=2.3568 vocab=1.1817 token=1.2787 | 6.17 steps/s | VRAM=26.30 Go [2026-07-18 04:00:39] step 620350/1000000 | loss=2.1467 vocab=0.8139 token=1.2518 | 6.15 steps/s | VRAM=28.59 Go [2026-07-18 04:00:47] step 620400/1000000 | loss=2.2970 vocab=0.8203 token=1.0331 | 6.15 steps/s | VRAM=32.87 Go [2026-07-18 04:00:55] step 620450/1000000 | loss=2.2612 vocab=0.9560 token=1.4655 | 6.13 steps/s | VRAM=26.28 Go [2026-07-18 04:01:04] step 620500/1000000 | loss=2.3082 vocab=0.5406 token=1.0593 | 6.15 steps/s | VRAM=28.03 Go [2026-07-18 04:01:12] step 620550/1000000 | loss=2.2484 vocab=1.0110 token=1.4902 | 6.13 steps/s | VRAM=27.04 Go [2026-07-18 04:01:20] step 620600/1000000 | loss=2.0346 vocab=0.7308 token=1.1645 | 6.16 steps/s | VRAM=31.61 Go [2026-07-18 04:01:28] step 620650/1000000 | loss=2.0960 vocab=0.5302 token=1.1749 | 6.13 steps/s | VRAM=27.93 Go [2026-07-18 04:01:36] step 620700/1000000 | loss=2.2444 vocab=0.7380 token=1.1275 | 6.13 steps/s | VRAM=29.30 Go [2026-07-18 04:01:44] step 620750/1000000 | loss=2.2001 vocab=0.8898 token=1.3203 | 6.15 steps/s | VRAM=31.39 Go [2026-07-18 04:01:52] step 620800/1000000 | loss=2.1736 vocab=0.8313 token=0.8128 | 6.13 steps/s | VRAM=29.29 Go [2026-07-18 04:02:01] step 620850/1000000 | loss=2.3117 vocab=1.2805 token=1.6106 | 6.16 steps/s | VRAM=32.33 Go [2026-07-18 04:02:09] step 620900/1000000 | loss=2.2777 vocab=0.6728 token=1.3945 | 6.13 steps/s | VRAM=29.67 Go [2026-07-18 04:02:17] step 620950/1000000 | loss=2.3113 vocab=0.5709 token=1.2902 | 6.14 steps/s | VRAM=31.33 Go [2026-07-18 04:02:25] step 621000/1000000 | loss=2.1230 vocab=0.6231 token=1.0540 | 6.13 steps/s | VRAM=32.41 Go [2026-07-18 04:02:33] step 621050/1000000 | loss=2.2682 vocab=0.8792 token=1.4451 | 6.14 steps/s | VRAM=28.38 Go [2026-07-18 04:02:41] step 621100/1000000 | loss=2.2434 vocab=1.3133 token=1.3242 | 6.13 steps/s | VRAM=32.64 Go [2026-07-18 04:02:49] step 621150/1000000 | loss=2.1748 vocab=0.6620 token=1.1237 | 6.15 steps/s | VRAM=31.60 Go [2026-07-18 04:02:58] step 621200/1000000 | loss=2.2218 vocab=0.8355 token=1.7197 | 6.10 steps/s | VRAM=26.52 Go [2026-07-18 04:03:06] step 621250/1000000 | loss=2.1580 vocab=0.8713 token=1.1503 | 6.16 steps/s | VRAM=29.15 Go [2026-07-18 04:03:14] step 621300/1000000 | loss=2.2020 vocab=0.7599 token=1.0301 | 6.13 steps/s | VRAM=28.34 Go [2026-07-18 04:03:22] step 621350/1000000 | loss=2.3253 vocab=0.8497 token=1.1714 | 6.14 steps/s | VRAM=28.35 Go [2026-07-18 04:03:30] step 621400/1000000 | loss=2.2648 vocab=0.8534 token=1.5370 | 6.16 steps/s | VRAM=26.07 Go [2026-07-18 04:03:38] step 621450/1000000 | loss=2.2851 vocab=0.6932 token=1.1577 | 6.14 steps/s | VRAM=29.46 Go [2026-07-18 04:03:46] step 621500/1000000 | loss=2.2584 vocab=0.6290 token=1.2866 | 6.14 steps/s | VRAM=32.61 Go [2026-07-18 04:03:55] step 621550/1000000 | loss=2.3145 vocab=0.8786 token=1.3564 | 6.12 steps/s | VRAM=28.86 Go [2026-07-18 04:04:03] step 621600/1000000 | loss=2.3162 vocab=0.9891 token=1.1008 | 6.14 steps/s | VRAM=32.99 Go [2026-07-18 04:04:11] step 621650/1000000 | loss=2.3075 vocab=0.9692 token=1.5137 | 6.15 steps/s | VRAM=32.30 Go [2026-07-18 04:04:19] step 621700/1000000 | loss=2.2951 vocab=1.2734 token=1.4923 | 6.13 steps/s | VRAM=28.64 Go [2026-07-18 04:04:27] step 621750/1000000 | loss=2.1637 vocab=0.9191 token=1.4855 | 6.11 steps/s | VRAM=26.03 Go [2026-07-18 04:04:35] step 621800/1000000 | loss=2.1983 vocab=1.7941 token=0.9814 | 6.15 steps/s | VRAM=31.68 Go [2026-07-18 04:04:44] step 621850/1000000 | loss=2.1809 vocab=0.9994 token=1.1211 | 6.12 steps/s | VRAM=26.07 Go [2026-07-18 04:04:52] step 621900/1000000 | loss=2.1316 vocab=0.7268 token=1.2707 | 6.13 steps/s | VRAM=28.61 Go [2026-07-18 04:05:00] step 621950/1000000 | loss=2.1683 vocab=1.7471 token=1.5689 | 6.15 steps/s | VRAM=32.27 Go [2026-07-18 04:05:08] step 622000/1000000 | loss=2.2299 vocab=1.3006 token=1.1988 | 6.15 steps/s | VRAM=28.98 Go [2026-07-18 04:05:16] step 622050/1000000 | loss=2.2569 vocab=0.7568 token=1.3147 | 6.14 steps/s | VRAM=32.17 Go [2026-07-18 04:05:24] step 622100/1000000 | loss=2.1811 vocab=1.1004 token=1.3439 | 6.15 steps/s | VRAM=28.31 Go [2026-07-18 04:05:32] step 622150/1000000 | loss=2.1502 vocab=0.8595 token=1.3835 | 6.17 steps/s | VRAM=29.04 Go [2026-07-18 04:05:40] step 622200/1000000 | loss=2.3748 vocab=1.6696 token=1.1828 | 6.17 steps/s | VRAM=32.04 Go [2026-07-18 04:05:49] step 622250/1000000 | loss=2.3958 vocab=0.9238 token=1.1552 | 6.13 steps/s | VRAM=32.60 Go [2026-07-18 04:05:57] step 622300/1000000 | loss=2.1768 vocab=0.7146 token=1.2748 | 6.12 steps/s | VRAM=28.68 Go [2026-07-18 04:06:05] step 622350/1000000 | loss=2.2077 vocab=1.5669 token=1.5972 | 6.11 steps/s | VRAM=29.08 Go [2026-07-18 04:06:13] step 622400/1000000 | loss=2.2602 vocab=1.1505 token=1.7779 | 6.11 steps/s | VRAM=28.51 Go [2026-07-18 04:06:21] step 622450/1000000 | loss=2.2277 vocab=0.4045 token=1.0635 | 6.13 steps/s | VRAM=32.93 Go [2026-07-18 04:06:30] step 622500/1000000 | loss=2.2444 vocab=1.1459 token=1.2781 | 6.11 steps/s | VRAM=25.96 Go [2026-07-18 04:06:38] step 622550/1000000 | loss=2.2777 vocab=1.1388 token=1.3522 | 6.16 steps/s | VRAM=22.78 Go [2026-07-18 04:06:46] step 622600/1000000 | loss=2.2480 vocab=0.5456 token=1.1078 | 6.15 steps/s | VRAM=28.37 Go [2026-07-18 04:06:54] step 622650/1000000 | loss=2.1925 vocab=1.2866 token=1.9251 | 6.16 steps/s | VRAM=28.76 Go [2026-07-18 04:07:02] step 622700/1000000 | loss=2.1249 vocab=0.5258 token=0.7892 | 6.15 steps/s | VRAM=32.93 Go [2026-07-18 04:07:10] step 622750/1000000 | loss=2.1806 vocab=0.8616 token=1.6184 | 6.15 steps/s | VRAM=25.37 Go [2026-07-18 04:07:18] step 622800/1000000 | loss=2.2942 vocab=1.1456 token=1.7338 | 6.16 steps/s | VRAM=29.44 Go [2026-07-18 04:07:26] step 622850/1000000 | loss=2.1076 vocab=1.1624 token=0.8369 | 6.19 steps/s | VRAM=31.78 Go [2026-07-18 04:07:34] step 622900/1000000 | loss=2.0857 vocab=0.7474 token=0.8721 | 6.12 steps/s | VRAM=28.02 Go [2026-07-18 04:07:43] step 622950/1000000 | loss=2.1869 vocab=0.5027 token=1.1847 | 6.12 steps/s | VRAM=25.87 Go [2026-07-18 04:07:51] step 623000/1000000 | loss=2.2520 vocab=1.0555 token=1.2730 | 6.14 steps/s | VRAM=31.61 Go [2026-07-18 04:07:59] step 623050/1000000 | loss=2.2915 vocab=1.3663 token=1.7034 | 6.12 steps/s | VRAM=27.77 Go [2026-07-18 04:08:07] step 623100/1000000 | loss=2.2784 vocab=0.4957 token=0.9230 | 6.10 steps/s | VRAM=30.32 Go [2026-07-18 04:08:15] step 623150/1000000 | loss=2.1755 vocab=0.8862 token=1.3777 | 6.15 steps/s | VRAM=31.05 Go [2026-07-18 04:08:23] step 623200/1000000 | loss=2.1174 vocab=1.3089 token=1.7825 | 6.15 steps/s | VRAM=32.21 Go [2026-07-18 04:08:32] step 623250/1000000 | loss=2.1752 vocab=0.4983 token=0.7933 | 6.15 steps/s | VRAM=31.93 Go [2026-07-18 04:08:40] step 623300/1000000 | loss=2.2473 vocab=0.9386 token=1.2542 | 6.12 steps/s | VRAM=32.94 Go [2026-07-18 04:08:48] step 623350/1000000 | loss=2.1730 vocab=0.8476 token=1.4816 | 6.18 steps/s | VRAM=31.15 Go [2026-07-18 04:08:56] step 623400/1000000 | loss=2.1387 vocab=0.6814 token=1.0172 | 6.12 steps/s | VRAM=30.43 Go [2026-07-18 04:09:04] step 623450/1000000 | loss=2.3161 vocab=0.9088 token=1.8054 | 6.16 steps/s | VRAM=30.72 Go [2026-07-18 04:09:12] step 623500/1000000 | loss=2.1269 vocab=0.7320 token=1.7625 | 6.12 steps/s | VRAM=30.73 Go [2026-07-18 04:09:20] step 623550/1000000 | loss=2.2274 vocab=1.4196 token=1.5249 | 6.16 steps/s | VRAM=32.08 Go [2026-07-18 04:09:29] step 623600/1000000 | loss=2.1503 vocab=0.7505 token=1.2716 | 6.14 steps/s | VRAM=25.90 Go [2026-07-18 04:09:37] step 623650/1000000 | loss=2.2084 vocab=0.9592 token=1.6977 | 6.12 steps/s | VRAM=26.33 Go [2026-07-18 04:09:45] step 623700/1000000 | loss=2.1536 vocab=0.5240 token=1.3391 | 6.15 steps/s | VRAM=32.14 Go [2026-07-18 04:09:53] step 623750/1000000 | loss=2.2000 vocab=1.2718 token=1.3073 | 6.15 steps/s | VRAM=27.99 Go [2026-07-18 04:10:01] step 623800/1000000 | loss=2.2198 vocab=1.0034 token=1.0597 | 6.13 steps/s | VRAM=26.20 Go [2026-07-18 04:10:09] step 623850/1000000 | loss=2.3674 vocab=1.6532 token=1.5600 | 6.17 steps/s | VRAM=32.08 Go [2026-07-18 04:10:17] step 623900/1000000 | loss=2.2036 vocab=0.9023 token=0.9518 | 6.16 steps/s | VRAM=31.97 Go [2026-07-18 04:10:25] step 623950/1000000 | loss=2.2039 vocab=0.8789 token=1.3328 | 6.14 steps/s | VRAM=29.13 Go [2026-07-18 04:10:34] step 624000/1000000 | loss=2.1243 vocab=1.0213 token=1.6445 | 6.14 steps/s | VRAM=28.49 Go [2026-07-18 04:10:42] step 624050/1000000 | loss=2.2170 vocab=1.1684 token=1.2570 | 6.14 steps/s | VRAM=31.74 Go [2026-07-18 04:10:50] step 624100/1000000 | loss=2.2203 vocab=0.9925 token=1.0343 | 6.16 steps/s | VRAM=31.83 Go [2026-07-18 04:10:58] step 624150/1000000 | loss=2.1642 vocab=1.0662 token=0.7990 | 6.13 steps/s | VRAM=28.30 Go [2026-07-18 04:11:06] step 624200/1000000 | loss=2.1656 vocab=0.7559 token=1.8862 | 6.12 steps/s | VRAM=26.47 Go [2026-07-18 04:11:14] step 624250/1000000 | loss=2.1746 vocab=0.6335 token=1.0506 | 6.13 steps/s | VRAM=32.55 Go [2026-07-18 04:11:23] step 624300/1000000 | loss=2.0945 vocab=0.5447 token=1.1455 | 6.15 steps/s | VRAM=31.69 Go [2026-07-18 04:11:31] step 624350/1000000 | loss=2.2400 vocab=0.8111 token=1.3677 | 6.13 steps/s | VRAM=28.02 Go [2026-07-18 04:11:39] step 624400/1000000 | loss=2.2928 vocab=0.4166 token=0.9778 | 6.11 steps/s | VRAM=28.30 Go [2026-07-18 04:11:47] step 624450/1000000 | loss=2.2204 vocab=0.8385 token=1.1249 | 6.13 steps/s | VRAM=26.16 Go [2026-07-18 04:11:55] step 624500/1000000 | loss=2.1311 vocab=0.4288 token=1.0993 | 6.14 steps/s | VRAM=27.88 Go [2026-07-18 04:12:03] step 624550/1000000 | loss=2.1881 vocab=1.3749 token=1.4287 | 6.15 steps/s | VRAM=28.20 Go [2026-07-18 04:12:11] step 624600/1000000 | loss=2.1169 vocab=1.3513 token=1.5231 | 6.18 steps/s | VRAM=32.36 Go [2026-07-18 04:12:20] step 624650/1000000 | loss=2.1972 vocab=1.0682 token=1.2202 | 6.14 steps/s | VRAM=28.56 Go [2026-07-18 04:12:28] step 624700/1000000 | loss=2.1756 vocab=1.4202 token=1.1881 | 6.13 steps/s | VRAM=25.64 Go [2026-07-18 04:12:36] step 624750/1000000 | loss=2.3162 vocab=0.6746 token=1.3866 | 6.15 steps/s | VRAM=25.08 Go [2026-07-18 04:12:44] step 624800/1000000 | loss=2.3501 vocab=0.7318 token=1.4620 | 6.15 steps/s | VRAM=31.37 Go [2026-07-18 04:12:52] step 624850/1000000 | loss=2.2129 vocab=1.3073 token=1.6228 | 6.11 steps/s | VRAM=29.25 Go [2026-07-18 04:13:00] step 624900/1000000 | loss=2.1422 vocab=0.7413 token=1.3213 | 6.16 steps/s | VRAM=28.08 Go [2026-07-18 04:13:08] step 624950/1000000 | loss=2.3039 vocab=0.6386 token=1.2393 | 6.14 steps/s | VRAM=32.12 Go [2026-07-18 04:13:17] step 625000/1000000 | loss=2.2612 vocab=0.3100 token=0.9662 | 6.12 steps/s | VRAM=28.74 Go [2026-07-18 04:13:25] step 625050/1000000 | loss=2.2745 vocab=1.1824 token=1.2281 | 6.18 steps/s | VRAM=29.26 Go [2026-07-18 04:13:33] step 625100/1000000 | loss=2.1643 vocab=1.2557 token=1.4337 | 6.14 steps/s | VRAM=26.58 Go [2026-07-18 04:13:41] step 625150/1000000 | loss=2.1129 vocab=0.7708 token=0.9741 | 6.15 steps/s | VRAM=31.32 Go [2026-07-18 04:13:49] step 625200/1000000 | loss=2.0751 vocab=0.7083 token=1.3228 | 6.15 steps/s | VRAM=25.84 Go [2026-07-18 04:13:57] step 625250/1000000 | loss=2.1670 vocab=0.6574 token=1.0786 | 6.11 steps/s | VRAM=25.97 Go [2026-07-18 04:14:05] step 625300/1000000 | loss=2.2420 vocab=1.0679 token=1.6106 | 6.15 steps/s | VRAM=28.63 Go [2026-07-18 04:14:14] step 625350/1000000 | loss=2.2656 vocab=0.9603 token=1.4417 | 6.14 steps/s | VRAM=29.09 Go [2026-07-18 04:14:22] step 625400/1000000 | loss=2.3109 vocab=1.0070 token=1.3069 | 6.11 steps/s | VRAM=26.61 Go [2026-07-18 04:14:30] step 625450/1000000 | loss=2.2623 vocab=0.7621 token=1.1292 | 6.15 steps/s | VRAM=28.31 Go [2026-07-18 04:14:38] step 625500/1000000 | loss=2.2964 vocab=1.2677 token=1.1140 | 6.10 steps/s | VRAM=28.93 Go [2026-07-18 04:14:46] step 625550/1000000 | loss=2.1328 vocab=0.7741 token=1.5734 | 6.14 steps/s | VRAM=28.84 Go [2026-07-18 04:14:54] step 625600/1000000 | loss=2.2767 vocab=0.7283 token=1.1109 | 6.14 steps/s | VRAM=28.48 Go [2026-07-18 04:15:03] step 625650/1000000 | loss=2.2881 vocab=0.9167 token=1.9480 | 6.12 steps/s | VRAM=27.58 Go [2026-07-18 04:15:11] step 625700/1000000 | loss=2.1995 vocab=0.9513 token=1.1123 | 6.13 steps/s | VRAM=28.27 Go [2026-07-18 04:15:19] step 625750/1000000 | loss=2.2232 vocab=0.8213 token=1.2947 | 6.12 steps/s | VRAM=29.43 Go [2026-07-18 04:15:27] step 625800/1000000 | loss=2.1547 vocab=0.8716 token=1.4991 | 6.15 steps/s | VRAM=31.64 Go [2026-07-18 04:15:35] step 625850/1000000 | loss=2.2342 vocab=0.7093 token=1.6381 | 6.13 steps/s | VRAM=28.57 Go [2026-07-18 04:15:43] step 625900/1000000 | loss=2.2416 vocab=0.6408 token=1.6333 | 6.14 steps/s | VRAM=30.29 Go [2026-07-18 04:15:51] step 625950/1000000 | loss=2.2055 vocab=0.7631 token=1.2449 | 6.16 steps/s | VRAM=28.74 Go [2026-07-18 04:16:00] step 626000/1000000 | loss=2.2394 vocab=1.4391 token=1.2382 | 6.15 steps/s | VRAM=32.48 Go [2026-07-18 04:16:08] step 626050/1000000 | loss=2.2159 vocab=0.7337 token=0.9197 | 6.14 steps/s | VRAM=29.36 Go [2026-07-18 04:16:16] step 626100/1000000 | loss=2.2254 vocab=0.6512 token=1.4199 | 6.15 steps/s | VRAM=30.06 Go [2026-07-18 04:16:24] step 626150/1000000 | loss=2.2503 vocab=1.2337 token=1.4711 | 6.15 steps/s | VRAM=28.31 Go [2026-07-18 04:16:32] step 626200/1000000 | loss=2.1933 vocab=1.1075 token=1.4419 | 6.15 steps/s | VRAM=28.32 Go [2026-07-18 04:16:40] step 626250/1000000 | loss=2.2545 vocab=0.8979 token=1.6627 | 6.17 steps/s | VRAM=29.08 Go [2026-07-18 04:16:48] step 626300/1000000 | loss=2.2423 vocab=0.8877 token=1.4311 | 6.18 steps/s | VRAM=25.04 Go [2026-07-18 04:16:56] step 626350/1000000 | loss=2.2607 vocab=0.7060 token=1.1653 | 6.19 steps/s | VRAM=25.33 Go [2026-07-18 04:17:04] step 626400/1000000 | loss=2.1193 vocab=1.0008 token=1.5943 | 6.14 steps/s | VRAM=25.76 Go [2026-07-18 04:17:13] step 626450/1000000 | loss=2.1212 vocab=0.7748 token=1.6667 | 6.17 steps/s | VRAM=30.37 Go [2026-07-18 04:17:21] step 626500/1000000 | loss=2.3103 vocab=0.9982 token=1.1961 | 6.16 steps/s | VRAM=31.62 Go [2026-07-18 04:17:29] step 626550/1000000 | loss=2.2004 vocab=1.1181 token=1.6821 | 6.16 steps/s | VRAM=29.86 Go [2026-07-18 04:17:37] step 626600/1000000 | loss=2.2473 vocab=1.1150 token=1.6045 | 6.16 steps/s | VRAM=31.56 Go [2026-07-18 04:17:45] step 626650/1000000 | loss=2.2337 vocab=1.3888 token=1.3873 | 6.14 steps/s | VRAM=32.02 Go [2026-07-18 04:17:53] step 626700/1000000 | loss=2.1567 vocab=0.7417 token=1.2864 | 6.17 steps/s | VRAM=31.24 Go [2026-07-18 04:18:01] step 626750/1000000 | loss=2.2579 vocab=0.9224 token=1.6665 | 6.14 steps/s | VRAM=28.81 Go [2026-07-18 04:18:09] step 626800/1000000 | loss=2.1479 vocab=1.0044 token=1.2766 | 6.16 steps/s | VRAM=28.87 Go [2026-07-18 04:18:18] step 626850/1000000 | loss=2.1365 vocab=0.8043 token=1.5300 | 6.16 steps/s | VRAM=32.65 Go [2026-07-18 04:18:26] step 626900/1000000 | loss=2.2178 vocab=0.6724 token=1.3088 | 6.14 steps/s | VRAM=26.33 Go [2026-07-18 04:18:34] step 626950/1000000 | loss=2.2107 vocab=0.8720 token=1.3465 | 6.16 steps/s | VRAM=26.77 Go [2026-07-18 04:18:42] step 627000/1000000 | loss=2.1899 vocab=0.6514 token=1.3064 | 6.18 steps/s | VRAM=28.73 Go [2026-07-18 04:18:50] step 627050/1000000 | loss=2.1874 vocab=1.1223 token=1.0393 | 6.19 steps/s | VRAM=28.79 Go [2026-07-18 04:18:58] step 627100/1000000 | loss=2.2436 vocab=0.4534 token=0.6959 | 6.12 steps/s | VRAM=28.05 Go [2026-07-18 04:19:06] step 627150/1000000 | loss=2.1551 vocab=0.9411 token=1.2393 | 6.16 steps/s | VRAM=32.00 Go [2026-07-18 04:19:14] step 627200/1000000 | loss=2.2471 vocab=1.2660 token=1.3030 | 6.16 steps/s | VRAM=28.65 Go [2026-07-18 04:19:22] step 627250/1000000 | loss=2.2482 vocab=1.0160 token=1.4050 | 6.18 steps/s | VRAM=25.85 Go [2026-07-18 04:19:31] step 627300/1000000 | loss=2.1334 vocab=1.0255 token=1.5520 | 6.14 steps/s | VRAM=30.56 Go [2026-07-18 04:19:39] step 627350/1000000 | loss=2.2031 vocab=0.8239 token=0.9138 | 6.16 steps/s | VRAM=31.71 Go [2026-07-18 04:19:47] step 627400/1000000 | loss=2.2777 vocab=0.7068 token=1.2132 | 6.12 steps/s | VRAM=29.47 Go [2026-07-18 04:19:55] step 627450/1000000 | loss=2.2032 vocab=0.8929 token=1.2784 | 6.17 steps/s | VRAM=28.41 Go [2026-07-18 04:20:03] step 627500/1000000 | loss=2.1870 vocab=1.2355 token=1.5400 | 6.16 steps/s | VRAM=29.49 Go [2026-07-18 04:20:11] step 627550/1000000 | loss=2.1999 vocab=0.6185 token=1.1224 | 6.16 steps/s | VRAM=32.48 Go [2026-07-18 04:20:19] step 627600/1000000 | loss=2.0616 vocab=0.8332 token=1.4812 | 6.12 steps/s | VRAM=28.80 Go [2026-07-18 04:20:28] step 627650/1000000 | loss=2.2755 vocab=1.3089 token=1.2139 | 6.13 steps/s | VRAM=29.54 Go [2026-07-18 04:20:36] step 627700/1000000 | loss=2.2852 vocab=1.2006 token=1.4976 | 6.17 steps/s | VRAM=25.66 Go [2026-07-18 04:20:44] step 627750/1000000 | loss=2.3276 vocab=0.9732 token=1.2996 | 6.18 steps/s | VRAM=32.40 Go [2026-07-18 04:20:52] step 627800/1000000 | loss=2.2044 vocab=0.7716 token=1.4501 | 6.13 steps/s | VRAM=26.41 Go [2026-07-18 04:21:00] step 627850/1000000 | loss=2.2031 vocab=0.5595 token=1.0057 | 6.14 steps/s | VRAM=28.66 Go [2026-07-18 04:21:08] step 627900/1000000 | loss=2.2242 vocab=1.0317 token=1.5493 | 6.16 steps/s | VRAM=32.80 Go [2026-07-18 04:21:16] step 627950/1000000 | loss=2.2537 vocab=1.0183 token=1.2027 | 6.12 steps/s | VRAM=32.00 Go [2026-07-18 04:21:25] step 628000/1000000 | loss=2.0968 vocab=0.8275 token=0.9086 | 6.13 steps/s | VRAM=32.23 Go [2026-07-18 04:21:33] step 628050/1000000 | loss=2.2451 vocab=1.0220 token=1.6724 | 6.15 steps/s | VRAM=28.64 Go [2026-07-18 04:21:41] step 628100/1000000 | loss=2.2508 vocab=0.6497 token=1.3339 | 6.14 steps/s | VRAM=26.03 Go [2026-07-18 04:21:49] step 628150/1000000 | loss=2.2412 vocab=0.9231 token=1.4310 | 6.13 steps/s | VRAM=26.43 Go [2026-07-18 04:21:57] step 628200/1000000 | loss=2.2241 vocab=0.8232 token=0.8813 | 6.17 steps/s | VRAM=31.97 Go [2026-07-18 04:22:05] step 628250/1000000 | loss=2.2529 vocab=0.8506 token=1.1755 | 6.15 steps/s | VRAM=29.18 Go [2026-07-18 04:22:13] step 628300/1000000 | loss=2.1988 vocab=0.9895 token=1.4661 | 6.15 steps/s | VRAM=28.52 Go [2026-07-18 04:22:21] step 628350/1000000 | loss=2.1981 vocab=0.9247 token=1.1999 | 6.15 steps/s | VRAM=31.49 Go [2026-07-18 04:22:30] step 628400/1000000 | loss=2.2637 vocab=0.7052 token=1.6430 | 6.15 steps/s | VRAM=32.60 Go [2026-07-18 04:22:38] step 628450/1000000 | loss=2.2014 vocab=0.6695 token=1.2974 | 6.13 steps/s | VRAM=29.10 Go [2026-07-18 04:22:46] step 628500/1000000 | loss=2.1512 vocab=1.1545 token=1.5420 | 6.13 steps/s | VRAM=29.14 Go [2026-07-18 04:22:54] step 628550/1000000 | loss=2.2116 vocab=0.6351 token=0.9145 | 6.15 steps/s | VRAM=28.55 Go [2026-07-18 04:23:02] step 628600/1000000 | loss=2.2891 vocab=0.9588 token=1.4318 | 6.16 steps/s | VRAM=32.35 Go [2026-07-18 04:23:10] step 628650/1000000 | loss=2.1373 vocab=0.5390 token=1.2407 | 6.14 steps/s | VRAM=28.81 Go [2026-07-18 04:23:18] step 628700/1000000 | loss=2.1607 vocab=0.6219 token=1.7078 | 6.15 steps/s | VRAM=28.36 Go [2026-07-18 04:23:27] step 628750/1000000 | loss=2.2129 vocab=1.0899 token=1.9333 | 6.17 steps/s | VRAM=31.56 Go [2026-07-18 04:23:35] step 628800/1000000 | loss=2.0662 vocab=0.5938 token=1.2891 | 6.14 steps/s | VRAM=29.70 Go [2026-07-18 04:23:43] step 628850/1000000 | loss=2.2086 vocab=1.2262 token=1.3198 | 6.13 steps/s | VRAM=26.28 Go [2026-07-18 04:23:51] step 628900/1000000 | loss=2.3209 vocab=0.6769 token=1.4274 | 6.13 steps/s | VRAM=26.61 Go [2026-07-18 04:23:59] step 628950/1000000 | loss=2.1567 vocab=0.9780 token=1.0582 | 6.14 steps/s | VRAM=25.06 Go [2026-07-18 04:24:07] step 629000/1000000 | loss=2.2075 vocab=1.0676 token=1.4292 | 6.11 steps/s | VRAM=28.56 Go [2026-07-18 04:24:16] step 629050/1000000 | loss=2.2406 vocab=1.0990 token=1.4810 | 6.12 steps/s | VRAM=26.49 Go [2026-07-18 04:24:24] step 629100/1000000 | loss=2.2997 vocab=0.8845 token=1.5011 | 6.15 steps/s | VRAM=24.81 Go [2026-07-18 04:24:32] step 629150/1000000 | loss=2.2481 vocab=1.0476 token=1.0500 | 6.14 steps/s | VRAM=32.98 Go [2026-07-18 04:24:40] step 629200/1000000 | loss=2.2681 vocab=1.1133 token=1.4286 | 6.16 steps/s | VRAM=27.58 Go [2026-07-18 04:24:48] step 629250/1000000 | loss=2.1736 vocab=0.7684 token=1.1581 | 6.16 steps/s | VRAM=31.27 Go [2026-07-18 04:24:56] step 629300/1000000 | loss=2.1644 vocab=1.1207 token=1.4916 | 6.14 steps/s | VRAM=29.01 Go [2026-07-18 04:25:04] step 629350/1000000 | loss=2.1948 vocab=1.0544 token=1.2018 | 6.14 steps/s | VRAM=28.70 Go [2026-07-18 04:25:12] step 629400/1000000 | loss=2.2255 vocab=0.6393 token=1.1790 | 6.11 steps/s | VRAM=23.99 Go [2026-07-18 04:25:21] step 629450/1000000 | loss=2.3150 vocab=1.2624 token=1.2708 | 6.18 steps/s | VRAM=27.37 Go [2026-07-18 04:25:29] step 629500/1000000 | loss=2.2050 vocab=0.9125 token=1.2794 | 6.09 steps/s | VRAM=27.01 Go [2026-07-18 04:25:37] step 629550/1000000 | loss=2.2407 vocab=0.7979 token=1.2706 | 6.17 steps/s | VRAM=31.83 Go [2026-07-18 04:25:45] step 629600/1000000 | loss=2.3087 vocab=1.2849 token=1.8974 | 6.13 steps/s | VRAM=29.18 Go [2026-07-18 04:25:53] step 629650/1000000 | loss=2.1591 vocab=1.1271 token=1.4342 | 6.15 steps/s | VRAM=27.40 Go [2026-07-18 04:26:01] step 629700/1000000 | loss=2.2966 vocab=0.7121 token=1.2448 | 6.15 steps/s | VRAM=25.99 Go [2026-07-18 04:26:09] step 629750/1000000 | loss=2.1862 vocab=0.9458 token=0.9302 | 6.17 steps/s | VRAM=31.57 Go [2026-07-18 04:26:18] step 629800/1000000 | loss=2.2729 vocab=1.0983 token=1.6536 | 6.13 steps/s | VRAM=29.74 Go [2026-07-18 04:26:26] step 629850/1000000 | loss=2.1915 vocab=1.0479 token=0.8813 | 6.12 steps/s | VRAM=31.58 Go [2026-07-18 04:26:34] step 629900/1000000 | loss=2.2067 vocab=0.7511 token=1.0861 | 6.16 steps/s | VRAM=32.46 Go [2026-07-18 04:26:42] step 629950/1000000 | loss=2.1128 vocab=0.5812 token=1.2914 | 6.15 steps/s | VRAM=25.63 Go [2026-07-18 04:26:50] step 630000/1000000 | loss=2.2135 vocab=1.1264 token=1.9218 | 6.13 steps/s | VRAM=28.95 Go [2026-07-18 04:26:58] step 630050/1000000 | loss=2.1555 vocab=0.7904 token=1.2515 | 6.12 steps/s | VRAM=32.87 Go [2026-07-18 04:27:06] step 630100/1000000 | loss=2.2138 vocab=0.8479 token=1.3988 | 6.15 steps/s | VRAM=31.99 Go [2026-07-18 04:27:15] step 630150/1000000 | loss=2.1698 vocab=0.9419 token=1.2070 | 6.14 steps/s | VRAM=32.70 Go [2026-07-18 04:27:23] step 630200/1000000 | loss=2.2851 vocab=0.9235 token=1.4722 | 6.12 steps/s | VRAM=24.78 Go [2026-07-18 04:27:31] step 630250/1000000 | loss=2.1550 vocab=0.7882 token=1.4401 | 6.13 steps/s | VRAM=25.50 Go [2026-07-18 04:27:39] step 630300/1000000 | loss=2.2576 vocab=1.2338 token=1.8491 | 6.12 steps/s | VRAM=28.34 Go [2026-07-18 04:27:47] step 630350/1000000 | loss=2.2516 vocab=1.0670 token=1.5844 | 6.14 steps/s | VRAM=32.32 Go [2026-07-18 04:27:55] step 630400/1000000 | loss=2.1508 vocab=0.7798 token=1.3779 | 6.14 steps/s | VRAM=27.30 Go [2026-07-18 04:28:04] step 630450/1000000 | loss=2.1992 vocab=0.6714 token=1.1306 | 6.12 steps/s | VRAM=30.27 Go [2026-07-18 04:28:12] step 630500/1000000 | loss=2.2481 vocab=0.6528 token=1.4703 | 6.15 steps/s | VRAM=27.93 Go [2026-07-18 04:28:20] step 630550/1000000 | loss=2.1686 vocab=0.4699 token=1.3500 | 6.13 steps/s | VRAM=27.15 Go [2026-07-18 04:28:28] step 630600/1000000 | loss=2.3268 vocab=0.8472 token=1.3211 | 6.14 steps/s | VRAM=27.16 Go [2026-07-18 04:28:36] step 630650/1000000 | loss=2.2085 vocab=0.6738 token=1.4982 | 6.16 steps/s | VRAM=27.61 Go [2026-07-18 04:28:44] step 630700/1000000 | loss=2.1931 vocab=1.4471 token=1.7311 | 6.15 steps/s | VRAM=27.61 Go [2026-07-18 04:28:52] step 630750/1000000 | loss=2.2205 vocab=1.1514 token=0.9825 | 6.18 steps/s | VRAM=30.49 Go [2026-07-18 04:29:00] step 630800/1000000 | loss=2.2308 vocab=1.0098 token=1.0925 | 6.18 steps/s | VRAM=26.64 Go [2026-07-18 04:29:09] step 630850/1000000 | loss=2.2684 vocab=1.4774 token=1.8206 | 6.13 steps/s | VRAM=29.67 Go [2026-07-18 04:29:17] step 630900/1000000 | loss=2.2384 vocab=0.8649 token=1.1539 | 6.10 steps/s | VRAM=28.08 Go [2026-07-18 04:29:25] step 630950/1000000 | loss=2.2265 vocab=0.4541 token=1.0171 | 6.14 steps/s | VRAM=31.95 Go [2026-07-18 04:29:33] step 631000/1000000 | loss=2.2469 vocab=0.8463 token=1.0474 | 6.16 steps/s | VRAM=29.08 Go [2026-07-18 04:29:41] step 631050/1000000 | loss=2.1093 vocab=1.3359 token=1.6068 | 6.12 steps/s | VRAM=32.73 Go [2026-07-18 04:29:49] step 631100/1000000 | loss=2.1721 vocab=0.5869 token=1.3137 | 6.13 steps/s | VRAM=26.00 Go [2026-07-18 04:29:57] step 631150/1000000 | loss=2.2254 vocab=0.8964 token=1.0517 | 6.16 steps/s | VRAM=27.84 Go [2026-07-18 04:30:06] step 631200/1000000 | loss=2.2539 vocab=0.6405 token=1.0342 | 6.14 steps/s | VRAM=29.99 Go [2026-07-18 04:30:14] step 631250/1000000 | loss=2.1049 vocab=0.7154 token=1.4100 | 6.12 steps/s | VRAM=29.53 Go [2026-07-18 04:30:22] step 631300/1000000 | loss=2.2398 vocab=0.9990 token=1.3317 | 6.15 steps/s | VRAM=29.79 Go [2026-07-18 04:30:30] step 631350/1000000 | loss=2.1119 vocab=1.1016 token=1.2707 | 6.16 steps/s | VRAM=30.99 Go [2026-07-18 04:30:38] step 631400/1000000 | loss=2.2189 vocab=1.0764 token=1.2171 | 6.12 steps/s | VRAM=26.49 Go [2026-07-18 04:30:46] step 631450/1000000 | loss=2.2009 vocab=1.1112 token=1.4088 | 6.14 steps/s | VRAM=32.05 Go [2026-07-18 04:30:55] step 631500/1000000 | loss=2.2738 vocab=1.0754 token=1.3672 | 6.12 steps/s | VRAM=29.86 Go [2026-07-18 04:31:03] step 631550/1000000 | loss=2.2649 vocab=0.6100 token=1.3295 | 6.12 steps/s | VRAM=30.91 Go [2026-07-18 04:31:11] step 631600/1000000 | loss=2.2619 vocab=0.9081 token=1.6838 | 6.14 steps/s | VRAM=29.88 Go [2026-07-18 04:31:19] step 631650/1000000 | loss=2.2498 vocab=1.0430 token=1.2932 | 6.15 steps/s | VRAM=31.56 Go [2026-07-18 04:31:27] step 631700/1000000 | loss=2.1786 vocab=0.7823 token=1.5356 | 6.14 steps/s | VRAM=27.84 Go [2026-07-18 04:31:35] step 631750/1000000 | loss=2.1758 vocab=0.4907 token=1.2184 | 6.17 steps/s | VRAM=32.45 Go [2026-07-18 04:31:43] step 631800/1000000 | loss=2.2165 vocab=0.6383 token=1.2488 | 6.18 steps/s | VRAM=31.75 Go [2026-07-18 04:31:52] step 631850/1000000 | loss=2.2707 vocab=0.9437 token=0.7756 | 6.11 steps/s | VRAM=28.99 Go [2026-07-18 04:32:00] step 631900/1000000 | loss=2.1240 vocab=0.4302 token=0.9739 | 6.13 steps/s | VRAM=29.21 Go [2026-07-18 04:32:08] step 631950/1000000 | loss=2.1179 vocab=1.0284 token=1.1409 | 6.14 steps/s | VRAM=31.69 Go [2026-07-18 04:32:16] step 632000/1000000 | loss=2.2134 vocab=1.4016 token=1.8622 | 6.14 steps/s | VRAM=28.84 Go [2026-07-18 04:32:24] step 632050/1000000 | loss=2.2231 vocab=0.8371 token=1.4267 | 6.13 steps/s | VRAM=24.63 Go [2026-07-18 04:32:32] step 632100/1000000 | loss=2.2757 vocab=2.3739 token=1.3973 | 6.16 steps/s | VRAM=28.78 Go [2026-07-18 04:32:40] step 632150/1000000 | loss=2.2023 vocab=0.5151 token=1.0457 | 6.13 steps/s | VRAM=28.43 Go [2026-07-18 04:32:49] step 632200/1000000 | loss=2.2809 vocab=0.9754 token=1.3480 | 6.13 steps/s | VRAM=26.70 Go [2026-07-18 04:32:57] step 632250/1000000 | loss=2.2293 vocab=0.8407 token=1.3869 | 6.13 steps/s | VRAM=28.22 Go [2026-07-18 04:33:05] step 632300/1000000 | loss=2.2390 vocab=1.2004 token=1.4352 | 6.11 steps/s | VRAM=32.95 Go [2026-07-18 04:33:13] step 632350/1000000 | loss=2.1607 vocab=1.0596 token=1.5592 | 6.11 steps/s | VRAM=29.93 Go [2026-07-18 04:33:21] step 632400/1000000 | loss=2.3081 vocab=1.1873 token=1.7710 | 6.14 steps/s | VRAM=28.02 Go [2026-07-18 04:33:29] step 632450/1000000 | loss=2.1366 vocab=1.0285 token=1.3367 | 6.14 steps/s | VRAM=31.33 Go [2026-07-18 04:33:38] step 632500/1000000 | loss=2.2786 vocab=1.7718 token=1.5339 | 6.14 steps/s | VRAM=25.61 Go [2026-07-18 04:33:46] step 632550/1000000 | loss=2.2347 vocab=0.5570 token=1.0064 | 6.14 steps/s | VRAM=28.57 Go [2026-07-18 04:33:54] step 632600/1000000 | loss=2.2956 vocab=1.0492 token=1.2556 | 6.15 steps/s | VRAM=31.48 Go [2026-07-18 04:34:02] step 632650/1000000 | loss=2.1782 vocab=1.1244 token=1.7387 | 6.13 steps/s | VRAM=25.47 Go [2026-07-18 04:34:10] step 632700/1000000 | loss=2.2126 vocab=0.3423 token=0.7608 | 6.10 steps/s | VRAM=28.31 Go [2026-07-18 04:34:18] step 632750/1000000 | loss=2.1947 vocab=0.8737 token=1.3651 | 6.16 steps/s | VRAM=28.21 Go [2026-07-18 04:34:26] step 632800/1000000 | loss=2.2392 vocab=1.5428 token=1.7371 | 6.13 steps/s | VRAM=32.61 Go [2026-07-18 04:34:35] step 632850/1000000 | loss=2.3454 vocab=1.2578 token=1.2038 | 6.13 steps/s | VRAM=28.49 Go [2026-07-18 04:34:43] step 632900/1000000 | loss=2.1899 vocab=1.2838 token=1.6685 | 6.15 steps/s | VRAM=32.52 Go [2026-07-18 04:34:51] step 632950/1000000 | loss=2.1971 vocab=0.9673 token=1.2445 | 6.16 steps/s | VRAM=28.69 Go [2026-07-18 04:34:59] step 633000/1000000 | loss=2.1993 vocab=0.9376 token=1.2195 | 6.14 steps/s | VRAM=32.01 Go [2026-07-18 04:35:07] step 633050/1000000 | loss=2.3648 vocab=1.0299 token=1.2202 | 6.16 steps/s | VRAM=28.07 Go [2026-07-18 04:35:15] step 633100/1000000 | loss=2.2817 vocab=1.0159 token=1.1962 | 6.12 steps/s | VRAM=32.42 Go [2026-07-18 04:35:23] step 633150/1000000 | loss=2.1932 vocab=1.0173 token=1.3413 | 6.13 steps/s | VRAM=29.03 Go [2026-07-18 04:35:32] step 633200/1000000 | loss=2.1765 vocab=1.0070 token=1.3678 | 6.11 steps/s | VRAM=26.64 Go [2026-07-18 04:35:40] step 633250/1000000 | loss=2.2390 vocab=0.7366 token=1.2989 | 6.12 steps/s | VRAM=28.27 Go [2026-07-18 04:35:48] step 633300/1000000 | loss=2.1493 vocab=0.6240 token=0.8610 | 6.14 steps/s | VRAM=32.44 Go [2026-07-18 04:35:56] step 633350/1000000 | loss=2.2908 vocab=0.9854 token=1.3005 | 6.11 steps/s | VRAM=26.77 Go [2026-07-18 04:36:04] step 633400/1000000 | loss=2.2387 vocab=0.3537 token=1.4123 | 6.15 steps/s | VRAM=32.23 Go [2026-07-18 04:36:12] step 633450/1000000 | loss=2.2222 vocab=1.0230 token=1.4503 | 6.15 steps/s | VRAM=32.52 Go [2026-07-18 04:36:21] step 633500/1000000 | loss=2.1970 vocab=0.8042 token=1.1071 | 6.12 steps/s | VRAM=29.55 Go [2026-07-18 04:36:29] step 633550/1000000 | loss=2.1767 vocab=0.9888 token=1.3021 | 6.15 steps/s | VRAM=31.32 Go [2026-07-18 04:36:37] step 633600/1000000 | loss=2.2175 vocab=0.5738 token=1.2617 | 6.17 steps/s | VRAM=28.66 Go [2026-07-18 04:36:45] step 633650/1000000 | loss=2.2233 vocab=0.7261 token=1.4553 | 6.12 steps/s | VRAM=29.18 Go [2026-07-18 04:36:53] step 633700/1000000 | loss=2.3321 vocab=1.3209 token=1.4124 | 6.15 steps/s | VRAM=31.87 Go [2026-07-18 04:37:01] step 633750/1000000 | loss=2.2202 vocab=0.6036 token=1.3632 | 6.17 steps/s | VRAM=28.36 Go [2026-07-18 04:37:09] step 633800/1000000 | loss=2.1657 vocab=0.8388 token=1.1964 | 6.16 steps/s | VRAM=31.47 Go [2026-07-18 04:37:18] step 633850/1000000 | loss=2.1744 vocab=1.2050 token=1.5291 | 6.10 steps/s | VRAM=28.30 Go [2026-07-18 04:37:26] step 633900/1000000 | loss=2.2366 vocab=0.9005 token=1.2462 | 6.12 steps/s | VRAM=29.16 Go [2026-07-18 04:37:34] step 633950/1000000 | loss=2.2647 vocab=0.6781 token=1.1631 | 6.12 steps/s | VRAM=26.55 Go [2026-07-18 04:37:42] step 634000/1000000 | loss=2.2454 vocab=1.1145 token=1.5545 | 6.15 steps/s | VRAM=28.32 Go [2026-07-18 04:37:50] step 634050/1000000 | loss=2.1916 vocab=1.0419 token=1.7550 | 6.14 steps/s | VRAM=28.56 Go [2026-07-18 04:37:58] step 634100/1000000 | loss=2.1241 vocab=0.9813 token=1.3385 | 6.16 steps/s | VRAM=31.55 Go [2026-07-18 04:38:06] step 634150/1000000 | loss=2.2433 vocab=1.5850 token=1.5831 | 6.15 steps/s | VRAM=28.74 Go [2026-07-18 04:38:15] step 634200/1000000 | loss=2.1937 vocab=1.0181 token=1.3831 | 6.15 steps/s | VRAM=32.03 Go [2026-07-18 04:38:23] step 634250/1000000 | loss=2.1600 vocab=0.7892 token=1.5084 | 6.16 steps/s | VRAM=29.30 Go [2026-07-18 04:38:31] step 634300/1000000 | loss=2.2180 vocab=0.7413 token=1.1686 | 6.14 steps/s | VRAM=31.54 Go [2026-07-18 04:38:39] step 634350/1000000 | loss=2.1736 vocab=0.4583 token=0.9897 | 6.13 steps/s | VRAM=29.09 Go [2026-07-18 04:38:47] step 634400/1000000 | loss=2.2512 vocab=1.1955 token=1.1533 | 6.10 steps/s | VRAM=26.89 Go [2026-07-18 04:38:55] step 634450/1000000 | loss=2.2914 vocab=0.7446 token=1.6696 | 6.15 steps/s | VRAM=24.95 Go [2026-07-18 04:39:03] step 634500/1000000 | loss=2.2274 vocab=0.4300 token=1.1395 | 6.16 steps/s | VRAM=26.23 Go [2026-07-18 04:39:11] step 634550/1000000 | loss=2.2928 vocab=0.8422 token=1.3599 | 6.16 steps/s | VRAM=32.38 Go [2026-07-18 04:39:20] step 634600/1000000 | loss=2.1759 vocab=1.3843 token=1.6266 | 6.14 steps/s | VRAM=31.49 Go [2026-07-18 04:39:28] step 634650/1000000 | loss=2.3342 vocab=1.1375 token=1.6888 | 6.14 steps/s | VRAM=32.08 Go [2026-07-18 04:39:36] step 634700/1000000 | loss=2.2438 vocab=1.1487 token=1.4798 | 6.14 steps/s | VRAM=24.91 Go [2026-07-18 04:39:44] step 634750/1000000 | loss=2.2374 vocab=0.8680 token=1.3626 | 6.16 steps/s | VRAM=28.39 Go [2026-07-18 04:39:52] step 634800/1000000 | loss=2.1322 vocab=0.7186 token=1.0548 | 6.14 steps/s | VRAM=26.10 Go [2026-07-18 04:40:00] step 634850/1000000 | loss=2.1945 vocab=1.3220 token=1.2545 | 6.16 steps/s | VRAM=32.61 Go [2026-07-18 04:40:08] step 634900/1000000 | loss=2.3118 vocab=0.9236 token=1.5620 | 6.15 steps/s | VRAM=28.00 Go [2026-07-18 04:40:17] step 634950/1000000 | loss=2.0286 vocab=0.3976 token=1.1061 | 6.13 steps/s | VRAM=32.96 Go [2026-07-18 04:40:25] step 635000/1000000 | loss=2.2331 vocab=0.8150 token=1.3184 | 6.12 steps/s | VRAM=29.93 Go [2026-07-18 04:40:33] step 635050/1000000 | loss=2.3337 vocab=0.6426 token=1.1213 | 6.16 steps/s | VRAM=28.06 Go [2026-07-18 04:40:41] step 635100/1000000 | loss=2.2159 vocab=0.8642 token=1.2833 | 6.12 steps/s | VRAM=28.97 Go [2026-07-18 04:40:49] step 635150/1000000 | loss=2.1796 vocab=0.8944 token=1.4210 | 6.13 steps/s | VRAM=29.41 Go [2026-07-18 04:40:57] step 635200/1000000 | loss=2.1059 vocab=0.6371 token=0.9852 | 6.13 steps/s | VRAM=25.38 Go [2026-07-18 04:41:06] step 635250/1000000 | loss=2.3430 vocab=0.6969 token=1.4312 | 6.14 steps/s | VRAM=28.94 Go [2026-07-18 04:41:14] step 635300/1000000 | loss=2.1963 vocab=0.5315 token=0.7015 | 6.12 steps/s | VRAM=29.34 Go [2026-07-18 04:41:22] step 635350/1000000 | loss=2.2789 vocab=0.9884 token=1.7715 | 6.12 steps/s | VRAM=29.72 Go [2026-07-18 04:41:30] step 635400/1000000 | loss=2.3031 vocab=0.8667 token=1.5397 | 6.13 steps/s | VRAM=28.29 Go [2026-07-18 04:41:38] step 635450/1000000 | loss=2.2178 vocab=0.6464 token=1.2514 | 6.09 steps/s | VRAM=29.30 Go [2026-07-18 04:41:46] step 635500/1000000 | loss=2.1866 vocab=1.0636 token=1.1851 | 6.14 steps/s | VRAM=32.53 Go [2026-07-18 04:41:55] step 635550/1000000 | loss=2.3088 vocab=1.0553 token=1.3187 | 6.12 steps/s | VRAM=29.03 Go [2026-07-18 04:42:03] step 635600/1000000 | loss=2.2179 vocab=0.8334 token=1.3561 | 6.12 steps/s | VRAM=29.55 Go [2026-07-18 04:42:11] step 635650/1000000 | loss=2.1903 vocab=1.2917 token=1.4495 | 6.13 steps/s | VRAM=28.83 Go [2026-07-18 04:42:19] step 635700/1000000 | loss=2.2368 vocab=1.1741 token=1.4720 | 6.16 steps/s | VRAM=32.49 Go [2026-07-18 04:42:27] step 635750/1000000 | loss=2.1595 vocab=0.7893 token=1.4109 | 6.13 steps/s | VRAM=28.47 Go [2026-07-18 04:42:35] step 635800/1000000 | loss=2.1786 vocab=0.9512 token=1.1202 | 6.13 steps/s | VRAM=28.62 Go [2026-07-18 04:42:43] step 635850/1000000 | loss=2.1521 vocab=0.4021 token=1.1826 | 6.16 steps/s | VRAM=31.87 Go [2026-07-18 04:42:52] step 635900/1000000 | loss=2.3078 vocab=0.8399 token=1.2747 | 6.15 steps/s | VRAM=31.83 Go [2026-07-18 04:43:00] step 635950/1000000 | loss=2.1919 vocab=0.7190 token=1.0041 | 6.12 steps/s | VRAM=28.44 Go [2026-07-18 04:43:08] step 636000/1000000 | loss=2.1674 vocab=0.9292 token=1.3733 | 6.10 steps/s | VRAM=29.11 Go [2026-07-18 04:43:16] step 636050/1000000 | loss=2.2495 vocab=0.8209 token=1.3651 | 6.13 steps/s | VRAM=25.96 Go [2026-07-18 04:43:24] step 636100/1000000 | loss=2.1872 vocab=0.9656 token=1.4440 | 6.15 steps/s | VRAM=31.69 Go [2026-07-18 04:43:32] step 636150/1000000 | loss=2.2661 vocab=0.8521 token=1.3423 | 6.10 steps/s | VRAM=29.71 Go [2026-07-18 04:43:40] step 636200/1000000 | loss=2.1622 vocab=0.9077 token=0.5520 | 6.16 steps/s | VRAM=32.48 Go [2026-07-18 04:43:49] step 636250/1000000 | loss=2.1837 vocab=1.1660 token=1.5440 | 6.12 steps/s | VRAM=26.31 Go [2026-07-18 04:43:57] step 636300/1000000 | loss=2.2169 vocab=1.0703 token=1.7075 | 6.14 steps/s | VRAM=29.90 Go [2026-07-18 04:44:05] step 636350/1000000 | loss=2.2179 vocab=2.4971 token=1.8419 | 6.14 steps/s | VRAM=26.33 Go [2026-07-18 04:44:13] step 636400/1000000 | loss=2.1810 vocab=0.5750 token=1.0830 | 6.14 steps/s | VRAM=25.29 Go [2026-07-18 04:44:21] step 636450/1000000 | loss=2.2321 vocab=0.5959 token=1.1156 | 6.14 steps/s | VRAM=24.52 Go [2026-07-18 04:44:29] step 636500/1000000 | loss=2.2402 vocab=1.2317 token=1.3752 | 6.13 steps/s | VRAM=26.07 Go [2026-07-18 04:44:38] step 636550/1000000 | loss=2.1776 vocab=0.4103 token=0.7202 | 6.16 steps/s | VRAM=31.83 Go [2026-07-18 04:44:46] step 636600/1000000 | loss=2.1647 vocab=0.5872 token=0.9893 | 6.16 steps/s | VRAM=25.65 Go [2026-07-18 04:44:54] step 636650/1000000 | loss=2.2339 vocab=0.7075 token=1.4797 | 6.14 steps/s | VRAM=32.42 Go [2026-07-18 04:45:02] step 636700/1000000 | loss=2.2180 vocab=1.2356 token=1.4508 | 6.16 steps/s | VRAM=28.65 Go [2026-07-18 04:45:10] step 636750/1000000 | loss=2.1777 vocab=1.4914 token=1.0264 | 6.12 steps/s | VRAM=26.15 Go [2026-07-18 04:45:18] step 636800/1000000 | loss=2.1703 vocab=0.8445 token=1.2801 | 6.15 steps/s | VRAM=28.39 Go [2026-07-18 04:45:26] step 636850/1000000 | loss=2.2082 vocab=0.9468 token=1.1842 | 6.14 steps/s | VRAM=32.42 Go [2026-07-18 04:45:35] step 636900/1000000 | loss=2.2510 vocab=0.9801 token=1.3525 | 6.13 steps/s | VRAM=28.36 Go [2026-07-18 04:45:43] step 636950/1000000 | loss=2.1702 vocab=0.9332 token=1.1755 | 6.13 steps/s | VRAM=29.22 Go [2026-07-18 04:45:51] step 637000/1000000 | loss=2.2780 vocab=1.0866 token=1.5916 | 6.13 steps/s | VRAM=28.65 Go [2026-07-18 04:45:59] step 637050/1000000 | loss=2.2121 vocab=0.5857 token=1.2372 | 6.14 steps/s | VRAM=28.75 Go [2026-07-18 04:46:07] step 637100/1000000 | loss=2.2035 vocab=0.4764 token=1.0142 | 6.13 steps/s | VRAM=33.00 Go [2026-07-18 04:46:15] step 637150/1000000 | loss=2.2815 vocab=1.0530 token=1.8250 | 6.14 steps/s | VRAM=28.58 Go [2026-07-18 04:46:23] step 637200/1000000 | loss=2.2711 vocab=0.8579 token=1.4284 | 6.15 steps/s | VRAM=28.51 Go [2026-07-18 04:46:32] step 637250/1000000 | loss=2.1910 vocab=0.7760 token=1.2930 | 6.14 steps/s | VRAM=26.28 Go [2026-07-18 04:46:40] step 637300/1000000 | loss=2.2215 vocab=0.6559 token=1.3306 | 6.19 steps/s | VRAM=28.27 Go [2026-07-18 04:46:48] step 637350/1000000 | loss=2.2072 vocab=0.7760 token=1.4795 | 6.15 steps/s | VRAM=25.02 Go [2026-07-18 04:46:56] step 637400/1000000 | loss=2.2275 vocab=1.2418 token=1.5387 | 6.14 steps/s | VRAM=29.33 Go [2026-07-18 04:47:04] step 637450/1000000 | loss=2.2485 vocab=1.2896 token=1.4947 | 6.13 steps/s | VRAM=24.90 Go [2026-07-18 04:47:12] step 637500/1000000 | loss=2.2014 vocab=0.5207 token=1.1305 | 6.14 steps/s | VRAM=28.36 Go [2026-07-18 04:47:20] step 637550/1000000 | loss=2.2148 vocab=0.4690 token=0.9978 | 6.14 steps/s | VRAM=32.77 Go [2026-07-18 04:47:28] step 637600/1000000 | loss=2.2419 vocab=1.5206 token=1.5886 | 6.13 steps/s | VRAM=28.83 Go [2026-07-18 04:47:37] step 637650/1000000 | loss=2.1917 vocab=1.1248 token=1.3638 | 6.18 steps/s | VRAM=32.40 Go [2026-07-18 04:47:45] step 637700/1000000 | loss=2.2542 vocab=0.8649 token=1.0857 | 6.15 steps/s | VRAM=29.42 Go [2026-07-18 04:47:53] step 637750/1000000 | loss=2.1424 vocab=0.3977 token=0.9440 | 6.13 steps/s | VRAM=26.65 Go [2026-07-18 04:48:01] step 637800/1000000 | loss=2.1696 vocab=1.2368 token=1.8091 | 6.15 steps/s | VRAM=24.39 Go [2026-07-18 04:48:09] step 637850/1000000 | loss=2.2265 vocab=1.1146 token=1.4204 | 6.13 steps/s | VRAM=26.86 Go [2026-07-18 04:48:17] step 637900/1000000 | loss=2.1453 vocab=0.6374 token=1.4698 | 6.15 steps/s | VRAM=28.79 Go [2026-07-18 04:48:25] step 637950/1000000 | loss=2.3004 vocab=1.1962 token=1.2907 | 6.13 steps/s | VRAM=32.89 Go [2026-07-18 04:48:34] step 638000/1000000 | loss=2.2118 vocab=0.8237 token=1.7010 | 6.14 steps/s | VRAM=26.42 Go [2026-07-18 04:48:42] step 638050/1000000 | loss=2.0824 vocab=1.2869 token=1.2047 | 6.16 steps/s | VRAM=32.79 Go [2026-07-18 04:48:50] step 638100/1000000 | loss=2.1524 vocab=0.7179 token=1.4125 | 6.14 steps/s | VRAM=28.49 Go [2026-07-18 04:48:58] step 638150/1000000 | loss=2.2077 vocab=1.2166 token=1.5563 | 6.12 steps/s | VRAM=28.33 Go [2026-07-18 04:49:06] step 638200/1000000 | loss=2.1620 vocab=1.2183 token=1.6477 | 6.11 steps/s | VRAM=29.62 Go [2026-07-18 04:49:14] step 638250/1000000 | loss=2.1305 vocab=1.1879 token=1.7526 | 6.14 steps/s | VRAM=28.93 Go [2026-07-18 04:49:22] step 638300/1000000 | loss=2.2272 vocab=0.7731 token=1.1361 | 6.14 steps/s | VRAM=28.88 Go [2026-07-18 04:49:31] step 638350/1000000 | loss=2.2353 vocab=0.7240 token=0.8333 | 6.15 steps/s | VRAM=28.87 Go [2026-07-18 04:49:39] step 638400/1000000 | loss=2.2591 vocab=0.9938 token=1.2811 | 6.13 steps/s | VRAM=29.01 Go [2026-07-18 04:49:47] step 638450/1000000 | loss=2.2118 vocab=0.6201 token=1.4097 | 6.16 steps/s | VRAM=28.38 Go [2026-07-18 04:49:55] step 638500/1000000 | loss=2.2533 vocab=1.0786 token=1.3989 | 6.14 steps/s | VRAM=28.66 Go [2026-07-18 04:50:03] step 638550/1000000 | loss=2.1867 vocab=1.0918 token=1.4291 | 6.12 steps/s | VRAM=26.52 Go [2026-07-18 04:50:11] step 638600/1000000 | loss=2.2182 vocab=1.0464 token=1.3997 | 6.15 steps/s | VRAM=28.81 Go [2026-07-18 04:50:19] step 638650/1000000 | loss=2.2429 vocab=1.1969 token=1.4306 | 6.16 steps/s | VRAM=24.35 Go [2026-07-18 04:50:28] step 638700/1000000 | loss=2.1593 vocab=0.6488 token=1.2323 | 6.15 steps/s | VRAM=29.25 Go [2026-07-18 04:50:36] step 638750/1000000 | loss=2.1793 vocab=0.8109 token=1.1841 | 6.17 steps/s | VRAM=24.83 Go [2026-07-18 04:50:44] step 638800/1000000 | loss=2.2582 vocab=0.6246 token=1.0780 | 6.18 steps/s | VRAM=32.09 Go [2026-07-18 04:50:52] step 638850/1000000 | loss=2.0853 vocab=0.4909 token=1.2805 | 6.14 steps/s | VRAM=29.16 Go [2026-07-18 04:51:00] step 638900/1000000 | loss=2.3286 vocab=0.9081 token=1.7619 | 6.16 steps/s | VRAM=27.97 Go [2026-07-18 04:51:08] step 638950/1000000 | loss=2.2581 vocab=1.1952 token=0.9227 | 6.12 steps/s | VRAM=23.66 Go [2026-07-18 04:51:16] step 639000/1000000 | loss=2.2325 vocab=0.7541 token=1.3082 | 6.14 steps/s | VRAM=28.18 Go [2026-07-18 04:51:25] step 639050/1000000 | loss=2.2292 vocab=0.6680 token=1.1214 | 6.14 steps/s | VRAM=32.18 Go [2026-07-18 04:51:33] step 639100/1000000 | loss=2.3389 vocab=1.0016 token=1.3438 | 6.13 steps/s | VRAM=30.47 Go [2026-07-18 04:51:41] step 639150/1000000 | loss=2.1588 vocab=0.8852 token=1.5609 | 6.17 steps/s | VRAM=28.02 Go [2026-07-18 04:51:49] step 639200/1000000 | loss=2.1796 vocab=0.9421 token=1.3495 | 6.16 steps/s | VRAM=31.84 Go [2026-07-18 04:51:57] step 639250/1000000 | loss=2.2600 vocab=1.2195 token=1.9144 | 6.16 steps/s | VRAM=30.88 Go [2026-07-18 04:52:05] step 639300/1000000 | loss=2.1937 vocab=0.6702 token=1.1965 | 6.16 steps/s | VRAM=31.64 Go [2026-07-18 04:52:13] step 639350/1000000 | loss=2.1942 vocab=0.6248 token=1.1107 | 6.13 steps/s | VRAM=27.47 Go [2026-07-18 04:52:21] step 639400/1000000 | loss=2.2050 vocab=1.0703 token=1.4606 | 6.18 steps/s | VRAM=32.94 Go [2026-07-18 04:52:30] step 639450/1000000 | loss=2.2002 vocab=0.3682 token=1.1295 | 6.14 steps/s | VRAM=28.57 Go [2026-07-18 04:52:38] step 639500/1000000 | loss=2.1347 vocab=0.5958 token=1.0602 | 6.13 steps/s | VRAM=26.61 Go [2026-07-18 04:52:46] step 639550/1000000 | loss=2.2456 vocab=0.6035 token=1.4215 | 6.16 steps/s | VRAM=26.07 Go [2026-07-18 04:52:54] step 639600/1000000 | loss=2.2279 vocab=0.5689 token=1.1879 | 6.13 steps/s | VRAM=32.74 Go [2026-07-18 04:53:02] step 639650/1000000 | loss=2.1500 vocab=0.7881 token=1.1700 | 6.05 steps/s | VRAM=30.85 Go [2026-07-18 04:53:10] step 639700/1000000 | loss=2.0534 vocab=1.2163 token=1.5823 | 6.16 steps/s | VRAM=32.36 Go [2026-07-18 04:53:18] step 639750/1000000 | loss=2.1757 vocab=1.0708 token=1.2754 | 6.15 steps/s | VRAM=32.10 Go [2026-07-18 04:53:27] step 639800/1000000 | loss=2.2589 vocab=1.0217 token=1.2544 | 6.12 steps/s | VRAM=29.37 Go [2026-07-18 04:53:35] step 639850/1000000 | loss=2.2635 vocab=0.5622 token=1.1818 | 6.12 steps/s | VRAM=28.84 Go [2026-07-18 04:53:43] step 639900/1000000 | loss=2.1318 vocab=0.6664 token=1.3056 | 6.12 steps/s | VRAM=28.41 Go [2026-07-18 04:53:51] step 639950/1000000 | loss=2.2593 vocab=0.8913 token=1.3870 | 6.11 steps/s | VRAM=24.42 Go [2026-07-18 04:53:59] step 640000/1000000 | loss=2.2724 vocab=1.4702 token=1.5690 | 6.11 steps/s | VRAM=28.58 Go [2026-07-18 04:54:07] step 640050/1000000 | loss=2.1113 vocab=0.8183 token=1.3623 | 6.14 steps/s | VRAM=28.61 Go [2026-07-18 04:54:16] step 640100/1000000 | loss=2.2277 vocab=0.9551 token=0.9967 | 6.12 steps/s | VRAM=26.11 Go [2026-07-18 04:54:24] step 640150/1000000 | loss=2.0262 vocab=0.7428 token=1.2447 | 6.14 steps/s | VRAM=29.17 Go [2026-07-18 04:54:32] step 640200/1000000 | loss=2.1389 vocab=0.5969 token=1.3930 | 6.10 steps/s | VRAM=26.18 Go [2026-07-18 04:54:40] step 640250/1000000 | loss=2.2025 vocab=0.7249 token=0.8660 | 6.12 steps/s | VRAM=28.44 Go [2026-07-18 04:54:48] step 640300/1000000 | loss=2.2646 vocab=1.1751 token=1.5182 | 6.19 steps/s | VRAM=28.00 Go [2026-07-18 04:54:56] step 640350/1000000 | loss=2.1869 vocab=1.2918 token=1.6405 | 6.11 steps/s | VRAM=29.12 Go [2026-07-18 04:55:05] step 640400/1000000 | loss=2.3763 vocab=0.5366 token=1.0449 | 6.11 steps/s | VRAM=32.62 Go [2026-07-18 04:55:13] step 640450/1000000 | loss=2.1901 vocab=0.7093 token=1.1334 | 6.12 steps/s | VRAM=27.99 Go [2026-07-18 04:55:21] step 640500/1000000 | loss=2.1808 vocab=1.0861 token=1.6696 | 6.14 steps/s | VRAM=28.53 Go [2026-07-18 04:55:29] step 640550/1000000 | loss=2.2210 vocab=0.8910 token=1.3257 | 6.14 steps/s | VRAM=28.52 Go [2026-07-18 04:55:37] step 640600/1000000 | loss=2.1506 vocab=1.1405 token=1.4693 | 6.16 steps/s | VRAM=28.62 Go [2026-07-18 04:55:45] step 640650/1000000 | loss=2.2242 vocab=0.7170 token=1.2000 | 6.11 steps/s | VRAM=29.55 Go [2026-07-18 04:55:54] step 640700/1000000 | loss=2.2359 vocab=1.1163 token=1.6838 | 6.13 steps/s | VRAM=29.34 Go [2026-07-18 04:56:02] step 640750/1000000 | loss=2.2524 vocab=0.9053 token=1.2364 | 6.11 steps/s | VRAM=32.05 Go [2026-07-18 04:56:10] step 640800/1000000 | loss=2.2010 vocab=0.5634 token=1.3931 | 6.14 steps/s | VRAM=28.14 Go [2026-07-18 04:56:18] step 640850/1000000 | loss=2.1973 vocab=0.9465 token=1.0137 | 6.12 steps/s | VRAM=27.17 Go [2026-07-18 04:56:26] step 640900/1000000 | loss=2.2320 vocab=0.9298 token=1.3916 | 6.13 steps/s | VRAM=29.27 Go [2026-07-18 04:56:34] step 640950/1000000 | loss=2.1783 vocab=1.0585 token=1.0601 | 6.14 steps/s | VRAM=28.05 Go [2026-07-18 04:56:42] step 641000/1000000 | loss=2.1344 vocab=0.9219 token=1.3217 | 6.12 steps/s | VRAM=29.06 Go [2026-07-18 04:56:51] step 641050/1000000 | loss=2.2871 vocab=0.4199 token=0.9938 | 6.10 steps/s | VRAM=30.23 Go [2026-07-18 04:56:59] step 641100/1000000 | loss=2.1284 vocab=0.7694 token=1.1441 | 6.11 steps/s | VRAM=28.91 Go [2026-07-18 04:57:07] step 641150/1000000 | loss=2.1894 vocab=0.8377 token=1.4218 | 6.12 steps/s | VRAM=26.21 Go [2026-07-18 04:57:15] step 641200/1000000 | loss=2.0854 vocab=0.8639 token=1.5243 | 6.12 steps/s | VRAM=32.72 Go [2026-07-18 04:57:23] step 641250/1000000 | loss=2.2886 vocab=1.0427 token=1.5780 | 6.11 steps/s | VRAM=25.55 Go [2026-07-18 04:57:31] step 641300/1000000 | loss=2.1712 vocab=0.7236 token=1.1885 | 6.15 steps/s | VRAM=29.37 Go [2026-07-18 04:57:40] step 641350/1000000 | loss=2.1938 vocab=0.9964 token=1.4048 | 6.16 steps/s | VRAM=32.26 Go [2026-07-18 04:57:48] step 641400/1000000 | loss=2.1683 vocab=0.7492 token=1.2474 | 6.17 steps/s | VRAM=29.12 Go [2026-07-18 04:57:56] step 641450/1000000 | loss=2.2360 vocab=0.8923 token=1.3562 | 6.15 steps/s | VRAM=26.50 Go [2026-07-18 04:58:04] step 641500/1000000 | loss=2.3595 vocab=1.0628 token=1.3389 | 6.13 steps/s | VRAM=31.78 Go [2026-07-18 04:58:12] step 641550/1000000 | loss=2.1195 vocab=0.4444 token=0.9033 | 6.16 steps/s | VRAM=31.61 Go [2026-07-18 04:58:20] step 641600/1000000 | loss=2.2759 vocab=1.4611 token=1.8659 | 6.14 steps/s | VRAM=28.53 Go [2026-07-18 04:58:28] step 641650/1000000 | loss=2.1401 vocab=0.7182 token=1.1063 | 6.11 steps/s | VRAM=29.40 Go [2026-07-18 04:58:37] step 641700/1000000 | loss=2.1746 vocab=1.1818 token=1.2966 | 6.12 steps/s | VRAM=29.24 Go [2026-07-18 04:58:45] step 641750/1000000 | loss=2.3014 vocab=0.8678 token=1.4944 | 6.15 steps/s | VRAM=31.55 Go [2026-07-18 04:58:53] step 641800/1000000 | loss=2.1629 vocab=1.2755 token=1.3220 | 6.13 steps/s | VRAM=28.78 Go [2026-07-18 04:59:01] step 641850/1000000 | loss=2.1827 vocab=0.9221 token=1.1700 | 6.11 steps/s | VRAM=29.45 Go [2026-07-18 04:59:09] step 641900/1000000 | loss=2.1825 vocab=0.9602 token=1.6599 | 6.14 steps/s | VRAM=28.06 Go [2026-07-18 04:59:17] step 641950/1000000 | loss=2.2601 vocab=0.6183 token=1.2248 | 6.13 steps/s | VRAM=28.38 Go [2026-07-18 04:59:25] step 642000/1000000 | loss=2.2087 vocab=0.9620 token=1.2939 | 6.17 steps/s | VRAM=31.91 Go [2026-07-18 04:59:34] step 642050/1000000 | loss=2.1189 vocab=0.8447 token=1.4117 | 6.12 steps/s | VRAM=25.53 Go [2026-07-18 04:59:42] step 642100/1000000 | loss=2.2317 vocab=0.8921 token=1.4477 | 6.14 steps/s | VRAM=32.05 Go [2026-07-18 04:59:50] step 642150/1000000 | loss=2.0561 vocab=0.2869 token=0.9289 | 6.14 steps/s | VRAM=25.20 Go [2026-07-18 04:59:58] step 642200/1000000 | loss=2.1910 vocab=1.4748 token=1.4390 | 6.13 steps/s | VRAM=29.05 Go [2026-07-18 05:00:06] step 642250/1000000 | loss=2.2561 vocab=0.9040 token=1.4551 | 6.15 steps/s | VRAM=28.45 Go [2026-07-18 05:00:14] step 642300/1000000 | loss=2.1847 vocab=1.1559 token=1.6857 | 6.13 steps/s | VRAM=28.06 Go [2026-07-18 05:00:23] step 642350/1000000 | loss=2.1406 vocab=1.0854 token=1.4088 | 6.14 steps/s | VRAM=31.27 Go [2026-07-18 05:00:31] step 642400/1000000 | loss=2.2200 vocab=0.6362 token=1.6266 | 6.14 steps/s | VRAM=26.05 Go [2026-07-18 05:00:39] step 642450/1000000 | loss=2.2536 vocab=1.2411 token=1.5958 | 6.17 steps/s | VRAM=28.48 Go [2026-07-18 05:00:47] step 642500/1000000 | loss=2.1989 vocab=0.8804 token=1.5551 | 6.15 steps/s | VRAM=25.62 Go [2026-07-18 05:00:55] step 642550/1000000 | loss=2.2350 vocab=1.0144 token=1.3027 | 6.13 steps/s | VRAM=29.40 Go [2026-07-18 05:01:03] step 642600/1000000 | loss=2.0687 vocab=1.5292 token=1.5253 | 6.14 steps/s | VRAM=31.52 Go [2026-07-18 05:01:11] step 642650/1000000 | loss=2.2395 vocab=0.7957 token=1.5590 | 6.16 steps/s | VRAM=28.74 Go [2026-07-18 05:01:19] step 642700/1000000 | loss=2.1938 vocab=1.2722 token=1.6088 | 6.12 steps/s | VRAM=32.77 Go [2026-07-18 05:01:28] step 642750/1000000 | loss=2.2539 vocab=1.1971 token=1.4408 | 6.14 steps/s | VRAM=32.97 Go [2026-07-18 05:01:36] step 642800/1000000 | loss=2.1956 vocab=0.8295 token=1.2861 | 6.17 steps/s | VRAM=26.50 Go [2026-07-18 05:01:44] step 642850/1000000 | loss=2.1998 vocab=0.7823 token=1.1092 | 6.11 steps/s | VRAM=29.26 Go [2026-07-18 05:01:52] step 642900/1000000 | loss=2.3419 vocab=0.9544 token=1.4952 | 6.13 steps/s | VRAM=32.87 Go [2026-07-18 05:02:00] step 642950/1000000 | loss=2.2010 vocab=1.1387 token=1.0344 | 6.13 steps/s | VRAM=31.35 Go [2026-07-18 05:02:08] step 643000/1000000 | loss=2.1607 vocab=0.6295 token=1.0985 | 6.14 steps/s | VRAM=26.41 Go [2026-07-18 05:02:17] step 643050/1000000 | loss=2.2352 vocab=0.6215 token=1.2922 | 6.15 steps/s | VRAM=26.67 Go [2026-07-18 05:02:25] step 643100/1000000 | loss=2.2029 vocab=0.7177 token=1.0129 | 6.15 steps/s | VRAM=28.37 Go [2026-07-18 05:02:33] step 643150/1000000 | loss=2.2775 vocab=0.4732 token=1.3391 | 6.15 steps/s | VRAM=32.28 Go [2026-07-18 05:02:41] step 643200/1000000 | loss=2.1637 vocab=0.6230 token=1.3166 | 6.17 steps/s | VRAM=24.81 Go [2026-07-18 05:02:49] step 643250/1000000 | loss=2.2410 vocab=1.0392 token=1.2035 | 6.15 steps/s | VRAM=29.45 Go [2026-07-18 05:02:57] step 643300/1000000 | loss=2.1697 vocab=0.4667 token=1.3178 | 6.15 steps/s | VRAM=30.67 Go [2026-07-18 05:03:05] step 643350/1000000 | loss=2.2344 vocab=1.0141 token=1.6738 | 6.20 steps/s | VRAM=30.54 Go [2026-07-18 05:03:13] step 643400/1000000 | loss=2.1380 vocab=0.9288 token=1.1795 | 6.18 steps/s | VRAM=23.68 Go [2026-07-18 05:03:21] step 643450/1000000 | loss=2.1679 vocab=0.8364 token=1.2382 | 6.16 steps/s | VRAM=28.84 Go [2026-07-18 05:03:30] step 643500/1000000 | loss=2.1635 vocab=0.5858 token=0.9613 | 6.15 steps/s | VRAM=26.35 Go [2026-07-18 05:03:38] step 643550/1000000 | loss=2.2996 vocab=0.7569 token=1.4505 | 6.15 steps/s | VRAM=27.71 Go [2026-07-18 05:03:46] step 643600/1000000 | loss=2.3298 vocab=1.3129 token=1.5356 | 6.14 steps/s | VRAM=29.67 Go [2026-07-18 05:03:54] step 643650/1000000 | loss=2.3382 vocab=1.0840 token=1.5709 | 6.15 steps/s | VRAM=26.84 Go [2026-07-18 05:04:02] step 643700/1000000 | loss=2.1749 vocab=0.9452 token=1.2827 | 6.12 steps/s | VRAM=28.12 Go [2026-07-18 05:04:10] step 643750/1000000 | loss=2.1874 vocab=0.5896 token=1.2654 | 6.13 steps/s | VRAM=27.07 Go [2026-07-18 05:04:18] step 643800/1000000 | loss=2.3514 vocab=0.7701 token=0.8977 | 6.16 steps/s | VRAM=28.03 Go [2026-07-18 05:04:27] step 643850/1000000 | loss=2.2616 vocab=0.4921 token=1.4362 | 6.14 steps/s | VRAM=29.49 Go [2026-07-18 05:04:35] step 643900/1000000 | loss=2.2160 vocab=0.7152 token=1.2364 | 6.14 steps/s | VRAM=32.70 Go [2026-07-18 05:04:43] step 643950/1000000 | loss=2.2037 vocab=0.5919 token=0.8303 | 6.12 steps/s | VRAM=28.08 Go [2026-07-18 05:04:51] step 644000/1000000 | loss=2.3249 vocab=1.0570 token=1.4123 | 6.12 steps/s | VRAM=27.28 Go [2026-07-18 05:04:59] step 644050/1000000 | loss=2.1840 vocab=1.1745 token=1.3194 | 6.13 steps/s | VRAM=28.86 Go [2026-07-18 05:05:07] step 644100/1000000 | loss=2.2280 vocab=0.8208 token=1.0504 | 6.15 steps/s | VRAM=30.81 Go [2026-07-18 05:05:15] step 644150/1000000 | loss=2.2740 vocab=0.4180 token=1.0917 | 6.13 steps/s | VRAM=32.11 Go [2026-07-18 05:05:24] step 644200/1000000 | loss=2.1064 vocab=0.8889 token=1.7087 | 6.12 steps/s | VRAM=32.57 Go [2026-07-18 05:05:32] step 644250/1000000 | loss=2.2310 vocab=0.5838 token=1.4455 | 6.10 steps/s | VRAM=32.94 Go [2026-07-18 05:05:40] step 644300/1000000 | loss=2.2511 vocab=0.5620 token=1.2227 | 6.12 steps/s | VRAM=32.03 Go [2026-07-18 05:05:48] step 644350/1000000 | loss=2.1876 vocab=1.4047 token=1.6888 | 6.14 steps/s | VRAM=31.70 Go [2026-07-18 05:05:56] step 644400/1000000 | loss=2.1738 vocab=0.6656 token=1.5849 | 6.13 steps/s | VRAM=26.12 Go [2026-07-18 05:06:04] step 644450/1000000 | loss=2.1820 vocab=0.8948 token=1.3034 | 6.13 steps/s | VRAM=27.91 Go [2026-07-18 05:06:13] step 644500/1000000 | loss=2.3097 vocab=1.1294 token=1.6367 | 6.13 steps/s | VRAM=28.27 Go [2026-07-18 05:06:21] step 644550/1000000 | loss=2.3056 vocab=1.4065 token=1.4657 | 6.15 steps/s | VRAM=29.57 Go [2026-07-18 05:06:29] step 644600/1000000 | loss=2.1535 vocab=1.2004 token=1.7363 | 6.15 steps/s | VRAM=27.59 Go [2026-07-18 05:06:37] step 644650/1000000 | loss=2.2517 vocab=1.2511 token=1.5536 | 6.15 steps/s | VRAM=25.66 Go [2026-07-18 05:06:45] step 644700/1000000 | loss=2.2185 vocab=1.3083 token=1.1273 | 6.12 steps/s | VRAM=25.05 Go [2026-07-18 05:06:53] step 644750/1000000 | loss=2.1636 vocab=0.7414 token=1.2774 | 6.13 steps/s | VRAM=32.80 Go [2026-07-18 05:07:02] step 644800/1000000 | loss=2.2828 vocab=0.7910 token=1.4736 | 6.10 steps/s | VRAM=27.96 Go [2026-07-18 05:07:10] step 644850/1000000 | loss=2.2595 vocab=0.9437 token=1.4565 | 6.12 steps/s | VRAM=28.85 Go [2026-07-18 05:07:18] step 644900/1000000 | loss=2.2586 vocab=1.5435 token=1.5727 | 6.15 steps/s | VRAM=29.03 Go [2026-07-18 05:07:26] step 644950/1000000 | loss=2.3528 vocab=1.1790 token=1.1707 | 6.14 steps/s | VRAM=28.66 Go [2026-07-18 05:07:34] step 645000/1000000 | loss=2.2445 vocab=1.1787 token=1.4677 | 6.19 steps/s | VRAM=32.21 Go [2026-07-18 05:07:42] step 645050/1000000 | loss=2.1362 vocab=1.1745 token=1.8820 | 6.18 steps/s | VRAM=32.28 Go [2026-07-18 05:07:50] step 645100/1000000 | loss=2.1786 vocab=0.4408 token=1.1834 | 6.16 steps/s | VRAM=25.78 Go [2026-07-18 05:07:58] step 645150/1000000 | loss=2.1034 vocab=0.7307 token=0.8531 | 6.14 steps/s | VRAM=25.60 Go [2026-07-18 05:08:07] step 645200/1000000 | loss=2.2098 vocab=1.0697 token=1.1306 | 6.15 steps/s | VRAM=27.62 Go [2026-07-18 05:08:15] step 645250/1000000 | loss=2.1528 vocab=0.6325 token=1.0996 | 6.15 steps/s | VRAM=29.01 Go [2026-07-18 05:08:23] step 645300/1000000 | loss=2.2567 vocab=1.1692 token=1.4279 | 6.16 steps/s | VRAM=28.33 Go [2026-07-18 05:08:31] step 645350/1000000 | loss=2.0763 vocab=0.7575 token=1.4489 | 6.16 steps/s | VRAM=29.35 Go [2026-07-18 05:08:39] step 645400/1000000 | loss=2.1711 vocab=0.7253 token=1.4142 | 6.15 steps/s | VRAM=29.32 Go [2026-07-18 05:08:47] step 645450/1000000 | loss=2.0720 vocab=0.5755 token=1.3717 | 6.14 steps/s | VRAM=31.26 Go [2026-07-18 05:08:55] step 645500/1000000 | loss=2.1908 vocab=0.6180 token=1.1836 | 6.15 steps/s | VRAM=26.09 Go [2026-07-18 05:09:03] step 645550/1000000 | loss=2.1309 vocab=1.5018 token=1.2371 | 6.18 steps/s | VRAM=29.13 Go [2026-07-18 05:09:12] step 645600/1000000 | loss=2.2064 vocab=1.1168 token=1.9330 | 6.15 steps/s | VRAM=28.20 Go [2026-07-18 05:09:20] step 645650/1000000 | loss=2.1850 vocab=0.4251 token=0.9310 | 6.16 steps/s | VRAM=30.32 Go [2026-07-18 05:09:28] step 645700/1000000 | loss=2.1627 vocab=0.7779 token=1.1433 | 6.15 steps/s | VRAM=32.53 Go [2026-07-18 05:09:36] step 645750/1000000 | loss=2.1349 vocab=1.4241 token=1.5490 | 6.15 steps/s | VRAM=28.16 Go [2026-07-18 05:09:44] step 645800/1000000 | loss=2.2142 vocab=1.1850 token=1.6844 | 6.14 steps/s | VRAM=26.83 Go [2026-07-18 05:09:52] step 645850/1000000 | loss=2.1286 vocab=0.8730 token=1.3374 | 6.15 steps/s | VRAM=28.62 Go [2026-07-18 05:10:00] step 645900/1000000 | loss=2.2138 vocab=1.3215 token=1.4594 | 6.18 steps/s | VRAM=32.48 Go [2026-07-18 05:10:08] step 645950/1000000 | loss=2.2632 vocab=0.9331 token=1.4519 | 6.15 steps/s | VRAM=32.22 Go [2026-07-18 05:10:17] step 646000/1000000 | loss=2.2250 vocab=0.5750 token=1.0651 | 6.15 steps/s | VRAM=25.96 Go [2026-07-18 05:10:25] step 646050/1000000 | loss=2.1355 vocab=0.7759 token=1.4378 | 6.12 steps/s | VRAM=28.49 Go [2026-07-18 05:10:33] step 646100/1000000 | loss=2.1415 vocab=1.1708 token=1.4697 | 6.15 steps/s | VRAM=32.49 Go [2026-07-18 05:10:41] step 646150/1000000 | loss=2.2503 vocab=1.7380 token=1.2282 | 6.14 steps/s | VRAM=32.31 Go [2026-07-18 05:10:49] step 646200/1000000 | loss=2.2913 vocab=0.4520 token=1.1579 | 6.17 steps/s | VRAM=24.55 Go [2026-07-18 05:10:57] step 646250/1000000 | loss=2.1212 vocab=0.7330 token=1.6059 | 6.16 steps/s | VRAM=29.24 Go [2026-07-18 05:11:05] step 646300/1000000 | loss=2.2886 vocab=0.4425 token=1.1936 | 6.14 steps/s | VRAM=28.29 Go [2026-07-18 05:11:13] step 646350/1000000 | loss=2.2521 vocab=0.4216 token=1.0240 | 6.16 steps/s | VRAM=28.99 Go [2026-07-18 05:11:22] step 646400/1000000 | loss=2.2831 vocab=0.8321 token=0.8475 | 6.12 steps/s | VRAM=32.17 Go [2026-07-18 05:11:30] step 646450/1000000 | loss=2.1633 vocab=0.8831 token=1.1109 | 6.17 steps/s | VRAM=29.00 Go [2026-07-18 05:11:38] step 646500/1000000 | loss=2.2787 vocab=0.8388 token=1.3648 | 6.17 steps/s | VRAM=28.39 Go [2026-07-18 05:11:46] step 646550/1000000 | loss=2.2017 vocab=0.6658 token=1.8837 | 6.15 steps/s | VRAM=31.49 Go [2026-07-18 05:11:54] step 646600/1000000 | loss=2.2080 vocab=1.2231 token=1.5494 | 6.15 steps/s | VRAM=30.58 Go [2026-07-18 05:12:02] step 646650/1000000 | loss=2.2817 vocab=1.5865 token=1.6044 | 6.18 steps/s | VRAM=31.62 Go [2026-07-18 05:12:10] step 646700/1000000 | loss=2.2046 vocab=0.7381 token=1.1474 | 6.16 steps/s | VRAM=29.44 Go [2026-07-18 05:12:18] step 646750/1000000 | loss=2.1683 vocab=0.3480 token=0.8806 | 6.16 steps/s | VRAM=28.46 Go [2026-07-18 05:12:27] step 646800/1000000 | loss=2.1479 vocab=1.0706 token=1.5609 | 6.16 steps/s | VRAM=31.41 Go [2026-07-18 05:12:35] step 646850/1000000 | loss=2.2005 vocab=0.7352 token=0.8719 | 6.16 steps/s | VRAM=26.98 Go [2026-07-18 05:12:43] step 646900/1000000 | loss=2.2430 vocab=1.3584 token=1.4950 | 6.14 steps/s | VRAM=29.94 Go [2026-07-18 05:12:51] step 646950/1000000 | loss=2.3640 vocab=0.9703 token=1.4748 | 6.17 steps/s | VRAM=25.49 Go [2026-07-18 05:12:59] step 647000/1000000 | loss=2.1803 vocab=0.9534 token=1.3719 | 6.16 steps/s | VRAM=26.32 Go [2026-07-18 05:13:07] step 647050/1000000 | loss=2.1996 vocab=0.9731 token=1.3155 | 6.12 steps/s | VRAM=26.07 Go [2026-07-18 05:13:15] step 647100/1000000 | loss=2.3132 vocab=0.5746 token=0.8531 | 6.14 steps/s | VRAM=29.02 Go [2026-07-18 05:13:24] step 647150/1000000 | loss=2.2448 vocab=0.6321 token=0.9730 | 6.12 steps/s | VRAM=28.34 Go [2026-07-18 05:13:32] step 647200/1000000 | loss=2.3377 vocab=0.9058 token=1.1006 | 6.10 steps/s | VRAM=28.39 Go [2026-07-18 05:13:40] step 647250/1000000 | loss=2.2006 vocab=0.6273 token=0.9324 | 6.12 steps/s | VRAM=28.14 Go [2026-07-18 05:13:48] step 647300/1000000 | loss=2.2665 vocab=0.7701 token=1.2578 | 6.14 steps/s | VRAM=32.00 Go [2026-07-18 05:13:56] step 647350/1000000 | loss=2.1574 vocab=0.7119 token=1.3272 | 6.17 steps/s | VRAM=31.67 Go [2026-07-18 05:14:04] step 647400/1000000 | loss=2.2507 vocab=0.5269 token=1.2952 | 6.15 steps/s | VRAM=29.35 Go [2026-07-18 05:14:12] step 647450/1000000 | loss=2.3298 vocab=0.8504 token=1.4799 | 6.11 steps/s | VRAM=24.14 Go [2026-07-18 05:14:21] step 647500/1000000 | loss=2.1991 vocab=0.7701 token=0.9280 | 6.15 steps/s | VRAM=31.16 Go [2026-07-18 05:14:29] step 647550/1000000 | loss=2.2379 vocab=0.9708 token=1.4191 | 6.18 steps/s | VRAM=29.41 Go [2026-07-18 05:14:37] step 647600/1000000 | loss=2.3169 vocab=0.6852 token=1.3996 | 6.15 steps/s | VRAM=27.81 Go [2026-07-18 05:14:45] step 647650/1000000 | loss=2.2292 vocab=1.0499 token=1.2256 | 6.09 steps/s | VRAM=27.40 Go [2026-07-18 05:14:53] step 647700/1000000 | loss=2.1550 vocab=0.9464 token=1.0421 | 6.15 steps/s | VRAM=32.05 Go [2026-07-18 05:15:01] step 647750/1000000 | loss=2.2518 vocab=1.1164 token=1.6803 | 6.12 steps/s | VRAM=27.87 Go [2026-07-18 05:15:09] step 647800/1000000 | loss=2.1499 vocab=1.0774 token=1.4326 | 6.17 steps/s | VRAM=26.36 Go [2026-07-18 05:15:18] step 647850/1000000 | loss=2.2369 vocab=1.0741 token=1.0554 | 6.12 steps/s | VRAM=29.91 Go [2026-07-18 05:15:26] step 647900/1000000 | loss=2.2201 vocab=0.6855 token=1.2735 | 6.16 steps/s | VRAM=24.53 Go [2026-07-18 05:15:34] step 647950/1000000 | loss=2.1781 vocab=0.5779 token=1.0793 | 6.13 steps/s | VRAM=30.01 Go [2026-07-18 05:15:42] step 648000/1000000 | loss=2.2698 vocab=0.7874 token=1.3554 | 6.14 steps/s | VRAM=28.48 Go [2026-07-18 05:15:50] step 648050/1000000 | loss=2.1872 vocab=0.7271 token=1.3861 | 6.14 steps/s | VRAM=24.94 Go [2026-07-18 05:15:58] step 648100/1000000 | loss=2.2467 vocab=0.9722 token=1.4540 | 6.12 steps/s | VRAM=28.67 Go [2026-07-18 05:16:06] step 648150/1000000 | loss=2.2305 vocab=1.2211 token=1.6942 | 6.14 steps/s | VRAM=29.30 Go [2026-07-18 05:16:15] step 648200/1000000 | loss=2.2167 vocab=0.7474 token=1.3606 | 6.15 steps/s | VRAM=32.48 Go [2026-07-18 05:16:23] step 648250/1000000 | loss=2.1718 vocab=0.9913 token=1.2875 | 6.13 steps/s | VRAM=28.81 Go [2026-07-18 05:16:31] step 648300/1000000 | loss=2.2695 vocab=1.0366 token=0.9185 | 6.15 steps/s | VRAM=28.64 Go [2026-07-18 05:16:39] step 648350/1000000 | loss=2.1835 vocab=0.5870 token=1.4158 | 6.15 steps/s | VRAM=32.17 Go [2026-07-18 05:16:47] step 648400/1000000 | loss=2.2195 vocab=0.7619 token=1.0486 | 6.14 steps/s | VRAM=26.09 Go [2026-07-18 05:16:55] step 648450/1000000 | loss=2.2405 vocab=0.8541 token=1.3682 | 6.14 steps/s | VRAM=25.82 Go [2026-07-18 05:17:03] step 648500/1000000 | loss=2.3275 vocab=0.7382 token=1.5909 | 6.16 steps/s | VRAM=26.05 Go [2026-07-18 05:17:12] step 648550/1000000 | loss=2.2239 vocab=0.8341 token=1.6355 | 6.13 steps/s | VRAM=27.85 Go [2026-07-18 05:17:20] step 648600/1000000 | loss=2.3126 vocab=0.7773 token=1.2459 | 6.15 steps/s | VRAM=28.56 Go [2026-07-18 05:17:28] step 648650/1000000 | loss=2.2549 vocab=0.9944 token=1.3913 | 6.14 steps/s | VRAM=27.68 Go [2026-07-18 05:17:36] step 648700/1000000 | loss=2.2320 vocab=0.7347 token=1.5535 | 6.11 steps/s | VRAM=29.20 Go [2026-07-18 05:17:44] step 648750/1000000 | loss=2.1266 vocab=1.1988 token=1.5122 | 6.17 steps/s | VRAM=31.59 Go [2026-07-18 05:17:52] step 648800/1000000 | loss=2.2244 vocab=0.6402 token=1.2258 | 6.17 steps/s | VRAM=32.35 Go [2026-07-18 05:18:00] step 648850/1000000 | loss=2.1572 vocab=0.7566 token=0.9486 | 6.12 steps/s | VRAM=26.81 Go [2026-07-18 05:18:08] step 648900/1000000 | loss=2.2024 vocab=0.9169 token=1.3953 | 6.15 steps/s | VRAM=29.23 Go [2026-07-18 05:18:17] step 648950/1000000 | loss=2.1676 vocab=0.6098 token=1.0979 | 6.11 steps/s | VRAM=27.09 Go [2026-07-18 05:18:25] step 649000/1000000 | loss=2.1685 vocab=1.1825 token=1.2579 | 6.08 steps/s | VRAM=28.79 Go [2026-07-18 05:18:33] step 649050/1000000 | loss=2.2178 vocab=0.6843 token=1.3338 | 6.15 steps/s | VRAM=31.77 Go [2026-07-18 05:18:41] step 649100/1000000 | loss=2.1913 vocab=1.0824 token=1.0729 | 6.14 steps/s | VRAM=28.09 Go [2026-07-18 05:18:49] step 649150/1000000 | loss=2.2324 vocab=0.7071 token=1.6355 | 6.13 steps/s | VRAM=25.17 Go [2026-07-18 05:18:57] step 649200/1000000 | loss=2.2563 vocab=0.7608 token=1.2973 | 6.14 steps/s | VRAM=30.76 Go [2026-07-18 05:19:06] step 649250/1000000 | loss=2.2695 vocab=0.7406 token=1.1424 | 6.17 steps/s | VRAM=30.87 Go [2026-07-18 05:19:14] step 649300/1000000 | loss=2.3173 vocab=1.1840 token=1.4015 | 6.13 steps/s | VRAM=32.29 Go [2026-07-18 05:19:22] step 649350/1000000 | loss=2.3075 vocab=0.7709 token=1.3015 | 6.12 steps/s | VRAM=29.13 Go [2026-07-18 05:19:30] step 649400/1000000 | loss=2.2189 vocab=1.1262 token=1.3580 | 6.16 steps/s | VRAM=26.30 Go [2026-07-18 05:19:38] step 649450/1000000 | loss=2.2074 vocab=0.9850 token=1.2824 | 6.16 steps/s | VRAM=30.86 Go [2026-07-18 05:19:46] step 649500/1000000 | loss=2.2421 vocab=0.8249 token=1.4809 | 6.13 steps/s | VRAM=27.28 Go [2026-07-18 05:19:54] step 649550/1000000 | loss=2.2226 vocab=0.9176 token=1.2096 | 6.14 steps/s | VRAM=27.97 Go [2026-07-18 05:20:03] step 649600/1000000 | loss=2.2092 vocab=0.9821 token=0.9197 | 6.14 steps/s | VRAM=29.14 Go [2026-07-18 05:20:11] step 649650/1000000 | loss=2.1998 vocab=0.4310 token=0.8404 | 6.13 steps/s | VRAM=32.47 Go [2026-07-18 05:20:19] step 649700/1000000 | loss=2.2608 vocab=0.8856 token=1.4604 | 6.11 steps/s | VRAM=32.32 Go [2026-07-18 05:20:27] step 649750/1000000 | loss=2.0925 vocab=1.6524 token=1.2634 | 6.10 steps/s | VRAM=26.35 Go [2026-07-18 05:20:35] step 649800/1000000 | loss=2.1953 vocab=0.5746 token=1.1545 | 6.13 steps/s | VRAM=32.38 Go [2026-07-18 05:20:43] step 649850/1000000 | loss=2.2205 vocab=0.7805 token=1.1403 | 6.12 steps/s | VRAM=32.96 Go [2026-07-18 05:20:52] step 649900/1000000 | loss=2.2519 vocab=1.2691 token=1.4914 | 6.13 steps/s | VRAM=31.06 Go [2026-07-18 05:21:00] step 649950/1000000 | loss=2.2108 vocab=1.0963 token=1.6496 | 6.10 steps/s | VRAM=26.92 Go [2026-07-18 05:21:08] step 650000/1000000 | loss=2.1951 vocab=1.1343 token=1.6893 | 6.12 steps/s | VRAM=28.44 Go [step 650000] loss_token=1.6893 > min recent (1.0528) -> push saute [2026-07-18 05:21:16] step 650050/1000000 | loss=2.1930 vocab=0.5192 token=1.0671 | 6.15 steps/s | VRAM=31.74 Go [2026-07-18 05:21:24] step 650100/1000000 | loss=2.1241 vocab=1.0585 token=1.2665 | 6.12 steps/s | VRAM=29.44 Go [2026-07-18 05:21:32] step 650150/1000000 | loss=2.2847 vocab=0.8289 token=1.1354 | 6.14 steps/s | VRAM=25.74 Go [2026-07-18 05:21:41] step 650200/1000000 | loss=2.1593 vocab=0.9357 token=1.4817 | 6.12 steps/s | VRAM=29.26 Go [2026-07-18 05:21:49] step 650250/1000000 | loss=2.2729 vocab=0.9892 token=1.1754 | 6.13 steps/s | VRAM=28.57 Go [2026-07-18 05:21:57] step 650300/1000000 | loss=2.2095 vocab=0.9238 token=1.1132 | 6.10 steps/s | VRAM=26.49 Go [2026-07-18 05:22:05] step 650350/1000000 | loss=2.1812 vocab=0.5867 token=1.2481 | 6.12 steps/s | VRAM=32.37 Go [2026-07-18 05:22:13] step 650400/1000000 | loss=2.3168 vocab=0.5974 token=1.2312 | 6.10 steps/s | VRAM=25.56 Go [2026-07-18 05:22:21] step 650450/1000000 | loss=2.2427 vocab=1.1120 token=1.3515 | 6.11 steps/s | VRAM=30.44 Go [2026-07-18 05:22:30] step 650500/1000000 | loss=2.1770 vocab=0.9282 token=1.3292 | 6.16 steps/s | VRAM=28.21 Go [2026-07-18 05:22:38] step 650550/1000000 | loss=2.2091 vocab=0.8350 token=1.2847 | 6.14 steps/s | VRAM=32.60 Go [2026-07-18 05:22:46] step 650600/1000000 | loss=2.0956 vocab=0.5126 token=1.1025 | 6.11 steps/s | VRAM=29.83 Go [2026-07-18 05:22:54] step 650650/1000000 | loss=2.2584 vocab=1.0335 token=1.4122 | 6.10 steps/s | VRAM=29.36 Go [2026-07-18 05:23:02] step 650700/1000000 | loss=2.0982 vocab=0.9078 token=1.4719 | 6.10 steps/s | VRAM=28.98 Go [2026-07-18 05:23:10] step 650750/1000000 | loss=2.2434 vocab=0.9202 token=1.5200 | 6.15 steps/s | VRAM=28.37 Go [2026-07-18 05:23:19] step 650800/1000000 | loss=2.2027 vocab=1.6544 token=1.7722 | 6.12 steps/s | VRAM=29.18 Go [2026-07-18 05:23:27] step 650850/1000000 | loss=2.2352 vocab=0.9090 token=1.3770 | 6.12 steps/s | VRAM=30.08 Go [2026-07-18 05:23:35] step 650900/1000000 | loss=2.2601 vocab=1.0901 token=1.5370 | 6.14 steps/s | VRAM=27.97 Go [2026-07-18 05:23:43] step 650950/1000000 | loss=2.1864 vocab=0.3832 token=1.4217 | 6.14 steps/s | VRAM=25.01 Go [2026-07-18 05:23:51] step 651000/1000000 | loss=2.2625 vocab=0.8777 token=1.1635 | 6.16 steps/s | VRAM=28.88 Go [2026-07-18 05:23:59] step 651050/1000000 | loss=2.2043 vocab=0.7442 token=1.2283 | 6.11 steps/s | VRAM=23.44 Go [2026-07-18 05:24:07] step 651100/1000000 | loss=2.1220 vocab=0.9178 token=1.0688 | 6.17 steps/s | VRAM=25.20 Go [2026-07-18 05:24:16] step 651150/1000000 | loss=2.1725 vocab=0.8152 token=1.4050 | 6.15 steps/s | VRAM=28.31 Go [2026-07-18 05:24:24] step 651200/1000000 | loss=2.2025 vocab=1.3482 token=1.3078 | 6.16 steps/s | VRAM=28.22 Go [2026-07-18 05:24:32] step 651250/1000000 | loss=2.1083 vocab=0.7797 token=1.4379 | 6.13 steps/s | VRAM=25.95 Go [2026-07-18 05:24:40] step 651300/1000000 | loss=2.1906 vocab=1.0760 token=1.4240 | 6.15 steps/s | VRAM=25.80 Go [2026-07-18 05:24:48] step 651350/1000000 | loss=2.2359 vocab=0.9286 token=1.3064 | 6.15 steps/s | VRAM=32.01 Go [2026-07-18 05:24:56] step 651400/1000000 | loss=2.1875 vocab=1.0471 token=1.5712 | 6.13 steps/s | VRAM=31.50 Go [2026-07-18 05:25:04] step 651450/1000000 | loss=2.3036 vocab=0.5663 token=1.0303 | 6.13 steps/s | VRAM=28.81 Go [2026-07-18 05:25:13] step 651500/1000000 | loss=2.1171 vocab=0.6626 token=1.2127 | 6.11 steps/s | VRAM=29.80 Go [2026-07-18 05:25:21] step 651550/1000000 | loss=2.2212 vocab=0.8418 token=1.4222 | 6.12 steps/s | VRAM=32.62 Go [2026-07-18 05:25:29] step 651600/1000000 | loss=2.1878 vocab=1.1063 token=1.1034 | 6.14 steps/s | VRAM=25.59 Go [2026-07-18 05:25:37] step 651650/1000000 | loss=2.2178 vocab=0.8277 token=1.4913 | 6.14 steps/s | VRAM=29.34 Go [2026-07-18 05:25:45] step 651700/1000000 | loss=2.1166 vocab=0.9250 token=1.0388 | 6.10 steps/s | VRAM=27.79 Go [2026-07-18 05:25:53] step 651750/1000000 | loss=2.2256 vocab=0.9658 token=1.0539 | 6.11 steps/s | VRAM=28.68 Go [2026-07-18 05:26:02] step 651800/1000000 | loss=2.3940 vocab=1.1127 token=1.3856 | 6.14 steps/s | VRAM=29.11 Go [2026-07-18 05:26:10] step 651850/1000000 | loss=2.2056 vocab=0.6844 token=1.1172 | 6.15 steps/s | VRAM=31.92 Go [2026-07-18 05:26:18] step 651900/1000000 | loss=2.0647 vocab=0.4557 token=1.1208 | 6.13 steps/s | VRAM=25.93 Go [2026-07-18 05:26:26] step 651950/1000000 | loss=2.1240 vocab=1.1343 token=1.4647 | 6.13 steps/s | VRAM=29.50 Go [2026-07-18 05:26:34] step 652000/1000000 | loss=2.1049 vocab=0.6721 token=1.1639 | 6.12 steps/s | VRAM=25.96 Go [2026-07-18 05:26:42] step 652050/1000000 | loss=2.2976 vocab=0.8163 token=1.4565 | 6.14 steps/s | VRAM=26.25 Go [2026-07-18 05:26:51] step 652100/1000000 | loss=2.1844 vocab=1.0043 token=1.2088 | 6.14 steps/s | VRAM=27.74 Go [2026-07-18 05:26:59] step 652150/1000000 | loss=2.1695 vocab=0.8994 token=1.7702 | 6.15 steps/s | VRAM=28.79 Go [2026-07-18 05:27:07] step 652200/1000000 | loss=2.2621 vocab=0.6027 token=1.0110 | 6.12 steps/s | VRAM=26.37 Go [2026-07-18 05:27:15] step 652250/1000000 | loss=2.1963 vocab=0.9527 token=1.2429 | 6.15 steps/s | VRAM=26.37 Go [2026-07-18 05:27:23] step 652300/1000000 | loss=2.2697 vocab=0.8053 token=1.4041 | 6.16 steps/s | VRAM=28.96 Go [2026-07-18 05:27:31] step 652350/1000000 | loss=2.3167 vocab=0.8566 token=1.2759 | 6.14 steps/s | VRAM=28.38 Go [2026-07-18 05:27:39] step 652400/1000000 | loss=2.2413 vocab=0.9997 token=1.0885 | 6.15 steps/s | VRAM=28.54 Go [2026-07-18 05:27:48] step 652450/1000000 | loss=2.1819 vocab=0.9771 token=1.4527 | 6.14 steps/s | VRAM=28.08 Go [2026-07-18 05:27:56] step 652500/1000000 | loss=2.1348 vocab=1.2688 token=1.2966 | 6.13 steps/s | VRAM=32.28 Go [2026-07-18 05:28:04] step 652550/1000000 | loss=2.0925 vocab=1.0095 token=1.5318 | 6.14 steps/s | VRAM=29.25 Go [2026-07-18 05:28:12] step 652600/1000000 | loss=2.3435 vocab=0.8111 token=1.0929 | 6.16 steps/s | VRAM=32.98 Go [2026-07-18 05:28:20] step 652650/1000000 | loss=2.3123 vocab=0.6200 token=1.2772 | 6.14 steps/s | VRAM=29.16 Go [2026-07-18 05:28:28] step 652700/1000000 | loss=2.1068 vocab=0.7414 token=1.1752 | 6.13 steps/s | VRAM=28.89 Go [2026-07-18 05:28:36] step 652750/1000000 | loss=2.2219 vocab=0.8092 token=1.3295 | 6.15 steps/s | VRAM=29.20 Go [2026-07-18 05:28:45] step 652800/1000000 | loss=2.0248 vocab=0.7973 token=0.8986 | 6.13 steps/s | VRAM=28.89 Go [2026-07-18 05:28:53] step 652850/1000000 | loss=2.2107 vocab=0.9502 token=1.4968 | 6.15 steps/s | VRAM=31.24 Go [2026-07-18 05:29:01] step 652900/1000000 | loss=2.2579 vocab=0.7630 token=1.3404 | 6.15 steps/s | VRAM=32.97 Go [2026-07-18 05:29:09] step 652950/1000000 | loss=2.3263 vocab=0.4750 token=0.8329 | 6.16 steps/s | VRAM=31.56 Go [2026-07-18 05:29:17] step 653000/1000000 | loss=2.2969 vocab=0.5401 token=1.1784 | 6.13 steps/s | VRAM=29.39 Go [2026-07-18 05:29:25] step 653050/1000000 | loss=2.2042 vocab=1.2267 token=1.4406 | 6.16 steps/s | VRAM=28.27 Go [2026-07-18 05:29:33] step 653100/1000000 | loss=2.2031 vocab=0.5439 token=1.1969 | 6.16 steps/s | VRAM=28.33 Go [2026-07-18 05:29:41] step 653150/1000000 | loss=2.1078 vocab=0.9753 token=1.2121 | 6.15 steps/s | VRAM=28.18 Go [2026-07-18 05:29:50] step 653200/1000000 | loss=2.2248 vocab=0.8329 token=0.9975 | 6.13 steps/s | VRAM=32.79 Go [2026-07-18 05:29:58] step 653250/1000000 | loss=2.2658 vocab=1.0848 token=1.2636 | 6.14 steps/s | VRAM=24.84 Go [2026-07-18 05:30:06] step 653300/1000000 | loss=2.0882 vocab=1.1809 token=1.0802 | 6.15 steps/s | VRAM=32.77 Go [2026-07-18 05:30:14] step 653350/1000000 | loss=2.1960 vocab=0.8418 token=1.1770 | 6.15 steps/s | VRAM=28.66 Go [2026-07-18 05:30:22] step 653400/1000000 | loss=2.2823 vocab=0.8892 token=1.4694 | 6.12 steps/s | VRAM=25.51 Go [2026-07-18 05:30:30] step 653450/1000000 | loss=2.2807 vocab=0.7850 token=1.2730 | 6.15 steps/s | VRAM=24.65 Go [2026-07-18 05:30:38] step 653500/1000000 | loss=2.1313 vocab=0.7305 token=1.0466 | 6.12 steps/s | VRAM=25.76 Go [2026-07-18 05:30:47] step 653550/1000000 | loss=2.2714 vocab=0.8091 token=1.4053 | 6.11 steps/s | VRAM=26.46 Go [2026-07-18 05:30:55] step 653600/1000000 | loss=2.2522 vocab=0.7618 token=1.1252 | 6.13 steps/s | VRAM=28.45 Go [2026-07-18 05:31:03] step 653650/1000000 | loss=2.2440 vocab=0.6848 token=0.9112 | 6.15 steps/s | VRAM=29.04 Go [2026-07-18 05:31:11] step 653700/1000000 | loss=2.2458 vocab=0.6854 token=1.5431 | 6.14 steps/s | VRAM=28.77 Go [2026-07-18 05:31:19] step 653750/1000000 | loss=2.2923 vocab=1.1373 token=1.0064 | 6.17 steps/s | VRAM=28.60 Go [2026-07-18 05:31:27] step 653800/1000000 | loss=2.1856 vocab=0.5989 token=1.2075 | 6.12 steps/s | VRAM=28.31 Go [2026-07-18 05:31:36] step 653850/1000000 | loss=2.2838 vocab=0.6712 token=0.9034 | 6.12 steps/s | VRAM=30.34 Go [2026-07-18 05:31:44] step 653900/1000000 | loss=2.2222 vocab=0.5182 token=0.6562 | 6.14 steps/s | VRAM=28.32 Go [2026-07-18 05:31:52] step 653950/1000000 | loss=2.2734 vocab=1.1050 token=1.2821 | 6.12 steps/s | VRAM=26.56 Go [2026-07-18 05:32:00] step 654000/1000000 | loss=2.1598 vocab=0.8480 token=1.5252 | 6.15 steps/s | VRAM=30.64 Go [2026-07-18 05:32:08] step 654050/1000000 | loss=2.1597 vocab=1.1734 token=1.5832 | 6.13 steps/s | VRAM=28.50 Go [2026-07-18 05:32:16] step 654100/1000000 | loss=2.2153 vocab=0.6368 token=0.9407 | 6.14 steps/s | VRAM=26.32 Go [2026-07-18 05:32:24] step 654150/1000000 | loss=2.2503 vocab=0.8440 token=1.1711 | 6.15 steps/s | VRAM=28.14 Go [2026-07-18 05:32:33] step 654200/1000000 | loss=2.1831 vocab=0.9306 token=1.4273 | 6.15 steps/s | VRAM=32.90 Go [2026-07-18 05:32:41] step 654250/1000000 | loss=2.1289 vocab=1.0224 token=1.9286 | 6.15 steps/s | VRAM=29.33 Go [2026-07-18 05:32:49] step 654300/1000000 | loss=2.2059 vocab=0.8766 token=1.2413 | 6.19 steps/s | VRAM=25.10 Go [2026-07-18 05:32:57] step 654350/1000000 | loss=2.2380 vocab=1.0238 token=1.7868 | 6.15 steps/s | VRAM=29.59 Go [2026-07-18 05:33:05] step 654400/1000000 | loss=2.1605 vocab=1.1748 token=1.5114 | 6.18 steps/s | VRAM=29.36 Go [2026-07-18 05:33:13] step 654450/1000000 | loss=2.1760 vocab=0.9167 token=1.1922 | 6.17 steps/s | VRAM=30.71 Go [2026-07-18 05:33:21] step 654500/1000000 | loss=2.1867 vocab=0.7368 token=1.2057 | 6.15 steps/s | VRAM=26.36 Go [2026-07-18 05:33:29] step 654550/1000000 | loss=2.1953 vocab=0.9804 token=1.2310 | 6.14 steps/s | VRAM=28.92 Go [2026-07-18 05:33:38] step 654600/1000000 | loss=2.2293 vocab=0.8247 token=1.5079 | 6.13 steps/s | VRAM=32.65 Go [2026-07-18 05:33:46] step 654650/1000000 | loss=2.1515 vocab=0.7188 token=1.4235 | 6.15 steps/s | VRAM=28.25 Go [2026-07-18 05:33:54] step 654700/1000000 | loss=2.1932 vocab=0.6868 token=1.2441 | 6.16 steps/s | VRAM=24.93 Go [2026-07-18 05:34:02] step 654750/1000000 | loss=2.1841 vocab=0.9392 token=1.5051 | 6.14 steps/s | VRAM=27.44 Go [2026-07-18 05:34:10] step 654800/1000000 | loss=2.2777 vocab=0.7498 token=1.3942 | 6.16 steps/s | VRAM=28.00 Go [2026-07-18 05:34:18] step 654850/1000000 | loss=2.2441 vocab=1.0422 token=1.4055 | 6.16 steps/s | VRAM=26.00 Go [2026-07-18 05:34:26] step 654900/1000000 | loss=2.1929 vocab=1.8034 token=1.6872 | 6.12 steps/s | VRAM=28.14 Go [2026-07-18 05:34:34] step 654950/1000000 | loss=2.1925 vocab=1.1967 token=1.5865 | 6.13 steps/s | VRAM=28.11 Go [2026-07-18 05:34:43] step 655000/1000000 | loss=2.1405 vocab=1.0786 token=1.2934 | 6.15 steps/s | VRAM=24.81 Go [2026-07-18 05:34:51] step 655050/1000000 | loss=2.3536 vocab=1.0050 token=1.5786 | 6.14 steps/s | VRAM=28.24 Go [2026-07-18 05:34:59] step 655100/1000000 | loss=2.1768 vocab=0.8837 token=1.5952 | 6.15 steps/s | VRAM=26.46 Go [2026-07-18 05:35:07] step 655150/1000000 | loss=2.1882 vocab=0.6605 token=0.9401 | 6.16 steps/s | VRAM=31.45 Go [2026-07-18 05:35:15] step 655200/1000000 | loss=2.2969 vocab=1.2170 token=1.1608 | 6.12 steps/s | VRAM=28.89 Go [2026-07-18 05:35:23] step 655250/1000000 | loss=2.1506 vocab=0.7297 token=1.3767 | 6.14 steps/s | VRAM=29.05 Go [2026-07-18 05:35:31] step 655300/1000000 | loss=2.3102 vocab=0.4505 token=1.0335 | 6.13 steps/s | VRAM=29.62 Go [2026-07-18 05:35:40] step 655350/1000000 | loss=2.1809 vocab=0.9762 token=1.6351 | 6.11 steps/s | VRAM=28.35 Go [2026-07-18 05:35:48] step 655400/1000000 | loss=2.2150 vocab=0.5495 token=1.3109 | 6.17 steps/s | VRAM=27.45 Go [2026-07-18 05:35:56] step 655450/1000000 | loss=2.3405 vocab=1.8148 token=1.5203 | 6.13 steps/s | VRAM=32.72 Go [2026-07-18 05:36:04] step 655500/1000000 | loss=2.2117 vocab=0.5346 token=1.0881 | 6.16 steps/s | VRAM=32.21 Go [2026-07-18 05:36:12] step 655550/1000000 | loss=2.1298 vocab=0.7562 token=1.1815 | 6.12 steps/s | VRAM=25.96 Go [2026-07-18 05:36:20] step 655600/1000000 | loss=2.1789 vocab=0.8562 token=0.9978 | 6.15 steps/s | VRAM=32.48 Go [2026-07-18 05:36:28] step 655650/1000000 | loss=2.2733 vocab=1.7715 token=1.5627 | 6.17 steps/s | VRAM=32.19 Go [2026-07-18 05:36:37] step 655700/1000000 | loss=2.2593 vocab=1.2601 token=1.6181 | 6.14 steps/s | VRAM=32.90 Go [2026-07-18 05:36:45] step 655750/1000000 | loss=2.2868 vocab=0.7375 token=1.3920 | 6.14 steps/s | VRAM=29.20 Go [2026-07-18 05:36:53] step 655800/1000000 | loss=2.2251 vocab=0.7179 token=1.7071 | 6.15 steps/s | VRAM=32.41 Go [2026-07-18 05:37:01] step 655850/1000000 | loss=2.1426 vocab=1.0000 token=1.3219 | 6.14 steps/s | VRAM=28.26 Go [2026-07-18 05:37:09] step 655900/1000000 | loss=2.1586 vocab=0.9681 token=1.7653 | 6.14 steps/s | VRAM=25.78 Go [2026-07-18 05:37:17] step 655950/1000000 | loss=2.1992 vocab=0.9999 token=1.4239 | 6.13 steps/s | VRAM=29.24 Go [2026-07-18 05:37:25] step 656000/1000000 | loss=2.2443 vocab=1.3200 token=1.3556 | 6.14 steps/s | VRAM=32.47 Go [2026-07-18 05:37:34] step 656050/1000000 | loss=2.1870 vocab=0.9493 token=1.1497 | 6.17 steps/s | VRAM=28.48 Go [2026-07-18 05:37:42] step 656100/1000000 | loss=2.0857 vocab=0.7725 token=1.0739 | 6.14 steps/s | VRAM=29.53 Go [2026-07-18 05:37:50] step 656150/1000000 | loss=2.1866 vocab=1.1022 token=1.5052 | 6.15 steps/s | VRAM=28.61 Go [2026-07-18 05:37:58] step 656200/1000000 | loss=2.2076 vocab=0.9318 token=1.7099 | 6.16 steps/s | VRAM=24.04 Go [2026-07-18 05:38:06] step 656250/1000000 | loss=2.2417 vocab=0.8272 token=1.0746 | 6.16 steps/s | VRAM=28.89 Go [2026-07-18 05:38:14] step 656300/1000000 | loss=2.2485 vocab=0.8916 token=1.2038 | 6.17 steps/s | VRAM=28.02 Go [2026-07-18 05:38:22] step 656350/1000000 | loss=2.2691 vocab=1.0299 token=1.4299 | 6.18 steps/s | VRAM=24.64 Go [2026-07-18 05:38:30] step 656400/1000000 | loss=2.1526 vocab=0.8145 token=1.6122 | 6.18 steps/s | VRAM=28.08 Go [2026-07-18 05:38:38] step 656450/1000000 | loss=2.1440 vocab=1.1018 token=1.3109 | 6.17 steps/s | VRAM=27.56 Go [2026-07-18 05:38:47] step 656500/1000000 | loss=2.2191 vocab=1.0272 token=0.8846 | 6.17 steps/s | VRAM=28.38 Go [2026-07-18 05:38:55] step 656550/1000000 | loss=2.2626 vocab=0.8134 token=1.3722 | 6.14 steps/s | VRAM=28.47 Go [2026-07-18 05:39:03] step 656600/1000000 | loss=2.2679 vocab=0.8989 token=1.2555 | 6.18 steps/s | VRAM=31.67 Go [2026-07-18 05:39:11] step 656650/1000000 | loss=2.2868 vocab=1.1856 token=1.5349 | 6.14 steps/s | VRAM=28.55 Go [2026-07-18 05:39:19] step 656700/1000000 | loss=2.2466 vocab=1.0313 token=1.5111 | 6.14 steps/s | VRAM=29.54 Go [2026-07-18 05:39:27] step 656750/1000000 | loss=2.1540 vocab=1.6008 token=1.6268 | 6.13 steps/s | VRAM=28.66 Go [2026-07-18 05:39:35] step 656800/1000000 | loss=2.2157 vocab=0.9991 token=1.4292 | 6.17 steps/s | VRAM=30.66 Go [2026-07-18 05:39:43] step 656850/1000000 | loss=2.2350 vocab=0.9050 token=1.2662 | 6.14 steps/s | VRAM=29.37 Go [2026-07-18 05:39:52] step 656900/1000000 | loss=2.1945 vocab=1.1087 token=1.5071 | 6.18 steps/s | VRAM=28.80 Go [2026-07-18 05:40:00] step 656950/1000000 | loss=2.1595 vocab=0.7354 token=1.3433 | 6.14 steps/s | VRAM=26.95 Go [2026-07-18 05:40:08] step 657000/1000000 | loss=2.2627 vocab=1.1419 token=1.7277 | 6.15 steps/s | VRAM=28.80 Go [2026-07-18 05:40:16] step 657050/1000000 | loss=2.1536 vocab=0.6215 token=1.2631 | 6.16 steps/s | VRAM=27.51 Go [2026-07-18 05:40:24] step 657100/1000000 | loss=2.2891 vocab=0.9465 token=1.3171 | 6.17 steps/s | VRAM=23.71 Go [2026-07-18 05:40:32] step 657150/1000000 | loss=2.1743 vocab=0.4432 token=1.3493 | 6.16 steps/s | VRAM=25.76 Go [2026-07-18 05:40:40] step 657200/1000000 | loss=2.1755 vocab=0.8535 token=1.3591 | 6.14 steps/s | VRAM=32.58 Go [2026-07-18 05:40:48] step 657250/1000000 | loss=2.2040 vocab=0.7427 token=1.0473 | 6.14 steps/s | VRAM=32.32 Go [2026-07-18 05:40:57] step 657300/1000000 | loss=2.2697 vocab=1.1383 token=1.2908 | 6.16 steps/s | VRAM=32.24 Go [2026-07-18 05:41:05] step 657350/1000000 | loss=2.2216 vocab=0.6232 token=0.8452 | 6.16 steps/s | VRAM=25.76 Go [2026-07-18 05:41:13] step 657400/1000000 | loss=2.1813 vocab=0.5889 token=1.2190 | 6.11 steps/s | VRAM=28.78 Go [2026-07-18 05:41:21] step 657450/1000000 | loss=2.1984 vocab=1.1922 token=1.0910 | 6.15 steps/s | VRAM=26.25 Go [2026-07-18 05:41:29] step 657500/1000000 | loss=2.1292 vocab=0.4530 token=0.8238 | 6.14 steps/s | VRAM=31.75 Go [2026-07-18 05:41:37] step 657550/1000000 | loss=2.1943 vocab=0.6507 token=1.3586 | 6.14 steps/s | VRAM=28.86 Go [2026-07-18 05:41:45] step 657600/1000000 | loss=2.3337 vocab=0.8330 token=1.5673 | 6.17 steps/s | VRAM=29.16 Go [2026-07-18 05:41:54] step 657650/1000000 | loss=2.1824 vocab=0.7038 token=1.3832 | 6.18 steps/s | VRAM=27.93 Go [2026-07-18 05:42:02] step 657700/1000000 | loss=2.2024 vocab=0.8612 token=1.2074 | 6.16 steps/s | VRAM=28.97 Go [2026-07-18 05:42:10] step 657750/1000000 | loss=2.1735 vocab=0.8508 token=1.2446 | 6.16 steps/s | VRAM=31.24 Go [2026-07-18 05:42:18] step 657800/1000000 | loss=2.1980 vocab=0.7342 token=1.6111 | 6.14 steps/s | VRAM=27.49 Go [2026-07-18 05:42:26] step 657850/1000000 | loss=2.1859 vocab=0.8307 token=1.5912 | 6.17 steps/s | VRAM=27.14 Go [2026-07-18 05:42:34] step 657900/1000000 | loss=2.1395 vocab=0.6489 token=0.9492 | 6.16 steps/s | VRAM=26.07 Go [2026-07-18 05:42:42] step 657950/1000000 | loss=2.1849 vocab=0.5108 token=1.0206 | 6.15 steps/s | VRAM=28.11 Go [2026-07-18 05:42:50] step 658000/1000000 | loss=2.3172 vocab=0.6188 token=1.0924 | 6.11 steps/s | VRAM=32.35 Go [2026-07-18 05:42:59] step 658050/1000000 | loss=2.1892 vocab=1.1550 token=1.6331 | 6.18 steps/s | VRAM=31.88 Go [2026-07-18 05:43:07] step 658100/1000000 | loss=2.2199 vocab=0.9428 token=1.3691 | 6.10 steps/s | VRAM=28.59 Go [2026-07-18 05:43:15] step 658150/1000000 | loss=2.1660 vocab=0.6552 token=1.3635 | 6.14 steps/s | VRAM=32.28 Go [2026-07-18 05:43:23] step 658200/1000000 | loss=2.2135 vocab=0.7137 token=1.5176 | 6.15 steps/s | VRAM=31.91 Go [2026-07-18 05:43:31] step 658250/1000000 | loss=2.2904 vocab=0.8952 token=1.1164 | 6.14 steps/s | VRAM=28.93 Go [2026-07-18 05:43:39] step 658300/1000000 | loss=2.2483 vocab=0.5205 token=0.9342 | 6.17 steps/s | VRAM=31.31 Go [2026-07-18 05:43:47] step 658350/1000000 | loss=2.2573 vocab=0.9592 token=1.1865 | 6.11 steps/s | VRAM=33.01 Go [2026-07-18 05:43:56] step 658400/1000000 | loss=2.1899 vocab=1.0019 token=1.2823 | 6.14 steps/s | VRAM=27.81 Go [2026-07-18 05:44:04] step 658450/1000000 | loss=2.2652 vocab=0.8896 token=1.3499 | 6.13 steps/s | VRAM=28.04 Go [2026-07-18 05:44:12] step 658500/1000000 | loss=2.2503 vocab=1.2715 token=1.4914 | 6.13 steps/s | VRAM=27.86 Go [2026-07-18 05:44:20] step 658550/1000000 | loss=2.2175 vocab=1.0537 token=1.6386 | 6.11 steps/s | VRAM=28.76 Go [2026-07-18 05:44:28] step 658600/1000000 | loss=2.2070 vocab=0.8167 token=1.2346 | 6.13 steps/s | VRAM=28.97 Go [2026-07-18 05:44:36] step 658650/1000000 | loss=2.2208 vocab=0.8299 token=1.2550 | 6.13 steps/s | VRAM=29.15 Go [2026-07-18 05:44:45] step 658700/1000000 | loss=2.2495 vocab=0.9672 token=1.3024 | 6.13 steps/s | VRAM=26.30 Go [2026-07-18 05:44:53] step 658750/1000000 | loss=2.3193 vocab=1.3597 token=1.3910 | 6.13 steps/s | VRAM=28.53 Go [2026-07-18 05:45:01] step 658800/1000000 | loss=2.1879 vocab=0.4350 token=0.8894 | 6.10 steps/s | VRAM=26.79 Go [2026-07-18 05:45:09] step 658850/1000000 | loss=2.2051 vocab=0.5466 token=1.1112 | 6.14 steps/s | VRAM=28.62 Go [2026-07-18 05:45:17] step 658900/1000000 | loss=2.2612 vocab=1.0941 token=1.1372 | 6.15 steps/s | VRAM=29.04 Go [2026-07-18 05:45:25] step 658950/1000000 | loss=2.1378 vocab=0.9646 token=1.3782 | 6.15 steps/s | VRAM=32.35 Go [2026-07-18 05:45:33] step 659000/1000000 | loss=2.2863 vocab=0.7464 token=1.1831 | 6.12 steps/s | VRAM=25.80 Go [2026-07-18 05:45:42] step 659050/1000000 | loss=2.1965 vocab=0.6764 token=1.4392 | 6.12 steps/s | VRAM=29.11 Go [2026-07-18 05:45:50] step 659100/1000000 | loss=2.1228 vocab=0.8643 token=0.9132 | 6.15 steps/s | VRAM=32.54 Go [2026-07-18 05:45:58] step 659150/1000000 | loss=2.1732 vocab=0.7628 token=1.2988 | 6.13 steps/s | VRAM=26.04 Go [2026-07-18 05:46:06] step 659200/1000000 | loss=2.2398 vocab=0.8405 token=1.4784 | 6.14 steps/s | VRAM=28.47 Go [2026-07-18 05:46:14] step 659250/1000000 | loss=2.2297 vocab=0.4735 token=1.3412 | 6.12 steps/s | VRAM=29.70 Go [2026-07-18 05:46:22] step 659300/1000000 | loss=2.2380 vocab=1.4587 token=1.4938 | 6.12 steps/s | VRAM=25.59 Go [2026-07-18 05:46:31] step 659350/1000000 | loss=2.1969 vocab=1.2246 token=1.4461 | 6.14 steps/s | VRAM=32.08 Go [2026-07-18 05:46:39] step 659400/1000000 | loss=2.2747 vocab=0.8795 token=0.8880 | 6.10 steps/s | VRAM=26.59 Go [2026-07-18 05:46:47] step 659450/1000000 | loss=2.2734 vocab=1.1136 token=1.5036 | 6.13 steps/s | VRAM=28.91 Go [2026-07-18 05:46:55] step 659500/1000000 | loss=2.1742 vocab=0.5637 token=1.1577 | 6.12 steps/s | VRAM=29.26 Go [2026-07-18 05:47:03] step 659550/1000000 | loss=2.2849 vocab=0.8775 token=1.2513 | 6.15 steps/s | VRAM=31.44 Go [2026-07-18 05:47:11] step 659600/1000000 | loss=2.3037 vocab=1.1840 token=1.7429 | 6.14 steps/s | VRAM=26.92 Go [2026-07-18 05:47:19] step 659650/1000000 | loss=2.1834 vocab=0.5277 token=1.1964 | 6.12 steps/s | VRAM=32.30 Go [2026-07-18 05:47:28] step 659700/1000000 | loss=2.2491 vocab=1.0783 token=1.7749 | 6.10 steps/s | VRAM=29.86 Go [2026-07-18 05:47:36] step 659750/1000000 | loss=2.2930 vocab=1.1813 token=1.3980 | 6.15 steps/s | VRAM=32.26 Go [2026-07-18 05:47:44] step 659800/1000000 | loss=2.1425 vocab=0.9547 token=1.5183 | 6.12 steps/s | VRAM=28.50 Go [2026-07-18 05:47:52] step 659850/1000000 | loss=2.1483 vocab=0.6800 token=0.9755 | 6.12 steps/s | VRAM=31.54 Go [2026-07-18 05:48:00] step 659900/1000000 | loss=2.1544 vocab=0.9204 token=1.7051 | 6.16 steps/s | VRAM=32.70 Go [2026-07-18 05:48:08] step 659950/1000000 | loss=2.2404 vocab=1.0818 token=1.5236 | 6.11 steps/s | VRAM=29.49 Go [2026-07-18 05:48:17] step 660000/1000000 | loss=2.1271 vocab=0.6342 token=1.1788 | 6.14 steps/s | VRAM=32.06 Go [2026-07-18 05:48:25] step 660050/1000000 | loss=2.2160 vocab=1.0051 token=0.8014 | 6.15 steps/s | VRAM=31.09 Go [2026-07-18 05:48:33] step 660100/1000000 | loss=2.2174 vocab=0.6617 token=1.2895 | 6.14 steps/s | VRAM=26.95 Go [2026-07-18 05:48:41] step 660150/1000000 | loss=2.1897 vocab=0.8046 token=1.5759 | 6.17 steps/s | VRAM=28.86 Go [2026-07-18 05:48:49] step 660200/1000000 | loss=2.1777 vocab=1.3508 token=1.1761 | 6.16 steps/s | VRAM=22.78 Go [2026-07-18 05:48:57] step 660250/1000000 | loss=2.2611 vocab=0.7733 token=1.2732 | 6.15 steps/s | VRAM=32.19 Go [2026-07-18 05:49:05] step 660300/1000000 | loss=2.0895 vocab=0.9778 token=0.9923 | 6.13 steps/s | VRAM=26.76 Go [2026-07-18 05:49:14] step 660350/1000000 | loss=2.2149 vocab=0.9342 token=1.3872 | 6.15 steps/s | VRAM=27.88 Go [2026-07-18 05:49:22] step 660400/1000000 | loss=2.2674 vocab=0.9492 token=1.2510 | 6.12 steps/s | VRAM=28.94 Go [2026-07-18 05:49:30] step 660450/1000000 | loss=2.1518 vocab=0.4839 token=1.0362 | 6.13 steps/s | VRAM=28.02 Go [2026-07-18 05:49:38] step 660500/1000000 | loss=2.1256 vocab=0.6826 token=0.9238 | 6.17 steps/s | VRAM=32.39 Go [2026-07-18 05:49:46] step 660550/1000000 | loss=2.2393 vocab=0.7935 token=1.1240 | 6.12 steps/s | VRAM=25.79 Go [2026-07-18 05:49:54] step 660600/1000000 | loss=2.1578 vocab=0.5375 token=1.2472 | 6.15 steps/s | VRAM=32.59 Go [2026-07-18 05:50:02] step 660650/1000000 | loss=2.2728 vocab=1.6593 token=1.5363 | 6.14 steps/s | VRAM=29.30 Go [2026-07-18 05:50:11] step 660700/1000000 | loss=2.1416 vocab=1.1456 token=1.1145 | 6.14 steps/s | VRAM=28.27 Go [2026-07-18 05:50:19] step 660750/1000000 | loss=2.1682 vocab=1.0910 token=1.2895 | 6.16 steps/s | VRAM=29.48 Go [2026-07-18 05:50:27] step 660800/1000000 | loss=2.0987 vocab=1.1396 token=1.0814 | 6.14 steps/s | VRAM=29.36 Go [2026-07-18 05:50:35] step 660850/1000000 | loss=2.1836 vocab=1.5626 token=1.3519 | 6.13 steps/s | VRAM=29.25 Go [2026-07-18 05:50:43] step 660900/1000000 | loss=2.1459 vocab=0.7829 token=1.2680 | 6.14 steps/s | VRAM=29.04 Go [2026-07-18 05:50:51] step 660950/1000000 | loss=2.2604 vocab=0.4842 token=0.8312 | 6.10 steps/s | VRAM=27.51 Go [2026-07-18 05:50:59] step 661000/1000000 | loss=2.2111 vocab=0.6017 token=1.4626 | 6.16 steps/s | VRAM=31.52 Go [2026-07-18 05:51:08] step 661050/1000000 | loss=2.3026 vocab=1.2352 token=1.1161 | 6.12 steps/s | VRAM=24.18 Go [2026-07-18 05:51:16] step 661100/1000000 | loss=2.1686 vocab=0.6964 token=1.3809 | 6.17 steps/s | VRAM=28.55 Go [2026-07-18 05:51:24] step 661150/1000000 | loss=2.2204 vocab=1.0639 token=1.5035 | 6.15 steps/s | VRAM=29.10 Go [2026-07-18 05:51:32] step 661200/1000000 | loss=2.1265 vocab=1.0219 token=1.5170 | 6.16 steps/s | VRAM=32.07 Go [2026-07-18 05:51:40] step 661250/1000000 | loss=2.2517 vocab=1.4849 token=1.1784 | 6.11 steps/s | VRAM=28.21 Go [2026-07-18 05:51:48] step 661300/1000000 | loss=2.2836 vocab=1.2081 token=1.5305 | 6.13 steps/s | VRAM=30.24 Go [2026-07-18 05:51:56] step 661350/1000000 | loss=2.1642 vocab=0.4933 token=1.0559 | 6.16 steps/s | VRAM=28.42 Go [2026-07-18 05:52:05] step 661400/1000000 | loss=2.1741 vocab=0.5915 token=1.0575 | 6.17 steps/s | VRAM=31.41 Go [2026-07-18 05:52:13] step 661450/1000000 | loss=2.3297 vocab=0.7428 token=1.3127 | 6.15 steps/s | VRAM=25.99 Go [2026-07-18 05:52:21] step 661500/1000000 | loss=2.1075 vocab=0.9888 token=1.1395 | 6.13 steps/s | VRAM=28.09 Go [2026-07-18 05:52:29] step 661550/1000000 | loss=2.1992 vocab=1.1656 token=1.4469 | 6.17 steps/s | VRAM=28.06 Go [2026-07-18 05:52:37] step 661600/1000000 | loss=2.2357 vocab=0.3844 token=0.7358 | 6.14 steps/s | VRAM=26.85 Go [2026-07-18 05:52:45] step 661650/1000000 | loss=2.2673 vocab=0.6244 token=1.3285 | 6.17 steps/s | VRAM=31.43 Go [2026-07-18 05:52:53] step 661700/1000000 | loss=2.2498 vocab=1.5378 token=1.6218 | 6.15 steps/s | VRAM=28.70 Go [2026-07-18 05:53:01] step 661750/1000000 | loss=2.2336 vocab=0.8751 token=1.2404 | 6.15 steps/s | VRAM=31.77 Go [2026-07-18 05:53:10] step 661800/1000000 | loss=2.1909 vocab=0.8350 token=1.2570 | 6.13 steps/s | VRAM=29.25 Go [2026-07-18 05:53:18] step 661850/1000000 | loss=2.2171 vocab=1.2765 token=1.4811 | 6.13 steps/s | VRAM=29.22 Go [2026-07-18 05:53:26] step 661900/1000000 | loss=2.0642 vocab=0.3640 token=0.9087 | 6.18 steps/s | VRAM=30.53 Go [2026-07-18 05:53:34] step 661950/1000000 | loss=2.2641 vocab=1.1386 token=1.3544 | 6.18 steps/s | VRAM=26.18 Go [2026-07-18 05:53:42] step 662000/1000000 | loss=2.2148 vocab=1.0478 token=1.2062 | 6.16 steps/s | VRAM=28.17 Go [2026-07-18 05:53:50] step 662050/1000000 | loss=2.3137 vocab=0.9014 token=0.8834 | 6.15 steps/s | VRAM=26.81 Go [2026-07-18 05:53:58] step 662100/1000000 | loss=2.3662 vocab=0.6498 token=1.5409 | 6.16 steps/s | VRAM=29.19 Go [2026-07-18 05:54:06] step 662150/1000000 | loss=2.2467 vocab=1.3309 token=1.0425 | 6.15 steps/s | VRAM=28.09 Go [2026-07-18 05:54:15] step 662200/1000000 | loss=2.2143 vocab=0.6625 token=1.1807 | 6.14 steps/s | VRAM=28.87 Go [2026-07-18 05:54:23] step 662250/1000000 | loss=2.2161 vocab=0.8019 token=1.0501 | 6.16 steps/s | VRAM=28.64 Go [2026-07-18 05:54:31] step 662300/1000000 | loss=2.2073 vocab=0.7368 token=1.4041 | 6.13 steps/s | VRAM=29.83 Go [2026-07-18 05:54:39] step 662350/1000000 | loss=2.1858 vocab=1.0250 token=1.0244 | 6.14 steps/s | VRAM=28.01 Go [2026-07-18 05:54:47] step 662400/1000000 | loss=2.0608 vocab=0.3587 token=1.1589 | 6.14 steps/s | VRAM=31.98 Go [2026-07-18 05:54:55] step 662450/1000000 | loss=2.1096 vocab=1.0620 token=1.3058 | 6.14 steps/s | VRAM=23.51 Go [2026-07-18 05:55:03] step 662500/1000000 | loss=2.2009 vocab=1.1541 token=1.8047 | 6.16 steps/s | VRAM=28.77 Go [2026-07-18 05:55:11] step 662550/1000000 | loss=2.1917 vocab=0.4025 token=1.2735 | 6.16 steps/s | VRAM=28.32 Go [2026-07-18 05:55:20] step 662600/1000000 | loss=2.2238 vocab=0.8207 token=1.0893 | 6.15 steps/s | VRAM=29.84 Go [2026-07-18 05:55:28] step 662650/1000000 | loss=2.2599 vocab=1.1559 token=1.5123 | 6.13 steps/s | VRAM=28.95 Go [2026-07-18 05:55:36] step 662700/1000000 | loss=2.1708 vocab=0.5896 token=1.1636 | 6.16 steps/s | VRAM=28.21 Go [2026-07-18 05:55:44] step 662750/1000000 | loss=2.2900 vocab=1.3338 token=1.4932 | 6.17 steps/s | VRAM=27.72 Go [2026-07-18 05:55:52] step 662800/1000000 | loss=2.2937 vocab=0.7503 token=1.1533 | 6.13 steps/s | VRAM=27.94 Go [2026-07-18 05:56:00] step 662850/1000000 | loss=2.2504 vocab=1.1348 token=1.1852 | 6.16 steps/s | VRAM=26.86 Go [2026-07-18 05:56:08] step 662900/1000000 | loss=2.2686 vocab=0.7407 token=0.9927 | 6.18 steps/s | VRAM=32.35 Go [2026-07-18 05:56:17] step 662950/1000000 | loss=2.3105 vocab=0.8768 token=1.4052 | 6.15 steps/s | VRAM=29.49 Go [2026-07-18 05:56:25] step 663000/1000000 | loss=2.2152 vocab=1.1516 token=1.3495 | 6.16 steps/s | VRAM=26.64 Go [2026-07-18 05:56:33] step 663050/1000000 | loss=2.1939 vocab=0.9917 token=1.5539 | 6.17 steps/s | VRAM=28.42 Go [2026-07-18 05:56:41] step 663100/1000000 | loss=2.1974 vocab=1.1875 token=1.3811 | 6.16 steps/s | VRAM=26.10 Go [2026-07-18 05:56:49] step 663150/1000000 | loss=2.2301 vocab=0.8613 token=1.3864 | 6.16 steps/s | VRAM=29.21 Go [2026-07-18 05:56:57] step 663200/1000000 | loss=2.3120 vocab=0.7248 token=1.1651 | 6.16 steps/s | VRAM=32.43 Go [2026-07-18 05:57:05] step 663250/1000000 | loss=2.2803 vocab=0.6336 token=1.5639 | 6.15 steps/s | VRAM=29.15 Go [2026-07-18 05:57:13] step 663300/1000000 | loss=2.2646 vocab=1.5508 token=1.4666 | 6.16 steps/s | VRAM=28.90 Go [2026-07-18 05:57:21] step 663350/1000000 | loss=2.2107 vocab=1.1965 token=1.5967 | 6.15 steps/s | VRAM=32.67 Go [2026-07-18 05:57:30] step 663400/1000000 | loss=2.1226 vocab=0.7033 token=1.2423 | 6.15 steps/s | VRAM=30.02 Go [2026-07-18 05:57:38] step 663450/1000000 | loss=2.2640 vocab=0.6025 token=1.5289 | 6.14 steps/s | VRAM=28.62 Go [2026-07-18 05:57:46] step 663500/1000000 | loss=2.3025 vocab=1.0656 token=1.4328 | 6.15 steps/s | VRAM=28.99 Go [2026-07-18 05:57:54] step 663550/1000000 | loss=2.2062 vocab=0.7934 token=1.1320 | 6.14 steps/s | VRAM=32.65 Go [2026-07-18 05:58:02] step 663600/1000000 | loss=2.2339 vocab=0.7830 token=1.7692 | 6.16 steps/s | VRAM=28.57 Go [2026-07-18 05:58:10] step 663650/1000000 | loss=2.1685 vocab=1.2015 token=1.8293 | 6.15 steps/s | VRAM=29.33 Go [2026-07-18 05:58:18] step 663700/1000000 | loss=2.2030 vocab=0.8494 token=1.0628 | 6.14 steps/s | VRAM=30.62 Go [2026-07-18 05:58:27] step 663750/1000000 | loss=2.1719 vocab=0.7936 token=1.1498 | 6.15 steps/s | VRAM=25.71 Go [2026-07-18 05:58:35] step 663800/1000000 | loss=2.2097 vocab=0.6842 token=0.9648 | 6.17 steps/s | VRAM=24.46 Go [2026-07-18 05:58:43] step 663850/1000000 | loss=2.2014 vocab=0.6143 token=0.8326 | 6.13 steps/s | VRAM=30.06 Go [2026-07-18 05:58:51] step 663900/1000000 | loss=2.1932 vocab=0.8275 token=1.2866 | 6.14 steps/s | VRAM=32.49 Go [2026-07-18 05:58:59] step 663950/1000000 | loss=2.1913 vocab=0.7590 token=1.2659 | 6.16 steps/s | VRAM=29.60 Go [2026-07-18 05:59:07] step 664000/1000000 | loss=2.2628 vocab=0.5828 token=1.6150 | 6.16 steps/s | VRAM=26.03 Go [2026-07-18 05:59:15] step 664050/1000000 | loss=2.1383 vocab=0.8797 token=1.3116 | 6.17 steps/s | VRAM=31.24 Go [2026-07-18 05:59:23] step 664100/1000000 | loss=2.1862 vocab=0.8911 token=1.4207 | 6.16 steps/s | VRAM=29.79 Go [2026-07-18 05:59:32] step 664150/1000000 | loss=2.2504 vocab=0.5395 token=1.3082 | 6.13 steps/s | VRAM=28.39 Go [2026-07-18 05:59:40] step 664200/1000000 | loss=2.2795 vocab=0.9345 token=1.4115 | 6.13 steps/s | VRAM=27.06 Go [2026-07-18 05:59:48] step 664250/1000000 | loss=2.2417 vocab=0.7923 token=1.3948 | 6.18 steps/s | VRAM=28.17 Go [2026-07-18 05:59:56] step 664300/1000000 | loss=2.1672 vocab=1.2659 token=1.0682 | 6.16 steps/s | VRAM=25.55 Go [2026-07-18 06:00:04] step 664350/1000000 | loss=2.2231 vocab=0.7451 token=0.9336 | 6.18 steps/s | VRAM=31.36 Go [2026-07-18 06:00:12] step 664400/1000000 | loss=2.1503 vocab=1.2871 token=1.3090 | 6.14 steps/s | VRAM=32.96 Go [2026-07-18 06:00:20] step 664450/1000000 | loss=2.2875 vocab=1.3562 token=1.6775 | 6.14 steps/s | VRAM=24.65 Go [2026-07-18 06:00:28] step 664500/1000000 | loss=2.2103 vocab=0.8360 token=1.1197 | 6.16 steps/s | VRAM=32.14 Go [2026-07-18 06:00:36] step 664550/1000000 | loss=2.2801 vocab=0.7887 token=1.2269 | 6.15 steps/s | VRAM=29.45 Go [2026-07-18 06:00:45] step 664600/1000000 | loss=2.1986 vocab=0.7188 token=1.6500 | 6.17 steps/s | VRAM=27.36 Go [2026-07-18 06:00:53] step 664650/1000000 | loss=2.1405 vocab=0.9567 token=1.5997 | 6.17 steps/s | VRAM=27.65 Go [2026-07-18 06:01:01] step 664700/1000000 | loss=2.2951 vocab=1.1483 token=1.4889 | 6.11 steps/s | VRAM=24.10 Go [2026-07-18 06:01:09] step 664750/1000000 | loss=2.0976 vocab=0.8041 token=1.2750 | 6.16 steps/s | VRAM=24.38 Go [2026-07-18 06:01:17] step 664800/1000000 | loss=2.1948 vocab=0.7386 token=1.5583 | 6.17 steps/s | VRAM=29.23 Go [2026-07-18 06:01:25] step 664850/1000000 | loss=2.2306 vocab=1.1311 token=1.2319 | 6.13 steps/s | VRAM=29.28 Go [2026-07-18 06:01:33] step 664900/1000000 | loss=2.2804 vocab=1.3849 token=1.4996 | 6.16 steps/s | VRAM=27.76 Go [2026-07-18 06:01:42] step 664950/1000000 | loss=2.2115 vocab=0.8835 token=1.3756 | 6.12 steps/s | VRAM=30.30 Go [2026-07-18 06:01:50] step 665000/1000000 | loss=2.2781 vocab=1.0055 token=1.2530 | 6.15 steps/s | VRAM=28.86 Go [2026-07-18 06:01:58] step 665050/1000000 | loss=2.1876 vocab=1.3384 token=1.3436 | 6.12 steps/s | VRAM=26.49 Go [2026-07-18 06:02:06] step 665100/1000000 | loss=2.1924 vocab=1.2303 token=1.1597 | 6.12 steps/s | VRAM=29.69 Go [2026-07-18 06:02:14] step 665150/1000000 | loss=2.1835 vocab=0.9321 token=1.3026 | 6.15 steps/s | VRAM=24.73 Go [2026-07-18 06:02:22] step 665200/1000000 | loss=2.1456 vocab=0.7043 token=1.0160 | 6.17 steps/s | VRAM=31.76 Go [2026-07-18 06:02:30] step 665250/1000000 | loss=2.2352 vocab=0.8959 token=1.1300 | 6.11 steps/s | VRAM=26.24 Go [2026-07-18 06:02:39] step 665300/1000000 | loss=2.3031 vocab=0.6220 token=1.1358 | 6.13 steps/s | VRAM=28.21 Go [2026-07-18 06:02:47] step 665350/1000000 | loss=2.1531 vocab=0.7700 token=1.0156 | 6.14 steps/s | VRAM=27.05 Go [2026-07-18 06:02:55] step 665400/1000000 | loss=2.1986 vocab=0.6469 token=1.0933 | 6.15 steps/s | VRAM=32.84 Go [2026-07-18 06:03:03] step 665450/1000000 | loss=2.1180 vocab=0.8304 token=1.4069 | 6.17 steps/s | VRAM=25.66 Go [2026-07-18 06:03:11] step 665500/1000000 | loss=2.1753 vocab=0.8373 token=1.3764 | 6.13 steps/s | VRAM=28.72 Go [2026-07-18 06:03:19] step 665550/1000000 | loss=2.2297 vocab=0.9716 token=1.5222 | 6.14 steps/s | VRAM=28.89 Go [2026-07-18 06:03:27] step 665600/1000000 | loss=2.1380 vocab=0.7724 token=1.0506 | 6.14 steps/s | VRAM=28.92 Go [2026-07-18 06:03:36] step 665650/1000000 | loss=2.2989 vocab=1.0537 token=1.5455 | 6.13 steps/s | VRAM=29.47 Go [2026-07-18 06:03:44] step 665700/1000000 | loss=2.1983 vocab=0.9491 token=1.0309 | 6.16 steps/s | VRAM=24.48 Go [2026-07-18 06:03:52] step 665750/1000000 | loss=2.1626 vocab=0.6820 token=1.0626 | 6.15 steps/s | VRAM=25.16 Go [2026-07-18 06:04:00] step 665800/1000000 | loss=2.1998 vocab=1.1502 token=0.7975 | 6.16 steps/s | VRAM=32.64 Go [2026-07-18 06:04:08] step 665850/1000000 | loss=2.3170 vocab=1.4442 token=1.7944 | 6.16 steps/s | VRAM=28.68 Go [2026-07-18 06:04:16] step 665900/1000000 | loss=2.2057 vocab=1.0940 token=1.4658 | 6.17 steps/s | VRAM=31.98 Go [2026-07-18 06:04:24] step 665950/1000000 | loss=2.2110 vocab=0.5955 token=1.2704 | 6.15 steps/s | VRAM=27.28 Go [2026-07-18 06:04:32] step 666000/1000000 | loss=2.2412 vocab=0.8145 token=0.9150 | 6.17 steps/s | VRAM=31.59 Go [2026-07-18 06:04:41] step 666050/1000000 | loss=2.3005 vocab=1.0916 token=1.3528 | 6.17 steps/s | VRAM=32.09 Go [2026-07-18 06:04:49] step 666100/1000000 | loss=2.2086 vocab=0.9071 token=1.1737 | 6.16 steps/s | VRAM=27.99 Go [2026-07-18 06:04:57] step 666150/1000000 | loss=2.1939 vocab=0.7756 token=1.3401 | 6.11 steps/s | VRAM=32.75 Go [2026-07-18 06:05:05] step 666200/1000000 | loss=2.2107 vocab=0.4156 token=1.1834 | 6.14 steps/s | VRAM=29.52 Go [2026-07-18 06:05:13] step 666250/1000000 | loss=2.2238 vocab=0.5690 token=1.1855 | 6.14 steps/s | VRAM=26.29 Go [2026-07-18 06:05:21] step 666300/1000000 | loss=2.2994 vocab=1.0357 token=1.1909 | 6.13 steps/s | VRAM=28.50 Go [2026-07-18 06:05:29] step 666350/1000000 | loss=2.1646 vocab=0.8210 token=1.3787 | 6.12 steps/s | VRAM=28.75 Go [2026-07-18 06:05:38] step 666400/1000000 | loss=2.3162 vocab=1.3913 token=1.4401 | 6.14 steps/s | VRAM=27.84 Go [2026-07-18 06:05:46] step 666450/1000000 | loss=2.3650 vocab=1.0018 token=1.6042 | 6.13 steps/s | VRAM=28.57 Go [2026-07-18 06:05:54] step 666500/1000000 | loss=2.1628 vocab=0.9983 token=1.2181 | 6.11 steps/s | VRAM=29.58 Go [2026-07-18 06:06:02] step 666550/1000000 | loss=2.1878 vocab=1.0711 token=1.4977 | 6.15 steps/s | VRAM=31.26 Go [2026-07-18 06:06:10] step 666600/1000000 | loss=2.2024 vocab=0.8811 token=1.5071 | 6.12 steps/s | VRAM=32.96 Go [2026-07-18 06:06:18] step 666650/1000000 | loss=2.1732 vocab=0.7533 token=1.5606 | 6.14 steps/s | VRAM=28.95 Go [2026-07-18 06:06:27] step 666700/1000000 | loss=2.2804 vocab=0.9058 token=1.0831 | 6.14 steps/s | VRAM=29.19 Go [2026-07-18 06:06:35] step 666750/1000000 | loss=2.2540 vocab=1.3565 token=1.5604 | 6.16 steps/s | VRAM=32.32 Go [2026-07-18 06:06:43] step 666800/1000000 | loss=2.2845 vocab=0.9561 token=1.5269 | 6.13 steps/s | VRAM=29.17 Go [2026-07-18 06:06:51] step 666850/1000000 | loss=2.0999 vocab=1.0317 token=1.1512 | 6.14 steps/s | VRAM=32.36 Go [2026-07-18 06:06:59] step 666900/1000000 | loss=2.1701 vocab=0.8524 token=1.0728 | 6.12 steps/s | VRAM=26.80 Go [2026-07-18 06:07:07] step 666950/1000000 | loss=2.0998 vocab=0.6373 token=1.1448 | 6.12 steps/s | VRAM=29.04 Go [2026-07-18 06:07:15] step 667000/1000000 | loss=2.3296 vocab=0.6411 token=0.9948 | 6.15 steps/s | VRAM=27.33 Go [2026-07-18 06:07:24] step 667050/1000000 | loss=2.1256 vocab=0.4452 token=0.9756 | 6.13 steps/s | VRAM=28.66 Go [2026-07-18 06:07:32] step 667100/1000000 | loss=2.3476 vocab=1.2204 token=1.6114 | 6.10 steps/s | VRAM=32.65 Go [2026-07-18 06:07:40] step 667150/1000000 | loss=2.2535 vocab=0.6464 token=1.1148 | 6.13 steps/s | VRAM=25.96 Go [2026-07-18 06:07:48] step 667200/1000000 | loss=2.1428 vocab=0.7113 token=1.2718 | 6.13 steps/s | VRAM=28.12 Go [2026-07-18 06:07:56] step 667250/1000000 | loss=2.4004 vocab=1.7444 token=1.4749 | 6.15 steps/s | VRAM=29.14 Go [2026-07-18 06:08:04] step 667300/1000000 | loss=2.2793 vocab=1.5900 token=1.5992 | 6.14 steps/s | VRAM=31.83 Go [2026-07-18 06:08:12] step 667350/1000000 | loss=2.1184 vocab=0.6178 token=1.1280 | 6.15 steps/s | VRAM=31.54 Go [2026-07-18 06:08:21] step 667400/1000000 | loss=2.2237 vocab=1.1108 token=1.7718 | 6.14 steps/s | VRAM=28.98 Go [2026-07-18 06:08:29] step 667450/1000000 | loss=2.2396 vocab=0.6469 token=1.0785 | 6.15 steps/s | VRAM=28.71 Go [2026-07-18 06:08:37] step 667500/1000000 | loss=2.1706 vocab=0.5400 token=1.3495 | 6.14 steps/s | VRAM=29.49 Go [2026-07-18 06:08:45] step 667550/1000000 | loss=2.1549 vocab=0.5779 token=0.9073 | 6.14 steps/s | VRAM=32.17 Go [2026-07-18 06:08:53] step 667600/1000000 | loss=2.1771 vocab=1.4494 token=0.9822 | 6.10 steps/s | VRAM=27.92 Go [2026-07-18 06:09:01] step 667650/1000000 | loss=2.2142 vocab=1.0998 token=1.3980 | 6.16 steps/s | VRAM=31.79 Go [2026-07-18 06:09:10] step 667700/1000000 | loss=2.1792 vocab=0.4600 token=0.8910 | 6.13 steps/s | VRAM=28.65 Go [2026-07-18 06:09:18] step 667750/1000000 | loss=2.2735 vocab=1.2824 token=1.4094 | 6.14 steps/s | VRAM=29.67 Go [2026-07-18 06:09:26] step 667800/1000000 | loss=2.1693 vocab=1.6485 token=1.4667 | 6.16 steps/s | VRAM=24.86 Go [2026-07-18 06:09:34] step 667850/1000000 | loss=2.1837 vocab=0.9204 token=1.2070 | 6.14 steps/s | VRAM=29.68 Go [2026-07-18 06:09:42] step 667900/1000000 | loss=2.1855 vocab=0.7653 token=1.0166 | 6.12 steps/s | VRAM=28.87 Go [2026-07-18 06:09:50] step 667950/1000000 | loss=2.1478 vocab=0.5924 token=0.6831 | 6.15 steps/s | VRAM=29.15 Go [2026-07-18 06:09:58] step 668000/1000000 | loss=2.2747 vocab=1.2050 token=1.3512 | 6.12 steps/s | VRAM=29.37 Go [2026-07-18 06:10:07] step 668050/1000000 | loss=2.1629 vocab=0.9131 token=1.1604 | 6.10 steps/s | VRAM=29.31 Go [2026-07-18 06:10:15] step 668100/1000000 | loss=2.1661 vocab=1.2415 token=1.4323 | 6.14 steps/s | VRAM=31.36 Go [2026-07-18 06:10:23] step 668150/1000000 | loss=2.1821 vocab=1.3099 token=1.1525 | 6.13 steps/s | VRAM=25.64 Go [2026-07-18 06:10:31] step 668200/1000000 | loss=2.3223 vocab=0.9742 token=1.0805 | 6.16 steps/s | VRAM=32.90 Go [2026-07-18 06:10:39] step 668250/1000000 | loss=2.2711 vocab=1.3433 token=1.5425 | 6.10 steps/s | VRAM=28.78 Go [2026-07-18 06:10:47] step 668300/1000000 | loss=2.1776 vocab=1.3117 token=1.4439 | 6.15 steps/s | VRAM=31.99 Go [2026-07-18 06:10:56] step 668350/1000000 | loss=2.1870 vocab=0.9252 token=1.2863 | 6.11 steps/s | VRAM=27.15 Go [2026-07-18 06:11:04] step 668400/1000000 | loss=2.2046 vocab=0.8856 token=1.2352 | 6.16 steps/s | VRAM=27.99 Go [2026-07-18 06:11:12] step 668450/1000000 | loss=2.1409 vocab=0.6902 token=1.2578 | 6.15 steps/s | VRAM=27.33 Go [2026-07-18 06:11:20] step 668500/1000000 | loss=2.2703 vocab=0.2505 token=0.9488 | 6.13 steps/s | VRAM=25.30 Go [2026-07-18 06:11:28] step 668550/1000000 | loss=2.3074 vocab=0.7625 token=1.4281 | 6.13 steps/s | VRAM=24.75 Go [2026-07-18 06:11:36] step 668600/1000000 | loss=2.2887 vocab=1.0454 token=1.5390 | 6.17 steps/s | VRAM=27.59 Go [2026-07-18 06:11:44] step 668650/1000000 | loss=2.2572 vocab=1.3424 token=1.3332 | 6.10 steps/s | VRAM=26.83 Go [2026-07-18 06:11:53] step 668700/1000000 | loss=2.2272 vocab=0.8032 token=1.0627 | 6.14 steps/s | VRAM=31.53 Go [2026-07-18 06:12:01] step 668750/1000000 | loss=2.3076 vocab=1.1022 token=1.5816 | 6.15 steps/s | VRAM=25.73 Go [2026-07-18 06:12:09] step 668800/1000000 | loss=2.2034 vocab=0.9621 token=1.0540 | 6.15 steps/s | VRAM=26.00 Go [2026-07-18 06:12:17] step 668850/1000000 | loss=2.1881 vocab=0.5100 token=0.8438 | 6.14 steps/s | VRAM=25.49 Go [2026-07-18 06:12:25] step 668900/1000000 | loss=2.1283 vocab=0.8424 token=1.0913 | 6.16 steps/s | VRAM=32.28 Go [2026-07-18 06:12:33] step 668950/1000000 | loss=2.1628 vocab=0.7329 token=1.4552 | 6.17 steps/s | VRAM=25.14 Go [2026-07-18 06:12:41] step 669000/1000000 | loss=2.1453 vocab=0.8155 token=1.8766 | 6.16 steps/s | VRAM=28.87 Go [2026-07-18 06:12:49] step 669050/1000000 | loss=2.2220 vocab=1.1165 token=1.1581 | 6.15 steps/s | VRAM=32.82 Go [2026-07-18 06:12:58] step 669100/1000000 | loss=2.1364 vocab=0.8134 token=1.5347 | 6.15 steps/s | VRAM=28.82 Go [2026-07-18 06:13:06] step 669150/1000000 | loss=2.1563 vocab=0.9825 token=1.3621 | 6.15 steps/s | VRAM=31.36 Go [2026-07-18 06:13:14] step 669200/1000000 | loss=2.2565 vocab=0.7136 token=1.5158 | 6.15 steps/s | VRAM=28.28 Go [2026-07-18 06:13:22] step 669250/1000000 | loss=2.2037 vocab=0.8868 token=1.0281 | 6.11 steps/s | VRAM=29.76 Go [2026-07-18 06:13:30] step 669300/1000000 | loss=2.1746 vocab=0.8709 token=1.0725 | 6.15 steps/s | VRAM=32.05 Go [2026-07-18 06:13:38] step 669350/1000000 | loss=2.1358 vocab=0.9701 token=1.3481 | 6.15 steps/s | VRAM=29.07 Go [2026-07-18 06:13:46] step 669400/1000000 | loss=2.2854 vocab=1.1880 token=1.4616 | 6.13 steps/s | VRAM=26.33 Go [2026-07-18 06:13:55] step 669450/1000000 | loss=2.2074 vocab=0.6847 token=1.0469 | 6.13 steps/s | VRAM=28.75 Go [2026-07-18 06:14:03] step 669500/1000000 | loss=2.4116 vocab=1.0720 token=1.7419 | 6.14 steps/s | VRAM=31.65 Go [2026-07-18 06:14:11] step 669550/1000000 | loss=2.2657 vocab=0.7598 token=1.4363 | 6.09 steps/s | VRAM=28.81 Go [2026-07-18 06:14:19] step 669600/1000000 | loss=2.2379 vocab=1.1158 token=1.4612 | 6.10 steps/s | VRAM=29.44 Go [2026-07-18 06:14:27] step 669650/1000000 | loss=2.1837 vocab=1.0525 token=1.1759 | 6.13 steps/s | VRAM=30.65 Go [2026-07-18 06:14:35] step 669700/1000000 | loss=2.2270 vocab=0.9837 token=0.9367 | 6.13 steps/s | VRAM=28.60 Go [2026-07-18 06:14:44] step 669750/1000000 | loss=2.2829 vocab=0.9489 token=1.7240 | 6.15 steps/s | VRAM=29.07 Go [2026-07-18 06:14:52] step 669800/1000000 | loss=2.1632 vocab=1.0144 token=1.5514 | 6.17 steps/s | VRAM=27.53 Go [2026-07-18 06:15:00] step 669850/1000000 | loss=2.2984 vocab=1.0029 token=1.6857 | 6.16 steps/s | VRAM=28.95 Go [2026-07-18 06:15:08] step 669900/1000000 | loss=2.1782 vocab=1.0004 token=1.5285 | 6.15 steps/s | VRAM=29.24 Go [2026-07-18 06:15:16] step 669950/1000000 | loss=2.1410 vocab=0.7514 token=1.0602 | 6.18 steps/s | VRAM=32.60 Go [2026-07-18 06:15:24] step 670000/1000000 | loss=2.1761 vocab=1.1156 token=1.8696 | 6.12 steps/s | VRAM=27.72 Go [2026-07-18 06:15:32] step 670050/1000000 | loss=2.3456 vocab=0.9841 token=1.6727 | 6.18 steps/s | VRAM=28.39 Go [2026-07-18 06:15:40] step 670100/1000000 | loss=2.2613 vocab=0.7510 token=1.3940 | 6.13 steps/s | VRAM=32.99 Go [2026-07-18 06:15:49] step 670150/1000000 | loss=2.1770 vocab=0.9333 token=0.8883 | 6.16 steps/s | VRAM=24.15 Go [2026-07-18 06:15:57] step 670200/1000000 | loss=2.2450 vocab=1.3924 token=1.7446 | 6.16 steps/s | VRAM=27.98 Go [2026-07-18 06:16:05] step 670250/1000000 | loss=2.1893 vocab=0.7214 token=0.9964 | 6.14 steps/s | VRAM=29.10 Go [2026-07-18 06:16:13] step 670300/1000000 | loss=2.2891 vocab=0.5848 token=1.1244 | 6.17 steps/s | VRAM=29.74 Go [2026-07-18 06:16:21] step 670350/1000000 | loss=2.1928 vocab=0.5413 token=1.0490 | 6.17 steps/s | VRAM=24.78 Go [2026-07-18 06:16:29] step 670400/1000000 | loss=2.1786 vocab=1.2231 token=1.5649 | 6.13 steps/s | VRAM=29.13 Go [2026-07-18 06:16:37] step 670450/1000000 | loss=2.1411 vocab=0.8495 token=1.3760 | 6.15 steps/s | VRAM=25.20 Go [2026-07-18 06:16:45] step 670500/1000000 | loss=2.2541 vocab=1.4582 token=1.2403 | 6.17 steps/s | VRAM=28.98 Go [2026-07-18 06:16:53] step 670550/1000000 | loss=2.1096 vocab=1.2768 token=1.2483 | 6.17 steps/s | VRAM=28.20 Go [2026-07-18 06:17:02] step 670600/1000000 | loss=2.1601 vocab=1.1812 token=1.5269 | 6.16 steps/s | VRAM=25.88 Go [2026-07-18 06:17:10] step 670650/1000000 | loss=2.1918 vocab=0.6457 token=1.2983 | 6.14 steps/s | VRAM=27.98 Go [2026-07-18 06:17:18] step 670700/1000000 | loss=2.1571 vocab=0.7790 token=1.4970 | 6.15 steps/s | VRAM=29.15 Go [2026-07-18 06:17:26] step 670750/1000000 | loss=2.3161 vocab=0.5243 token=1.6701 | 6.15 steps/s | VRAM=28.90 Go [2026-07-18 06:17:34] step 670800/1000000 | loss=2.2393 vocab=0.5727 token=1.1047 | 6.16 steps/s | VRAM=28.84 Go [2026-07-18 06:17:42] step 670850/1000000 | loss=2.1299 vocab=0.9344 token=1.3962 | 6.13 steps/s | VRAM=26.52 Go [2026-07-18 06:17:50] step 670900/1000000 | loss=2.2018 vocab=1.2871 token=1.4487 | 6.18 steps/s | VRAM=32.12 Go [2026-07-18 06:17:59] step 670950/1000000 | loss=2.1865 vocab=1.0612 token=1.8290 | 6.16 steps/s | VRAM=31.45 Go [2026-07-18 06:18:07] step 671000/1000000 | loss=2.2789 vocab=1.5069 token=1.6774 | 6.15 steps/s | VRAM=30.18 Go [2026-07-18 06:18:15] step 671050/1000000 | loss=2.2371 vocab=1.1313 token=1.6314 | 6.14 steps/s | VRAM=27.32 Go [2026-07-18 06:18:23] step 671100/1000000 | loss=2.1685 vocab=1.1583 token=1.7115 | 6.16 steps/s | VRAM=24.84 Go [2026-07-18 06:18:31] step 671150/1000000 | loss=2.2753 vocab=1.4293 token=1.6811 | 6.14 steps/s | VRAM=28.44 Go [2026-07-18 06:18:39] step 671200/1000000 | loss=2.2394 vocab=0.7601 token=1.3566 | 6.13 steps/s | VRAM=31.06 Go [2026-07-18 06:18:47] step 671250/1000000 | loss=2.3020 vocab=0.7291 token=1.1997 | 6.10 steps/s | VRAM=27.33 Go [2026-07-18 06:18:56] step 671300/1000000 | loss=2.1775 vocab=0.7998 token=1.2122 | 6.15 steps/s | VRAM=28.15 Go [2026-07-18 06:19:04] step 671350/1000000 | loss=2.0851 vocab=0.9477 token=1.5207 | 6.13 steps/s | VRAM=28.85 Go [2026-07-18 06:19:12] step 671400/1000000 | loss=2.3512 vocab=1.7271 token=1.7512 | 6.13 steps/s | VRAM=29.63 Go [2026-07-18 06:19:20] step 671450/1000000 | loss=2.2123 vocab=1.1565 token=1.2953 | 6.17 steps/s | VRAM=25.34 Go [2026-07-18 06:19:28] step 671500/1000000 | loss=2.2401 vocab=0.8516 token=1.6220 | 6.15 steps/s | VRAM=32.62 Go [2026-07-18 06:19:36] step 671550/1000000 | loss=2.1738 vocab=1.0077 token=1.8812 | 6.16 steps/s | VRAM=26.00 Go [2026-07-18 06:19:44] step 671600/1000000 | loss=2.2904 vocab=1.1655 token=1.5041 | 6.14 steps/s | VRAM=30.45 Go [2026-07-18 06:19:52] step 671650/1000000 | loss=2.1471 vocab=1.1224 token=1.6378 | 6.15 steps/s | VRAM=28.45 Go [2026-07-18 06:20:01] step 671700/1000000 | loss=2.1244 vocab=0.7483 token=1.6009 | 6.12 steps/s | VRAM=29.12 Go [2026-07-18 06:20:09] step 671750/1000000 | loss=2.2201 vocab=1.0123 token=1.4154 | 6.12 steps/s | VRAM=25.24 Go [2026-07-18 06:20:17] step 671800/1000000 | loss=2.2133 vocab=1.0209 token=1.3260 | 6.15 steps/s | VRAM=28.04 Go [2026-07-18 06:20:25] step 671850/1000000 | loss=2.2161 vocab=0.6328 token=1.0670 | 6.13 steps/s | VRAM=28.59 Go [2026-07-18 06:20:33] step 671900/1000000 | loss=2.2417 vocab=0.8393 token=1.3267 | 6.14 steps/s | VRAM=29.00 Go [2026-07-18 06:20:41] step 671950/1000000 | loss=2.1984 vocab=1.4503 token=1.6073 | 6.12 steps/s | VRAM=32.70 Go [2026-07-18 06:20:50] step 672000/1000000 | loss=2.2055 vocab=0.8685 token=1.1033 | 6.13 steps/s | VRAM=28.48 Go [2026-07-18 06:20:58] step 672050/1000000 | loss=2.2162 vocab=0.6051 token=1.1513 | 6.16 steps/s | VRAM=29.52 Go [2026-07-18 06:21:06] step 672100/1000000 | loss=2.1791 vocab=0.7132 token=1.0921 | 6.14 steps/s | VRAM=29.40 Go [2026-07-18 06:21:14] step 672150/1000000 | loss=2.2576 vocab=0.7479 token=1.3444 | 6.15 steps/s | VRAM=25.24 Go [2026-07-18 06:21:22] step 672200/1000000 | loss=2.2515 vocab=0.9037 token=1.5660 | 6.13 steps/s | VRAM=29.02 Go [2026-07-18 06:21:30] step 672250/1000000 | loss=2.2125 vocab=1.0164 token=1.6854 | 6.15 steps/s | VRAM=32.84 Go [2026-07-18 06:21:38] step 672300/1000000 | loss=2.1948 vocab=0.6524 token=1.2966 | 6.12 steps/s | VRAM=29.61 Go [2026-07-18 06:21:47] step 672350/1000000 | loss=2.2110 vocab=0.8472 token=1.0409 | 6.15 steps/s | VRAM=29.67 Go [2026-07-18 06:21:55] step 672400/1000000 | loss=2.2882 vocab=1.2669 token=1.8144 | 6.17 steps/s | VRAM=31.17 Go [2026-07-18 06:22:03] step 672450/1000000 | loss=2.3302 vocab=1.2226 token=1.2219 | 6.11 steps/s | VRAM=30.22 Go [2026-07-18 06:22:11] step 672500/1000000 | loss=2.1258 vocab=0.6206 token=1.3286 | 6.13 steps/s | VRAM=28.87 Go [2026-07-18 06:22:19] step 672550/1000000 | loss=2.2536 vocab=0.5902 token=1.3674 | 6.18 steps/s | VRAM=26.24 Go [2026-07-18 06:22:27] step 672600/1000000 | loss=2.1740 vocab=0.7870 token=1.2983 | 6.18 steps/s | VRAM=32.32 Go [2026-07-18 06:22:35] step 672650/1000000 | loss=2.1913 vocab=0.6468 token=1.2430 | 6.17 steps/s | VRAM=29.27 Go [2026-07-18 06:22:43] step 672700/1000000 | loss=2.1955 vocab=0.6926 token=1.0280 | 6.17 steps/s | VRAM=31.10 Go [2026-07-18 06:22:51] step 672750/1000000 | loss=2.2185 vocab=1.2574 token=1.4272 | 6.18 steps/s | VRAM=24.84 Go [2026-07-18 06:23:00] step 672800/1000000 | loss=2.2176 vocab=0.4977 token=1.3007 | 6.12 steps/s | VRAM=29.90 Go [2026-07-18 06:23:08] step 672850/1000000 | loss=2.2332 vocab=0.8621 token=1.2754 | 6.17 steps/s | VRAM=31.66 Go [2026-07-18 06:23:16] step 672900/1000000 | loss=2.2917 vocab=0.6745 token=1.1135 | 6.15 steps/s | VRAM=26.26 Go [2026-07-18 06:23:24] step 672950/1000000 | loss=2.1587 vocab=0.5334 token=1.3656 | 6.16 steps/s | VRAM=29.21 Go [2026-07-18 06:23:32] step 673000/1000000 | loss=2.1750 vocab=0.8061 token=1.4265 | 6.16 steps/s | VRAM=28.85 Go [2026-07-18 06:23:40] step 673050/1000000 | loss=2.2374 vocab=1.0398 token=1.5020 | 6.14 steps/s | VRAM=26.23 Go [2026-07-18 06:23:48] step 673100/1000000 | loss=2.1578 vocab=0.7875 token=1.1035 | 6.16 steps/s | VRAM=26.89 Go [2026-07-18 06:23:56] step 673150/1000000 | loss=2.1494 vocab=1.0435 token=1.4952 | 6.13 steps/s | VRAM=29.19 Go [2026-07-18 06:24:05] step 673200/1000000 | loss=2.2300 vocab=0.8714 token=0.7371 | 6.17 steps/s | VRAM=32.66 Go [2026-07-18 06:24:13] step 673250/1000000 | loss=2.1663 vocab=1.4105 token=1.3107 | 6.16 steps/s | VRAM=30.08 Go [2026-07-18 06:24:21] step 673300/1000000 | loss=2.1676 vocab=0.9668 token=1.4195 | 6.17 steps/s | VRAM=25.29 Go [2026-07-18 06:24:29] step 673350/1000000 | loss=2.1638 vocab=1.0735 token=1.5693 | 6.17 steps/s | VRAM=32.09 Go [2026-07-18 06:24:37] step 673400/1000000 | loss=2.2080 vocab=0.6907 token=1.2954 | 6.17 steps/s | VRAM=29.54 Go [2026-07-18 06:24:45] step 673450/1000000 | loss=2.2462 vocab=1.2784 token=1.6817 | 6.19 steps/s | VRAM=31.46 Go [2026-07-18 06:24:53] step 673500/1000000 | loss=2.1689 vocab=1.8486 token=1.8351 | 6.14 steps/s | VRAM=27.95 Go [2026-07-18 06:25:01] step 673550/1000000 | loss=2.2500 vocab=1.0370 token=1.5362 | 6.16 steps/s | VRAM=25.84 Go [2026-07-18 06:25:10] step 673600/1000000 | loss=2.2393 vocab=1.0968 token=1.2249 | 6.14 steps/s | VRAM=29.88 Go [2026-07-18 06:25:18] step 673650/1000000 | loss=2.1905 vocab=0.6769 token=1.2371 | 6.16 steps/s | VRAM=27.76 Go [2026-07-18 06:25:26] step 673700/1000000 | loss=2.2322 vocab=1.2149 token=1.5835 | 6.16 steps/s | VRAM=24.88 Go [2026-07-18 06:25:34] step 673750/1000000 | loss=2.3112 vocab=1.4252 token=2.0447 | 6.13 steps/s | VRAM=28.41 Go [2026-07-18 06:25:42] step 673800/1000000 | loss=2.1413 vocab=0.3477 token=0.9573 | 6.15 steps/s | VRAM=26.77 Go [2026-07-18 06:25:50] step 673850/1000000 | loss=2.1804 vocab=1.2929 token=1.8003 | 6.12 steps/s | VRAM=29.29 Go [2026-07-18 06:25:58] step 673900/1000000 | loss=2.3161 vocab=1.2098 token=1.4232 | 6.16 steps/s | VRAM=31.56 Go [2026-07-18 06:26:06] step 673950/1000000 | loss=2.2114 vocab=0.8013 token=1.3805 | 6.14 steps/s | VRAM=32.98 Go [2026-07-18 06:26:15] step 674000/1000000 | loss=2.2043 vocab=0.5786 token=0.9782 | 6.13 steps/s | VRAM=29.46 Go [2026-07-18 06:26:23] step 674050/1000000 | loss=2.1216 vocab=0.5789 token=0.8925 | 6.16 steps/s | VRAM=24.32 Go [2026-07-18 06:26:31] step 674100/1000000 | loss=2.0995 vocab=0.8514 token=0.8888 | 6.12 steps/s | VRAM=27.81 Go [2026-07-18 06:26:39] step 674150/1000000 | loss=2.1733 vocab=1.3415 token=1.5471 | 6.11 steps/s | VRAM=26.47 Go [2026-07-18 06:26:47] step 674200/1000000 | loss=2.2618 vocab=0.3824 token=1.2289 | 6.12 steps/s | VRAM=25.57 Go [2026-07-18 06:26:55] step 674250/1000000 | loss=2.2545 vocab=1.0583 token=1.2404 | 6.14 steps/s | VRAM=29.48 Go [2026-07-18 06:27:04] step 674300/1000000 | loss=2.1874 vocab=1.5650 token=1.3790 | 6.16 steps/s | VRAM=28.84 Go [2026-07-18 06:27:12] step 674350/1000000 | loss=2.2566 vocab=1.1234 token=1.4441 | 6.13 steps/s | VRAM=28.58 Go [2026-07-18 06:27:20] step 674400/1000000 | loss=2.1576 vocab=0.4792 token=1.1273 | 6.14 steps/s | VRAM=28.27 Go [2026-07-18 06:27:28] step 674450/1000000 | loss=2.1021 vocab=0.4062 token=1.3591 | 6.14 steps/s | VRAM=32.73 Go [2026-07-18 06:27:36] step 674500/1000000 | loss=2.1439 vocab=0.6785 token=1.2376 | 6.09 steps/s | VRAM=29.43 Go [2026-07-18 06:27:44] step 674550/1000000 | loss=2.2037 vocab=1.3665 token=1.0705 | 6.15 steps/s | VRAM=28.08 Go [2026-07-18 06:27:52] step 674600/1000000 | loss=2.2342 vocab=0.9860 token=1.1536 | 6.15 steps/s | VRAM=24.74 Go [2026-07-18 06:28:01] step 674650/1000000 | loss=2.2022 vocab=0.7454 token=1.1944 | 6.12 steps/s | VRAM=26.11 Go [2026-07-18 06:28:09] step 674700/1000000 | loss=2.1901 vocab=0.4714 token=1.0951 | 6.13 steps/s | VRAM=25.22 Go [2026-07-18 06:28:17] step 674750/1000000 | loss=2.2246 vocab=1.0027 token=1.4264 | 6.14 steps/s | VRAM=29.54 Go [2026-07-18 06:28:25] step 674800/1000000 | loss=2.2429 vocab=0.6376 token=1.1292 | 6.16 steps/s | VRAM=28.62 Go [2026-07-18 06:28:33] step 674850/1000000 | loss=2.2799 vocab=1.2588 token=1.3695 | 6.15 steps/s | VRAM=32.45 Go [2026-07-18 06:28:41] step 674900/1000000 | loss=2.3905 vocab=0.4583 token=1.2633 | 6.14 steps/s | VRAM=31.76 Go [2026-07-18 06:28:49] step 674950/1000000 | loss=2.3673 vocab=1.6810 token=1.6626 | 6.15 steps/s | VRAM=27.86 Go [2026-07-18 06:28:58] step 675000/1000000 | loss=2.2163 vocab=0.8912 token=1.2205 | 6.12 steps/s | VRAM=29.39 Go [2026-07-18 06:29:06] step 675050/1000000 | loss=2.2362 vocab=0.6857 token=1.2646 | 6.13 steps/s | VRAM=28.09 Go [2026-07-18 06:29:14] step 675100/1000000 | loss=2.1215 vocab=0.7999 token=1.1632 | 6.14 steps/s | VRAM=28.03 Go [2026-07-18 06:29:22] step 675150/1000000 | loss=2.2380 vocab=1.1475 token=1.4631 | 6.17 steps/s | VRAM=24.74 Go [2026-07-18 06:29:30] step 675200/1000000 | loss=2.3095 vocab=0.7476 token=1.4367 | 6.15 steps/s | VRAM=31.20 Go [2026-07-18 06:29:38] step 675250/1000000 | loss=2.1133 vocab=0.4981 token=1.1118 | 6.14 steps/s | VRAM=31.88 Go [2026-07-18 06:29:46] step 675300/1000000 | loss=2.1582 vocab=1.1652 token=1.6545 | 6.14 steps/s | VRAM=30.01 Go [2026-07-18 06:29:55] step 675350/1000000 | loss=2.2434 vocab=0.7778 token=1.3365 | 6.16 steps/s | VRAM=28.34 Go [2026-07-18 06:30:03] step 675400/1000000 | loss=2.2883 vocab=0.9055 token=1.3668 | 6.19 steps/s | VRAM=31.91 Go [2026-07-18 06:30:11] step 675450/1000000 | loss=2.1097 vocab=0.8627 token=1.3556 | 6.15 steps/s | VRAM=28.44 Go [2026-07-18 06:30:19] step 675500/1000000 | loss=2.1535 vocab=0.5755 token=1.0262 | 6.14 steps/s | VRAM=32.15 Go [2026-07-18 06:30:27] step 675550/1000000 | loss=2.2288 vocab=0.7107 token=1.3798 | 6.14 steps/s | VRAM=28.77 Go [2026-07-18 06:30:35] step 675600/1000000 | loss=2.1585 vocab=1.1094 token=1.5736 | 6.14 steps/s | VRAM=25.50 Go [2026-07-18 06:30:43] step 675650/1000000 | loss=2.1898 vocab=0.7071 token=0.8595 | 6.12 steps/s | VRAM=28.13 Go [2026-07-18 06:30:51] step 675700/1000000 | loss=2.1678 vocab=1.2827 token=1.4569 | 6.15 steps/s | VRAM=25.30 Go [2026-07-18 06:31:00] step 675750/1000000 | loss=2.1124 vocab=1.2124 token=1.4671 | 6.15 steps/s | VRAM=28.64 Go [2026-07-18 06:31:08] step 675800/1000000 | loss=2.1564 vocab=0.8886 token=1.2802 | 6.17 steps/s | VRAM=32.07 Go [2026-07-18 06:31:16] step 675850/1000000 | loss=2.1109 vocab=1.3360 token=1.6406 | 6.15 steps/s | VRAM=29.37 Go [2026-07-18 06:31:24] step 675900/1000000 | loss=2.2639 vocab=0.9316 token=1.3485 | 6.14 steps/s | VRAM=31.50 Go [2026-07-18 06:31:32] step 675950/1000000 | loss=2.0423 vocab=0.3891 token=0.8590 | 6.13 steps/s | VRAM=27.92 Go [2026-07-18 06:31:40] step 676000/1000000 | loss=2.2022 vocab=0.3252 token=1.2092 | 6.15 steps/s | VRAM=31.41 Go [2026-07-18 06:31:48] step 676050/1000000 | loss=2.1576 vocab=0.8047 token=1.2030 | 6.13 steps/s | VRAM=25.34 Go [2026-07-18 06:31:57] step 676100/1000000 | loss=2.1307 vocab=0.6300 token=1.3565 | 6.15 steps/s | VRAM=29.25 Go [2026-07-18 06:32:05] step 676150/1000000 | loss=2.0791 vocab=1.0367 token=1.4027 | 6.13 steps/s | VRAM=28.06 Go [2026-07-18 06:32:13] step 676200/1000000 | loss=2.2720 vocab=0.6310 token=1.1057 | 6.14 steps/s | VRAM=28.43 Go [2026-07-18 06:32:21] step 676250/1000000 | loss=2.1884 vocab=1.0394 token=1.1529 | 6.11 steps/s | VRAM=28.43 Go [2026-07-18 06:32:29] step 676300/1000000 | loss=2.1689 vocab=0.9386 token=1.3807 | 6.16 steps/s | VRAM=30.48 Go [2026-07-18 06:32:37] step 676350/1000000 | loss=2.2267 vocab=0.7956 token=1.3251 | 6.10 steps/s | VRAM=24.19 Go [2026-07-18 06:32:46] step 676400/1000000 | loss=2.0753 vocab=0.7359 token=1.0099 | 6.11 steps/s | VRAM=27.27 Go [2026-07-18 06:32:54] step 676450/1000000 | loss=2.1953 vocab=0.6109 token=1.3025 | 6.11 steps/s | VRAM=29.02 Go [2026-07-18 06:33:02] step 676500/1000000 | loss=2.1837 vocab=1.4748 token=1.4743 | 6.13 steps/s | VRAM=31.58 Go [2026-07-18 06:33:10] step 676550/1000000 | loss=2.1312 vocab=0.9351 token=1.0777 | 6.11 steps/s | VRAM=32.22 Go [2026-07-18 06:33:18] step 676600/1000000 | loss=2.1517 vocab=1.3769 token=1.3561 | 6.14 steps/s | VRAM=32.35 Go [2026-07-18 06:33:26] step 676650/1000000 | loss=2.1753 vocab=0.7188 token=1.1287 | 6.11 steps/s | VRAM=31.09 Go [2026-07-18 06:33:35] step 676700/1000000 | loss=2.2138 vocab=1.3232 token=1.7071 | 6.15 steps/s | VRAM=28.95 Go [2026-07-18 06:33:43] step 676750/1000000 | loss=2.1818 vocab=0.8024 token=1.3300 | 6.13 steps/s | VRAM=32.42 Go [2026-07-18 06:33:51] step 676800/1000000 | loss=2.2911 vocab=1.5455 token=1.3534 | 6.14 steps/s | VRAM=28.24 Go [2026-07-18 06:33:59] step 676850/1000000 | loss=2.2822 vocab=1.6150 token=1.3902 | 6.14 steps/s | VRAM=28.98 Go [2026-07-18 06:34:07] step 676900/1000000 | loss=2.2812 vocab=0.5489 token=1.1701 | 6.15 steps/s | VRAM=32.36 Go [2026-07-18 06:34:15] step 676950/1000000 | loss=2.3133 vocab=0.9507 token=1.5534 | 6.12 steps/s | VRAM=25.23 Go [2026-07-18 06:34:23] step 677000/1000000 | loss=2.2791 vocab=1.3607 token=1.7646 | 6.14 steps/s | VRAM=32.62 Go [2026-07-18 06:34:32] step 677050/1000000 | loss=2.1781 vocab=1.1809 token=1.0956 | 6.13 steps/s | VRAM=26.17 Go [2026-07-18 06:34:40] step 677100/1000000 | loss=2.1618 vocab=0.3577 token=0.8178 | 6.16 steps/s | VRAM=30.94 Go [2026-07-18 06:34:48] step 677150/1000000 | loss=2.2480 vocab=0.5151 token=1.0683 | 6.11 steps/s | VRAM=29.83 Go [2026-07-18 06:34:56] step 677200/1000000 | loss=2.2145 vocab=0.6427 token=1.4526 | 6.15 steps/s | VRAM=30.20 Go [2026-07-18 06:35:04] step 677250/1000000 | loss=2.1922 vocab=0.4032 token=1.0590 | 6.13 steps/s | VRAM=29.33 Go [2026-07-18 06:35:12] step 677300/1000000 | loss=2.0963 vocab=0.6209 token=0.9472 | 6.13 steps/s | VRAM=32.46 Go [2026-07-18 06:35:20] step 677350/1000000 | loss=2.2082 vocab=1.0147 token=1.0563 | 6.14 steps/s | VRAM=29.35 Go [2026-07-18 06:35:29] step 677400/1000000 | loss=2.2041 vocab=1.1089 token=0.9387 | 6.09 steps/s | VRAM=28.81 Go [2026-07-18 06:35:37] step 677450/1000000 | loss=2.1780 vocab=0.4665 token=0.9142 | 6.13 steps/s | VRAM=25.87 Go [2026-07-18 06:35:45] step 677500/1000000 | loss=2.0950 vocab=1.1535 token=0.9984 | 6.15 steps/s | VRAM=26.07 Go [2026-07-18 06:35:53] step 677550/1000000 | loss=2.2605 vocab=0.6193 token=0.9170 | 6.15 steps/s | VRAM=31.55 Go [2026-07-18 06:36:01] step 677600/1000000 | loss=2.1856 vocab=0.4704 token=0.9002 | 6.15 steps/s | VRAM=25.47 Go [2026-07-18 06:36:09] step 677650/1000000 | loss=2.1169 vocab=0.7543 token=1.0206 | 6.14 steps/s | VRAM=28.88 Go [2026-07-18 06:36:18] step 677700/1000000 | loss=2.2310 vocab=0.8875 token=1.2935 | 6.12 steps/s | VRAM=25.26 Go [2026-07-18 06:36:26] step 677750/1000000 | loss=2.1469 vocab=1.2087 token=1.4420 | 6.13 steps/s | VRAM=32.18 Go [2026-07-18 06:36:34] step 677800/1000000 | loss=2.2274 vocab=1.7264 token=1.4130 | 6.14 steps/s | VRAM=26.03 Go [2026-07-18 06:36:42] step 677850/1000000 | loss=2.2263 vocab=0.8528 token=1.4689 | 6.13 steps/s | VRAM=28.41 Go [2026-07-18 06:36:50] step 677900/1000000 | loss=2.1777 vocab=0.7061 token=1.2960 | 6.16 steps/s | VRAM=24.68 Go [2026-07-18 06:36:58] step 677950/1000000 | loss=2.2494 vocab=0.7245 token=1.1359 | 6.16 steps/s | VRAM=28.18 Go [2026-07-18 06:37:06] step 678000/1000000 | loss=2.2014 vocab=0.2944 token=1.1490 | 6.10 steps/s | VRAM=28.15 Go [2026-07-18 06:37:15] step 678050/1000000 | loss=2.2542 vocab=0.6508 token=1.1629 | 6.14 steps/s | VRAM=30.86 Go [2026-07-18 06:37:23] step 678100/1000000 | loss=2.2535 vocab=0.6198 token=1.5012 | 6.12 steps/s | VRAM=29.04 Go [2026-07-18 06:37:31] step 678150/1000000 | loss=2.2150 vocab=0.9493 token=1.7105 | 6.18 steps/s | VRAM=28.63 Go [2026-07-18 06:37:39] step 678200/1000000 | loss=2.2377 vocab=0.5005 token=1.3053 | 6.16 steps/s | VRAM=29.13 Go [2026-07-18 06:37:47] step 678250/1000000 | loss=2.1765 vocab=1.1956 token=1.4455 | 6.16 steps/s | VRAM=28.54 Go [2026-07-18 06:37:55] step 678300/1000000 | loss=2.1419 vocab=0.9662 token=1.4260 | 6.17 steps/s | VRAM=29.10 Go [2026-07-18 06:38:03] step 678350/1000000 | loss=2.2135 vocab=0.5793 token=1.3778 | 6.14 steps/s | VRAM=28.56 Go [2026-07-18 06:38:11] step 678400/1000000 | loss=2.3092 vocab=0.7935 token=1.2849 | 6.16 steps/s | VRAM=32.03 Go [2026-07-18 06:38:20] step 678450/1000000 | loss=2.2087 vocab=0.7306 token=0.9975 | 6.13 steps/s | VRAM=26.60 Go [2026-07-18 06:38:28] step 678500/1000000 | loss=2.2149 vocab=0.8701 token=1.1220 | 6.18 steps/s | VRAM=31.22 Go [2026-07-18 06:38:36] step 678550/1000000 | loss=2.2186 vocab=0.7348 token=1.4137 | 6.14 steps/s | VRAM=27.99 Go [2026-07-18 06:38:44] step 678600/1000000 | loss=2.1649 vocab=0.6633 token=1.2250 | 6.11 steps/s | VRAM=30.13 Go [2026-07-18 06:38:52] step 678650/1000000 | loss=2.1866 vocab=0.6498 token=1.5345 | 6.16 steps/s | VRAM=25.44 Go [2026-07-18 06:39:00] step 678700/1000000 | loss=2.2145 vocab=0.8283 token=1.0938 | 6.13 steps/s | VRAM=28.34 Go [2026-07-18 06:39:08] step 678750/1000000 | loss=2.1699 vocab=1.0016 token=1.1355 | 6.15 steps/s | VRAM=25.15 Go [2026-07-18 06:39:17] step 678800/1000000 | loss=2.3997 vocab=1.0661 token=1.1120 | 6.14 steps/s | VRAM=28.97 Go [2026-07-18 06:39:25] step 678850/1000000 | loss=2.2111 vocab=0.9205 token=1.5380 | 6.12 steps/s | VRAM=28.07 Go [2026-07-18 06:39:33] step 678900/1000000 | loss=2.1829 vocab=0.6500 token=1.0927 | 6.14 steps/s | VRAM=32.87 Go [2026-07-18 06:39:41] step 678950/1000000 | loss=2.2234 vocab=1.2095 token=1.5993 | 6.15 steps/s | VRAM=29.42 Go [2026-07-18 06:39:49] step 679000/1000000 | loss=2.2188 vocab=0.9900 token=1.3723 | 6.13 steps/s | VRAM=28.97 Go [2026-07-18 06:39:57] step 679050/1000000 | loss=2.2225 vocab=1.0568 token=1.3098 | 6.13 steps/s | VRAM=28.62 Go [2026-07-18 06:40:05] step 679100/1000000 | loss=2.2705 vocab=1.2303 token=1.5966 | 6.12 steps/s | VRAM=32.30 Go [2026-07-18 06:40:14] step 679150/1000000 | loss=2.1710 vocab=1.0341 token=0.9920 | 6.10 steps/s | VRAM=27.39 Go [2026-07-18 06:40:22] step 679200/1000000 | loss=2.1797 vocab=0.5639 token=1.1695 | 6.17 steps/s | VRAM=30.56 Go [2026-07-18 06:40:30] step 679250/1000000 | loss=2.2609 vocab=0.9207 token=1.1982 | 6.12 steps/s | VRAM=25.76 Go [2026-07-18 06:40:38] step 679300/1000000 | loss=2.1858 vocab=0.9907 token=0.9919 | 6.11 steps/s | VRAM=26.74 Go [2026-07-18 06:40:46] step 679350/1000000 | loss=2.1641 vocab=0.6052 token=1.5095 | 6.12 steps/s | VRAM=29.11 Go [2026-07-18 06:40:54] step 679400/1000000 | loss=2.2078 vocab=0.6948 token=1.1599 | 6.14 steps/s | VRAM=31.59 Go [2026-07-18 06:41:03] step 679450/1000000 | loss=2.2212 vocab=0.8567 token=1.3497 | 6.13 steps/s | VRAM=29.05 Go [2026-07-18 06:41:11] step 679500/1000000 | loss=2.2313 vocab=0.8460 token=1.3240 | 6.10 steps/s | VRAM=24.97 Go [2026-07-18 06:41:19] step 679550/1000000 | loss=2.2027 vocab=0.7388 token=1.1233 | 6.11 steps/s | VRAM=28.52 Go [2026-07-18 06:41:27] step 679600/1000000 | loss=2.0963 vocab=0.9480 token=1.2769 | 6.16 steps/s | VRAM=31.58 Go [2026-07-18 06:41:35] step 679650/1000000 | loss=2.1762 vocab=1.0029 token=0.9313 | 6.11 steps/s | VRAM=26.85 Go [2026-07-18 06:41:43] step 679700/1000000 | loss=2.1040 vocab=0.8363 token=0.9589 | 6.14 steps/s | VRAM=28.41 Go [2026-07-18 06:41:52] step 679750/1000000 | loss=2.2015 vocab=0.6588 token=1.0740 | 6.12 steps/s | VRAM=32.79 Go [2026-07-18 06:42:00] step 679800/1000000 | loss=2.2788 vocab=0.8819 token=1.0008 | 6.14 steps/s | VRAM=27.77 Go [2026-07-18 06:42:08] step 679850/1000000 | loss=2.1631 vocab=0.9571 token=1.1362 | 6.13 steps/s | VRAM=28.91 Go [2026-07-18 06:42:16] step 679900/1000000 | loss=2.2723 vocab=1.2830 token=1.5498 | 6.14 steps/s | VRAM=25.20 Go [2026-07-18 06:42:24] step 679950/1000000 | loss=2.1840 vocab=0.8326 token=1.1069 | 6.15 steps/s | VRAM=29.09 Go [2026-07-18 06:42:32] step 680000/1000000 | loss=2.1316 vocab=1.3489 token=1.6786 | 6.13 steps/s | VRAM=31.27 Go [2026-07-18 06:42:41] step 680050/1000000 | loss=2.2181 vocab=1.2280 token=1.2302 | 6.10 steps/s | VRAM=28.95 Go [2026-07-18 06:42:49] step 680100/1000000 | loss=2.1571 vocab=0.8201 token=1.2622 | 6.11 steps/s | VRAM=28.59 Go [2026-07-18 06:42:57] step 680150/1000000 | loss=2.1844 vocab=0.7633 token=1.2656 | 6.14 steps/s | VRAM=25.66 Go [2026-07-18 06:43:05] step 680200/1000000 | loss=2.1522 vocab=0.9209 token=1.3264 | 6.12 steps/s | VRAM=26.17 Go [2026-07-18 06:43:13] step 680250/1000000 | loss=2.1924 vocab=0.5720 token=1.0576 | 6.15 steps/s | VRAM=29.40 Go [2026-07-18 06:43:21] step 680300/1000000 | loss=2.2917 vocab=0.7511 token=1.2520 | 6.13 steps/s | VRAM=32.91 Go [2026-07-18 06:43:29] step 680350/1000000 | loss=2.2336 vocab=0.9859 token=1.1105 | 6.13 steps/s | VRAM=32.15 Go [2026-07-18 06:43:38] step 680400/1000000 | loss=2.2848 vocab=0.8977 token=1.1055 | 6.08 steps/s | VRAM=26.57 Go [2026-07-18 06:43:46] step 680450/1000000 | loss=2.1644 vocab=1.0358 token=1.4391 | 6.17 steps/s | VRAM=31.47 Go [2026-07-18 06:43:54] step 680500/1000000 | loss=2.1769 vocab=0.7780 token=1.3017 | 6.15 steps/s | VRAM=28.32 Go [2026-07-18 06:44:02] step 680550/1000000 | loss=2.3555 vocab=0.8696 token=0.8821 | 6.12 steps/s | VRAM=26.41 Go [2026-07-18 06:44:10] step 680600/1000000 | loss=2.0895 vocab=0.7944 token=1.1408 | 6.13 steps/s | VRAM=26.47 Go [2026-07-18 06:44:18] step 680650/1000000 | loss=2.3193 vocab=0.9171 token=1.2965 | 6.17 steps/s | VRAM=31.34 Go [2026-07-18 06:44:27] step 680700/1000000 | loss=2.1939 vocab=1.1769 token=1.2320 | 6.11 steps/s | VRAM=26.13 Go [2026-07-18 06:44:35] step 680750/1000000 | loss=2.0739 vocab=1.0469 token=1.0321 | 6.14 steps/s | VRAM=28.81 Go [2026-07-18 06:44:43] step 680800/1000000 | loss=2.1618 vocab=1.0094 token=1.6090 | 6.14 steps/s | VRAM=29.03 Go [2026-07-18 06:44:51] step 680850/1000000 | loss=2.1706 vocab=1.1969 token=1.7749 | 6.15 steps/s | VRAM=29.25 Go [2026-07-18 06:44:59] step 680900/1000000 | loss=2.1875 vocab=0.5687 token=1.1083 | 6.15 steps/s | VRAM=28.09 Go [2026-07-18 06:45:07] step 680950/1000000 | loss=2.2316 vocab=0.8954 token=1.5746 | 6.12 steps/s | VRAM=28.01 Go [2026-07-18 06:45:15] step 681000/1000000 | loss=2.1985 vocab=0.8935 token=1.6402 | 6.17 steps/s | VRAM=26.49 Go [2026-07-18 06:45:24] step 681050/1000000 | loss=2.2845 vocab=1.5751 token=1.4680 | 6.13 steps/s | VRAM=27.48 Go [2026-07-18 06:45:32] step 681100/1000000 | loss=2.2472 vocab=0.6061 token=1.1178 | 6.14 steps/s | VRAM=28.06 Go [2026-07-18 06:45:40] step 681150/1000000 | loss=2.3003 vocab=1.2431 token=1.6118 | 6.15 steps/s | VRAM=27.70 Go [2026-07-18 06:45:48] step 681200/1000000 | loss=2.2683 vocab=1.0032 token=1.4727 | 6.12 steps/s | VRAM=27.58 Go [2026-07-18 06:45:56] step 681250/1000000 | loss=2.2414 vocab=1.2274 token=1.3329 | 6.12 steps/s | VRAM=29.48 Go [2026-07-18 06:46:04] step 681300/1000000 | loss=2.2609 vocab=1.2787 token=1.5961 | 6.15 steps/s | VRAM=28.01 Go [2026-07-18 06:46:12] step 681350/1000000 | loss=2.2576 vocab=1.0778 token=1.3938 | 6.17 steps/s | VRAM=31.48 Go [2026-07-18 06:46:21] step 681400/1000000 | loss=2.1682 vocab=0.8363 token=1.2138 | 6.13 steps/s | VRAM=28.42 Go [2026-07-18 06:46:29] step 681450/1000000 | loss=2.1817 vocab=1.0070 token=1.4269 | 6.16 steps/s | VRAM=31.35 Go [2026-07-18 06:46:37] step 681500/1000000 | loss=2.1877 vocab=0.9087 token=1.3162 | 6.12 steps/s | VRAM=29.18 Go [2026-07-18 06:46:45] step 681550/1000000 | loss=2.1218 vocab=1.2382 token=1.2389 | 6.13 steps/s | VRAM=25.94 Go [2026-07-18 06:46:53] step 681600/1000000 | loss=2.2233 vocab=1.5586 token=1.6582 | 6.16 steps/s | VRAM=24.57 Go [2026-07-18 06:47:01] step 681650/1000000 | loss=2.1767 vocab=0.7012 token=1.4961 | 6.15 steps/s | VRAM=28.31 Go [2026-07-18 06:47:09] step 681700/1000000 | loss=2.2163 vocab=1.3700 token=1.8404 | 6.16 steps/s | VRAM=31.85 Go [2026-07-18 06:47:18] step 681750/1000000 | loss=2.2822 vocab=1.2659 token=1.6153 | 6.10 steps/s | VRAM=30.55 Go [2026-07-18 06:47:26] step 681800/1000000 | loss=2.1674 vocab=1.2421 token=1.1415 | 6.18 steps/s | VRAM=28.71 Go [2026-07-18 06:47:34] step 681850/1000000 | loss=2.2406 vocab=1.0920 token=1.1948 | 6.15 steps/s | VRAM=27.79 Go [2026-07-18 06:47:42] step 681900/1000000 | loss=2.2040 vocab=0.9730 token=1.5243 | 6.16 steps/s | VRAM=32.46 Go [2026-07-18 06:47:50] step 681950/1000000 | loss=2.1990 vocab=1.1417 token=1.1816 | 6.15 steps/s | VRAM=25.56 Go [2026-07-18 06:47:58] step 682000/1000000 | loss=2.2427 vocab=1.4115 token=1.3533 | 6.16 steps/s | VRAM=32.07 Go [2026-07-18 06:48:06] step 682050/1000000 | loss=2.1934 vocab=0.6285 token=1.2304 | 6.16 steps/s | VRAM=31.35 Go [2026-07-18 06:48:14] step 682100/1000000 | loss=2.1277 vocab=1.1519 token=1.5045 | 6.15 steps/s | VRAM=26.76 Go [2026-07-18 06:48:22] step 682150/1000000 | loss=2.2378 vocab=0.2797 token=0.9336 | 6.16 steps/s | VRAM=32.47 Go [2026-07-18 06:48:31] step 682200/1000000 | loss=2.1418 vocab=1.4611 token=1.1720 | 6.13 steps/s | VRAM=28.91 Go [2026-07-18 06:48:39] step 682250/1000000 | loss=2.2146 vocab=0.9591 token=1.3459 | 6.14 steps/s | VRAM=32.00 Go [2026-07-18 06:48:47] step 682300/1000000 | loss=2.2089 vocab=0.6355 token=1.4465 | 6.13 steps/s | VRAM=28.41 Go [2026-07-18 06:48:55] step 682350/1000000 | loss=2.2616 vocab=0.7227 token=1.1602 | 6.13 steps/s | VRAM=26.61 Go [2026-07-18 06:49:03] step 682400/1000000 | loss=2.2150 vocab=0.8152 token=1.4115 | 6.13 steps/s | VRAM=32.86 Go [2026-07-18 06:49:11] step 682450/1000000 | loss=2.2472 vocab=0.9706 token=1.4658 | 6.11 steps/s | VRAM=25.06 Go [2026-07-18 06:49:20] step 682500/1000000 | loss=2.2407 vocab=1.0494 token=1.4363 | 6.15 steps/s | VRAM=25.69 Go [2026-07-18 06:49:28] step 682550/1000000 | loss=2.2844 vocab=1.1221 token=1.2914 | 6.14 steps/s | VRAM=25.25 Go [2026-07-18 06:49:36] step 682600/1000000 | loss=2.2018 vocab=0.6926 token=1.0703 | 6.15 steps/s | VRAM=32.53 Go [2026-07-18 06:49:44] step 682650/1000000 | loss=2.1801 vocab=1.5879 token=1.3227 | 6.13 steps/s | VRAM=28.77 Go [2026-07-18 06:49:52] step 682700/1000000 | loss=2.1948 vocab=0.8077 token=1.6526 | 6.14 steps/s | VRAM=26.89 Go [2026-07-18 06:50:00] step 682750/1000000 | loss=2.1843 vocab=0.8609 token=1.2994 | 6.14 steps/s | VRAM=31.12 Go [2026-07-18 06:50:08] step 682800/1000000 | loss=2.1701 vocab=1.0287 token=1.5367 | 6.12 steps/s | VRAM=30.07 Go [2026-07-18 06:50:17] step 682850/1000000 | loss=2.2435 vocab=0.9188 token=1.3942 | 6.13 steps/s | VRAM=28.97 Go [2026-07-18 06:50:25] step 682900/1000000 | loss=2.2995 vocab=0.6446 token=1.0391 | 6.12 steps/s | VRAM=28.48 Go [2026-07-18 06:50:33] step 682950/1000000 | loss=2.2946 vocab=0.7450 token=1.4020 | 6.12 steps/s | VRAM=28.62 Go [2026-07-18 06:50:41] step 683000/1000000 | loss=2.2218 vocab=0.6647 token=0.9634 | 6.15 steps/s | VRAM=28.42 Go [2026-07-18 06:50:49] step 683050/1000000 | loss=2.1615 vocab=0.7507 token=1.2945 | 6.14 steps/s | VRAM=28.37 Go [2026-07-18 06:50:57] step 683100/1000000 | loss=2.2038 vocab=1.1233 token=1.6135 | 6.11 steps/s | VRAM=32.90 Go [2026-07-18 06:51:06] step 683150/1000000 | loss=2.2215 vocab=1.1367 token=1.3495 | 6.11 steps/s | VRAM=29.02 Go [2026-07-18 06:51:14] step 683200/1000000 | loss=2.2757 vocab=0.7539 token=1.2714 | 6.14 steps/s | VRAM=30.83 Go [2026-07-18 06:51:22] step 683250/1000000 | loss=2.2411 vocab=0.8818 token=1.5319 | 6.14 steps/s | VRAM=26.48 Go [2026-07-18 06:51:30] step 683300/1000000 | loss=2.2034 vocab=0.5929 token=1.4419 | 6.15 steps/s | VRAM=31.35 Go [2026-07-18 06:51:38] step 683350/1000000 | loss=2.1684 vocab=0.6569 token=1.1272 | 6.12 steps/s | VRAM=32.11 Go [2026-07-18 06:51:46] step 683400/1000000 | loss=2.2795 vocab=1.2796 token=1.4108 | 6.10 steps/s | VRAM=26.28 Go [2026-07-18 06:51:55] step 683450/1000000 | loss=2.3114 vocab=0.6018 token=0.8537 | 6.15 steps/s | VRAM=32.11 Go [2026-07-18 06:52:03] step 683500/1000000 | loss=2.1831 vocab=1.0433 token=1.2620 | 6.12 steps/s | VRAM=28.91 Go [2026-07-18 06:52:11] step 683550/1000000 | loss=2.2035 vocab=1.3529 token=1.7496 | 6.14 steps/s | VRAM=24.92 Go [2026-07-18 06:52:19] step 683600/1000000 | loss=2.2587 vocab=1.2411 token=1.6771 | 6.11 steps/s | VRAM=28.92 Go [2026-07-18 06:52:27] step 683650/1000000 | loss=2.2117 vocab=0.9265 token=1.4385 | 6.12 steps/s | VRAM=26.34 Go [2026-07-18 06:52:35] step 683700/1000000 | loss=2.3016 vocab=0.8637 token=1.1505 | 6.14 steps/s | VRAM=28.16 Go [2026-07-18 06:52:43] step 683750/1000000 | loss=2.1448 vocab=1.2335 token=1.2455 | 6.13 steps/s | VRAM=28.68 Go [2026-07-18 06:52:52] step 683800/1000000 | loss=2.1482 vocab=0.7424 token=0.9174 | 6.14 steps/s | VRAM=28.35 Go [2026-07-18 06:53:00] step 683850/1000000 | loss=2.2001 vocab=1.1811 token=1.3045 | 6.14 steps/s | VRAM=31.85 Go [2026-07-18 06:53:08] step 683900/1000000 | loss=2.1893 vocab=1.0228 token=1.5495 | 6.11 steps/s | VRAM=25.69 Go [2026-07-18 06:53:16] step 683950/1000000 | loss=2.1912 vocab=0.9920 token=1.4226 | 6.13 steps/s | VRAM=28.29 Go [2026-07-18 06:53:24] step 684000/1000000 | loss=2.1690 vocab=0.5500 token=0.7504 | 6.12 steps/s | VRAM=32.47 Go [2026-07-18 06:53:32] step 684050/1000000 | loss=2.2659 vocab=1.2139 token=1.7094 | 6.11 steps/s | VRAM=30.10 Go [2026-07-18 06:53:41] step 684100/1000000 | loss=2.1689 vocab=1.2565 token=1.4585 | 6.14 steps/s | VRAM=31.80 Go [2026-07-18 06:53:49] step 684150/1000000 | loss=2.1663 vocab=0.7485 token=1.1236 | 6.13 steps/s | VRAM=26.79 Go [2026-07-18 06:53:57] step 684200/1000000 | loss=2.3121 vocab=1.0963 token=0.9019 | 6.15 steps/s | VRAM=27.52 Go [2026-07-18 06:54:05] step 684250/1000000 | loss=2.1350 vocab=0.6132 token=1.4253 | 6.10 steps/s | VRAM=29.09 Go [2026-07-18 06:54:13] step 684300/1000000 | loss=2.1517 vocab=0.9068 token=1.6033 | 6.12 steps/s | VRAM=30.67 Go [2026-07-18 06:54:21] step 684350/1000000 | loss=2.3000 vocab=1.1545 token=1.4533 | 6.13 steps/s | VRAM=28.30 Go [2026-07-18 06:54:30] step 684400/1000000 | loss=2.2399 vocab=1.1095 token=1.5676 | 6.14 steps/s | VRAM=31.87 Go [2026-07-18 06:54:38] step 684450/1000000 | loss=2.1156 vocab=0.9042 token=0.7105 | 6.15 steps/s | VRAM=25.63 Go [2026-07-18 06:54:46] step 684500/1000000 | loss=2.1374 vocab=0.5800 token=0.8715 | 6.14 steps/s | VRAM=31.92 Go [2026-07-18 06:54:54] step 684550/1000000 | loss=2.1747 vocab=1.0044 token=1.4514 | 6.13 steps/s | VRAM=28.48 Go [2026-07-18 06:55:02] step 684600/1000000 | loss=2.2687 vocab=0.9301 token=1.2525 | 6.15 steps/s | VRAM=28.30 Go [2026-07-18 06:55:10] step 684650/1000000 | loss=2.1946 vocab=1.0284 token=0.9610 | 6.13 steps/s | VRAM=26.45 Go [2026-07-18 06:55:18] step 684700/1000000 | loss=2.1575 vocab=0.7067 token=1.2167 | 6.16 steps/s | VRAM=31.40 Go [2026-07-18 06:55:26] step 684750/1000000 | loss=2.1655 vocab=0.7813 token=1.1560 | 6.16 steps/s | VRAM=31.57 Go [2026-07-18 06:55:35] step 684800/1000000 | loss=2.2753 vocab=1.1387 token=1.4414 | 6.11 steps/s | VRAM=23.46 Go [2026-07-18 06:55:43] step 684850/1000000 | loss=2.2260 vocab=0.7920 token=1.3252 | 6.14 steps/s | VRAM=24.73 Go [2026-07-18 06:55:51] step 684900/1000000 | loss=2.2406 vocab=0.9142 token=1.2699 | 6.11 steps/s | VRAM=32.22 Go [2026-07-18 06:55:59] step 684950/1000000 | loss=2.1876 vocab=0.7693 token=1.6405 | 6.15 steps/s | VRAM=32.91 Go [2026-07-18 06:56:07] step 685000/1000000 | loss=2.2396 vocab=0.6315 token=1.4834 | 6.10 steps/s | VRAM=29.76 Go [2026-07-18 06:56:15] step 685050/1000000 | loss=2.2964 vocab=0.7903 token=1.2109 | 6.15 steps/s | VRAM=31.56 Go [2026-07-18 06:56:24] step 685100/1000000 | loss=2.1849 vocab=1.3115 token=1.4640 | 6.10 steps/s | VRAM=27.10 Go [2026-07-18 06:56:32] step 685150/1000000 | loss=2.2067 vocab=1.0794 token=1.4489 | 6.14 steps/s | VRAM=32.09 Go [2026-07-18 06:56:40] step 685200/1000000 | loss=2.2254 vocab=1.0422 token=1.3962 | 6.11 steps/s | VRAM=28.80 Go [2026-07-18 06:56:48] step 685250/1000000 | loss=2.0650 vocab=0.7322 token=1.0595 | 6.14 steps/s | VRAM=32.76 Go [2026-07-18 06:56:56] step 685300/1000000 | loss=2.2239 vocab=0.7694 token=1.2062 | 6.15 steps/s | VRAM=28.11 Go [2026-07-18 06:57:04] step 685350/1000000 | loss=2.2173 vocab=0.7582 token=1.3496 | 6.15 steps/s | VRAM=28.90 Go [2026-07-18 06:57:13] step 685400/1000000 | loss=2.1239 vocab=1.9785 token=1.8454 | 6.13 steps/s | VRAM=30.26 Go [2026-07-18 06:57:21] step 685450/1000000 | loss=2.2146 vocab=1.2154 token=1.3619 | 6.12 steps/s | VRAM=28.61 Go [2026-07-18 06:57:29] step 685500/1000000 | loss=2.1799 vocab=1.1168 token=1.0923 | 6.17 steps/s | VRAM=30.84 Go [2026-07-18 06:57:37] step 685550/1000000 | loss=2.2079 vocab=1.2202 token=1.1184 | 6.13 steps/s | VRAM=31.45 Go [2026-07-18 06:57:45] step 685600/1000000 | loss=2.2453 vocab=1.1088 token=1.3750 | 6.13 steps/s | VRAM=27.78 Go [2026-07-18 06:57:53] step 685650/1000000 | loss=2.2450 vocab=0.8498 token=1.3085 | 6.14 steps/s | VRAM=32.69 Go [2026-07-18 06:58:01] step 685700/1000000 | loss=2.0625 vocab=0.6414 token=1.0310 | 6.14 steps/s | VRAM=26.03 Go [2026-07-18 06:58:10] step 685750/1000000 | loss=2.1643 vocab=0.9870 token=1.0631 | 6.08 steps/s | VRAM=27.08 Go [2026-07-18 06:58:18] step 685800/1000000 | loss=2.1410 vocab=0.9658 token=1.5821 | 6.12 steps/s | VRAM=32.41 Go [2026-07-18 06:58:26] step 685850/1000000 | loss=2.2111 vocab=1.3027 token=1.2116 | 6.13 steps/s | VRAM=27.90 Go [2026-07-18 06:58:34] step 685900/1000000 | loss=2.1996 vocab=0.8301 token=1.0663 | 6.14 steps/s | VRAM=28.74 Go [2026-07-18 06:58:42] step 685950/1000000 | loss=2.2610 vocab=0.5349 token=1.2718 | 6.13 steps/s | VRAM=28.26 Go [2026-07-18 06:58:50] step 686000/1000000 | loss=2.1313 vocab=0.7345 token=1.3250 | 6.13 steps/s | VRAM=27.56 Go [2026-07-18 06:58:59] step 686050/1000000 | loss=2.0870 vocab=0.9170 token=1.3056 | 6.14 steps/s | VRAM=31.09 Go [2026-07-18 06:59:07] step 686100/1000000 | loss=2.2323 vocab=1.8371 token=1.5739 | 6.10 steps/s | VRAM=28.16 Go [2026-07-18 06:59:15] step 686150/1000000 | loss=2.1518 vocab=0.7647 token=1.1909 | 6.17 steps/s | VRAM=30.18 Go [2026-07-18 06:59:23] step 686200/1000000 | loss=2.1618 vocab=1.1425 token=1.3214 | 6.13 steps/s | VRAM=32.39 Go [2026-07-18 06:59:31] step 686250/1000000 | loss=2.1917 vocab=0.7914 token=1.1348 | 6.11 steps/s | VRAM=27.25 Go [2026-07-18 06:59:39] step 686300/1000000 | loss=2.1764 vocab=0.6520 token=1.4877 | 6.14 steps/s | VRAM=31.43 Go [2026-07-18 06:59:48] step 686350/1000000 | loss=2.0827 vocab=0.7833 token=1.4709 | 6.11 steps/s | VRAM=27.24 Go [2026-07-18 06:59:56] step 686400/1000000 | loss=2.0927 vocab=1.1274 token=1.8726 | 6.15 steps/s | VRAM=32.22 Go [2026-07-18 07:00:04] step 686450/1000000 | loss=2.1933 vocab=1.3577 token=1.6871 | 6.12 steps/s | VRAM=28.10 Go [2026-07-18 07:00:12] step 686500/1000000 | loss=2.1546 vocab=0.9760 token=1.2458 | 6.09 steps/s | VRAM=29.00 Go [2026-07-18 07:00:20] step 686550/1000000 | loss=2.2004 vocab=1.1244 token=1.3614 | 6.14 steps/s | VRAM=32.13 Go [2026-07-18 07:00:28] step 686600/1000000 | loss=2.2369 vocab=0.7673 token=0.7651 | 6.14 steps/s | VRAM=31.94 Go [2026-07-18 07:00:36] step 686650/1000000 | loss=2.2413 vocab=1.4233 token=1.1698 | 6.12 steps/s | VRAM=32.85 Go [2026-07-18 07:00:45] step 686700/1000000 | loss=2.1689 vocab=1.0174 token=1.0349 | 6.11 steps/s | VRAM=28.67 Go [2026-07-18 07:00:53] step 686750/1000000 | loss=2.2419 vocab=0.7188 token=1.3460 | 6.13 steps/s | VRAM=29.19 Go [2026-07-18 07:01:01] step 686800/1000000 | loss=2.2056 vocab=1.0463 token=1.5500 | 6.12 steps/s | VRAM=25.73 Go [2026-07-18 07:01:09] step 686850/1000000 | loss=2.1475 vocab=0.8426 token=1.0912 | 6.13 steps/s | VRAM=28.78 Go [2026-07-18 07:01:17] step 686900/1000000 | loss=2.1729 vocab=0.8093 token=1.0471 | 6.13 steps/s | VRAM=23.33 Go [2026-07-18 07:01:25] step 686950/1000000 | loss=2.2789 vocab=0.7907 token=1.3486 | 6.15 steps/s | VRAM=31.76 Go [2026-07-18 07:01:34] step 687000/1000000 | loss=2.1913 vocab=1.1868 token=1.6043 | 6.12 steps/s | VRAM=31.83 Go [2026-07-18 07:01:42] step 687050/1000000 | loss=2.1438 vocab=0.6325 token=1.3362 | 6.12 steps/s | VRAM=32.13 Go [2026-07-18 07:01:50] step 687100/1000000 | loss=2.2051 vocab=0.7426 token=1.4871 | 6.11 steps/s | VRAM=25.21 Go [2026-07-18 07:01:58] step 687150/1000000 | loss=2.2102 vocab=0.4838 token=0.8851 | 6.15 steps/s | VRAM=29.82 Go [2026-07-18 07:02:06] step 687200/1000000 | loss=2.2017 vocab=0.3927 token=1.1684 | 6.11 steps/s | VRAM=32.86 Go [2026-07-18 07:02:14] step 687250/1000000 | loss=2.1165 vocab=0.6896 token=1.1703 | 6.12 steps/s | VRAM=27.31 Go [2026-07-18 07:02:23] step 687300/1000000 | loss=2.2540 vocab=1.3649 token=1.8307 | 6.13 steps/s | VRAM=29.10 Go [2026-07-18 07:02:31] step 687350/1000000 | loss=2.2215 vocab=1.2632 token=1.5199 | 6.13 steps/s | VRAM=28.89 Go [2026-07-18 07:02:39] step 687400/1000000 | loss=2.1766 vocab=0.7453 token=1.2120 | 6.14 steps/s | VRAM=25.99 Go [2026-07-18 07:02:47] step 687450/1000000 | loss=2.2131 vocab=0.6261 token=1.4762 | 6.12 steps/s | VRAM=30.05 Go [2026-07-18 07:02:55] step 687500/1000000 | loss=2.2190 vocab=0.8540 token=1.4136 | 6.12 steps/s | VRAM=26.44 Go [2026-07-18 07:03:03] step 687550/1000000 | loss=2.1529 vocab=1.0714 token=1.5778 | 6.14 steps/s | VRAM=28.17 Go [2026-07-18 07:03:12] step 687600/1000000 | loss=2.2455 vocab=1.2512 token=1.3834 | 6.12 steps/s | VRAM=28.74 Go [2026-07-18 07:03:20] step 687650/1000000 | loss=2.2346 vocab=0.7445 token=1.7299 | 6.15 steps/s | VRAM=25.00 Go [2026-07-18 07:03:28] step 687700/1000000 | loss=2.2740 vocab=0.4022 token=1.0677 | 6.13 steps/s | VRAM=26.13 Go [2026-07-18 07:03:36] step 687750/1000000 | loss=2.2793 vocab=1.3456 token=1.2651 | 6.10 steps/s | VRAM=29.69 Go [2026-07-18 07:03:44] step 687800/1000000 | loss=2.2055 vocab=1.0214 token=1.3202 | 6.11 steps/s | VRAM=26.41 Go [2026-07-18 07:03:52] step 687850/1000000 | loss=2.2357 vocab=1.4286 token=1.6392 | 6.15 steps/s | VRAM=27.22 Go [2026-07-18 07:04:01] step 687900/1000000 | loss=2.1062 vocab=0.7331 token=0.9875 | 6.12 steps/s | VRAM=26.84 Go [2026-07-18 07:04:09] step 687950/1000000 | loss=2.1652 vocab=1.3288 token=1.3769 | 6.11 steps/s | VRAM=27.14 Go [2026-07-18 07:04:17] step 688000/1000000 | loss=2.0582 vocab=0.5318 token=0.8534 | 6.09 steps/s | VRAM=29.09 Go [2026-07-18 07:04:25] step 688050/1000000 | loss=2.1791 vocab=0.9414 token=1.6137 | 6.13 steps/s | VRAM=24.48 Go [2026-07-18 07:04:33] step 688100/1000000 | loss=2.2396 vocab=0.6937 token=1.4454 | 6.13 steps/s | VRAM=26.32 Go [2026-07-18 07:04:41] step 688150/1000000 | loss=2.1999 vocab=0.6932 token=1.2602 | 6.15 steps/s | VRAM=31.76 Go [2026-07-18 07:04:50] step 688200/1000000 | loss=2.1404 vocab=0.7186 token=1.1185 | 6.10 steps/s | VRAM=25.31 Go [2026-07-18 07:04:58] step 688250/1000000 | loss=2.1706 vocab=1.0422 token=1.2156 | 6.11 steps/s | VRAM=30.20 Go [2026-07-18 07:05:06] step 688300/1000000 | loss=2.2469 vocab=0.7415 token=1.4680 | 6.15 steps/s | VRAM=28.58 Go [2026-07-18 07:05:14] step 688350/1000000 | loss=2.1559 vocab=0.5728 token=1.2196 | 6.14 steps/s | VRAM=28.06 Go [2026-07-18 07:05:22] step 688400/1000000 | loss=2.1032 vocab=0.8680 token=1.5373 | 6.14 steps/s | VRAM=29.36 Go [2026-07-18 07:05:30] step 688450/1000000 | loss=2.1748 vocab=1.2309 token=1.8405 | 6.14 steps/s | VRAM=32.95 Go [2026-07-18 07:05:38] step 688500/1000000 | loss=2.1306 vocab=1.0376 token=1.5240 | 6.15 steps/s | VRAM=32.22 Go [2026-07-18 07:05:47] step 688550/1000000 | loss=2.2106 vocab=0.9905 token=1.0413 | 6.12 steps/s | VRAM=24.10 Go [2026-07-18 07:05:55] step 688600/1000000 | loss=2.1509 vocab=0.8574 token=1.3977 | 6.13 steps/s | VRAM=28.16 Go [2026-07-18 07:06:03] step 688650/1000000 | loss=2.0892 vocab=0.7566 token=0.9381 | 6.16 steps/s | VRAM=29.11 Go [2026-07-18 07:06:11] step 688700/1000000 | loss=2.1766 vocab=0.7754 token=0.8945 | 6.13 steps/s | VRAM=27.61 Go [2026-07-18 07:06:19] step 688750/1000000 | loss=2.1674 vocab=0.8849 token=1.2081 | 6.18 steps/s | VRAM=32.27 Go [2026-07-18 07:06:27] step 688800/1000000 | loss=2.1841 vocab=0.8227 token=1.3311 | 6.18 steps/s | VRAM=24.60 Go [2026-07-18 07:06:35] step 688850/1000000 | loss=2.2748 vocab=1.1067 token=1.5836 | 6.14 steps/s | VRAM=26.20 Go [2026-07-18 07:06:44] step 688900/1000000 | loss=2.2185 vocab=1.3201 token=1.3675 | 6.13 steps/s | VRAM=28.95 Go [2026-07-18 07:06:52] step 688950/1000000 | loss=2.1569 vocab=0.4547 token=1.0945 | 6.13 steps/s | VRAM=29.08 Go [2026-07-18 07:07:00] step 689000/1000000 | loss=2.1952 vocab=1.0350 token=1.5176 | 6.14 steps/s | VRAM=28.62 Go [2026-07-18 07:07:08] step 689050/1000000 | loss=2.2065 vocab=0.8713 token=1.7085 | 6.15 steps/s | VRAM=25.61 Go [2026-07-18 07:07:16] step 689100/1000000 | loss=2.2504 vocab=0.9364 token=1.4330 | 6.16 steps/s | VRAM=27.95 Go [2026-07-18 07:07:24] step 689150/1000000 | loss=2.1864 vocab=0.8930 token=1.3793 | 6.12 steps/s | VRAM=29.04 Go [2026-07-18 07:07:32] step 689200/1000000 | loss=2.2484 vocab=1.0256 token=1.1357 | 6.11 steps/s | VRAM=26.43 Go [2026-07-18 07:07:41] step 689250/1000000 | loss=2.1920 vocab=0.8265 token=1.0330 | 6.14 steps/s | VRAM=32.39 Go [2026-07-18 07:07:49] step 689300/1000000 | loss=2.1931 vocab=1.1228 token=1.1934 | 6.13 steps/s | VRAM=30.02 Go [2026-07-18 07:07:57] step 689350/1000000 | loss=2.1727 vocab=0.8851 token=1.1893 | 6.14 steps/s | VRAM=31.00 Go [2026-07-18 07:08:05] step 689400/1000000 | loss=2.2125 vocab=0.9700 token=1.5603 | 6.16 steps/s | VRAM=28.65 Go [2026-07-18 07:08:13] step 689450/1000000 | loss=2.2261 vocab=0.8015 token=1.2738 | 6.12 steps/s | VRAM=28.75 Go [2026-07-18 07:08:21] step 689500/1000000 | loss=2.1283 vocab=1.3746 token=1.7069 | 6.14 steps/s | VRAM=29.40 Go [2026-07-18 07:08:30] step 689550/1000000 | loss=2.2271 vocab=0.9841 token=1.5443 | 6.08 steps/s | VRAM=27.72 Go [2026-07-18 07:08:38] step 689600/1000000 | loss=2.1797 vocab=1.0314 token=1.4679 | 6.15 steps/s | VRAM=31.20 Go [2026-07-18 07:08:46] step 689650/1000000 | loss=2.1952 vocab=0.8568 token=1.4598 | 6.12 steps/s | VRAM=25.91 Go [2026-07-18 07:08:54] step 689700/1000000 | loss=2.1110 vocab=1.0464 token=1.0058 | 6.15 steps/s | VRAM=26.25 Go [2026-07-18 07:09:02] step 689750/1000000 | loss=2.1674 vocab=0.9451 token=1.1932 | 6.12 steps/s | VRAM=32.75 Go [2026-07-18 07:09:10] step 689800/1000000 | loss=2.1899 vocab=1.1854 token=1.3280 | 6.12 steps/s | VRAM=28.61 Go [2026-07-18 07:09:18] step 689850/1000000 | loss=2.1979 vocab=1.0999 token=1.6871 | 6.13 steps/s | VRAM=26.30 Go [2026-07-18 07:09:27] step 689900/1000000 | loss=2.1553 vocab=1.0291 token=1.2477 | 6.14 steps/s | VRAM=32.72 Go [2026-07-18 07:09:35] step 689950/1000000 | loss=2.2619 vocab=0.5935 token=1.5714 | 6.15 steps/s | VRAM=32.22 Go [2026-07-18 07:09:43] step 690000/1000000 | loss=2.1916 vocab=1.0650 token=1.1061 | 6.12 steps/s | VRAM=29.17 Go [2026-07-18 07:09:51] step 690050/1000000 | loss=2.1967 vocab=1.0813 token=1.4654 | 6.13 steps/s | VRAM=28.47 Go [2026-07-18 07:09:59] step 690100/1000000 | loss=2.2505 vocab=0.8731 token=1.2017 | 6.14 steps/s | VRAM=26.35 Go [2026-07-18 07:10:07] step 690150/1000000 | loss=2.2169 vocab=0.8474 token=1.3218 | 6.15 steps/s | VRAM=28.24 Go [2026-07-18 07:10:15] step 690200/1000000 | loss=2.2965 vocab=0.6358 token=1.7108 | 6.16 steps/s | VRAM=28.70 Go [2026-07-18 07:10:24] step 690250/1000000 | loss=2.1933 vocab=1.2730 token=1.4922 | 6.14 steps/s | VRAM=28.52 Go [2026-07-18 07:10:32] step 690300/1000000 | loss=2.1284 vocab=0.7269 token=1.3415 | 6.13 steps/s | VRAM=26.16 Go [2026-07-18 07:10:40] step 690350/1000000 | loss=2.2134 vocab=1.0886 token=1.3987 | 6.14 steps/s | VRAM=28.53 Go [2026-07-18 07:10:48] step 690400/1000000 | loss=2.1846 vocab=0.8495 token=1.0282 | 6.16 steps/s | VRAM=28.48 Go [2026-07-18 07:10:56] step 690450/1000000 | loss=2.1736 vocab=0.9406 token=1.2832 | 6.13 steps/s | VRAM=32.75 Go [2026-07-18 07:11:04] step 690500/1000000 | loss=2.2395 vocab=0.9182 token=1.1505 | 6.11 steps/s | VRAM=28.91 Go [2026-07-18 07:11:12] step 690550/1000000 | loss=2.2257 vocab=0.7083 token=1.2657 | 6.15 steps/s | VRAM=30.13 Go [2026-07-18 07:11:21] step 690600/1000000 | loss=2.2101 vocab=1.3863 token=1.3465 | 6.14 steps/s | VRAM=30.38 Go [2026-07-18 07:11:29] step 690650/1000000 | loss=2.2209 vocab=1.5036 token=1.5671 | 6.16 steps/s | VRAM=28.27 Go [2026-07-18 07:11:37] step 690700/1000000 | loss=2.2521 vocab=0.8784 token=1.3552 | 6.17 steps/s | VRAM=32.24 Go [2026-07-18 07:11:45] step 690750/1000000 | loss=2.2140 vocab=1.2799 token=1.4604 | 6.14 steps/s | VRAM=25.92 Go [2026-07-18 07:11:53] step 690800/1000000 | loss=2.1829 vocab=1.1996 token=1.4395 | 6.14 steps/s | VRAM=24.88 Go [2026-07-18 07:12:01] step 690850/1000000 | loss=2.2183 vocab=0.7615 token=1.1624 | 6.13 steps/s | VRAM=28.77 Go [2026-07-18 07:12:09] step 690900/1000000 | loss=2.2639 vocab=0.5609 token=1.1364 | 6.12 steps/s | VRAM=28.33 Go [2026-07-18 07:12:18] step 690950/1000000 | loss=2.2656 vocab=1.2416 token=1.3413 | 6.14 steps/s | VRAM=28.86 Go [2026-07-18 07:12:26] step 691000/1000000 | loss=2.2168 vocab=0.6117 token=1.5253 | 6.15 steps/s | VRAM=28.14 Go [2026-07-18 07:12:34] step 691050/1000000 | loss=2.1190 vocab=0.4556 token=0.9912 | 6.14 steps/s | VRAM=30.85 Go [2026-07-18 07:12:42] step 691100/1000000 | loss=2.3287 vocab=0.6102 token=0.9341 | 6.16 steps/s | VRAM=29.82 Go [2026-07-18 07:12:50] step 691150/1000000 | loss=2.1817 vocab=0.5250 token=1.1846 | 6.12 steps/s | VRAM=28.19 Go [2026-07-18 07:12:58] step 691200/1000000 | loss=2.1766 vocab=1.0706 token=1.5021 | 6.14 steps/s | VRAM=30.90 Go [2026-07-18 07:13:06] step 691250/1000000 | loss=2.2572 vocab=1.3417 token=1.6216 | 6.14 steps/s | VRAM=32.03 Go [2026-07-18 07:13:15] step 691300/1000000 | loss=2.2005 vocab=0.8227 token=1.2303 | 6.13 steps/s | VRAM=28.52 Go [2026-07-18 07:13:23] step 691350/1000000 | loss=2.1269 vocab=0.8714 token=0.8488 | 6.08 steps/s | VRAM=29.15 Go [2026-07-18 07:13:31] step 691400/1000000 | loss=2.1070 vocab=1.0095 token=1.3643 | 6.13 steps/s | VRAM=26.09 Go [2026-07-18 07:13:39] step 691450/1000000 | loss=2.1989 vocab=0.7949 token=1.4299 | 6.14 steps/s | VRAM=31.76 Go [2026-07-18 07:13:47] step 691500/1000000 | loss=2.2418 vocab=1.1915 token=1.1560 | 6.13 steps/s | VRAM=30.19 Go [2026-07-18 07:13:55] step 691550/1000000 | loss=2.1767 vocab=1.0963 token=0.9231 | 6.17 steps/s | VRAM=32.06 Go [2026-07-18 07:14:04] step 691600/1000000 | loss=2.1242 vocab=0.5368 token=1.3525 | 6.14 steps/s | VRAM=29.41 Go [2026-07-18 07:14:12] step 691650/1000000 | loss=2.2771 vocab=0.6196 token=1.1142 | 6.12 steps/s | VRAM=28.66 Go [2026-07-18 07:14:20] step 691700/1000000 | loss=2.2069 vocab=0.9545 token=1.6135 | 6.12 steps/s | VRAM=29.33 Go [2026-07-18 07:14:28] step 691750/1000000 | loss=2.1467 vocab=0.4646 token=1.1402 | 6.10 steps/s | VRAM=24.18 Go [2026-07-18 07:14:36] step 691800/1000000 | loss=2.1810 vocab=0.5935 token=0.9354 | 6.13 steps/s | VRAM=28.51 Go [2026-07-18 07:14:44] step 691850/1000000 | loss=2.1473 vocab=0.6263 token=0.9516 | 6.13 steps/s | VRAM=23.79 Go [2026-07-18 07:14:52] step 691900/1000000 | loss=2.1769 vocab=0.8525 token=1.1268 | 6.17 steps/s | VRAM=24.85 Go [2026-07-18 07:15:01] step 691950/1000000 | loss=2.1949 vocab=0.9563 token=1.5302 | 6.13 steps/s | VRAM=32.90 Go [2026-07-18 07:15:09] step 692000/1000000 | loss=2.1649 vocab=0.9738 token=1.3962 | 6.13 steps/s | VRAM=32.14 Go [2026-07-18 07:15:17] step 692050/1000000 | loss=2.2248 vocab=0.7949 token=0.9838 | 6.12 steps/s | VRAM=29.52 Go [2026-07-18 07:15:25] step 692100/1000000 | loss=2.0914 vocab=0.8271 token=1.5229 | 6.15 steps/s | VRAM=28.95 Go [2026-07-18 07:15:33] step 692150/1000000 | loss=2.2883 vocab=1.1220 token=1.8653 | 6.14 steps/s | VRAM=26.26 Go [2026-07-18 07:15:41] step 692200/1000000 | loss=2.2172 vocab=0.5358 token=0.8842 | 6.13 steps/s | VRAM=28.15 Go [2026-07-18 07:15:50] step 692250/1000000 | loss=2.3167 vocab=1.0825 token=1.2525 | 6.16 steps/s | VRAM=32.29 Go [2026-07-18 07:15:58] step 692300/1000000 | loss=2.1344 vocab=0.8945 token=1.3235 | 6.10 steps/s | VRAM=28.84 Go [2026-07-18 07:16:06] step 692350/1000000 | loss=2.1881 vocab=1.1670 token=1.5363 | 6.14 steps/s | VRAM=28.08 Go [2026-07-18 07:16:14] step 692400/1000000 | loss=2.1035 vocab=0.9415 token=1.2193 | 6.12 steps/s | VRAM=25.90 Go [2026-07-18 07:16:22] step 692450/1000000 | loss=2.2385 vocab=1.1615 token=1.1276 | 6.16 steps/s | VRAM=31.44 Go [2026-07-18 07:16:30] step 692500/1000000 | loss=2.3710 vocab=1.0113 token=1.6168 | 6.14 steps/s | VRAM=27.44 Go [2026-07-18 07:16:38] step 692550/1000000 | loss=2.0273 vocab=1.0895 token=1.1084 | 6.12 steps/s | VRAM=28.48 Go [2026-07-18 07:16:47] step 692600/1000000 | loss=2.2592 vocab=1.5048 token=1.6180 | 6.14 steps/s | VRAM=32.17 Go [2026-07-18 07:16:55] step 692650/1000000 | loss=2.1707 vocab=1.1232 token=1.4406 | 6.14 steps/s | VRAM=24.91 Go [2026-07-18 07:17:03] step 692700/1000000 | loss=2.2325 vocab=0.7273 token=1.2677 | 6.16 steps/s | VRAM=25.91 Go [2026-07-18 07:17:11] step 692750/1000000 | loss=2.2740 vocab=0.9394 token=1.0992 | 6.13 steps/s | VRAM=26.92 Go [2026-07-18 07:17:19] step 692800/1000000 | loss=2.1903 vocab=0.3705 token=1.0697 | 6.17 steps/s | VRAM=31.48 Go [2026-07-18 07:17:27] step 692850/1000000 | loss=2.1430 vocab=0.9698 token=1.6418 | 6.13 steps/s | VRAM=28.28 Go [2026-07-18 07:17:35] step 692900/1000000 | loss=2.2539 vocab=0.8801 token=1.7529 | 6.18 steps/s | VRAM=24.71 Go [2026-07-18 07:17:44] step 692950/1000000 | loss=2.2261 vocab=0.7317 token=1.4448 | 6.15 steps/s | VRAM=31.55 Go [2026-07-18 07:17:52] step 693000/1000000 | loss=2.2774 vocab=0.8426 token=1.3224 | 6.16 steps/s | VRAM=29.31 Go [2026-07-18 07:18:00] step 693050/1000000 | loss=2.1784 vocab=0.7258 token=1.1386 | 6.14 steps/s | VRAM=31.91 Go [2026-07-18 07:18:08] step 693100/1000000 | loss=2.2316 vocab=0.9874 token=1.1281 | 6.12 steps/s | VRAM=26.13 Go [2026-07-18 07:18:16] step 693150/1000000 | loss=2.2312 vocab=0.7691 token=1.0430 | 6.14 steps/s | VRAM=28.33 Go [2026-07-18 07:18:24] step 693200/1000000 | loss=2.1724 vocab=0.8626 token=1.2017 | 6.14 steps/s | VRAM=31.79 Go [2026-07-18 07:18:32] step 693250/1000000 | loss=2.1539 vocab=1.4072 token=1.3240 | 6.18 steps/s | VRAM=30.09 Go [2026-07-18 07:18:40] step 693300/1000000 | loss=2.1504 vocab=0.7675 token=1.1243 | 6.15 steps/s | VRAM=27.38 Go [2026-07-18 07:18:49] step 693350/1000000 | loss=2.2063 vocab=0.9543 token=1.1476 | 6.14 steps/s | VRAM=28.18 Go [2026-07-18 07:18:57] step 693400/1000000 | loss=2.2863 vocab=1.2469 token=1.5540 | 6.12 steps/s | VRAM=26.53 Go [2026-07-18 07:19:05] step 693450/1000000 | loss=2.2026 vocab=1.1157 token=1.1630 | 6.13 steps/s | VRAM=25.96 Go [2026-07-18 07:19:13] step 693500/1000000 | loss=2.1702 vocab=0.7520 token=1.5956 | 6.13 steps/s | VRAM=26.63 Go [2026-07-18 07:19:21] step 693550/1000000 | loss=2.1703 vocab=0.6856 token=1.1394 | 6.15 steps/s | VRAM=25.51 Go [2026-07-18 07:19:29] step 693600/1000000 | loss=2.3031 vocab=1.8438 token=1.2505 | 6.13 steps/s | VRAM=32.35 Go [2026-07-18 07:19:38] step 693650/1000000 | loss=2.3399 vocab=0.8128 token=1.2666 | 6.12 steps/s | VRAM=32.43 Go [2026-07-18 07:19:46] step 693700/1000000 | loss=2.1184 vocab=1.0389 token=1.4670 | 6.12 steps/s | VRAM=29.97 Go [2026-07-18 07:19:54] step 693750/1000000 | loss=2.2151 vocab=1.1425 token=1.5417 | 6.14 steps/s | VRAM=25.85 Go [2026-07-18 07:20:02] step 693800/1000000 | loss=2.2843 vocab=0.5399 token=1.1817 | 6.15 steps/s | VRAM=32.55 Go [2026-07-18 07:20:10] step 693850/1000000 | loss=2.2247 vocab=1.0098 token=1.4330 | 6.17 steps/s | VRAM=32.21 Go [2026-07-18 07:20:18] step 693900/1000000 | loss=2.1978 vocab=0.8937 token=1.1968 | 6.18 steps/s | VRAM=25.23 Go [2026-07-18 07:20:26] step 693950/1000000 | loss=2.3144 vocab=0.7636 token=1.4541 | 6.14 steps/s | VRAM=28.74 Go [2026-07-18 07:20:34] step 694000/1000000 | loss=2.1920 vocab=1.4110 token=1.6146 | 6.14 steps/s | VRAM=29.82 Go [2026-07-18 07:20:43] step 694050/1000000 | loss=2.1636 vocab=0.8522 token=1.3893 | 6.12 steps/s | VRAM=28.24 Go [2026-07-18 07:20:51] step 694100/1000000 | loss=2.2364 vocab=0.6377 token=1.0520 | 6.14 steps/s | VRAM=28.53 Go [2026-07-18 07:20:59] step 694150/1000000 | loss=2.2162 vocab=0.5571 token=1.2263 | 6.13 steps/s | VRAM=31.20 Go [2026-07-18 07:21:07] step 694200/1000000 | loss=2.1910 vocab=0.8400 token=1.1147 | 6.12 steps/s | VRAM=28.81 Go [2026-07-18 07:21:15] step 694250/1000000 | loss=2.2042 vocab=0.1507 token=0.6946 | 6.15 steps/s | VRAM=29.09 Go [2026-07-18 07:21:23] step 694300/1000000 | loss=2.1238 vocab=0.4501 token=1.2227 | 6.18 steps/s | VRAM=28.08 Go [2026-07-18 07:21:31] step 694350/1000000 | loss=2.2029 vocab=0.8345 token=1.2412 | 6.15 steps/s | VRAM=31.73 Go [2026-07-18 07:21:40] step 694400/1000000 | loss=2.2548 vocab=1.2506 token=1.6638 | 6.13 steps/s | VRAM=26.33 Go [2026-07-18 07:21:48] step 694450/1000000 | loss=2.3336 vocab=1.0726 token=1.3846 | 6.15 steps/s | VRAM=28.70 Go [2026-07-18 07:21:56] step 694500/1000000 | loss=2.0778 vocab=0.7304 token=1.1055 | 6.15 steps/s | VRAM=25.00 Go [2026-07-18 07:22:04] step 694550/1000000 | loss=2.2619 vocab=0.9491 token=1.7424 | 6.13 steps/s | VRAM=28.39 Go [2026-07-18 07:22:12] step 694600/1000000 | loss=2.2083 vocab=0.3874 token=0.9345 | 6.19 steps/s | VRAM=31.72 Go [2026-07-18 07:22:20] step 694650/1000000 | loss=2.2873 vocab=0.6007 token=1.1272 | 6.13 steps/s | VRAM=28.92 Go [2026-07-18 07:22:28] step 694700/1000000 | loss=2.3207 vocab=0.6991 token=1.1386 | 6.12 steps/s | VRAM=25.87 Go [2026-07-18 07:22:37] step 694750/1000000 | loss=2.1241 vocab=1.1283 token=1.7447 | 6.10 steps/s | VRAM=25.95 Go [2026-07-18 07:22:45] step 694800/1000000 | loss=2.2739 vocab=1.1743 token=1.2941 | 6.12 steps/s | VRAM=28.83 Go [2026-07-18 07:22:53] step 694850/1000000 | loss=2.3113 vocab=0.9361 token=1.2511 | 6.14 steps/s | VRAM=29.37 Go [2026-07-18 07:23:01] step 694900/1000000 | loss=2.1841 vocab=1.5747 token=1.5087 | 6.13 steps/s | VRAM=26.00 Go [2026-07-18 07:23:09] step 694950/1000000 | loss=2.3124 vocab=1.2855 token=1.2577 | 6.14 steps/s | VRAM=29.12 Go [2026-07-18 07:23:17] step 695000/1000000 | loss=2.1196 vocab=1.0479 token=1.3514 | 6.15 steps/s | VRAM=29.09 Go [2026-07-18 07:23:26] step 695050/1000000 | loss=2.0902 vocab=1.0747 token=1.4824 | 6.12 steps/s | VRAM=29.24 Go [2026-07-18 07:23:34] step 695100/1000000 | loss=2.3579 vocab=1.0096 token=1.2879 | 6.14 steps/s | VRAM=26.25 Go [2026-07-18 07:23:42] step 695150/1000000 | loss=2.2247 vocab=0.9554 token=1.5660 | 6.15 steps/s | VRAM=28.88 Go [2026-07-18 07:23:50] step 695200/1000000 | loss=2.1136 vocab=0.7907 token=1.3533 | 6.14 steps/s | VRAM=32.31 Go [2026-07-18 07:23:58] step 695250/1000000 | loss=2.0544 vocab=0.7183 token=1.4829 | 6.14 steps/s | VRAM=25.76 Go [2026-07-18 07:24:06] step 695300/1000000 | loss=2.2213 vocab=1.1261 token=1.5511 | 6.13 steps/s | VRAM=29.56 Go [2026-07-18 07:24:14] step 695350/1000000 | loss=2.1910 vocab=1.4572 token=1.8219 | 6.11 steps/s | VRAM=26.37 Go [2026-07-18 07:24:23] step 695400/1000000 | loss=2.2990 vocab=0.8224 token=1.2544 | 6.14 steps/s | VRAM=29.34 Go [2026-07-18 07:24:31] step 695450/1000000 | loss=2.1698 vocab=1.0162 token=1.4964 | 6.12 steps/s | VRAM=26.53 Go [2026-07-18 07:24:39] step 695500/1000000 | loss=2.2397 vocab=0.6760 token=1.4423 | 6.18 steps/s | VRAM=29.35 Go [2026-07-18 07:24:47] step 695550/1000000 | loss=2.1691 vocab=0.8765 token=1.3168 | 6.12 steps/s | VRAM=26.19 Go [2026-07-18 07:24:55] step 695600/1000000 | loss=2.1978 vocab=1.2784 token=1.3731 | 6.14 steps/s | VRAM=29.40 Go [2026-07-18 07:25:03] step 695650/1000000 | loss=2.2563 vocab=1.0057 token=1.0460 | 6.14 steps/s | VRAM=29.00 Go [2026-07-18 07:25:11] step 695700/1000000 | loss=2.2176 vocab=0.4695 token=0.9667 | 6.15 steps/s | VRAM=28.61 Go [2026-07-18 07:25:20] step 695750/1000000 | loss=2.3477 vocab=1.1626 token=1.5424 | 6.13 steps/s | VRAM=25.90 Go [2026-07-18 07:25:28] step 695800/1000000 | loss=2.2077 vocab=0.6348 token=1.2864 | 5.76 steps/s | VRAM=26.24 Go [2026-07-18 07:25:36] step 695850/1000000 | loss=2.0927 vocab=0.5304 token=1.3318 | 6.17 steps/s | VRAM=28.14 Go [2026-07-18 07:25:45] step 695900/1000000 | loss=2.2410 vocab=1.3119 token=1.1270 | 6.12 steps/s | VRAM=26.03 Go [2026-07-18 07:25:53] step 695950/1000000 | loss=2.3375 vocab=0.9367 token=1.0936 | 6.14 steps/s | VRAM=31.39 Go [2026-07-18 07:26:01] step 696000/1000000 | loss=2.1862 vocab=1.0668 token=1.5204 | 6.14 steps/s | VRAM=28.81 Go [2026-07-18 07:26:09] step 696050/1000000 | loss=2.1772 vocab=0.7033 token=1.4953 | 6.13 steps/s | VRAM=25.86 Go [2026-07-18 07:26:17] step 696100/1000000 | loss=2.1885 vocab=0.6428 token=0.8782 | 6.11 steps/s | VRAM=28.46 Go [2026-07-18 07:26:25] step 696150/1000000 | loss=2.1548 vocab=0.7494 token=1.2649 | 6.16 steps/s | VRAM=27.72 Go [2026-07-18 07:26:34] step 696200/1000000 | loss=2.1412 vocab=1.0110 token=1.3520 | 6.13 steps/s | VRAM=23.89 Go [2026-07-18 07:26:42] step 696250/1000000 | loss=2.3099 vocab=0.4204 token=1.3525 | 6.15 steps/s | VRAM=31.09 Go [2026-07-18 07:26:50] step 696300/1000000 | loss=2.1970 vocab=1.1039 token=1.2410 | 6.16 steps/s | VRAM=28.08 Go [2026-07-18 07:26:58] step 696350/1000000 | loss=2.1022 vocab=0.8301 token=1.3665 | 6.11 steps/s | VRAM=26.18 Go [2026-07-18 07:27:06] step 696400/1000000 | loss=2.1232 vocab=1.0555 token=1.8206 | 6.14 steps/s | VRAM=32.21 Go [2026-07-18 07:27:14] step 696450/1000000 | loss=2.0550 vocab=1.1217 token=1.6144 | 6.12 steps/s | VRAM=28.14 Go [2026-07-18 07:27:22] step 696500/1000000 | loss=2.3017 vocab=0.6674 token=1.6300 | 6.12 steps/s | VRAM=25.56 Go [2026-07-18 07:27:31] step 696550/1000000 | loss=2.2073 vocab=1.0752 token=1.7481 | 6.15 steps/s | VRAM=28.90 Go [2026-07-18 07:27:39] step 696600/1000000 | loss=2.3105 vocab=0.8434 token=1.1834 | 6.14 steps/s | VRAM=26.06 Go [2026-07-18 07:27:47] step 696650/1000000 | loss=2.2218 vocab=0.9153 token=1.6621 | 6.14 steps/s | VRAM=26.14 Go [2026-07-18 07:27:55] step 696700/1000000 | loss=2.1470 vocab=0.7814 token=1.9862 | 6.15 steps/s | VRAM=32.74 Go [2026-07-18 07:28:03] step 696750/1000000 | loss=2.1551 vocab=0.8654 token=1.1874 | 6.15 steps/s | VRAM=28.74 Go [2026-07-18 07:28:11] step 696800/1000000 | loss=2.2998 vocab=0.7625 token=1.3672 | 6.13 steps/s | VRAM=28.54 Go [2026-07-18 07:28:19] step 696850/1000000 | loss=2.1197 vocab=1.3849 token=1.3322 | 6.14 steps/s | VRAM=30.13 Go [2026-07-18 07:28:28] step 696900/1000000 | loss=2.1930 vocab=0.4750 token=0.9652 | 6.09 steps/s | VRAM=29.36 Go [2026-07-18 07:28:36] step 696950/1000000 | loss=2.0812 vocab=0.9406 token=1.5169 | 6.11 steps/s | VRAM=28.60 Go [2026-07-18 07:28:44] step 697000/1000000 | loss=2.2553 vocab=1.0134 token=1.6179 | 6.13 steps/s | VRAM=25.85 Go [2026-07-18 07:28:52] step 697050/1000000 | loss=2.2584 vocab=1.2429 token=1.2021 | 6.18 steps/s | VRAM=31.96 Go [2026-07-18 07:29:00] step 697100/1000000 | loss=2.2497 vocab=1.0820 token=1.3483 | 6.14 steps/s | VRAM=28.39 Go [2026-07-18 07:29:08] step 697150/1000000 | loss=2.2477 vocab=0.9079 token=1.2331 | 6.13 steps/s | VRAM=26.82 Go [2026-07-18 07:29:16] step 697200/1000000 | loss=2.0999 vocab=0.3747 token=1.2831 | 6.17 steps/s | VRAM=29.16 Go [2026-07-18 07:29:25] step 697250/1000000 | loss=2.2184 vocab=0.7761 token=1.3186 | 6.17 steps/s | VRAM=31.14 Go [2026-07-18 07:29:33] step 697300/1000000 | loss=2.1290 vocab=1.3391 token=1.5742 | 6.17 steps/s | VRAM=29.39 Go [2026-07-18 07:29:41] step 697350/1000000 | loss=2.1502 vocab=0.8583 token=1.1413 | 6.13 steps/s | VRAM=31.85 Go [2026-07-18 07:29:49] step 697400/1000000 | loss=2.1579 vocab=0.5656 token=1.3064 | 6.16 steps/s | VRAM=32.17 Go [2026-07-18 07:29:57] step 697450/1000000 | loss=2.2658 vocab=0.6203 token=1.5963 | 6.13 steps/s | VRAM=32.91 Go [2026-07-18 07:30:05] step 697500/1000000 | loss=2.0992 vocab=0.8650 token=1.3342 | 6.16 steps/s | VRAM=28.16 Go [2026-07-18 07:30:13] step 697550/1000000 | loss=2.2721 vocab=0.4924 token=1.1023 | 6.16 steps/s | VRAM=27.92 Go [2026-07-18 07:30:21] step 697600/1000000 | loss=2.2407 vocab=0.8833 token=1.0364 | 6.15 steps/s | VRAM=32.19 Go [2026-07-18 07:30:30] step 697650/1000000 | loss=2.1177 vocab=0.8024 token=1.4114 | 6.14 steps/s | VRAM=31.66 Go [2026-07-18 07:30:38] step 697700/1000000 | loss=2.1606 vocab=0.7379 token=1.2322 | 6.10 steps/s | VRAM=28.36 Go [2026-07-18 07:30:46] step 697750/1000000 | loss=2.1952 vocab=0.7556 token=1.0901 | 6.10 steps/s | VRAM=29.95 Go [2026-07-18 07:30:54] step 697800/1000000 | loss=2.1099 vocab=0.7381 token=1.2253 | 6.12 steps/s | VRAM=30.09 Go [2026-07-18 07:31:02] step 697850/1000000 | loss=2.1515 vocab=0.5374 token=0.7333 | 6.15 steps/s | VRAM=31.96 Go [2026-07-18 07:31:10] step 697900/1000000 | loss=2.2786 vocab=0.8413 token=1.5148 | 6.14 steps/s | VRAM=32.27 Go [2026-07-18 07:31:19] step 697950/1000000 | loss=2.0985 vocab=0.6917 token=1.3402 | 6.14 steps/s | VRAM=31.39 Go [2026-07-18 07:31:27] step 698000/1000000 | loss=2.2194 vocab=0.9312 token=0.9881 | 6.15 steps/s | VRAM=29.20 Go [2026-07-18 07:31:35] step 698050/1000000 | loss=2.1852 vocab=0.9167 token=1.4021 | 6.14 steps/s | VRAM=31.75 Go [2026-07-18 07:31:43] step 698100/1000000 | loss=2.0559 vocab=0.4703 token=0.9716 | 6.10 steps/s | VRAM=26.27 Go [2026-07-18 07:31:51] step 698150/1000000 | loss=2.0924 vocab=0.7058 token=1.2218 | 6.13 steps/s | VRAM=30.04 Go [2026-07-18 07:31:59] step 698200/1000000 | loss=2.1996 vocab=0.8757 token=1.5254 | 6.16 steps/s | VRAM=31.54 Go [2026-07-18 07:32:07] step 698250/1000000 | loss=2.2390 vocab=0.9482 token=1.2544 | 6.13 steps/s | VRAM=25.84 Go [2026-07-18 07:32:16] step 698300/1000000 | loss=2.1621 vocab=0.7500 token=1.1108 | 6.15 steps/s | VRAM=31.61 Go [2026-07-18 07:32:24] step 698350/1000000 | loss=2.2745 vocab=0.5760 token=1.8739 | 6.14 steps/s | VRAM=27.90 Go [2026-07-18 07:32:32] step 698400/1000000 | loss=2.1947 vocab=0.4675 token=1.2126 | 6.11 steps/s | VRAM=29.51 Go [2026-07-18 07:32:40] step 698450/1000000 | loss=2.1467 vocab=0.9357 token=1.4086 | 6.14 steps/s | VRAM=31.35 Go [2026-07-18 07:32:48] step 698500/1000000 | loss=2.1917 vocab=0.5825 token=1.5199 | 6.17 steps/s | VRAM=27.01 Go [2026-07-18 07:32:56] step 698550/1000000 | loss=2.1365 vocab=0.5378 token=0.8825 | 6.11 steps/s | VRAM=30.09 Go [2026-07-18 07:33:04] step 698600/1000000 | loss=2.2815 vocab=0.5987 token=1.3614 | 6.14 steps/s | VRAM=25.21 Go [2026-07-18 07:33:13] step 698650/1000000 | loss=2.2039 vocab=1.0965 token=1.1876 | 6.14 steps/s | VRAM=27.98 Go [2026-07-18 07:33:21] step 698700/1000000 | loss=2.1911 vocab=1.0158 token=1.1653 | 6.14 steps/s | VRAM=29.27 Go [2026-07-18 07:33:29] step 698750/1000000 | loss=2.1938 vocab=1.6251 token=1.5944 | 6.10 steps/s | VRAM=28.37 Go [2026-07-18 07:33:37] step 698800/1000000 | loss=2.3078 vocab=0.8955 token=1.2184 | 6.12 steps/s | VRAM=25.27 Go [2026-07-18 07:33:45] step 698850/1000000 | loss=2.2430 vocab=0.7900 token=1.1337 | 6.13 steps/s | VRAM=28.25 Go [2026-07-18 07:33:53] step 698900/1000000 | loss=2.1541 vocab=0.9662 token=1.1196 | 6.14 steps/s | VRAM=28.60 Go [2026-07-18 07:34:02] step 698950/1000000 | loss=2.1944 vocab=1.1309 token=1.6674 | 6.11 steps/s | VRAM=29.29 Go [2026-07-18 07:34:10] step 699000/1000000 | loss=2.0807 vocab=0.8555 token=1.0355 | 6.15 steps/s | VRAM=32.26 Go [2026-07-18 07:34:18] step 699050/1000000 | loss=2.2169 vocab=0.8794 token=1.5702 | 6.15 steps/s | VRAM=28.47 Go [2026-07-18 07:34:26] step 699100/1000000 | loss=2.2329 vocab=0.9277 token=0.7133 | 6.13 steps/s | VRAM=33.00 Go [2026-07-18 07:34:34] step 699150/1000000 | loss=2.1163 vocab=0.3979 token=0.7031 | 6.18 steps/s | VRAM=28.89 Go [2026-07-18 07:34:42] step 699200/1000000 | loss=2.1842 vocab=1.6077 token=1.3366 | 6.14 steps/s | VRAM=25.10 Go [2026-07-18 07:34:50] step 699250/1000000 | loss=2.1848 vocab=0.8013 token=1.2060 | 6.12 steps/s | VRAM=28.83 Go [2026-07-18 07:34:59] step 699300/1000000 | loss=2.2140 vocab=0.7181 token=1.2517 | 6.12 steps/s | VRAM=28.97 Go [2026-07-18 07:35:07] step 699350/1000000 | loss=2.1383 vocab=0.4969 token=0.9298 | 6.14 steps/s | VRAM=29.14 Go [2026-07-18 07:35:15] step 699400/1000000 | loss=2.2247 vocab=0.7265 token=1.2431 | 6.11 steps/s | VRAM=26.19 Go [2026-07-18 07:35:23] step 699450/1000000 | loss=2.1428 vocab=0.5670 token=1.3413 | 6.12 steps/s | VRAM=28.16 Go [2026-07-18 07:35:31] step 699500/1000000 | loss=2.1934 vocab=1.0247 token=1.3042 | 6.14 steps/s | VRAM=31.53 Go [2026-07-18 07:35:39] step 699550/1000000 | loss=2.1132 vocab=0.5494 token=1.3103 | 6.16 steps/s | VRAM=25.13 Go [2026-07-18 07:35:48] step 699600/1000000 | loss=2.2411 vocab=1.0113 token=1.5363 | 6.15 steps/s | VRAM=29.62 Go [2026-07-18 07:35:56] step 699650/1000000 | loss=2.1927 vocab=0.5878 token=1.3244 | 6.12 steps/s | VRAM=24.67 Go [2026-07-18 07:36:04] step 699700/1000000 | loss=2.0952 vocab=1.1446 token=1.7063 | 6.11 steps/s | VRAM=26.13 Go [2026-07-18 07:36:12] step 699750/1000000 | loss=2.1534 vocab=0.9719 token=1.5566 | 6.15 steps/s | VRAM=25.33 Go [2026-07-18 07:36:20] step 699800/1000000 | loss=2.2521 vocab=0.6665 token=1.1853 | 6.16 steps/s | VRAM=27.77 Go [2026-07-18 07:36:28] step 699850/1000000 | loss=2.1480 vocab=1.0366 token=1.3971 | 6.12 steps/s | VRAM=29.46 Go [2026-07-18 07:36:36] step 699900/1000000 | loss=2.0895 vocab=0.7178 token=1.1391 | 6.14 steps/s | VRAM=28.44 Go [2026-07-18 07:36:45] step 699950/1000000 | loss=2.1960 vocab=1.2000 token=1.3835 | 6.14 steps/s | VRAM=26.23 Go [2026-07-18 07:36:53] step 700000/1000000 | loss=2.2743 vocab=1.2242 token=1.5906 | 6.13 steps/s | VRAM=27.85 Go [push] step 700000 OK (best-of-2, loss_token=1.5906) [2026-07-18 07:38:20] step 700050/1000000 | loss=2.1338 vocab=0.4497 token=0.8688 | 0.57 steps/s | VRAM=29.35 Go [2026-07-18 07:38:28] step 700100/1000000 | loss=2.2358 vocab=1.2513 token=1.2581 | 6.25 steps/s | VRAM=25.09 Go [2026-07-18 07:38:36] step 700150/1000000 | loss=2.0604 vocab=0.7402 token=1.3830 | 6.19 steps/s | VRAM=27.69 Go [2026-07-18 07:38:44] step 700200/1000000 | loss=2.2694 vocab=0.6692 token=1.3156 | 6.17 steps/s | VRAM=30.04 Go [2026-07-18 07:38:52] step 700250/1000000 | loss=2.2087 vocab=0.6594 token=1.0013 | 6.18 steps/s | VRAM=27.66 Go [2026-07-18 07:39:00] step 700300/1000000 | loss=2.2110 vocab=1.1237 token=1.4473 | 6.16 steps/s | VRAM=29.21 Go [2026-07-18 07:39:08] step 700350/1000000 | loss=2.2911 vocab=0.3858 token=1.3348 | 6.17 steps/s | VRAM=29.40 Go [2026-07-18 07:39:16] step 700400/1000000 | loss=2.0415 vocab=0.6199 token=1.3630 | 6.17 steps/s | VRAM=25.57 Go [2026-07-18 07:39:24] step 700450/1000000 | loss=2.1619 vocab=1.1577 token=1.3636 | 6.16 steps/s | VRAM=32.83 Go [2026-07-18 07:39:33] step 700500/1000000 | loss=2.1854 vocab=1.0446 token=1.4219 | 6.13 steps/s | VRAM=29.10 Go [2026-07-18 07:39:41] step 700550/1000000 | loss=2.1234 vocab=0.7941 token=1.2428 | 6.18 steps/s | VRAM=28.14 Go [2026-07-18 07:39:49] step 700600/1000000 | loss=2.2062 vocab=1.1686 token=1.5414 | 6.15 steps/s | VRAM=32.60 Go [2026-07-18 07:39:57] step 700650/1000000 | loss=2.1860 vocab=1.2135 token=1.3891 | 6.16 steps/s | VRAM=26.12 Go [2026-07-18 07:40:05] step 700700/1000000 | loss=2.1179 vocab=0.7536 token=1.2090 | 6.15 steps/s | VRAM=28.72 Go [2026-07-18 07:40:13] step 700750/1000000 | loss=2.2317 vocab=0.4908 token=1.0392 | 6.15 steps/s | VRAM=28.47 Go [2026-07-18 07:40:21] step 700800/1000000 | loss=2.3315 vocab=0.5243 token=1.1748 | 6.12 steps/s | VRAM=26.34 Go [2026-07-18 07:40:29] step 700850/1000000 | loss=2.2248 vocab=1.2045 token=1.2997 | 6.16 steps/s | VRAM=24.95 Go [2026-07-18 07:40:38] step 700900/1000000 | loss=2.2402 vocab=0.8388 token=1.0772 | 6.09 steps/s | VRAM=26.37 Go [2026-07-18 07:40:46] step 700950/1000000 | loss=2.2283 vocab=1.3195 token=1.6544 | 6.16 steps/s | VRAM=28.28 Go [2026-07-18 07:40:54] step 701000/1000000 | loss=2.1988 vocab=1.3313 token=1.5949 | 6.15 steps/s | VRAM=32.54 Go [2026-07-18 07:41:02] step 701050/1000000 | loss=2.1715 vocab=0.6817 token=0.8010 | 6.12 steps/s | VRAM=22.82 Go [2026-07-18 07:41:10] step 701100/1000000 | loss=2.1440 vocab=1.2982 token=1.0195 | 6.15 steps/s | VRAM=33.03 Go [2026-07-18 07:41:18] step 701150/1000000 | loss=2.1881 vocab=1.0203 token=1.4673 | 6.13 steps/s | VRAM=26.39 Go [2026-07-18 07:41:27] step 701200/1000000 | loss=2.2082 vocab=0.5880 token=1.3763 | 6.11 steps/s | VRAM=26.64 Go [2026-07-18 07:41:35] step 701250/1000000 | loss=2.0872 vocab=1.0117 token=1.2359 | 6.15 steps/s | VRAM=30.78 Go [2026-07-18 07:41:43] step 701300/1000000 | loss=2.1403 vocab=0.5910 token=1.1973 | 6.15 steps/s | VRAM=25.00 Go [2026-07-18 07:41:51] step 701350/1000000 | loss=2.0719 vocab=1.1265 token=1.1063 | 6.15 steps/s | VRAM=30.20 Go [2026-07-18 07:41:59] step 701400/1000000 | loss=2.2255 vocab=0.6637 token=1.1041 | 6.14 steps/s | VRAM=28.66 Go [2026-07-18 07:42:07] step 701450/1000000 | loss=2.2540 vocab=1.2382 token=1.7633 | 6.13 steps/s | VRAM=28.61 Go [2026-07-18 07:42:15] step 701500/1000000 | loss=2.1373 vocab=0.5414 token=1.0681 | 6.15 steps/s | VRAM=32.48 Go [2026-07-18 07:42:24] step 701550/1000000 | loss=2.1258 vocab=0.8759 token=1.7483 | 6.12 steps/s | VRAM=28.86 Go [2026-07-18 07:42:32] step 701600/1000000 | loss=2.1012 vocab=1.0691 token=1.1180 | 6.16 steps/s | VRAM=28.66 Go [2026-07-18 07:42:40] step 701650/1000000 | loss=2.0744 vocab=0.6040 token=1.2526 | 6.13 steps/s | VRAM=29.53 Go [2026-07-18 07:42:48] step 701700/1000000 | loss=2.1903 vocab=1.1126 token=1.2733 | 6.16 steps/s | VRAM=28.28 Go [2026-07-18 07:42:56] step 701750/1000000 | loss=2.2530 vocab=1.0056 token=0.6845 | 6.15 steps/s | VRAM=29.39 Go [2026-07-18 07:43:04] step 701800/1000000 | loss=2.2006 vocab=0.6963 token=1.0772 | 6.11 steps/s | VRAM=28.82 Go [2026-07-18 07:43:12] step 701850/1000000 | loss=2.2207 vocab=0.8712 token=1.5299 | 6.14 steps/s | VRAM=25.09 Go [2026-07-18 07:43:21] step 701900/1000000 | loss=2.1725 vocab=0.4936 token=1.2371 | 6.15 steps/s | VRAM=32.95 Go [2026-07-18 07:43:29] step 701950/1000000 | loss=2.1059 vocab=0.7371 token=1.3003 | 6.14 steps/s | VRAM=28.38 Go [2026-07-18 07:43:37] step 702000/1000000 | loss=2.1960 vocab=0.6033 token=1.1173 | 6.14 steps/s | VRAM=26.53 Go [2026-07-18 07:43:45] step 702050/1000000 | loss=2.1911 vocab=0.3822 token=1.0217 | 6.13 steps/s | VRAM=28.48 Go [2026-07-18 07:43:53] step 702100/1000000 | loss=2.1898 vocab=1.2016 token=1.6361 | 6.15 steps/s | VRAM=27.22 Go [2026-07-18 07:44:01] step 702150/1000000 | loss=2.1301 vocab=0.9390 token=1.3195 | 6.13 steps/s | VRAM=28.76 Go [2026-07-18 07:44:10] step 702200/1000000 | loss=2.2709 vocab=1.0556 token=1.4042 | 6.11 steps/s | VRAM=27.03 Go [2026-07-18 07:44:18] step 702250/1000000 | loss=2.2449 vocab=0.7978 token=1.1141 | 6.14 steps/s | VRAM=24.95 Go [2026-07-18 07:44:26] step 702300/1000000 | loss=2.2326 vocab=1.2370 token=1.3948 | 6.16 steps/s | VRAM=27.84 Go [2026-07-18 07:44:34] step 702350/1000000 | loss=2.1941 vocab=1.1495 token=1.1985 | 6.15 steps/s | VRAM=29.81 Go [2026-07-18 07:44:42] step 702400/1000000 | loss=2.1982 vocab=1.2325 token=1.6133 | 6.14 steps/s | VRAM=29.73 Go [2026-07-18 07:44:50] step 702450/1000000 | loss=2.2282 vocab=0.9227 token=1.4009 | 6.13 steps/s | VRAM=31.41 Go [2026-07-18 07:44:58] step 702500/1000000 | loss=2.1692 vocab=0.7765 token=1.5948 | 6.14 steps/s | VRAM=27.93 Go [2026-07-18 07:45:06] step 702550/1000000 | loss=2.2483 vocab=0.7443 token=1.4094 | 6.17 steps/s | VRAM=31.78 Go [2026-07-18 07:45:15] step 702600/1000000 | loss=2.1415 vocab=0.7830 token=1.4421 | 6.16 steps/s | VRAM=31.34 Go [2026-07-18 07:45:23] step 702650/1000000 | loss=2.1378 vocab=1.0015 token=1.3279 | 6.13 steps/s | VRAM=29.48 Go [2026-07-18 07:45:31] step 702700/1000000 | loss=2.1435 vocab=0.4920 token=0.9387 | 6.14 steps/s | VRAM=29.32 Go [2026-07-18 07:45:39] step 702750/1000000 | loss=2.2727 vocab=0.8223 token=1.6755 | 6.14 steps/s | VRAM=31.80 Go [2026-07-18 07:45:47] step 702800/1000000 | loss=2.2257 vocab=1.2510 token=0.8699 | 6.12 steps/s | VRAM=29.84 Go [2026-07-18 07:45:55] step 702850/1000000 | loss=2.1273 vocab=0.8800 token=0.9985 | 6.13 steps/s | VRAM=25.74 Go [2026-07-18 07:46:03] step 702900/1000000 | loss=2.2103 vocab=0.7320 token=1.3460 | 6.13 steps/s | VRAM=28.58 Go [2026-07-18 07:46:12] step 702950/1000000 | loss=2.1899 vocab=1.1359 token=1.6864 | 6.13 steps/s | VRAM=28.72 Go [2026-07-18 07:46:20] step 703000/1000000 | loss=2.2265 vocab=0.9487 token=1.4252 | 6.13 steps/s | VRAM=31.63 Go [2026-07-18 07:46:28] step 703050/1000000 | loss=2.1169 vocab=0.7294 token=1.4565 | 6.16 steps/s | VRAM=31.13 Go [2026-07-18 07:46:36] step 703100/1000000 | loss=2.1369 vocab=0.9586 token=1.6901 | 6.15 steps/s | VRAM=26.44 Go [2026-07-18 07:46:44] step 703150/1000000 | loss=2.2440 vocab=0.9884 token=1.3810 | 6.14 steps/s | VRAM=29.18 Go [2026-07-18 07:46:52] step 703200/1000000 | loss=2.1641 vocab=1.0781 token=1.4093 | 6.17 steps/s | VRAM=28.14 Go [2026-07-18 07:47:00] step 703250/1000000 | loss=2.1540 vocab=0.5257 token=1.1865 | 6.14 steps/s | VRAM=28.79 Go [2026-07-18 07:47:09] step 703300/1000000 | loss=2.3924 vocab=0.6008 token=1.2042 | 6.14 steps/s | VRAM=26.48 Go [2026-07-18 07:47:17] step 703350/1000000 | loss=2.2698 vocab=0.9612 token=1.3378 | 6.12 steps/s | VRAM=30.01 Go [2026-07-18 07:47:25] step 703400/1000000 | loss=2.1258 vocab=1.0269 token=1.6566 | 6.14 steps/s | VRAM=32.79 Go [2026-07-18 07:47:33] step 703450/1000000 | loss=2.2079 vocab=1.1271 token=1.5134 | 6.15 steps/s | VRAM=28.78 Go [2026-07-18 07:47:41] step 703500/1000000 | loss=2.1185 vocab=0.9950 token=1.3994 | 6.14 steps/s | VRAM=32.05 Go [2026-07-18 07:47:49] step 703550/1000000 | loss=2.2494 vocab=0.7654 token=1.1738 | 6.17 steps/s | VRAM=27.71 Go [2026-07-18 07:47:57] step 703600/1000000 | loss=2.2133 vocab=0.7649 token=1.2362 | 6.14 steps/s | VRAM=28.44 Go [2026-07-18 07:48:06] step 703650/1000000 | loss=2.1536 vocab=0.5380 token=1.0253 | 6.14 steps/s | VRAM=28.74 Go [2026-07-18 07:48:14] step 703700/1000000 | loss=2.1423 vocab=0.8066 token=1.2700 | 6.12 steps/s | VRAM=28.19 Go [2026-07-18 07:48:22] step 703750/1000000 | loss=2.1848 vocab=0.8231 token=1.3963 | 6.14 steps/s | VRAM=32.67 Go [2026-07-18 07:48:30] step 703800/1000000 | loss=2.0979 vocab=0.9136 token=0.9303 | 6.12 steps/s | VRAM=31.77 Go [2026-07-18 07:48:38] step 703850/1000000 | loss=2.1446 vocab=1.1211 token=1.4213 | 6.14 steps/s | VRAM=24.59 Go [2026-07-18 07:48:46] step 703900/1000000 | loss=2.2265 vocab=1.0692 token=1.8850 | 6.14 steps/s | VRAM=30.62 Go [2026-07-18 07:48:55] step 703950/1000000 | loss=2.2001 vocab=0.3617 token=0.8972 | 6.13 steps/s | VRAM=27.79 Go [2026-07-18 07:49:03] step 704000/1000000 | loss=2.1312 vocab=0.4644 token=0.7948 | 6.18 steps/s | VRAM=32.67 Go [2026-07-18 07:49:11] step 704050/1000000 | loss=2.2044 vocab=0.9330 token=1.5185 | 6.17 steps/s | VRAM=27.52 Go [2026-07-18 07:49:19] step 704100/1000000 | loss=2.1952 vocab=1.0174 token=1.1329 | 6.11 steps/s | VRAM=28.27 Go [2026-07-18 07:49:27] step 704150/1000000 | loss=2.2031 vocab=0.7029 token=1.4682 | 6.12 steps/s | VRAM=29.61 Go [2026-07-18 07:49:35] step 704200/1000000 | loss=2.1493 vocab=0.4230 token=0.9340 | 6.13 steps/s | VRAM=26.62 Go [2026-07-18 07:49:43] step 704250/1000000 | loss=2.2193 vocab=0.7187 token=0.9975 | 6.15 steps/s | VRAM=28.27 Go [2026-07-18 07:49:51] step 704300/1000000 | loss=2.2324 vocab=0.9728 token=1.4253 | 6.17 steps/s | VRAM=31.06 Go [2026-07-18 07:50:00] step 704350/1000000 | loss=2.2589 vocab=0.7460 token=1.3762 | 6.15 steps/s | VRAM=32.22 Go [2026-07-18 07:50:08] step 704400/1000000 | loss=2.1773 vocab=0.7448 token=1.3435 | 6.16 steps/s | VRAM=25.89 Go [2026-07-18 07:50:16] step 704450/1000000 | loss=2.2668 vocab=0.9205 token=1.5780 | 6.14 steps/s | VRAM=26.06 Go [2026-07-18 07:50:24] step 704500/1000000 | loss=2.1518 vocab=1.1504 token=1.2617 | 6.14 steps/s | VRAM=25.05 Go [2026-07-18 07:50:32] step 704550/1000000 | loss=2.2653 vocab=0.9028 token=1.6093 | 6.12 steps/s | VRAM=28.25 Go [2026-07-18 07:50:40] step 704600/1000000 | loss=2.2685 vocab=0.4882 token=1.3209 | 6.15 steps/s | VRAM=28.68 Go [2026-07-18 07:50:48] step 704650/1000000 | loss=2.2886 vocab=0.6754 token=0.8688 | 6.13 steps/s | VRAM=29.19 Go [2026-07-18 07:50:57] step 704700/1000000 | loss=2.0916 vocab=0.6505 token=1.3190 | 6.16 steps/s | VRAM=32.17 Go [2026-07-18 07:51:05] step 704750/1000000 | loss=2.1301 vocab=0.8744 token=0.9220 | 6.14 steps/s | VRAM=29.73 Go [2026-07-18 07:51:13] step 704800/1000000 | loss=2.2396 vocab=0.7355 token=1.0902 | 6.16 steps/s | VRAM=30.90 Go [2026-07-18 07:51:21] step 704850/1000000 | loss=2.1597 vocab=1.0867 token=1.3141 | 6.15 steps/s | VRAM=25.27 Go [2026-07-18 07:51:29] step 704900/1000000 | loss=2.2320 vocab=0.8166 token=1.3011 | 6.13 steps/s | VRAM=26.66 Go [2026-07-18 07:51:37] step 704950/1000000 | loss=2.1606 vocab=0.7902 token=1.1248 | 6.15 steps/s | VRAM=28.46 Go [2026-07-18 07:51:45] step 705000/1000000 | loss=2.2546 vocab=0.4117 token=1.2072 | 6.15 steps/s | VRAM=32.09 Go [2026-07-18 07:51:54] step 705050/1000000 | loss=2.1191 vocab=0.4857 token=1.2558 | 6.14 steps/s | VRAM=31.98 Go [2026-07-18 07:52:02] step 705100/1000000 | loss=2.2649 vocab=0.8053 token=1.0185 | 6.15 steps/s | VRAM=28.67 Go [2026-07-18 07:52:10] step 705150/1000000 | loss=2.2084 vocab=1.3490 token=1.0479 | 6.15 steps/s | VRAM=26.50 Go [2026-07-18 07:52:18] step 705200/1000000 | loss=2.1948 vocab=1.2334 token=1.2085 | 6.15 steps/s | VRAM=28.60 Go [2026-07-18 07:52:26] step 705250/1000000 | loss=2.2546 vocab=0.8869 token=1.5010 | 6.16 steps/s | VRAM=28.73 Go [2026-07-18 07:52:34] step 705300/1000000 | loss=2.1861 vocab=1.2482 token=1.5315 | 6.18 steps/s | VRAM=32.05 Go [2026-07-18 07:52:42] step 705350/1000000 | loss=2.1520 vocab=1.1513 token=1.4964 | 6.16 steps/s | VRAM=32.14 Go [2026-07-18 07:52:50] step 705400/1000000 | loss=2.0926 vocab=0.9167 token=1.2870 | 6.13 steps/s | VRAM=28.74 Go [2026-07-18 07:52:59] step 705450/1000000 | loss=2.2793 vocab=0.8586 token=1.1762 | 6.14 steps/s | VRAM=29.10 Go [2026-07-18 07:53:07] step 705500/1000000 | loss=2.2329 vocab=0.6846 token=1.0521 | 6.14 steps/s | VRAM=32.24 Go [2026-07-18 07:53:15] step 705550/1000000 | loss=2.2016 vocab=1.6447 token=1.4684 | 6.13 steps/s | VRAM=26.06 Go [2026-07-18 07:53:23] step 705600/1000000 | loss=2.2056 vocab=1.0125 token=1.3424 | 6.14 steps/s | VRAM=29.07 Go [2026-07-18 07:53:31] step 705650/1000000 | loss=2.1302 vocab=1.1034 token=1.6361 | 6.13 steps/s | VRAM=31.08 Go [2026-07-18 07:53:39] step 705700/1000000 | loss=2.1676 vocab=0.8750 token=1.5520 | 6.12 steps/s | VRAM=26.06 Go [2026-07-18 07:53:47] step 705750/1000000 | loss=2.2157 vocab=1.1677 token=1.0688 | 6.14 steps/s | VRAM=29.93 Go [2026-07-18 07:53:56] step 705800/1000000 | loss=2.1242 vocab=0.7006 token=1.0542 | 6.16 steps/s | VRAM=30.74 Go [2026-07-18 07:54:04] step 705850/1000000 | loss=2.2064 vocab=0.6032 token=1.2777 | 6.14 steps/s | VRAM=32.53 Go [2026-07-18 07:54:12] step 705900/1000000 | loss=2.2194 vocab=0.8939 token=1.2279 | 6.13 steps/s | VRAM=29.12 Go [2026-07-18 07:54:20] step 705950/1000000 | loss=2.1560 vocab=0.7703 token=1.1411 | 6.14 steps/s | VRAM=29.63 Go [2026-07-18 07:54:28] step 706000/1000000 | loss=2.1691 vocab=0.8837 token=1.2674 | 6.16 steps/s | VRAM=31.20 Go [2026-07-18 07:54:36] step 706050/1000000 | loss=2.1543 vocab=1.3874 token=1.3218 | 6.11 steps/s | VRAM=28.63 Go [2026-07-18 07:54:44] step 706100/1000000 | loss=2.2453 vocab=1.1466 token=1.4730 | 6.17 steps/s | VRAM=30.87 Go [2026-07-18 07:54:53] step 706150/1000000 | loss=2.1270 vocab=0.8040 token=1.1504 | 6.15 steps/s | VRAM=31.76 Go [2026-07-18 07:55:01] step 706200/1000000 | loss=2.2496 vocab=0.9282 token=1.8469 | 6.13 steps/s | VRAM=28.97 Go [2026-07-18 07:55:09] step 706250/1000000 | loss=2.2109 vocab=0.9035 token=1.6820 | 6.14 steps/s | VRAM=29.82 Go [2026-07-18 07:55:17] step 706300/1000000 | loss=2.2068 vocab=1.2255 token=1.6856 | 6.14 steps/s | VRAM=28.54 Go [2026-07-18 07:55:25] step 706350/1000000 | loss=2.2511 vocab=1.3644 token=1.2229 | 6.13 steps/s | VRAM=29.18 Go [2026-07-18 07:55:33] step 706400/1000000 | loss=2.2243 vocab=0.9490 token=1.8360 | 6.15 steps/s | VRAM=31.26 Go [2026-07-18 07:55:41] step 706450/1000000 | loss=2.1517 vocab=0.7208 token=1.4232 | 6.11 steps/s | VRAM=29.04 Go [2026-07-18 07:55:50] step 706500/1000000 | loss=2.1600 vocab=0.6260 token=1.2094 | 6.12 steps/s | VRAM=26.62 Go [2026-07-18 07:55:58] step 706550/1000000 | loss=2.1446 vocab=0.8226 token=1.5215 | 6.15 steps/s | VRAM=25.76 Go [2026-07-18 07:56:06] step 706600/1000000 | loss=2.1438 vocab=1.0541 token=1.3930 | 6.14 steps/s | VRAM=28.83 Go [2026-07-18 07:56:14] step 706650/1000000 | loss=2.2287 vocab=0.7613 token=1.6461 | 6.15 steps/s | VRAM=32.61 Go [2026-07-18 07:56:22] step 706700/1000000 | loss=2.2728 vocab=0.4967 token=1.0592 | 6.15 steps/s | VRAM=32.26 Go [2026-07-18 07:56:30] step 706750/1000000 | loss=2.1752 vocab=1.2215 token=1.6527 | 6.14 steps/s | VRAM=32.15 Go [2026-07-18 07:56:38] step 706800/1000000 | loss=2.2099 vocab=0.7295 token=1.0606 | 6.17 steps/s | VRAM=31.47 Go [2026-07-18 07:56:47] step 706850/1000000 | loss=2.1868 vocab=1.0563 token=1.1005 | 6.12 steps/s | VRAM=28.86 Go [2026-07-18 07:56:55] step 706900/1000000 | loss=2.2909 vocab=0.7328 token=1.1513 | 6.16 steps/s | VRAM=27.99 Go [2026-07-18 07:57:03] step 706950/1000000 | loss=2.1609 vocab=0.6690 token=1.1631 | 6.12 steps/s | VRAM=32.44 Go [2026-07-18 07:57:11] step 707000/1000000 | loss=2.2264 vocab=0.3972 token=1.0501 | 6.16 steps/s | VRAM=24.61 Go [2026-07-18 07:57:19] step 707050/1000000 | loss=2.1088 vocab=0.8790 token=1.1135 | 6.12 steps/s | VRAM=32.13 Go [2026-07-18 07:57:27] step 707100/1000000 | loss=2.1420 vocab=0.7141 token=1.5256 | 6.13 steps/s | VRAM=27.94 Go [2026-07-18 07:57:35] step 707150/1000000 | loss=2.1681 vocab=1.1628 token=1.8565 | 6.14 steps/s | VRAM=29.13 Go [2026-07-18 07:57:44] step 707200/1000000 | loss=2.2819 vocab=0.6074 token=1.4271 | 6.13 steps/s | VRAM=31.22 Go [2026-07-18 07:57:52] step 707250/1000000 | loss=1.9922 vocab=0.6571 token=1.3339 | 6.10 steps/s | VRAM=28.27 Go [2026-07-18 07:58:00] step 707300/1000000 | loss=2.1704 vocab=0.9460 token=1.5978 | 6.12 steps/s | VRAM=25.88 Go [2026-07-18 07:58:08] step 707350/1000000 | loss=2.1718 vocab=0.8411 token=1.4458 | 6.15 steps/s | VRAM=29.71 Go [2026-07-18 07:58:16] step 707400/1000000 | loss=2.0753 vocab=1.3176 token=1.2293 | 6.13 steps/s | VRAM=31.74 Go [2026-07-18 07:58:24] step 707450/1000000 | loss=2.2284 vocab=1.1058 token=1.4635 | 6.14 steps/s | VRAM=28.36 Go [2026-07-18 07:58:33] step 707500/1000000 | loss=2.1309 vocab=0.7124 token=1.2611 | 6.12 steps/s | VRAM=25.82 Go [2026-07-18 07:58:41] step 707550/1000000 | loss=2.2182 vocab=0.6659 token=1.2505 | 6.16 steps/s | VRAM=32.67 Go [2026-07-18 07:58:49] step 707600/1000000 | loss=2.1400 vocab=0.9556 token=1.3128 | 6.15 steps/s | VRAM=29.33 Go [2026-07-18 07:58:57] step 707650/1000000 | loss=2.2107 vocab=1.0210 token=1.2419 | 6.12 steps/s | VRAM=32.65 Go [2026-07-18 07:59:05] step 707700/1000000 | loss=2.1584 vocab=0.7145 token=1.0424 | 6.10 steps/s | VRAM=28.93 Go [2026-07-18 07:59:13] step 707750/1000000 | loss=2.2280 vocab=0.9432 token=1.2282 | 6.10 steps/s | VRAM=29.27 Go [2026-07-18 07:59:21] step 707800/1000000 | loss=2.2412 vocab=0.6484 token=1.3970 | 6.16 steps/s | VRAM=31.87 Go [2026-07-18 07:59:30] step 707850/1000000 | loss=2.1855 vocab=0.7403 token=1.3356 | 6.15 steps/s | VRAM=32.20 Go [2026-07-18 07:59:38] step 707900/1000000 | loss=2.2145 vocab=1.1746 token=1.5835 | 6.14 steps/s | VRAM=31.17 Go [2026-07-18 07:59:46] step 707950/1000000 | loss=2.1938 vocab=1.4391 token=1.2985 | 6.16 steps/s | VRAM=32.20 Go [2026-07-18 07:59:54] step 708000/1000000 | loss=2.0412 vocab=1.0793 token=1.3352 | 6.12 steps/s | VRAM=26.06 Go [2026-07-18 08:00:02] step 708050/1000000 | loss=2.1253 vocab=1.0233 token=1.6075 | 6.16 steps/s | VRAM=25.90 Go [2026-07-18 08:00:10] step 708100/1000000 | loss=2.0662 vocab=1.1863 token=1.5129 | 6.18 steps/s | VRAM=32.22 Go [2026-07-18 08:00:18] step 708150/1000000 | loss=2.1665 vocab=0.8928 token=1.2125 | 6.10 steps/s | VRAM=26.50 Go [2026-07-18 08:00:27] step 708200/1000000 | loss=2.1325 vocab=0.6562 token=1.1333 | 6.14 steps/s | VRAM=28.49 Go [2026-07-18 08:00:35] step 708250/1000000 | loss=2.1603 vocab=0.9519 token=1.3410 | 6.13 steps/s | VRAM=25.88 Go [2026-07-18 08:00:43] step 708300/1000000 | loss=2.2404 vocab=1.2105 token=1.1337 | 6.14 steps/s | VRAM=28.21 Go [2026-07-18 08:00:51] step 708350/1000000 | loss=2.2295 vocab=1.0164 token=1.6769 | 6.15 steps/s | VRAM=28.92 Go [2026-07-18 08:00:59] step 708400/1000000 | loss=2.1645 vocab=1.0152 token=1.2839 | 6.12 steps/s | VRAM=26.67 Go [2026-07-18 08:01:07] step 708450/1000000 | loss=2.1977 vocab=0.9121 token=1.3892 | 6.18 steps/s | VRAM=31.98 Go [2026-07-18 08:01:15] step 708500/1000000 | loss=2.2145 vocab=1.2531 token=1.6706 | 6.16 steps/s | VRAM=32.32 Go [2026-07-18 08:01:24] step 708550/1000000 | loss=2.1140 vocab=0.7771 token=1.3308 | 6.13 steps/s | VRAM=24.90 Go [2026-07-18 08:01:32] step 708600/1000000 | loss=2.1585 vocab=1.0964 token=1.0873 | 6.16 steps/s | VRAM=30.88 Go [2026-07-18 08:01:40] step 708650/1000000 | loss=2.1939 vocab=1.0593 token=1.7814 | 6.14 steps/s | VRAM=32.02 Go [2026-07-18 08:01:48] step 708700/1000000 | loss=2.3414 vocab=1.6029 token=1.4301 | 6.16 steps/s | VRAM=31.28 Go [2026-07-18 08:01:56] step 708750/1000000 | loss=2.2944 vocab=1.1221 token=1.6902 | 6.13 steps/s | VRAM=29.31 Go [2026-07-18 08:02:04] step 708800/1000000 | loss=2.3547 vocab=0.6324 token=1.2119 | 6.13 steps/s | VRAM=29.29 Go [2026-07-18 08:02:12] step 708850/1000000 | loss=2.1580 vocab=0.9068 token=1.2286 | 6.14 steps/s | VRAM=31.39 Go [2026-07-18 08:02:21] step 708900/1000000 | loss=2.2226 vocab=1.1752 token=1.1053 | 6.12 steps/s | VRAM=25.87 Go [2026-07-18 08:02:29] step 708950/1000000 | loss=2.1360 vocab=0.9400 token=1.6501 | 6.15 steps/s | VRAM=28.26 Go [2026-07-18 08:02:37] step 709000/1000000 | loss=2.0845 vocab=0.6578 token=1.5462 | 6.14 steps/s | VRAM=32.26 Go [2026-07-18 08:02:45] step 709050/1000000 | loss=2.2182 vocab=1.2209 token=1.5742 | 6.17 steps/s | VRAM=25.88 Go [2026-07-18 08:02:53] step 709100/1000000 | loss=2.2358 vocab=1.3123 token=0.9804 | 6.14 steps/s | VRAM=28.72 Go [2026-07-18 08:03:01] step 709150/1000000 | loss=2.1404 vocab=0.8889 token=1.0158 | 6.14 steps/s | VRAM=28.83 Go [2026-07-18 08:03:09] step 709200/1000000 | loss=2.1301 vocab=0.8026 token=0.7566 | 6.15 steps/s | VRAM=29.17 Go [2026-07-18 08:03:17] step 709250/1000000 | loss=2.2120 vocab=1.2495 token=2.0165 | 6.15 steps/s | VRAM=28.21 Go [2026-07-18 08:03:26] step 709300/1000000 | loss=2.1370 vocab=1.1651 token=0.8394 | 6.14 steps/s | VRAM=29.46 Go [2026-07-18 08:03:34] step 709350/1000000 | loss=2.1001 vocab=0.9739 token=1.2406 | 6.15 steps/s | VRAM=31.28 Go [2026-07-18 08:03:42] step 709400/1000000 | loss=2.1273 vocab=0.4750 token=1.1519 | 6.13 steps/s | VRAM=28.87 Go [2026-07-18 08:03:50] step 709450/1000000 | loss=2.0433 vocab=0.8604 token=1.3999 | 6.14 steps/s | VRAM=28.78 Go [2026-07-18 08:03:58] step 709500/1000000 | loss=2.2012 vocab=0.1916 token=1.0146 | 6.17 steps/s | VRAM=28.65 Go [2026-07-18 08:04:06] step 709550/1000000 | loss=2.1303 vocab=1.0274 token=1.0513 | 6.18 steps/s | VRAM=28.76 Go [2026-07-18 08:04:14] step 709600/1000000 | loss=2.1963 vocab=0.8149 token=1.2284 | 6.11 steps/s | VRAM=23.37 Go [2026-07-18 08:04:23] step 709650/1000000 | loss=2.2150 vocab=1.1058 token=1.6569 | 6.15 steps/s | VRAM=28.66 Go [2026-07-18 08:04:31] step 709700/1000000 | loss=2.1607 vocab=0.6860 token=1.3090 | 6.14 steps/s | VRAM=28.58 Go [2026-07-18 08:04:39] step 709750/1000000 | loss=2.1526 vocab=0.9013 token=1.3067 | 6.14 steps/s | VRAM=28.53 Go [2026-07-18 08:04:47] step 709800/1000000 | loss=2.1387 vocab=0.7810 token=1.2639 | 6.14 steps/s | VRAM=26.65 Go [2026-07-18 08:04:55] step 709850/1000000 | loss=2.3011 vocab=1.2519 token=1.4610 | 6.17 steps/s | VRAM=28.83 Go [2026-07-18 08:05:03] step 709900/1000000 | loss=2.2521 vocab=1.2442 token=1.3584 | 6.15 steps/s | VRAM=30.89 Go [2026-07-18 08:05:11] step 709950/1000000 | loss=2.2664 vocab=1.6395 token=1.6445 | 6.16 steps/s | VRAM=29.40 Go [2026-07-18 08:05:19] step 710000/1000000 | loss=2.1970 vocab=0.6094 token=0.9757 | 6.16 steps/s | VRAM=28.56 Go [2026-07-18 08:05:28] step 710050/1000000 | loss=2.2629 vocab=1.1713 token=1.5795 | 6.15 steps/s | VRAM=31.90 Go [2026-07-18 08:05:36] step 710100/1000000 | loss=2.2095 vocab=1.0412 token=1.0454 | 6.16 steps/s | VRAM=25.90 Go [2026-07-18 08:05:44] step 710150/1000000 | loss=2.1725 vocab=0.7047 token=1.3117 | 6.18 steps/s | VRAM=28.31 Go [2026-07-18 08:05:52] step 710200/1000000 | loss=2.3087 vocab=0.6375 token=1.0839 | 6.16 steps/s | VRAM=29.74 Go [2026-07-18 08:06:00] step 710250/1000000 | loss=2.1442 vocab=0.9016 token=0.7594 | 6.18 steps/s | VRAM=28.09 Go [2026-07-18 08:06:08] step 710300/1000000 | loss=2.2135 vocab=0.9548 token=1.2869 | 6.14 steps/s | VRAM=28.69 Go [2026-07-18 08:06:16] step 710350/1000000 | loss=2.1512 vocab=0.9515 token=1.0107 | 6.17 steps/s | VRAM=28.91 Go [2026-07-18 08:06:24] step 710400/1000000 | loss=2.2858 vocab=1.5459 token=1.4494 | 6.15 steps/s | VRAM=32.27 Go [2026-07-18 08:06:32] step 710450/1000000 | loss=2.3068 vocab=0.6841 token=0.9939 | 6.19 steps/s | VRAM=31.86 Go [2026-07-18 08:06:41] step 710500/1000000 | loss=2.1465 vocab=0.7431 token=0.7609 | 6.15 steps/s | VRAM=28.56 Go [2026-07-18 08:06:49] step 710550/1000000 | loss=2.1948 vocab=1.1128 token=1.5728 | 6.13 steps/s | VRAM=29.49 Go [2026-07-18 08:06:57] step 710600/1000000 | loss=2.1814 vocab=0.6549 token=1.0920 | 6.15 steps/s | VRAM=32.81 Go [2026-07-18 08:07:05] step 710650/1000000 | loss=2.2083 vocab=0.6865 token=1.2054 | 6.14 steps/s | VRAM=26.06 Go [2026-07-18 08:07:13] step 710700/1000000 | loss=2.2512 vocab=0.9939 token=1.3879 | 6.15 steps/s | VRAM=27.48 Go [2026-07-18 08:07:21] step 710750/1000000 | loss=2.2618 vocab=0.7809 token=1.0397 | 6.16 steps/s | VRAM=30.33 Go [2026-07-18 08:07:29] step 710800/1000000 | loss=2.1142 vocab=0.5971 token=1.1742 | 6.16 steps/s | VRAM=31.59 Go [2026-07-18 08:07:38] step 710850/1000000 | loss=2.1461 vocab=0.9254 token=1.2535 | 6.13 steps/s | VRAM=29.23 Go [2026-07-18 08:07:46] step 710900/1000000 | loss=2.2492 vocab=1.2539 token=1.3092 | 6.16 steps/s | VRAM=32.03 Go [2026-07-18 08:07:54] step 710950/1000000 | loss=2.2089 vocab=1.0992 token=1.6964 | 6.11 steps/s | VRAM=29.51 Go [2026-07-18 08:08:02] step 711000/1000000 | loss=2.2005 vocab=0.7993 token=1.1260 | 6.15 steps/s | VRAM=28.44 Go [2026-07-18 08:08:10] step 711050/1000000 | loss=2.3073 vocab=1.7517 token=1.4124 | 6.15 steps/s | VRAM=32.51 Go [2026-07-18 08:08:18] step 711100/1000000 | loss=2.2274 vocab=1.1347 token=1.1412 | 6.12 steps/s | VRAM=26.15 Go [2026-07-18 08:08:26] step 711150/1000000 | loss=2.1490 vocab=0.9745 token=1.3361 | 6.15 steps/s | VRAM=30.26 Go [2026-07-18 08:08:35] step 711200/1000000 | loss=2.2443 vocab=0.6058 token=1.1388 | 6.14 steps/s | VRAM=29.39 Go [2026-07-18 08:08:43] step 711250/1000000 | loss=2.1591 vocab=0.7049 token=1.2532 | 6.11 steps/s | VRAM=28.54 Go [2026-07-18 08:08:51] step 711300/1000000 | loss=2.2631 vocab=0.8684 token=1.2469 | 6.14 steps/s | VRAM=25.72 Go [2026-07-18 08:08:59] step 711350/1000000 | loss=2.2277 vocab=1.3886 token=1.7574 | 6.16 steps/s | VRAM=32.74 Go [2026-07-18 08:09:07] step 711400/1000000 | loss=2.1319 vocab=0.9469 token=1.3295 | 6.12 steps/s | VRAM=28.82 Go [2026-07-18 08:09:15] step 711450/1000000 | loss=2.1705 vocab=0.9647 token=1.0175 | 6.14 steps/s | VRAM=28.59 Go [2026-07-18 08:09:23] step 711500/1000000 | loss=2.1486 vocab=0.8802 token=1.9389 | 6.13 steps/s | VRAM=27.87 Go [2026-07-18 08:09:32] step 711550/1000000 | loss=2.1025 vocab=0.8745 token=1.3969 | 6.17 steps/s | VRAM=28.45 Go [2026-07-18 08:09:40] step 711600/1000000 | loss=2.1001 vocab=0.7255 token=1.1273 | 6.12 steps/s | VRAM=26.39 Go [2026-07-18 08:09:48] step 711650/1000000 | loss=2.2841 vocab=0.7992 token=1.3995 | 6.14 steps/s | VRAM=26.15 Go [2026-07-18 08:09:56] step 711700/1000000 | loss=2.1396 vocab=0.6057 token=1.2119 | 6.14 steps/s | VRAM=28.51 Go [2026-07-18 08:10:04] step 711750/1000000 | loss=2.1306 vocab=0.8556 token=1.7276 | 6.19 steps/s | VRAM=31.62 Go [2026-07-18 08:10:12] step 711800/1000000 | loss=2.2495 vocab=1.3461 token=1.3361 | 6.10 steps/s | VRAM=26.70 Go [2026-07-18 08:10:20] step 711850/1000000 | loss=2.1707 vocab=0.8072 token=1.3187 | 6.17 steps/s | VRAM=28.71 Go [2026-07-18 08:10:29] step 711900/1000000 | loss=2.1569 vocab=0.7344 token=1.2688 | 6.14 steps/s | VRAM=29.86 Go [2026-07-18 08:10:37] step 711950/1000000 | loss=2.1845 vocab=1.0294 token=1.1518 | 6.15 steps/s | VRAM=28.03 Go [2026-07-18 08:10:45] step 712000/1000000 | loss=2.1833 vocab=0.8654 token=1.1804 | 6.17 steps/s | VRAM=29.14 Go [2026-07-18 08:10:53] step 712050/1000000 | loss=2.1291 vocab=1.2887 token=1.1737 | 6.13 steps/s | VRAM=28.88 Go [2026-07-18 08:11:01] step 712100/1000000 | loss=2.2237 vocab=0.7732 token=1.2778 | 6.14 steps/s | VRAM=29.20 Go [2026-07-18 08:11:09] step 712150/1000000 | loss=2.1441 vocab=0.5618 token=1.2692 | 6.15 steps/s | VRAM=32.41 Go [2026-07-18 08:11:17] step 712200/1000000 | loss=2.2503 vocab=1.1823 token=1.1912 | 6.12 steps/s | VRAM=29.20 Go [2026-07-18 08:11:26] step 712250/1000000 | loss=2.2335 vocab=0.6184 token=1.3551 | 6.14 steps/s | VRAM=32.31 Go [2026-07-18 08:11:34] step 712300/1000000 | loss=2.2051 vocab=1.1739 token=1.5326 | 6.12 steps/s | VRAM=27.55 Go [2026-07-18 08:11:42] step 712350/1000000 | loss=2.1710 vocab=0.7146 token=1.2401 | 6.13 steps/s | VRAM=27.69 Go [2026-07-18 08:11:50] step 712400/1000000 | loss=2.2118 vocab=1.0079 token=1.2591 | 6.15 steps/s | VRAM=28.58 Go [2026-07-18 08:11:58] step 712450/1000000 | loss=2.1750 vocab=0.7623 token=1.2001 | 6.13 steps/s | VRAM=26.10 Go [2026-07-18 08:12:06] step 712500/1000000 | loss=2.2400 vocab=0.6129 token=1.1738 | 6.09 steps/s | VRAM=28.70 Go [2026-07-18 08:12:14] step 712550/1000000 | loss=2.1918 vocab=0.6075 token=1.2360 | 6.15 steps/s | VRAM=29.95 Go [2026-07-18 08:12:23] step 712600/1000000 | loss=2.3201 vocab=1.2298 token=1.0430 | 6.14 steps/s | VRAM=32.06 Go [2026-07-18 08:12:31] step 712650/1000000 | loss=2.2457 vocab=1.1583 token=1.6592 | 6.14 steps/s | VRAM=29.40 Go [2026-07-18 08:12:39] step 712700/1000000 | loss=2.1253 vocab=0.6092 token=0.8886 | 6.13 steps/s | VRAM=28.57 Go [2026-07-18 08:12:47] step 712750/1000000 | loss=2.0953 vocab=0.8346 token=0.9642 | 6.11 steps/s | VRAM=32.84 Go [2026-07-18 08:12:55] step 712800/1000000 | loss=2.1699 vocab=1.1265 token=1.8619 | 6.13 steps/s | VRAM=32.24 Go [2026-07-18 08:13:03] step 712850/1000000 | loss=2.1181 vocab=0.8493 token=1.4974 | 6.16 steps/s | VRAM=32.10 Go [2026-07-18 08:13:12] step 712900/1000000 | loss=2.2233 vocab=0.8039 token=1.4827 | 6.13 steps/s | VRAM=29.01 Go [2026-07-18 08:13:20] step 712950/1000000 | loss=2.1450 vocab=0.8040 token=1.1189 | 6.14 steps/s | VRAM=25.47 Go [2026-07-18 08:13:28] step 713000/1000000 | loss=2.2194 vocab=0.6847 token=0.6383 | 6.13 steps/s | VRAM=26.12 Go [2026-07-18 08:13:36] step 713050/1000000 | loss=2.2002 vocab=1.0608 token=1.1903 | 6.17 steps/s | VRAM=29.32 Go [2026-07-18 08:13:44] step 713100/1000000 | loss=2.1993 vocab=0.8352 token=1.1186 | 6.16 steps/s | VRAM=28.43 Go [2026-07-18 08:13:52] step 713150/1000000 | loss=2.1582 vocab=0.9715 token=1.5239 | 6.15 steps/s | VRAM=28.34 Go [2026-07-18 08:14:00] step 713200/1000000 | loss=2.1746 vocab=0.5107 token=1.1032 | 6.13 steps/s | VRAM=26.24 Go [2026-07-18 08:14:09] step 713250/1000000 | loss=2.1205 vocab=0.7136 token=0.9241 | 6.12 steps/s | VRAM=29.30 Go [2026-07-18 08:14:17] step 713300/1000000 | loss=2.2380 vocab=1.0682 token=1.7576 | 6.11 steps/s | VRAM=26.32 Go [2026-07-18 08:14:25] step 713350/1000000 | loss=2.1695 vocab=1.2200 token=1.5073 | 6.14 steps/s | VRAM=27.47 Go [2026-07-18 08:14:33] step 713400/1000000 | loss=2.1613 vocab=1.1457 token=1.6363 | 6.18 steps/s | VRAM=27.69 Go [2026-07-18 08:14:41] step 713450/1000000 | loss=2.2173 vocab=0.7375 token=1.2272 | 6.16 steps/s | VRAM=28.61 Go [2026-07-18 08:14:49] step 713500/1000000 | loss=2.2389 vocab=0.9331 token=1.0712 | 6.11 steps/s | VRAM=27.12 Go [2026-07-18 08:14:57] step 713550/1000000 | loss=2.1553 vocab=0.6959 token=1.2914 | 6.11 steps/s | VRAM=29.67 Go [2026-07-18 08:15:06] step 713600/1000000 | loss=2.2328 vocab=0.9978 token=1.1617 | 6.11 steps/s | VRAM=29.04 Go [2026-07-18 08:15:14] step 713650/1000000 | loss=2.2720 vocab=1.6257 token=1.1221 | 6.14 steps/s | VRAM=29.93 Go [2026-07-18 08:15:22] step 713700/1000000 | loss=2.2767 vocab=1.0189 token=1.3540 | 6.11 steps/s | VRAM=28.68 Go [2026-07-18 08:15:30] step 713750/1000000 | loss=2.1887 vocab=0.5507 token=1.3074 | 6.13 steps/s | VRAM=30.35 Go [2026-07-18 08:15:38] step 713800/1000000 | loss=2.2254 vocab=1.1601 token=1.8615 | 6.15 steps/s | VRAM=28.83 Go [2026-07-18 08:15:46] step 713850/1000000 | loss=2.3847 vocab=0.4782 token=1.0108 | 6.15 steps/s | VRAM=25.45 Go [2026-07-18 08:15:54] step 713900/1000000 | loss=2.2077 vocab=0.9866 token=1.7037 | 6.14 steps/s | VRAM=26.29 Go [2026-07-18 08:16:03] step 713950/1000000 | loss=2.2002 vocab=1.0989 token=1.1582 | 6.14 steps/s | VRAM=28.21 Go [2026-07-18 08:16:11] step 714000/1000000 | loss=2.1825 vocab=1.3133 token=1.3114 | 6.12 steps/s | VRAM=26.35 Go [2026-07-18 08:16:19] step 714050/1000000 | loss=2.2216 vocab=0.7995 token=1.3426 | 6.13 steps/s | VRAM=32.18 Go [2026-07-18 08:16:27] step 714100/1000000 | loss=2.1747 vocab=0.9109 token=1.4266 | 6.17 steps/s | VRAM=31.37 Go [2026-07-18 08:16:35] step 714150/1000000 | loss=2.1701 vocab=1.4888 token=1.2955 | 6.11 steps/s | VRAM=28.83 Go [2026-07-18 08:16:43] step 714200/1000000 | loss=2.2591 vocab=1.4784 token=1.8998 | 6.16 steps/s | VRAM=29.10 Go [2026-07-18 08:16:51] step 714250/1000000 | loss=2.1421 vocab=1.0663 token=1.5178 | 6.15 steps/s | VRAM=26.45 Go [2026-07-18 08:17:00] step 714300/1000000 | loss=2.2277 vocab=0.8831 token=1.6468 | 6.14 steps/s | VRAM=28.75 Go [2026-07-18 08:17:08] step 714350/1000000 | loss=2.1123 vocab=0.8782 token=1.0688 | 6.11 steps/s | VRAM=29.30 Go [2026-07-18 08:17:16] step 714400/1000000 | loss=2.1841 vocab=1.2542 token=1.9320 | 6.10 steps/s | VRAM=29.61 Go [2026-07-18 08:17:24] step 714450/1000000 | loss=2.1037 vocab=0.2458 token=0.8257 | 6.16 steps/s | VRAM=32.05 Go [2026-07-18 08:17:32] step 714500/1000000 | loss=2.1568 vocab=0.9305 token=1.3565 | 6.11 steps/s | VRAM=26.43 Go [2026-07-18 08:17:40] step 714550/1000000 | loss=2.1591 vocab=1.2109 token=1.4849 | 6.13 steps/s | VRAM=28.75 Go [2026-07-18 08:17:49] step 714600/1000000 | loss=2.2375 vocab=0.7879 token=1.2494 | 6.12 steps/s | VRAM=24.88 Go [2026-07-18 08:17:57] step 714650/1000000 | loss=2.2365 vocab=0.6917 token=1.2527 | 6.16 steps/s | VRAM=32.11 Go [2026-07-18 08:18:05] step 714700/1000000 | loss=2.2299 vocab=1.1252 token=1.5330 | 6.13 steps/s | VRAM=30.51 Go [2026-07-18 08:18:13] step 714750/1000000 | loss=2.1950 vocab=1.0392 token=1.7041 | 6.15 steps/s | VRAM=31.65 Go [2026-07-18 08:18:21] step 714800/1000000 | loss=2.3545 vocab=0.7624 token=1.0128 | 6.14 steps/s | VRAM=25.92 Go [2026-07-18 08:18:29] step 714850/1000000 | loss=2.2279 vocab=0.7864 token=1.1617 | 6.15 steps/s | VRAM=25.74 Go [2026-07-18 08:18:37] step 714900/1000000 | loss=2.2373 vocab=0.6746 token=1.1904 | 6.12 steps/s | VRAM=26.21 Go [2026-07-18 08:18:46] step 714950/1000000 | loss=2.1937 vocab=0.9618 token=1.1771 | 6.15 steps/s | VRAM=28.84 Go [2026-07-18 08:18:54] step 715000/1000000 | loss=2.1773 vocab=0.6140 token=0.8718 | 6.12 steps/s | VRAM=26.59 Go [2026-07-18 08:19:02] step 715050/1000000 | loss=2.2194 vocab=0.4061 token=1.3338 | 6.16 steps/s | VRAM=28.65 Go [2026-07-18 08:19:10] step 715100/1000000 | loss=2.1920 vocab=0.9236 token=1.5275 | 6.15 steps/s | VRAM=30.17 Go [2026-07-18 08:19:18] step 715150/1000000 | loss=2.1544 vocab=1.5823 token=1.4337 | 6.14 steps/s | VRAM=28.08 Go [2026-07-18 08:19:26] step 715200/1000000 | loss=2.1134 vocab=0.9611 token=1.1978 | 6.15 steps/s | VRAM=25.11 Go [2026-07-18 08:19:34] step 715250/1000000 | loss=2.2532 vocab=0.9175 token=1.1434 | 6.13 steps/s | VRAM=26.77 Go [2026-07-18 08:19:43] step 715300/1000000 | loss=2.3050 vocab=0.8407 token=1.4128 | 6.15 steps/s | VRAM=27.94 Go [2026-07-18 08:19:51] step 715350/1000000 | loss=2.1459 vocab=0.6705 token=1.2876 | 6.12 steps/s | VRAM=29.92 Go [2026-07-18 08:19:59] step 715400/1000000 | loss=2.1475 vocab=1.1507 token=1.4877 | 6.12 steps/s | VRAM=26.47 Go [2026-07-18 08:20:07] step 715450/1000000 | loss=2.2225 vocab=0.6952 token=1.4075 | 6.15 steps/s | VRAM=32.48 Go [2026-07-18 08:20:15] step 715500/1000000 | loss=2.1929 vocab=0.8863 token=1.3402 | 6.15 steps/s | VRAM=29.62 Go [2026-07-18 08:20:23] step 715550/1000000 | loss=2.2401 vocab=1.3842 token=1.4617 | 6.13 steps/s | VRAM=26.15 Go [2026-07-18 08:20:31] step 715600/1000000 | loss=2.2480 vocab=0.7057 token=1.1313 | 6.15 steps/s | VRAM=31.05 Go [2026-07-18 08:20:40] step 715650/1000000 | loss=2.2391 vocab=1.4017 token=1.2674 | 6.16 steps/s | VRAM=28.60 Go [2026-07-18 08:20:48] step 715700/1000000 | loss=2.0991 vocab=1.1181 token=1.2452 | 6.14 steps/s | VRAM=27.75 Go [2026-07-18 08:20:56] step 715750/1000000 | loss=2.2286 vocab=1.2589 token=1.6050 | 6.11 steps/s | VRAM=28.75 Go [2026-07-18 08:21:04] step 715800/1000000 | loss=2.2297 vocab=1.0129 token=0.8952 | 6.13 steps/s | VRAM=29.00 Go [2026-07-18 08:21:12] step 715850/1000000 | loss=2.2374 vocab=0.9040 token=1.4587 | 6.14 steps/s | VRAM=32.48 Go [2026-07-18 08:21:20] step 715900/1000000 | loss=2.3278 vocab=0.9057 token=1.1388 | 6.14 steps/s | VRAM=27.64 Go [2026-07-18 08:21:29] step 715950/1000000 | loss=2.1631 vocab=1.0190 token=1.5282 | 6.13 steps/s | VRAM=28.80 Go [2026-07-18 08:21:37] step 716000/1000000 | loss=2.1966 vocab=0.7808 token=1.4188 | 6.14 steps/s | VRAM=30.94 Go [2026-07-18 08:21:45] step 716050/1000000 | loss=2.2315 vocab=1.3049 token=1.7598 | 6.11 steps/s | VRAM=22.76 Go [2026-07-18 08:21:53] step 716100/1000000 | loss=2.2409 vocab=0.9632 token=1.1157 | 6.14 steps/s | VRAM=32.90 Go [2026-07-18 08:22:01] step 716150/1000000 | loss=2.1377 vocab=0.7153 token=1.1849 | 6.14 steps/s | VRAM=31.34 Go [2026-07-18 08:22:09] step 716200/1000000 | loss=2.3156 vocab=0.6469 token=1.1049 | 6.12 steps/s | VRAM=25.91 Go [2026-07-18 08:22:17] step 716250/1000000 | loss=2.1133 vocab=0.4194 token=1.0711 | 6.14 steps/s | VRAM=32.38 Go [2026-07-18 08:22:26] step 716300/1000000 | loss=2.1691 vocab=0.8232 token=1.5530 | 6.12 steps/s | VRAM=31.78 Go [2026-07-18 08:22:34] step 716350/1000000 | loss=2.2561 vocab=0.7127 token=0.8173 | 6.16 steps/s | VRAM=31.49 Go [2026-07-18 08:22:42] step 716400/1000000 | loss=2.1596 vocab=0.7597 token=1.5552 | 6.13 steps/s | VRAM=31.75 Go [2026-07-18 08:22:50] step 716450/1000000 | loss=2.0544 vocab=0.5504 token=1.2924 | 6.15 steps/s | VRAM=31.43 Go [2026-07-18 08:22:58] step 716500/1000000 | loss=2.1421 vocab=0.7079 token=1.1075 | 6.15 steps/s | VRAM=28.06 Go [2026-07-18 08:23:06] step 716550/1000000 | loss=2.1935 vocab=0.6007 token=0.9896 | 6.12 steps/s | VRAM=32.47 Go [2026-07-18 08:23:14] step 716600/1000000 | loss=2.2836 vocab=0.7232 token=1.1777 | 6.16 steps/s | VRAM=32.16 Go [2026-07-18 08:23:23] step 716650/1000000 | loss=2.2814 vocab=1.6637 token=1.6996 | 6.11 steps/s | VRAM=29.17 Go [2026-07-18 08:23:31] step 716700/1000000 | loss=2.2862 vocab=0.6775 token=1.2379 | 6.11 steps/s | VRAM=28.87 Go [2026-07-18 08:23:39] step 716750/1000000 | loss=2.2287 vocab=0.3528 token=0.9950 | 6.15 steps/s | VRAM=25.75 Go [2026-07-18 08:23:47] step 716800/1000000 | loss=2.1675 vocab=0.7599 token=1.2021 | 6.18 steps/s | VRAM=24.90 Go [2026-07-18 08:23:55] step 716850/1000000 | loss=2.1223 vocab=1.0172 token=1.4467 | 6.13 steps/s | VRAM=28.69 Go [2026-07-18 08:24:03] step 716900/1000000 | loss=2.1355 vocab=1.1258 token=1.4964 | 6.15 steps/s | VRAM=28.74 Go [2026-07-18 08:24:11] step 716950/1000000 | loss=2.1119 vocab=1.1584 token=0.9741 | 6.15 steps/s | VRAM=28.33 Go [2026-07-18 08:24:20] step 717000/1000000 | loss=2.2136 vocab=0.8846 token=1.6740 | 6.12 steps/s | VRAM=28.09 Go [2026-07-18 08:24:28] step 717050/1000000 | loss=2.1719 vocab=1.0717 token=1.6612 | 6.16 steps/s | VRAM=32.33 Go [2026-07-18 08:24:36] step 717100/1000000 | loss=2.1519 vocab=1.4360 token=1.5463 | 6.14 steps/s | VRAM=31.10 Go [2026-07-18 08:24:44] step 717150/1000000 | loss=2.1972 vocab=0.7943 token=1.1005 | 6.15 steps/s | VRAM=32.04 Go [2026-07-18 08:24:52] step 717200/1000000 | loss=2.1237 vocab=1.1021 token=1.2993 | 6.10 steps/s | VRAM=24.82 Go [2026-07-18 08:25:00] step 717250/1000000 | loss=2.1249 vocab=1.3403 token=1.3377 | 6.12 steps/s | VRAM=32.55 Go [2026-07-18 08:25:09] step 717300/1000000 | loss=2.0850 vocab=0.7666 token=1.3151 | 6.12 steps/s | VRAM=28.16 Go [2026-07-18 08:25:17] step 717350/1000000 | loss=2.2676 vocab=0.6332 token=1.5998 | 6.10 steps/s | VRAM=32.21 Go [2026-07-18 08:25:25] step 717400/1000000 | loss=2.1880 vocab=1.0216 token=1.1878 | 6.11 steps/s | VRAM=28.91 Go [2026-07-18 08:25:33] step 717450/1000000 | loss=2.3663 vocab=0.6224 token=0.8404 | 6.10 steps/s | VRAM=26.34 Go [2026-07-18 08:25:41] step 717500/1000000 | loss=2.2146 vocab=1.5028 token=1.8281 | 6.15 steps/s | VRAM=28.25 Go [2026-07-18 08:25:49] step 717550/1000000 | loss=2.1597 vocab=1.1942 token=1.2276 | 6.13 steps/s | VRAM=32.47 Go [2026-07-18 08:25:58] step 717600/1000000 | loss=2.1765 vocab=0.9177 token=1.2902 | 6.11 steps/s | VRAM=32.19 Go [2026-07-18 08:26:06] step 717650/1000000 | loss=2.1992 vocab=0.6950 token=1.1063 | 6.13 steps/s | VRAM=29.13 Go [2026-07-18 08:26:14] step 717700/1000000 | loss=2.1410 vocab=0.6297 token=1.2759 | 6.14 steps/s | VRAM=32.02 Go [2026-07-18 08:26:22] step 717750/1000000 | loss=2.2071 vocab=1.4187 token=1.0713 | 6.14 steps/s | VRAM=32.75 Go [2026-07-18 08:26:30] step 717800/1000000 | loss=2.2835 vocab=1.2941 token=1.9004 | 6.09 steps/s | VRAM=26.43 Go [2026-07-18 08:26:38] step 717850/1000000 | loss=2.1941 vocab=0.5223 token=1.0620 | 6.13 steps/s | VRAM=29.06 Go [2026-07-18 08:26:47] step 717900/1000000 | loss=2.2600 vocab=0.7322 token=1.3929 | 6.13 steps/s | VRAM=25.43 Go [2026-07-18 08:26:55] step 717950/1000000 | loss=2.2557 vocab=0.8748 token=1.3079 | 6.13 steps/s | VRAM=28.29 Go [2026-07-18 08:27:03] step 718000/1000000 | loss=2.1292 vocab=0.5158 token=1.4288 | 6.13 steps/s | VRAM=32.21 Go [2026-07-18 08:27:11] step 718050/1000000 | loss=2.2163 vocab=1.1499 token=1.6687 | 6.10 steps/s | VRAM=29.28 Go [2026-07-18 08:27:19] step 718100/1000000 | loss=2.1517 vocab=0.8053 token=1.6332 | 6.12 steps/s | VRAM=28.80 Go [2026-07-18 08:27:27] step 718150/1000000 | loss=2.1126 vocab=0.8564 token=0.7735 | 6.11 steps/s | VRAM=26.03 Go [2026-07-18 08:27:36] step 718200/1000000 | loss=2.1236 vocab=1.1465 token=1.2942 | 6.10 steps/s | VRAM=26.00 Go [2026-07-18 08:27:44] step 718250/1000000 | loss=2.3100 vocab=1.0881 token=1.5619 | 6.10 steps/s | VRAM=29.16 Go [2026-07-18 08:27:52] step 718300/1000000 | loss=2.1500 vocab=0.8050 token=1.0746 | 6.12 steps/s | VRAM=26.55 Go [2026-07-18 08:28:00] step 718350/1000000 | loss=2.1753 vocab=0.8639 token=1.3054 | 6.13 steps/s | VRAM=28.84 Go [2026-07-18 08:28:08] step 718400/1000000 | loss=2.0898 vocab=0.6105 token=1.2544 | 6.14 steps/s | VRAM=31.45 Go [2026-07-18 08:28:16] step 718450/1000000 | loss=2.2492 vocab=1.5745 token=1.7886 | 6.16 steps/s | VRAM=32.03 Go [2026-07-18 08:28:24] step 718500/1000000 | loss=2.1994 vocab=0.5386 token=1.2750 | 6.15 steps/s | VRAM=28.32 Go [2026-07-18 08:28:33] step 718550/1000000 | loss=2.2756 vocab=0.9227 token=1.7206 | 6.15 steps/s | VRAM=32.28 Go [2026-07-18 08:28:41] step 718600/1000000 | loss=2.1932 vocab=0.6639 token=1.0360 | 6.10 steps/s | VRAM=28.48 Go [2026-07-18 08:28:49] step 718650/1000000 | loss=2.0844 vocab=0.5366 token=1.0041 | 6.13 steps/s | VRAM=32.47 Go [2026-07-18 08:28:57] step 718700/1000000 | loss=2.1342 vocab=1.0306 token=1.1730 | 6.13 steps/s | VRAM=31.94 Go [2026-07-18 08:29:05] step 718750/1000000 | loss=2.2186 vocab=0.7539 token=1.6019 | 6.12 steps/s | VRAM=26.04 Go [2026-07-18 08:29:13] step 718800/1000000 | loss=2.1826 vocab=0.7586 token=0.8977 | 6.14 steps/s | VRAM=29.54 Go [2026-07-18 08:29:22] step 718850/1000000 | loss=2.1401 vocab=0.9511 token=1.1366 | 6.12 steps/s | VRAM=28.88 Go [2026-07-18 08:29:30] step 718900/1000000 | loss=2.1373 vocab=1.4798 token=1.3220 | 6.10 steps/s | VRAM=28.77 Go [2026-07-18 08:29:38] step 718950/1000000 | loss=2.1464 vocab=0.8715 token=1.5339 | 6.13 steps/s | VRAM=28.68 Go [2026-07-18 08:29:46] step 719000/1000000 | loss=2.1551 vocab=0.8658 token=1.5313 | 6.11 steps/s | VRAM=26.29 Go [2026-07-18 08:29:54] step 719050/1000000 | loss=2.1747 vocab=1.0296 token=1.5016 | 6.12 steps/s | VRAM=28.87 Go [2026-07-18 08:30:02] step 719100/1000000 | loss=2.2131 vocab=0.8506 token=1.2463 | 6.14 steps/s | VRAM=32.34 Go [2026-07-18 08:30:11] step 719150/1000000 | loss=2.1583 vocab=0.7010 token=1.2377 | 6.15 steps/s | VRAM=27.53 Go [2026-07-18 08:30:19] step 719200/1000000 | loss=2.1375 vocab=1.0646 token=1.6434 | 6.14 steps/s | VRAM=28.18 Go [2026-07-18 08:30:27] step 719250/1000000 | loss=2.2472 vocab=1.1131 token=1.4007 | 6.12 steps/s | VRAM=25.63 Go [2026-07-18 08:30:35] step 719300/1000000 | loss=2.1510 vocab=0.5779 token=1.0436 | 6.12 steps/s | VRAM=31.92 Go [2026-07-18 08:30:43] step 719350/1000000 | loss=2.2401 vocab=1.0559 token=1.2241 | 6.12 steps/s | VRAM=31.14 Go [2026-07-18 08:30:51] step 719400/1000000 | loss=2.2333 vocab=0.9543 token=1.4068 | 6.14 steps/s | VRAM=31.43 Go [2026-07-18 08:31:00] step 719450/1000000 | loss=2.2211 vocab=0.7077 token=1.4660 | 6.13 steps/s | VRAM=28.43 Go [2026-07-18 08:31:08] step 719500/1000000 | loss=2.1062 vocab=1.0203 token=1.4354 | 6.13 steps/s | VRAM=32.41 Go [2026-07-18 08:31:16] step 719550/1000000 | loss=2.2554 vocab=0.8729 token=1.2105 | 6.14 steps/s | VRAM=31.73 Go [2026-07-18 08:31:24] step 719600/1000000 | loss=2.0772 vocab=0.8714 token=1.3141 | 6.08 steps/s | VRAM=30.04 Go [2026-07-18 08:31:32] step 719650/1000000 | loss=2.2239 vocab=0.9080 token=1.4641 | 6.09 steps/s | VRAM=29.19 Go [2026-07-18 08:31:40] step 719700/1000000 | loss=2.1271 vocab=0.8974 token=1.4224 | 6.15 steps/s | VRAM=32.28 Go [2026-07-18 08:31:49] step 719750/1000000 | loss=2.3384 vocab=1.0087 token=1.4222 | 6.13 steps/s | VRAM=25.87 Go [2026-07-18 08:31:57] step 719800/1000000 | loss=2.1459 vocab=1.1533 token=1.5930 | 6.14 steps/s | VRAM=24.93 Go [2026-07-18 08:32:05] step 719850/1000000 | loss=2.2349 vocab=0.9970 token=1.4434 | 6.14 steps/s | VRAM=28.75 Go [2026-07-18 08:32:13] step 719900/1000000 | loss=2.2387 vocab=0.9166 token=1.3275 | 6.12 steps/s | VRAM=29.11 Go [2026-07-18 08:32:21] step 719950/1000000 | loss=2.1268 vocab=0.6206 token=0.9871 | 6.10 steps/s | VRAM=26.05 Go [2026-07-18 08:32:29] step 720000/1000000 | loss=2.2692 vocab=0.9119 token=1.6433 | 6.13 steps/s | VRAM=29.24 Go [2026-07-18 08:32:37] step 720050/1000000 | loss=2.2102 vocab=1.9086 token=1.2035 | 6.15 steps/s | VRAM=30.30 Go [2026-07-18 08:32:46] step 720100/1000000 | loss=2.2948 vocab=0.7777 token=1.2111 | 6.10 steps/s | VRAM=28.88 Go [2026-07-18 08:32:54] step 720150/1000000 | loss=2.1235 vocab=0.6526 token=1.0412 | 6.11 steps/s | VRAM=24.26 Go [2026-07-18 08:33:02] step 720200/1000000 | loss=2.2204 vocab=1.1543 token=1.1852 | 6.11 steps/s | VRAM=29.24 Go [2026-07-18 08:33:10] step 720250/1000000 | loss=2.1963 vocab=1.0395 token=1.5500 | 6.15 steps/s | VRAM=31.41 Go [2026-07-18 08:33:18] step 720300/1000000 | loss=2.2770 vocab=0.3100 token=0.9904 | 6.14 steps/s | VRAM=29.23 Go [2026-07-18 08:33:27] step 720350/1000000 | loss=2.1486 vocab=0.9400 token=1.5846 | 6.12 steps/s | VRAM=26.63 Go [2026-07-18 08:33:35] step 720400/1000000 | loss=2.1697 vocab=0.7222 token=1.5444 | 6.12 steps/s | VRAM=28.26 Go [2026-07-18 08:33:43] step 720450/1000000 | loss=2.1308 vocab=0.7002 token=0.7196 | 6.09 steps/s | VRAM=26.73 Go [2026-07-18 08:33:51] step 720500/1000000 | loss=2.2266 vocab=0.7828 token=1.1288 | 6.10 steps/s | VRAM=29.02 Go [2026-07-18 08:33:59] step 720550/1000000 | loss=2.2199 vocab=0.7373 token=1.2118 | 6.11 steps/s | VRAM=31.89 Go [2026-07-18 08:34:07] step 720600/1000000 | loss=2.2243 vocab=0.7849 token=1.6427 | 6.11 steps/s | VRAM=25.79 Go [2026-07-18 08:34:16] step 720650/1000000 | loss=2.2832 vocab=0.8021 token=1.1307 | 6.11 steps/s | VRAM=25.93 Go [2026-07-18 08:34:24] step 720700/1000000 | loss=2.1653 vocab=1.0676 token=1.3425 | 6.15 steps/s | VRAM=32.32 Go [2026-07-18 08:34:32] step 720750/1000000 | loss=2.2243 vocab=1.0179 token=1.3754 | 6.11 steps/s | VRAM=28.38 Go [2026-07-18 08:34:40] step 720800/1000000 | loss=2.1474 vocab=1.2212 token=1.5601 | 6.14 steps/s | VRAM=25.08 Go [2026-07-18 08:34:48] step 720850/1000000 | loss=2.2241 vocab=0.6445 token=1.4471 | 6.11 steps/s | VRAM=29.58 Go [2026-07-18 08:34:56] step 720900/1000000 | loss=2.2055 vocab=0.6194 token=1.2244 | 6.11 steps/s | VRAM=31.99 Go [2026-07-18 08:35:05] step 720950/1000000 | loss=2.1319 vocab=0.9007 token=1.2167 | 6.17 steps/s | VRAM=26.37 Go [2026-07-18 08:35:13] step 721000/1000000 | loss=2.1609 vocab=0.8945 token=1.4692 | 6.14 steps/s | VRAM=29.33 Go [2026-07-18 08:35:21] step 721050/1000000 | loss=2.1823 vocab=0.9764 token=1.3830 | 6.17 steps/s | VRAM=25.65 Go [2026-07-18 08:35:29] step 721100/1000000 | loss=2.1792 vocab=0.5422 token=1.1015 | 6.15 steps/s | VRAM=28.62 Go [2026-07-18 08:35:37] step 721150/1000000 | loss=2.2045 vocab=0.3705 token=0.9915 | 6.13 steps/s | VRAM=25.36 Go [2026-07-18 08:35:45] step 721200/1000000 | loss=2.1443 vocab=1.0667 token=1.6417 | 6.14 steps/s | VRAM=28.26 Go [2026-07-18 08:35:53] step 721250/1000000 | loss=2.2302 vocab=0.7941 token=1.5132 | 6.15 steps/s | VRAM=32.48 Go [2026-07-18 08:36:02] step 721300/1000000 | loss=2.2025 vocab=0.9558 token=1.2565 | 6.14 steps/s | VRAM=26.36 Go [2026-07-18 08:36:10] step 721350/1000000 | loss=2.1858 vocab=0.5510 token=1.4237 | 6.17 steps/s | VRAM=27.65 Go [2026-07-18 08:36:18] step 721400/1000000 | loss=2.1716 vocab=0.8307 token=1.7128 | 6.15 steps/s | VRAM=31.01 Go [2026-07-18 08:36:26] step 721450/1000000 | loss=2.1901 vocab=0.6527 token=1.1323 | 6.12 steps/s | VRAM=25.31 Go [2026-07-18 08:36:34] step 721500/1000000 | loss=2.1078 vocab=0.9510 token=1.1091 | 6.13 steps/s | VRAM=29.14 Go [2026-07-18 08:36:42] step 721550/1000000 | loss=2.1801 vocab=1.0527 token=1.3490 | 6.14 steps/s | VRAM=29.18 Go [2026-07-18 08:36:50] step 721600/1000000 | loss=2.1666 vocab=0.8391 token=1.2732 | 6.17 steps/s | VRAM=32.16 Go [2026-07-18 08:36:59] step 721650/1000000 | loss=2.1572 vocab=0.7691 token=1.1227 | 6.09 steps/s | VRAM=28.75 Go [2026-07-18 08:37:07] step 721700/1000000 | loss=2.0986 vocab=1.1006 token=1.4721 | 6.10 steps/s | VRAM=26.56 Go [2026-07-18 08:37:15] step 721750/1000000 | loss=2.1983 vocab=0.8812 token=1.1845 | 6.15 steps/s | VRAM=30.31 Go [2026-07-18 08:37:23] step 721800/1000000 | loss=2.3005 vocab=1.1805 token=1.1035 | 6.14 steps/s | VRAM=28.85 Go [2026-07-18 08:37:31] step 721850/1000000 | loss=2.2789 vocab=0.7462 token=1.6439 | 6.13 steps/s | VRAM=29.31 Go [2026-07-18 08:37:39] step 721900/1000000 | loss=2.2090 vocab=0.6457 token=1.4368 | 6.17 steps/s | VRAM=31.85 Go [2026-07-18 08:37:47] step 721950/1000000 | loss=2.1545 vocab=0.6809 token=1.2577 | 6.12 steps/s | VRAM=25.91 Go [2026-07-18 08:37:56] step 722000/1000000 | loss=2.3534 vocab=0.8133 token=1.0884 | 6.12 steps/s | VRAM=28.87 Go [2026-07-18 08:38:04] step 722050/1000000 | loss=2.1097 vocab=1.1734 token=1.8765 | 6.15 steps/s | VRAM=28.18 Go [2026-07-18 08:38:12] step 722100/1000000 | loss=2.2485 vocab=0.7367 token=0.9117 | 6.15 steps/s | VRAM=32.24 Go [2026-07-18 08:38:20] step 722150/1000000 | loss=2.3112 vocab=0.9873 token=1.6433 | 6.14 steps/s | VRAM=28.72 Go [2026-07-18 08:38:28] step 722200/1000000 | loss=2.2297 vocab=1.0625 token=1.2677 | 6.17 steps/s | VRAM=28.46 Go [2026-07-18 08:38:36] step 722250/1000000 | loss=2.2172 vocab=0.5972 token=0.7920 | 6.16 steps/s | VRAM=31.72 Go [2026-07-18 08:38:44] step 722300/1000000 | loss=2.1754 vocab=1.3782 token=1.4168 | 6.14 steps/s | VRAM=26.24 Go [2026-07-18 08:38:53] step 722350/1000000 | loss=2.2217 vocab=0.4231 token=1.0523 | 6.12 steps/s | VRAM=26.14 Go [2026-07-18 08:39:01] step 722400/1000000 | loss=2.1668 vocab=1.1474 token=1.4270 | 6.12 steps/s | VRAM=26.15 Go [2026-07-18 08:39:09] step 722450/1000000 | loss=2.1565 vocab=0.8730 token=1.4487 | 6.16 steps/s | VRAM=28.00 Go [2026-07-18 08:39:17] step 722500/1000000 | loss=2.2351 vocab=0.7241 token=1.4535 | 6.15 steps/s | VRAM=27.84 Go [2026-07-18 08:39:25] step 722550/1000000 | loss=2.2174 vocab=0.5532 token=1.0123 | 6.11 steps/s | VRAM=29.18 Go [2026-07-18 08:39:33] step 722600/1000000 | loss=2.2478 vocab=1.2953 token=1.3150 | 6.10 steps/s | VRAM=32.58 Go [2026-07-18 08:39:42] step 722650/1000000 | loss=2.2573 vocab=0.9794 token=1.2578 | 6.15 steps/s | VRAM=32.18 Go [2026-07-18 08:39:50] step 722700/1000000 | loss=2.2591 vocab=0.8175 token=0.8682 | 6.12 steps/s | VRAM=28.60 Go [2026-07-18 08:39:58] step 722750/1000000 | loss=2.3083 vocab=1.6024 token=1.6624 | 6.13 steps/s | VRAM=24.74 Go [2026-07-18 08:40:06] step 722800/1000000 | loss=2.1744 vocab=1.4415 token=1.6184 | 6.15 steps/s | VRAM=31.42 Go [2026-07-18 08:40:14] step 722850/1000000 | loss=2.2845 vocab=0.9941 token=1.2124 | 6.12 steps/s | VRAM=28.18 Go [2026-07-18 08:40:22] step 722900/1000000 | loss=2.2591 vocab=1.0209 token=1.4054 | 6.15 steps/s | VRAM=28.02 Go [2026-07-18 08:40:30] step 722950/1000000 | loss=2.0357 vocab=0.8470 token=1.2478 | 6.14 steps/s | VRAM=29.22 Go [2026-07-18 08:40:39] step 723000/1000000 | loss=2.1302 vocab=0.5019 token=0.8155 | 6.14 steps/s | VRAM=32.52 Go [2026-07-18 08:40:47] step 723050/1000000 | loss=2.2314 vocab=0.7959 token=1.1906 | 6.14 steps/s | VRAM=26.45 Go [2026-07-18 08:40:55] step 723100/1000000 | loss=2.1533 vocab=0.7980 token=1.3938 | 6.14 steps/s | VRAM=28.83 Go [2026-07-18 08:41:03] step 723150/1000000 | loss=2.2884 vocab=0.6196 token=1.9151 | 6.14 steps/s | VRAM=32.94 Go [2026-07-18 08:41:11] step 723200/1000000 | loss=2.1723 vocab=0.7575 token=1.4213 | 6.18 steps/s | VRAM=30.31 Go [2026-07-18 08:41:19] step 723250/1000000 | loss=2.0736 vocab=0.6808 token=0.9140 | 6.13 steps/s | VRAM=29.12 Go [2026-07-18 08:41:27] step 723300/1000000 | loss=2.2346 vocab=0.9366 token=1.2688 | 6.15 steps/s | VRAM=27.58 Go [2026-07-18 08:41:36] step 723350/1000000 | loss=2.0773 vocab=0.6470 token=1.2992 | 6.14 steps/s | VRAM=26.39 Go [2026-07-18 08:41:44] step 723400/1000000 | loss=2.2489 vocab=0.9389 token=1.3865 | 6.11 steps/s | VRAM=29.20 Go [2026-07-18 08:41:52] step 723450/1000000 | loss=2.2458 vocab=0.6799 token=1.1890 | 6.12 steps/s | VRAM=26.71 Go [2026-07-18 08:42:00] step 723500/1000000 | loss=2.1613 vocab=1.3684 token=1.4439 | 6.13 steps/s | VRAM=28.55 Go [2026-07-18 08:42:08] step 723550/1000000 | loss=2.1603 vocab=0.8460 token=1.4291 | 6.14 steps/s | VRAM=25.44 Go [2026-07-18 08:42:16] step 723600/1000000 | loss=2.2132 vocab=0.9411 token=1.3368 | 6.15 steps/s | VRAM=32.64 Go [2026-07-18 08:42:24] step 723650/1000000 | loss=2.2139 vocab=1.2742 token=1.4889 | 6.13 steps/s | VRAM=28.64 Go [2026-07-18 08:42:33] step 723700/1000000 | loss=2.1733 vocab=1.4594 token=1.5804 | 6.13 steps/s | VRAM=31.92 Go [2026-07-18 08:42:41] step 723750/1000000 | loss=2.2462 vocab=0.4346 token=0.9716 | 6.13 steps/s | VRAM=31.60 Go [2026-07-18 08:42:49] step 723800/1000000 | loss=2.2200 vocab=1.1782 token=1.3191 | 6.10 steps/s | VRAM=26.47 Go [2026-07-18 08:42:57] step 723850/1000000 | loss=2.2210 vocab=1.0860 token=1.3217 | 6.15 steps/s | VRAM=29.17 Go [2026-07-18 08:43:05] step 723900/1000000 | loss=2.2354 vocab=0.8264 token=1.3354 | 6.12 steps/s | VRAM=28.80 Go [2026-07-18 08:43:13] step 723950/1000000 | loss=2.1619 vocab=0.5774 token=1.1693 | 6.14 steps/s | VRAM=32.18 Go [2026-07-18 08:43:22] step 724000/1000000 | loss=2.0989 vocab=0.8641 token=1.3409 | 6.07 steps/s | VRAM=26.34 Go [2026-07-18 08:43:30] step 724050/1000000 | loss=2.0929 vocab=0.8005 token=1.0010 | 6.12 steps/s | VRAM=26.58 Go [2026-07-18 08:43:38] step 724100/1000000 | loss=2.1493 vocab=0.9092 token=1.8186 | 6.14 steps/s | VRAM=29.77 Go [2026-07-18 08:43:46] step 724150/1000000 | loss=2.1691 vocab=0.9794 token=1.2102 | 6.14 steps/s | VRAM=31.82 Go [2026-07-18 08:43:54] step 724200/1000000 | loss=2.1889 vocab=0.5932 token=1.2656 | 6.14 steps/s | VRAM=26.46 Go [2026-07-18 08:44:02] step 724250/1000000 | loss=2.1658 vocab=0.9260 token=1.3325 | 6.16 steps/s | VRAM=27.48 Go [2026-07-18 08:44:11] step 724300/1000000 | loss=2.1788 vocab=0.8264 token=1.3201 | 6.11 steps/s | VRAM=29.23 Go [2026-07-18 08:44:19] step 724350/1000000 | loss=2.1512 vocab=1.0534 token=1.4582 | 6.12 steps/s | VRAM=28.12 Go [2026-07-18 08:44:27] step 724400/1000000 | loss=2.0478 vocab=0.6879 token=1.2341 | 6.14 steps/s | VRAM=28.59 Go [2026-07-18 08:44:35] step 724450/1000000 | loss=2.2597 vocab=0.5283 token=1.3297 | 6.10 steps/s | VRAM=26.72 Go [2026-07-18 08:44:43] step 724500/1000000 | loss=2.2709 vocab=0.6630 token=1.4204 | 6.12 steps/s | VRAM=26.25 Go [2026-07-18 08:44:51] step 724550/1000000 | loss=2.1862 vocab=1.2738 token=1.1465 | 6.13 steps/s | VRAM=29.78 Go [2026-07-18 08:45:00] step 724600/1000000 | loss=2.1719 vocab=1.2244 token=1.4311 | 6.16 steps/s | VRAM=31.69 Go [2026-07-18 08:45:08] step 724650/1000000 | loss=2.1815 vocab=0.8010 token=1.5122 | 6.15 steps/s | VRAM=32.59 Go [2026-07-18 08:45:16] step 724700/1000000 | loss=2.1584 vocab=1.0785 token=1.5389 | 6.12 steps/s | VRAM=28.79 Go [2026-07-18 08:45:24] step 724750/1000000 | loss=2.1771 vocab=0.5503 token=1.1800 | 6.15 steps/s | VRAM=26.12 Go [2026-07-18 08:45:32] step 724800/1000000 | loss=2.1898 vocab=1.1372 token=0.7633 | 6.15 steps/s | VRAM=31.67 Go [2026-07-18 08:45:40] step 724850/1000000 | loss=2.1668 vocab=0.7532 token=1.3803 | 6.13 steps/s | VRAM=29.00 Go [2026-07-18 08:45:48] step 724900/1000000 | loss=2.3025 vocab=1.0195 token=1.4518 | 6.12 steps/s | VRAM=31.60 Go [2026-07-18 08:45:57] step 724950/1000000 | loss=2.2194 vocab=1.1195 token=1.3746 | 6.12 steps/s | VRAM=25.57 Go [2026-07-18 08:46:05] step 725000/1000000 | loss=2.3504 vocab=1.5589 token=1.3622 | 6.14 steps/s | VRAM=25.74 Go [2026-07-18 08:46:13] step 725050/1000000 | loss=2.2263 vocab=1.0048 token=1.9815 | 6.16 steps/s | VRAM=25.36 Go [2026-07-18 08:46:21] step 725100/1000000 | loss=2.0949 vocab=1.2858 token=1.4914 | 6.17 steps/s | VRAM=29.38 Go [2026-07-18 08:46:29] step 725150/1000000 | loss=2.1558 vocab=0.9438 token=1.6267 | 6.11 steps/s | VRAM=29.69 Go [2026-07-18 08:46:37] step 725200/1000000 | loss=2.2147 vocab=0.6348 token=1.0168 | 6.15 steps/s | VRAM=28.71 Go [2026-07-18 08:46:45] step 725250/1000000 | loss=2.2388 vocab=0.6430 token=1.2415 | 6.12 steps/s | VRAM=28.09 Go [2026-07-18 08:46:54] step 725300/1000000 | loss=2.1864 vocab=0.5838 token=1.1328 | 6.15 steps/s | VRAM=28.46 Go [2026-07-18 08:47:02] step 725350/1000000 | loss=2.2829 vocab=0.5553 token=1.1108 | 6.14 steps/s | VRAM=32.88 Go [2026-07-18 08:47:10] step 725400/1000000 | loss=2.2640 vocab=0.8327 token=1.0554 | 6.14 steps/s | VRAM=29.25 Go [2026-07-18 08:47:18] step 725450/1000000 | loss=2.0367 vocab=0.5989 token=0.9770 | 6.12 steps/s | VRAM=28.11 Go [2026-07-18 08:47:26] step 725500/1000000 | loss=2.1153 vocab=0.6651 token=1.4106 | 6.09 steps/s | VRAM=26.31 Go [2026-07-18 08:47:34] step 725550/1000000 | loss=2.2167 vocab=1.0155 token=1.3065 | 6.13 steps/s | VRAM=31.96 Go [2026-07-18 08:47:43] step 725600/1000000 | loss=2.1757 vocab=0.6970 token=1.2820 | 6.10 steps/s | VRAM=28.97 Go [2026-07-18 08:47:51] step 725650/1000000 | loss=2.1431 vocab=0.5772 token=1.3844 | 6.13 steps/s | VRAM=28.10 Go [2026-07-18 08:47:59] step 725700/1000000 | loss=2.2000 vocab=1.0274 token=1.0855 | 6.15 steps/s | VRAM=29.32 Go [2026-07-18 08:48:07] step 725750/1000000 | loss=2.2591 vocab=0.9259 token=1.3790 | 6.13 steps/s | VRAM=32.13 Go [2026-07-18 08:48:15] step 725800/1000000 | loss=2.2793 vocab=1.2851 token=1.5376 | 6.15 steps/s | VRAM=31.21 Go [2026-07-18 08:48:23] step 725850/1000000 | loss=2.1485 vocab=1.3460 token=0.9954 | 6.15 steps/s | VRAM=30.93 Go [2026-07-18 08:48:31] step 725900/1000000 | loss=2.1866 vocab=0.8866 token=1.2866 | 6.14 steps/s | VRAM=25.30 Go [2026-07-18 08:48:40] step 725950/1000000 | loss=2.0693 vocab=0.6958 token=1.4647 | 6.13 steps/s | VRAM=28.19 Go [2026-07-18 08:48:48] step 726000/1000000 | loss=2.1261 vocab=1.1764 token=1.3393 | 6.16 steps/s | VRAM=30.98 Go [2026-07-18 08:48:56] step 726050/1000000 | loss=2.0426 vocab=0.9022 token=1.3023 | 6.13 steps/s | VRAM=26.36 Go [2026-07-18 08:49:04] step 726100/1000000 | loss=2.0829 vocab=0.5139 token=1.2904 | 6.14 steps/s | VRAM=26.74 Go [2026-07-18 08:49:12] step 726150/1000000 | loss=2.2355 vocab=0.7518 token=1.6352 | 6.14 steps/s | VRAM=32.27 Go [2026-07-18 08:49:20] step 726200/1000000 | loss=2.1475 vocab=1.3179 token=1.4428 | 6.14 steps/s | VRAM=25.28 Go [2026-07-18 08:49:28] step 726250/1000000 | loss=2.1532 vocab=0.6237 token=1.4321 | 6.16 steps/s | VRAM=32.97 Go [2026-07-18 08:49:37] step 726300/1000000 | loss=2.1935 vocab=0.6697 token=1.0620 | 6.14 steps/s | VRAM=28.35 Go [2026-07-18 08:49:45] step 726350/1000000 | loss=2.1983 vocab=1.8289 token=1.4570 | 6.15 steps/s | VRAM=25.52 Go [2026-07-18 08:49:53] step 726400/1000000 | loss=2.1831 vocab=0.6488 token=0.9863 | 6.14 steps/s | VRAM=32.24 Go [2026-07-18 08:50:01] step 726450/1000000 | loss=2.0912 vocab=0.8791 token=1.3129 | 6.14 steps/s | VRAM=32.32 Go [2026-07-18 08:50:09] step 726500/1000000 | loss=2.1978 vocab=0.6939 token=1.5120 | 6.14 steps/s | VRAM=26.16 Go [2026-07-18 08:50:17] step 726550/1000000 | loss=2.0935 vocab=0.8476 token=1.4589 | 6.16 steps/s | VRAM=25.92 Go [2026-07-18 08:50:25] step 726600/1000000 | loss=2.1079 vocab=0.4713 token=0.5824 | 6.15 steps/s | VRAM=26.06 Go [2026-07-18 08:50:33] step 726650/1000000 | loss=2.1507 vocab=0.6957 token=0.6822 | 6.18 steps/s | VRAM=32.34 Go [2026-07-18 08:50:42] step 726700/1000000 | loss=2.2909 vocab=1.0098 token=1.2906 | 6.17 steps/s | VRAM=31.95 Go [2026-07-18 08:50:50] step 726750/1000000 | loss=2.2149 vocab=1.2002 token=1.4148 | 6.17 steps/s | VRAM=28.38 Go [2026-07-18 08:50:58] step 726800/1000000 | loss=2.2606 vocab=0.8036 token=1.2058 | 6.14 steps/s | VRAM=31.84 Go [2026-07-18 08:51:06] step 726850/1000000 | loss=2.1526 vocab=0.7832 token=1.1753 | 6.15 steps/s | VRAM=28.54 Go [2026-07-18 08:51:14] step 726900/1000000 | loss=2.1694 vocab=0.7780 token=1.1636 | 6.17 steps/s | VRAM=32.08 Go [2026-07-18 08:51:22] step 726950/1000000 | loss=2.0819 vocab=0.4509 token=1.1235 | 6.12 steps/s | VRAM=32.61 Go [2026-07-18 08:51:30] step 727000/1000000 | loss=2.2334 vocab=1.1257 token=1.2420 | 6.15 steps/s | VRAM=32.95 Go [2026-07-18 08:51:38] step 727050/1000000 | loss=2.2248 vocab=0.9131 token=1.2741 | 6.14 steps/s | VRAM=24.48 Go [2026-07-18 08:51:47] step 727100/1000000 | loss=2.2604 vocab=0.5991 token=1.0734 | 6.13 steps/s | VRAM=32.64 Go [2026-07-18 08:51:55] step 727150/1000000 | loss=2.2221 vocab=0.9592 token=1.5313 | 6.11 steps/s | VRAM=25.44 Go [2026-07-18 08:52:03] step 727200/1000000 | loss=2.2282 vocab=1.1860 token=1.3717 | 6.16 steps/s | VRAM=24.36 Go [2026-07-18 08:52:11] step 727250/1000000 | loss=2.1690 vocab=0.7337 token=1.3607 | 6.15 steps/s | VRAM=29.11 Go [2026-07-18 08:52:19] step 727300/1000000 | loss=2.2488 vocab=0.6571 token=1.4003 | 6.17 steps/s | VRAM=32.25 Go [2026-07-18 08:52:27] step 727350/1000000 | loss=2.2117 vocab=1.2562 token=1.5327 | 6.11 steps/s | VRAM=24.39 Go [2026-07-18 08:52:35] step 727400/1000000 | loss=2.2934 vocab=0.7909 token=1.3597 | 6.17 steps/s | VRAM=31.92 Go [2026-07-18 08:52:44] step 727450/1000000 | loss=2.2229 vocab=0.8621 token=1.4012 | 6.14 steps/s | VRAM=25.86 Go [2026-07-18 08:52:52] step 727500/1000000 | loss=2.1912 vocab=0.7537 token=1.2521 | 6.15 steps/s | VRAM=28.36 Go [2026-07-18 09:16:34] step 727550/1000000 | loss=1.5396 vocab=0.8853 token=1.4605 | lr=1.00e-04 | 7.73 steps/s | VRAM=24.02 Go [2026-07-18 09:16:42] step 727600/1000000 | loss=2.2327 vocab=0.8378 token=1.6834 | lr=1.00e-04 | 6.27 steps/s | VRAM=28.00 Go [2026-07-18 09:16:49] step 727650/1000000 | loss=2.3176 vocab=0.8631 token=1.6538 | lr=1.00e-04 | 6.28 steps/s | VRAM=31.76 Go [2026-07-18 09:16:57] step 727700/1000000 | loss=2.2356 vocab=0.7329 token=1.7532 | lr=1.00e-04 | 6.28 steps/s | VRAM=32.31 Go [2026-07-18 09:17:06] step 727750/1000000 | loss=2.1573 vocab=0.5375 token=1.3260 | lr=1.00e-04 | 6.20 steps/s | VRAM=26.28 Go [2026-07-18 09:17:14] step 727800/1000000 | loss=2.2080 vocab=1.1559 token=1.1600 | lr=1.00e-04 | 6.25 steps/s | VRAM=29.38 Go [2026-07-18 09:17:22] step 727850/1000000 | loss=2.1453 vocab=0.6466 token=0.9675 | lr=1.00e-04 | 6.23 steps/s | VRAM=26.83 Go [2026-07-18 09:17:30] step 727900/1000000 | loss=2.2106 vocab=0.7138 token=0.9872 | lr=1.00e-04 | 6.18 steps/s | VRAM=28.65 Go [2026-07-18 09:17:38] step 727950/1000000 | loss=2.0837 vocab=0.5382 token=1.2112 | lr=1.00e-04 | 6.23 steps/s | VRAM=28.42 Go [2026-07-18 09:17:46] step 728000/1000000 | loss=2.2007 vocab=0.8466 token=1.0713 | lr=1.00e-04 | 6.20 steps/s | VRAM=32.59 Go [2026-07-18 09:17:54] step 728050/1000000 | loss=2.1797 vocab=1.0125 token=1.1961 | lr=1.00e-04 | 6.18 steps/s | VRAM=32.18 Go [2026-07-18 09:18:02] step 728100/1000000 | loss=2.1081 vocab=0.7323 token=0.9358 | lr=1.00e-04 | 6.20 steps/s | VRAM=31.92 Go [2026-07-18 09:18:10] step 728150/1000000 | loss=2.0947 vocab=0.7547 token=1.1401 | lr=1.00e-04 | 6.17 steps/s | VRAM=29.26 Go [2026-07-18 09:18:18] step 728200/1000000 | loss=2.2216 vocab=1.0588 token=1.3773 | lr=1.00e-04 | 6.18 steps/s | VRAM=29.15 Go [2026-07-18 09:18:26] step 728250/1000000 | loss=2.3141 vocab=1.0422 token=1.3615 | lr=1.00e-04 | 6.16 steps/s | VRAM=27.00 Go [2026-07-18 09:18:34] step 728300/1000000 | loss=2.1446 vocab=1.2524 token=1.1459 | lr=1.00e-04 | 6.18 steps/s | VRAM=24.80 Go [2026-07-18 09:18:42] step 728350/1000000 | loss=2.1598 vocab=1.0723 token=1.4431 | lr=1.00e-04 | 6.15 steps/s | VRAM=29.70 Go [2026-07-18 09:18:51] step 728400/1000000 | loss=2.1787 vocab=1.0718 token=1.4582 | lr=1.00e-04 | 6.12 steps/s | VRAM=29.69 Go [2026-07-18 09:18:59] step 728450/1000000 | loss=2.3188 vocab=1.4843 token=1.5273 | lr=1.00e-04 | 6.16 steps/s | VRAM=29.36 Go [2026-07-18 09:19:07] step 728500/1000000 | loss=2.1062 vocab=1.1379 token=1.4688 | lr=1.00e-04 | 6.14 steps/s | VRAM=28.71 Go [2026-07-18 09:19:15] step 728550/1000000 | loss=2.1833 vocab=0.7383 token=1.5214 | lr=1.00e-04 | 6.12 steps/s | VRAM=29.44 Go [2026-07-18 09:19:23] step 728600/1000000 | loss=2.1631 vocab=0.8785 token=1.4596 | lr=1.00e-04 | 6.14 steps/s | VRAM=25.17 Go [2026-07-18 09:19:31] step 728650/1000000 | loss=2.1634 vocab=1.2762 token=1.2249 | lr=1.00e-04 | 6.16 steps/s | VRAM=31.32 Go [2026-07-18 09:19:39] step 728700/1000000 | loss=2.1583 vocab=0.5291 token=1.0968 | lr=1.00e-04 | 6.11 steps/s | VRAM=26.52 Go [2026-07-18 09:19:48] step 728750/1000000 | loss=2.2584 vocab=0.9663 token=1.2516 | lr=1.00e-04 | 6.14 steps/s | VRAM=28.96 Go [2026-07-18 09:19:56] step 728800/1000000 | loss=2.1787 vocab=0.9087 token=1.7215 | lr=1.00e-04 | 6.14 steps/s | VRAM=26.54 Go [2026-07-18 09:20:04] step 728850/1000000 | loss=2.1379 vocab=0.7672 token=1.3325 | lr=1.00e-04 | 6.16 steps/s | VRAM=32.44 Go [2026-07-18 09:20:12] step 728900/1000000 | loss=2.0604 vocab=0.6897 token=1.4489 | lr=1.00e-04 | 6.14 steps/s | VRAM=28.89 Go [2026-07-18 09:20:20] step 728950/1000000 | loss=2.1991 vocab=1.1318 token=1.5586 | lr=1.00e-04 | 6.13 steps/s | VRAM=25.01 Go [2026-07-18 09:20:28] step 729000/1000000 | loss=2.1526 vocab=0.7540 token=1.0902 | lr=1.00e-04 | 6.13 steps/s | VRAM=32.12 Go [2026-07-18 09:20:36] step 729050/1000000 | loss=2.0777 vocab=0.6741 token=1.4381 | lr=1.00e-04 | 6.11 steps/s | VRAM=30.02 Go [2026-07-18 09:20:45] step 729100/1000000 | loss=2.2514 vocab=1.6816 token=1.3161 | lr=1.00e-04 | 6.13 steps/s | VRAM=25.70 Go [2026-07-18 09:20:53] step 729150/1000000 | loss=2.1259 vocab=1.0256 token=1.3067 | lr=1.00e-04 | 6.11 steps/s | VRAM=28.81 Go [2026-07-18 09:21:01] step 729200/1000000 | loss=2.1509 vocab=0.8490 token=1.6240 | lr=1.00e-04 | 6.11 steps/s | VRAM=26.39 Go [2026-07-18 09:21:09] step 729250/1000000 | loss=2.0660 vocab=0.8424 token=1.1575 | lr=1.00e-04 | 6.16 steps/s | VRAM=31.59 Go [2026-07-18 09:21:17] step 729300/1000000 | loss=2.1501 vocab=1.2654 token=1.3339 | lr=1.00e-04 | 6.13 steps/s | VRAM=26.75 Go [2026-07-18 09:21:25] step 729350/1000000 | loss=2.1804 vocab=1.4763 token=1.5895 | lr=1.00e-04 | 6.14 steps/s | VRAM=28.05 Go [2026-07-18 09:21:34] step 729400/1000000 | loss=2.1292 vocab=0.8183 token=1.3034 | lr=1.00e-04 | 6.14 steps/s | VRAM=27.38 Go [2026-07-18 09:21:42] step 729450/1000000 | loss=2.1843 vocab=1.0903 token=1.7371 | lr=1.00e-04 | 6.15 steps/s | VRAM=27.55 Go [2026-07-18 09:21:50] step 729500/1000000 | loss=2.2019 vocab=0.7702 token=1.3293 | lr=1.00e-04 | 6.10 steps/s | VRAM=28.53 Go [2026-07-18 09:21:58] step 729550/1000000 | loss=2.2228 vocab=0.9618 token=1.2988 | lr=1.00e-04 | 6.12 steps/s | VRAM=25.87 Go [2026-07-18 09:22:06] step 729600/1000000 | loss=2.2183 vocab=0.8413 token=1.5066 | lr=1.00e-04 | 6.14 steps/s | VRAM=29.04 Go [2026-07-18 09:22:14] step 729650/1000000 | loss=2.2513 vocab=0.7248 token=1.5071 | lr=1.00e-04 | 6.12 steps/s | VRAM=29.28 Go [2026-07-18 09:22:23] step 729700/1000000 | loss=2.2591 vocab=0.5099 token=1.2362 | lr=1.00e-04 | 6.13 steps/s | VRAM=22.99 Go [2026-07-18 09:22:31] step 729750/1000000 | loss=2.1693 vocab=0.8713 token=1.1409 | lr=1.00e-04 | 6.12 steps/s | VRAM=26.35 Go [2026-07-18 09:22:39] step 729800/1000000 | loss=2.1195 vocab=1.0998 token=1.3169 | lr=1.00e-04 | 6.14 steps/s | VRAM=29.04 Go [2026-07-18 09:22:47] step 729850/1000000 | loss=2.1374 vocab=0.7219 token=1.5315 | lr=1.00e-04 | 6.10 steps/s | VRAM=26.73 Go [2026-07-18 09:22:55] step 729900/1000000 | loss=2.2785 vocab=0.9983 token=1.2555 | lr=1.00e-04 | 6.13 steps/s | VRAM=28.78 Go [2026-07-18 09:23:03] step 729950/1000000 | loss=2.2423 vocab=0.5635 token=1.4136 | lr=1.00e-04 | 6.15 steps/s | VRAM=31.43 Go [2026-07-18 09:23:12] step 730000/1000000 | loss=2.2059 vocab=1.1011 token=1.3680 | lr=1.00e-04 | 6.12 steps/s | VRAM=32.14 Go [2026-07-18 09:23:20] step 730050/1000000 | loss=2.2891 vocab=1.2248 token=1.5644 | lr=1.00e-04 | 6.11 steps/s | VRAM=26.88 Go [2026-07-18 09:23:28] step 730100/1000000 | loss=2.2457 vocab=1.2418 token=1.1725 | lr=1.00e-04 | 6.14 steps/s | VRAM=31.25 Go [2026-07-18 09:23:36] step 730150/1000000 | loss=2.0620 vocab=0.7341 token=1.3992 | lr=1.00e-04 | 6.16 steps/s | VRAM=31.12 Go [2026-07-18 09:23:44] step 730200/1000000 | loss=2.2599 vocab=0.8636 token=1.2081 | lr=1.00e-04 | 6.17 steps/s | VRAM=28.46 Go [2026-07-18 09:23:52] step 730250/1000000 | loss=2.2544 vocab=1.4080 token=1.5570 | lr=1.00e-04 | 6.12 steps/s | VRAM=30.04 Go [2026-07-18 09:24:00] step 730300/1000000 | loss=2.1422 vocab=1.1001 token=1.3890 | lr=1.00e-04 | 6.12 steps/s | VRAM=26.45 Go [2026-07-18 09:24:09] step 730350/1000000 | loss=2.1273 vocab=0.7142 token=1.2073 | lr=1.00e-04 | 6.15 steps/s | VRAM=25.00 Go [2026-07-18 09:24:17] step 730400/1000000 | loss=2.2305 vocab=1.2226 token=1.3205 | lr=1.00e-04 | 6.14 steps/s | VRAM=28.85 Go [2026-07-18 09:24:25] step 730450/1000000 | loss=2.1832 vocab=0.7139 token=1.1020 | lr=1.00e-04 | 6.16 steps/s | VRAM=27.82 Go [2026-07-18 09:24:33] step 730500/1000000 | loss=2.1074 vocab=1.0803 token=1.3795 | lr=1.00e-04 | 6.15 steps/s | VRAM=29.34 Go [2026-07-18 09:24:41] step 730550/1000000 | loss=2.2073 vocab=0.9772 token=1.7002 | lr=1.00e-04 | 6.12 steps/s | VRAM=28.63 Go [2026-07-18 09:24:49] step 730600/1000000 | loss=2.2075 vocab=0.5709 token=1.2018 | lr=1.00e-04 | 6.15 steps/s | VRAM=26.17 Go [2026-07-18 09:24:57] step 730650/1000000 | loss=2.1185 vocab=1.4980 token=1.3213 | lr=1.00e-04 | 6.15 steps/s | VRAM=28.12 Go [2026-07-18 09:25:05] step 730700/1000000 | loss=2.2024 vocab=1.2148 token=1.2838 | lr=1.00e-04 | 6.15 steps/s | VRAM=29.09 Go [2026-07-18 09:25:14] step 730750/1000000 | loss=2.1722 vocab=0.5833 token=0.8300 | lr=1.00e-04 | 6.13 steps/s | VRAM=26.66 Go [2026-07-18 09:25:22] step 730800/1000000 | loss=2.1300 vocab=0.6169 token=0.8494 | lr=1.00e-04 | 6.15 steps/s | VRAM=32.38 Go [2026-07-18 09:25:30] step 730850/1000000 | loss=2.2298 vocab=0.7217 token=1.3857 | lr=1.00e-04 | 6.15 steps/s | VRAM=31.91 Go [2026-07-18 09:25:38] step 730900/1000000 | loss=2.1965 vocab=1.0248 token=1.1358 | lr=1.00e-04 | 6.12 steps/s | VRAM=28.74 Go [2026-07-18 09:25:46] step 730950/1000000 | loss=2.2345 vocab=0.8876 token=1.5393 | lr=1.00e-04 | 6.18 steps/s | VRAM=27.68 Go [2026-07-18 09:25:54] step 731000/1000000 | loss=2.1699 vocab=0.9749 token=1.3174 | lr=1.00e-04 | 6.13 steps/s | VRAM=28.97 Go [2026-07-18 09:26:02] step 731050/1000000 | loss=2.0776 vocab=0.8108 token=1.2115 | lr=1.00e-04 | 6.19 steps/s | VRAM=32.45 Go [2026-07-18 09:26:11] step 731100/1000000 | loss=2.1160 vocab=0.9881 token=1.2650 | lr=1.00e-04 | 6.13 steps/s | VRAM=26.38 Go [2026-07-18 09:26:19] step 731150/1000000 | loss=2.1950 vocab=0.7019 token=1.3675 | lr=1.00e-04 | 6.18 steps/s | VRAM=28.19 Go [2026-07-18 09:26:27] step 731200/1000000 | loss=2.1765 vocab=0.7292 token=1.3553 | lr=1.00e-04 | 6.15 steps/s | VRAM=31.69 Go [2026-07-18 09:26:35] step 731250/1000000 | loss=2.2024 vocab=0.7902 token=1.0935 | lr=1.00e-04 | 6.16 steps/s | VRAM=28.85 Go [2026-07-18 09:26:43] step 731300/1000000 | loss=2.2427 vocab=0.7045 token=1.3463 | lr=1.00e-04 | 6.16 steps/s | VRAM=31.26 Go [2026-07-18 09:26:51] step 731350/1000000 | loss=2.1516 vocab=0.9379 token=1.1077 | lr=1.00e-04 | 6.16 steps/s | VRAM=25.24 Go [2026-07-18 09:26:59] step 731400/1000000 | loss=2.1408 vocab=0.8162 token=0.8583 | lr=1.00e-04 | 6.14 steps/s | VRAM=28.13 Go [2026-07-18 09:27:07] step 731450/1000000 | loss=2.1872 vocab=0.7281 token=1.2917 | lr=9.99e-05 | 6.15 steps/s | VRAM=27.41 Go [2026-07-18 09:27:16] step 731500/1000000 | loss=2.1447 vocab=1.1422 token=1.2941 | lr=9.99e-05 | 6.11 steps/s | VRAM=29.93 Go [2026-07-18 09:27:24] step 731550/1000000 | loss=2.2505 vocab=0.4998 token=1.0020 | lr=9.99e-05 | 6.15 steps/s | VRAM=32.54 Go [2026-07-18 09:27:32] step 731600/1000000 | loss=2.1300 vocab=0.8571 token=1.2927 | lr=9.99e-05 | 6.10 steps/s | VRAM=26.63 Go [2026-07-18 09:27:40] step 731650/1000000 | loss=2.0552 vocab=0.7811 token=0.9404 | lr=9.99e-05 | 6.14 steps/s | VRAM=26.46 Go [2026-07-18 09:27:48] step 731700/1000000 | loss=2.1533 vocab=0.6664 token=1.0390 | lr=9.99e-05 | 6.19 steps/s | VRAM=31.90 Go [2026-07-18 09:27:56] step 731750/1000000 | loss=2.1477 vocab=0.4277 token=1.2186 | lr=9.99e-05 | 6.17 steps/s | VRAM=28.09 Go [2026-07-18 09:28:04] step 731800/1000000 | loss=2.2681 vocab=0.8611 token=1.5856 | lr=9.99e-05 | 6.14 steps/s | VRAM=27.71 Go [2026-07-18 09:28:13] step 731850/1000000 | loss=2.1774 vocab=1.0195 token=1.2702 | lr=9.99e-05 | 6.10 steps/s | VRAM=27.19 Go [2026-07-18 09:28:21] step 731900/1000000 | loss=2.1481 vocab=0.7953 token=1.1949 | lr=9.99e-05 | 6.14 steps/s | VRAM=25.88 Go [2026-07-18 09:28:29] step 731950/1000000 | loss=2.1740 vocab=1.2128 token=1.6560 | lr=9.99e-05 | 6.16 steps/s | VRAM=31.04 Go [2026-07-18 09:28:37] step 732000/1000000 | loss=2.1221 vocab=1.2591 token=1.2030 | lr=9.99e-05 | 6.17 steps/s | VRAM=32.27 Go [2026-07-18 09:28:45] step 732050/1000000 | loss=2.1566 vocab=0.8153 token=1.4972 | lr=9.99e-05 | 6.17 steps/s | VRAM=31.86 Go [2026-07-18 09:28:53] step 732100/1000000 | loss=2.0551 vocab=0.6445 token=0.6591 | lr=9.99e-05 | 6.13 steps/s | VRAM=28.24 Go [2026-07-18 09:29:01] step 732150/1000000 | loss=2.2373 vocab=0.8274 token=1.1766 | lr=9.99e-05 | 6.15 steps/s | VRAM=28.55 Go [2026-07-18 09:29:09] step 732200/1000000 | loss=2.0859 vocab=0.4954 token=1.0803 | lr=9.99e-05 | 6.15 steps/s | VRAM=27.40 Go [2026-07-18 09:29:18] step 732250/1000000 | loss=2.1647 vocab=1.1569 token=1.4590 | lr=9.99e-05 | 6.14 steps/s | VRAM=31.80 Go [2026-07-18 09:29:26] step 732300/1000000 | loss=2.0882 vocab=1.2729 token=1.6697 | lr=9.99e-05 | 6.12 steps/s | VRAM=28.30 Go [2026-07-18 09:29:34] step 732350/1000000 | loss=2.0710 vocab=0.4282 token=1.2052 | lr=9.99e-05 | 6.13 steps/s | VRAM=29.21 Go [2026-07-18 09:29:42] step 732400/1000000 | loss=2.2086 vocab=0.7242 token=1.4188 | lr=9.99e-05 | 6.15 steps/s | VRAM=33.02 Go [2026-07-18 09:29:50] step 732450/1000000 | loss=2.2297 vocab=0.9780 token=1.7383 | lr=9.99e-05 | 6.14 steps/s | VRAM=25.48 Go [2026-07-18 09:29:58] step 732500/1000000 | loss=2.3418 vocab=1.0533 token=1.1085 | lr=9.99e-05 | 6.15 steps/s | VRAM=30.47 Go [2026-07-18 09:30:06] step 732550/1000000 | loss=2.2149 vocab=1.3515 token=1.5263 | lr=9.99e-05 | 6.13 steps/s | VRAM=29.28 Go [2026-07-18 09:30:15] step 732600/1000000 | loss=2.1989 vocab=0.8798 token=1.3229 | lr=9.99e-05 | 6.15 steps/s | VRAM=29.46 Go [2026-07-18 09:30:23] step 732650/1000000 | loss=2.2057 vocab=1.3718 token=1.1810 | lr=9.99e-05 | 6.13 steps/s | VRAM=28.83 Go [2026-07-18 09:30:31] step 732700/1000000 | loss=2.1511 vocab=0.7690 token=1.5295 | lr=9.99e-05 | 6.11 steps/s | VRAM=28.97 Go [2026-07-18 09:30:39] step 732750/1000000 | loss=2.1344 vocab=0.7546 token=1.3879 | lr=9.99e-05 | 6.12 steps/s | VRAM=26.49 Go [2026-07-18 09:30:47] step 732800/1000000 | loss=2.1822 vocab=0.6541 token=1.1142 | lr=9.99e-05 | 6.13 steps/s | VRAM=28.83 Go [2026-07-18 09:30:55] step 732850/1000000 | loss=2.2292 vocab=0.7041 token=1.1189 | lr=9.99e-05 | 6.12 steps/s | VRAM=29.26 Go [2026-07-18 09:31:04] step 732900/1000000 | loss=2.1643 vocab=0.8978 token=1.4312 | lr=9.99e-05 | 6.14 steps/s | VRAM=28.82 Go [2026-07-18 09:31:12] step 732950/1000000 | loss=2.1332 vocab=1.0352 token=1.1277 | lr=9.99e-05 | 6.13 steps/s | VRAM=25.05 Go [2026-07-18 09:31:20] step 733000/1000000 | loss=2.2149 vocab=1.0809 token=1.3414 | lr=9.99e-05 | 6.14 steps/s | VRAM=29.15 Go [2026-07-18 09:31:28] step 733050/1000000 | loss=2.1374 vocab=0.7042 token=1.6186 | lr=9.99e-05 | 6.15 steps/s | VRAM=25.17 Go [2026-07-18 09:31:36] step 733100/1000000 | loss=2.1738 vocab=0.6064 token=1.2036 | lr=9.99e-05 | 6.15 steps/s | VRAM=26.58 Go [2026-07-18 09:31:44] step 733150/1000000 | loss=2.1127 vocab=0.9529 token=1.5499 | lr=9.99e-05 | 6.14 steps/s | VRAM=27.93 Go [2026-07-18 09:31:52] step 733200/1000000 | loss=2.2002 vocab=0.5895 token=1.1073 | lr=9.99e-05 | 6.11 steps/s | VRAM=27.52 Go [2026-07-18 09:32:01] step 733250/1000000 | loss=2.1715 vocab=1.2683 token=1.2848 | lr=9.99e-05 | 6.14 steps/s | VRAM=28.98 Go [2026-07-18 09:32:09] step 733300/1000000 | loss=2.1950 vocab=0.8991 token=1.4464 | lr=9.99e-05 | 6.14 steps/s | VRAM=25.77 Go [2026-07-18 09:32:17] step 733350/1000000 | loss=2.1491 vocab=1.4043 token=1.5572 | lr=9.99e-05 | 6.18 steps/s | VRAM=31.18 Go [2026-07-18 09:32:25] step 733400/1000000 | loss=2.2117 vocab=0.7658 token=1.0139 | lr=9.99e-05 | 6.14 steps/s | VRAM=28.79 Go [2026-07-18 09:32:33] step 733450/1000000 | loss=2.3070 vocab=0.9300 token=1.1984 | lr=9.99e-05 | 6.15 steps/s | VRAM=26.11 Go [2026-07-18 09:32:41] step 733500/1000000 | loss=2.1890 vocab=0.9139 token=1.2814 | lr=9.99e-05 | 6.17 steps/s | VRAM=28.33 Go [2026-07-18 09:32:49] step 733550/1000000 | loss=2.2262 vocab=0.8926 token=1.6907 | lr=9.99e-05 | 6.15 steps/s | VRAM=31.59 Go [2026-07-18 09:32:57] step 733600/1000000 | loss=2.1854 vocab=0.7585 token=1.0550 | lr=9.99e-05 | 6.17 steps/s | VRAM=32.46 Go [2026-07-18 09:33:06] step 733650/1000000 | loss=2.1542 vocab=0.5143 token=0.8139 | lr=9.99e-05 | 6.15 steps/s | VRAM=28.39 Go [2026-07-18 09:33:14] step 733700/1000000 | loss=2.1728 vocab=0.7356 token=1.3627 | lr=9.99e-05 | 6.13 steps/s | VRAM=27.15 Go [2026-07-18 09:33:22] step 733750/1000000 | loss=2.1380 vocab=1.2172 token=1.0423 | lr=9.99e-05 | 6.14 steps/s | VRAM=25.60 Go [2026-07-18 09:33:30] step 733800/1000000 | loss=2.1733 vocab=0.4298 token=1.1884 | lr=9.99e-05 | 6.13 steps/s | VRAM=24.98 Go [2026-07-18 09:33:38] step 733850/1000000 | loss=2.1342 vocab=0.5701 token=1.0982 | lr=9.99e-05 | 6.18 steps/s | VRAM=32.81 Go [2026-07-18 09:33:46] step 733900/1000000 | loss=2.1531 vocab=0.7875 token=1.1598 | lr=9.99e-05 | 6.12 steps/s | VRAM=26.24 Go [2026-07-18 09:33:55] step 733950/1000000 | loss=2.1880 vocab=0.8109 token=1.2536 | lr=9.99e-05 | 6.11 steps/s | VRAM=28.93 Go [2026-07-18 09:34:03] step 734000/1000000 | loss=2.2421 vocab=0.5836 token=0.8698 | lr=9.99e-05 | 6.14 steps/s | VRAM=25.81 Go [2026-07-18 09:34:11] step 734050/1000000 | loss=2.2082 vocab=0.9389 token=1.5841 | lr=9.99e-05 | 6.17 steps/s | VRAM=26.10 Go [2026-07-18 09:34:19] step 734100/1000000 | loss=2.1677 vocab=0.9949 token=1.5785 | lr=9.99e-05 | 6.12 steps/s | VRAM=26.37 Go [2026-07-18 09:34:27] step 734150/1000000 | loss=2.1401 vocab=0.8413 token=1.1486 | lr=9.99e-05 | 6.14 steps/s | VRAM=28.05 Go [2026-07-18 09:34:35] step 734200/1000000 | loss=2.1682 vocab=1.8422 token=1.2641 | lr=9.99e-05 | 6.11 steps/s | VRAM=28.90 Go [2026-07-18 09:34:43] step 734250/1000000 | loss=2.3586 vocab=0.5967 token=0.9788 | lr=9.99e-05 | 6.13 steps/s | VRAM=32.92 Go [2026-07-18 09:34:52] step 734300/1000000 | loss=1.9876 vocab=0.7271 token=0.9303 | lr=9.98e-05 | 6.16 steps/s | VRAM=24.83 Go [2026-07-18 09:35:00] step 734350/1000000 | loss=2.1834 vocab=0.9870 token=1.4915 | lr=9.98e-05 | 6.15 steps/s | VRAM=29.07 Go [2026-07-18 09:35:08] step 734400/1000000 | loss=2.1145 vocab=0.7926 token=1.3799 | lr=9.98e-05 | 6.13 steps/s | VRAM=26.61 Go [2026-07-18 09:35:16] step 734450/1000000 | loss=2.0935 vocab=1.0577 token=1.6624 | lr=9.98e-05 | 6.18 steps/s | VRAM=31.42 Go [2026-07-18 09:35:24] step 734500/1000000 | loss=2.1716 vocab=0.9019 token=1.2004 | lr=9.98e-05 | 6.15 steps/s | VRAM=31.58 Go [2026-07-18 09:35:32] step 734550/1000000 | loss=2.1200 vocab=0.7428 token=1.1697 | lr=9.98e-05 | 6.17 steps/s | VRAM=27.99 Go [2026-07-18 09:35:40] step 734600/1000000 | loss=2.1680 vocab=1.6831 token=1.1577 | lr=9.98e-05 | 6.17 steps/s | VRAM=32.35 Go [2026-07-18 09:35:48] step 734650/1000000 | loss=2.2033 vocab=0.9339 token=0.9304 | lr=9.98e-05 | 6.15 steps/s | VRAM=28.00 Go [2026-07-18 09:35:56] step 734700/1000000 | loss=2.1367 vocab=0.7476 token=1.1547 | lr=9.98e-05 | 6.13 steps/s | VRAM=23.40 Go [2026-07-18 09:36:05] step 734750/1000000 | loss=2.1711 vocab=0.5303 token=1.3154 | lr=9.98e-05 | 6.17 steps/s | VRAM=30.04 Go [2026-07-18 09:36:13] step 734800/1000000 | loss=2.2261 vocab=0.7584 token=1.1771 | lr=9.98e-05 | 6.15 steps/s | VRAM=32.59 Go [2026-07-18 09:36:21] step 734850/1000000 | loss=2.1825 vocab=1.1930 token=1.5432 | lr=9.98e-05 | 6.15 steps/s | VRAM=28.55 Go [2026-07-18 09:36:29] step 734900/1000000 | loss=2.2630 vocab=1.1627 token=1.5535 | lr=9.98e-05 | 6.15 steps/s | VRAM=28.54 Go [2026-07-18 09:36:37] step 734950/1000000 | loss=2.2102 vocab=0.9406 token=1.4975 | lr=9.98e-05 | 6.16 steps/s | VRAM=32.24 Go [2026-07-18 09:36:45] step 735000/1000000 | loss=2.2452 vocab=0.9724 token=1.2883 | lr=9.98e-05 | 6.15 steps/s | VRAM=28.63 Go [2026-07-18 09:36:53] step 735050/1000000 | loss=2.2260 vocab=0.8580 token=1.4403 | lr=9.98e-05 | 6.17 steps/s | VRAM=31.70 Go [2026-07-18 09:37:01] step 735100/1000000 | loss=2.1809 vocab=0.8459 token=1.1035 | lr=9.98e-05 | 6.18 steps/s | VRAM=28.37 Go [2026-07-18 09:37:10] step 735150/1000000 | loss=2.0775 vocab=0.6853 token=0.9674 | lr=9.98e-05 | 6.16 steps/s | VRAM=25.62 Go [2026-07-18 09:37:18] step 735200/1000000 | loss=2.1957 vocab=1.0672 token=1.6109 | lr=9.98e-05 | 6.14 steps/s | VRAM=24.75 Go [2026-07-18 09:37:26] step 735250/1000000 | loss=2.1397 vocab=0.4343 token=0.9449 | lr=9.98e-05 | 6.16 steps/s | VRAM=30.09 Go [2026-07-18 09:37:34] step 735300/1000000 | loss=2.2041 vocab=1.0163 token=1.3405 | lr=9.98e-05 | 6.15 steps/s | VRAM=26.46 Go [2026-07-18 09:37:42] step 735350/1000000 | loss=2.2587 vocab=0.8384 token=1.5185 | lr=9.98e-05 | 6.15 steps/s | VRAM=28.82 Go [2026-07-18 09:37:50] step 735400/1000000 | loss=2.2465 vocab=0.8299 token=1.0734 | lr=9.98e-05 | 6.14 steps/s | VRAM=26.28 Go [2026-07-18 09:37:58] step 735450/1000000 | loss=2.1678 vocab=0.9299 token=1.5830 | lr=9.98e-05 | 6.14 steps/s | VRAM=26.05 Go [2026-07-18 09:38:06] step 735500/1000000 | loss=2.1931 vocab=0.7088 token=1.1512 | lr=9.98e-05 | 6.16 steps/s | VRAM=31.66 Go [2026-07-18 09:38:15] step 735550/1000000 | loss=2.1454 vocab=0.6126 token=1.2658 | lr=9.98e-05 | 6.13 steps/s | VRAM=29.14 Go [2026-07-18 09:38:23] step 735600/1000000 | loss=2.1959 vocab=0.9851 token=1.3122 | lr=9.98e-05 | 6.13 steps/s | VRAM=29.33 Go [2026-07-18 09:38:31] step 735650/1000000 | loss=2.1483 vocab=0.6900 token=1.3007 | lr=9.98e-05 | 6.13 steps/s | VRAM=26.37 Go [2026-07-18 09:38:39] step 735700/1000000 | loss=2.1490 vocab=0.9755 token=1.5032 | lr=9.98e-05 | 6.18 steps/s | VRAM=31.27 Go [2026-07-18 09:38:47] step 735750/1000000 | loss=2.1835 vocab=0.9704 token=1.2280 | lr=9.98e-05 | 6.15 steps/s | VRAM=31.22 Go [2026-07-18 09:38:55] step 735800/1000000 | loss=2.2302 vocab=1.0906 token=1.2089 | lr=9.98e-05 | 6.13 steps/s | VRAM=29.77 Go [2026-07-18 09:39:03] step 735850/1000000 | loss=2.1507 vocab=0.4826 token=1.1332 | lr=9.98e-05 | 6.15 steps/s | VRAM=27.73 Go [2026-07-18 09:39:12] step 735900/1000000 | loss=2.1578 vocab=0.7637 token=1.3777 | lr=9.98e-05 | 6.14 steps/s | VRAM=29.21 Go [2026-07-18 09:39:20] step 735950/1000000 | loss=2.0970 vocab=0.4910 token=0.8964 | lr=9.98e-05 | 6.16 steps/s | VRAM=32.78 Go [2026-07-18 09:39:28] step 736000/1000000 | loss=2.1839 vocab=1.1116 token=1.2560 | lr=9.98e-05 | 6.09 steps/s | VRAM=26.60 Go [2026-07-18 09:39:36] step 736050/1000000 | loss=2.0908 vocab=0.7094 token=0.7230 | lr=9.98e-05 | 6.18 steps/s | VRAM=27.69 Go [2026-07-18 09:39:44] step 736100/1000000 | loss=2.0698 vocab=0.6238 token=1.5610 | lr=9.98e-05 | 6.13 steps/s | VRAM=29.59 Go [2026-07-18 09:39:52] step 736150/1000000 | loss=2.2001 vocab=0.8227 token=1.7793 | lr=9.98e-05 | 6.14 steps/s | VRAM=29.88 Go [2026-07-18 09:40:01] step 736200/1000000 | loss=2.2274 vocab=0.8760 token=1.1529 | lr=9.98e-05 | 6.13 steps/s | VRAM=27.96 Go [2026-07-18 09:40:09] step 736250/1000000 | loss=2.1406 vocab=0.8746 token=1.4117 | lr=9.97e-05 | 6.15 steps/s | VRAM=28.59 Go [2026-07-18 09:40:17] step 736300/1000000 | loss=2.0790 vocab=0.7020 token=1.5779 | lr=9.97e-05 | 6.19 steps/s | VRAM=31.95 Go [2026-07-18 09:40:25] step 736350/1000000 | loss=2.1650 vocab=0.5908 token=1.4012 | lr=9.97e-05 | 6.14 steps/s | VRAM=28.80 Go [2026-07-18 09:40:33] step 736400/1000000 | loss=2.2947 vocab=1.1221 token=1.5601 | lr=9.97e-05 | 6.12 steps/s | VRAM=28.71 Go [2026-07-18 09:40:41] step 736450/1000000 | loss=2.1749 vocab=1.0640 token=1.4838 | lr=9.97e-05 | 6.13 steps/s | VRAM=25.63 Go [2026-07-18 09:40:49] step 736500/1000000 | loss=2.1422 vocab=0.8324 token=1.3821 | lr=9.97e-05 | 6.14 steps/s | VRAM=29.47 Go [2026-07-18 09:40:57] step 736550/1000000 | loss=2.1645 vocab=1.2771 token=1.5380 | lr=9.97e-05 | 6.14 steps/s | VRAM=31.71 Go [2026-07-18 09:41:06] step 736600/1000000 | loss=2.0775 vocab=0.6383 token=0.9846 | lr=9.97e-05 | 6.16 steps/s | VRAM=28.69 Go [2026-07-18 09:41:14] step 736650/1000000 | loss=2.2520 vocab=0.9552 token=1.6666 | lr=9.97e-05 | 6.18 steps/s | VRAM=32.66 Go [2026-07-18 09:41:22] step 736700/1000000 | loss=2.1253 vocab=0.4195 token=1.1205 | lr=9.97e-05 | 6.14 steps/s | VRAM=29.26 Go [2026-07-18 09:41:30] step 736750/1000000 | loss=2.1209 vocab=0.9723 token=1.3281 | lr=9.97e-05 | 6.14 steps/s | VRAM=32.62 Go [2026-07-18 09:41:38] step 736800/1000000 | loss=2.2060 vocab=0.4454 token=1.2981 | lr=9.97e-05 | 6.14 steps/s | VRAM=32.78 Go [2026-07-18 09:41:46] step 736850/1000000 | loss=2.1990 vocab=0.8235 token=1.4557 | lr=9.97e-05 | 6.10 steps/s | VRAM=26.96 Go [2026-07-18 09:41:54] step 736900/1000000 | loss=2.1707 vocab=1.0780 token=1.1057 | lr=9.97e-05 | 6.18 steps/s | VRAM=30.89 Go [2026-07-18 09:42:03] step 736950/1000000 | loss=2.2210 vocab=1.4484 token=1.3833 | lr=9.97e-05 | 6.15 steps/s | VRAM=31.41 Go [2026-07-18 09:42:11] step 737000/1000000 | loss=2.3150 vocab=1.4156 token=1.6436 | lr=9.97e-05 | 6.15 steps/s | VRAM=29.17 Go [2026-07-18 09:42:19] step 737050/1000000 | loss=2.2304 vocab=0.6377 token=1.1474 | lr=9.97e-05 | 6.16 steps/s | VRAM=32.36 Go [2026-07-18 09:42:27] step 737100/1000000 | loss=2.2692 vocab=1.1610 token=1.3528 | lr=9.97e-05 | 6.13 steps/s | VRAM=25.55 Go [2026-07-18 09:42:35] step 737150/1000000 | loss=2.2626 vocab=0.9927 token=0.9621 | lr=9.97e-05 | 6.15 steps/s | VRAM=28.51 Go [2026-07-18 09:42:43] step 737200/1000000 | loss=2.0752 vocab=1.1076 token=1.4383 | lr=9.97e-05 | 6.15 steps/s | VRAM=29.54 Go [2026-07-18 09:42:51] step 737250/1000000 | loss=2.1840 vocab=1.2957 token=1.5459 | lr=9.97e-05 | 6.13 steps/s | VRAM=28.48 Go [2026-07-18 09:43:00] step 737300/1000000 | loss=2.2098 vocab=0.7696 token=0.9521 | lr=9.97e-05 | 6.11 steps/s | VRAM=26.19 Go [2026-07-18 09:43:08] step 737350/1000000 | loss=2.2506 vocab=0.9744 token=0.8214 | lr=9.97e-05 | 6.14 steps/s | VRAM=32.02 Go [2026-07-18 09:43:16] step 737400/1000000 | loss=2.2221 vocab=0.5207 token=1.2440 | lr=9.97e-05 | 6.12 steps/s | VRAM=26.37 Go [2026-07-18 09:43:24] step 737450/1000000 | loss=2.0730 vocab=0.4863 token=1.1750 | lr=9.97e-05 | 6.14 steps/s | VRAM=28.45 Go [2026-07-18 09:43:32] step 737500/1000000 | loss=2.1715 vocab=0.9154 token=1.5359 | lr=9.97e-05 | 6.16 steps/s | VRAM=27.72 Go [2026-07-18 09:43:40] step 737550/1000000 | loss=2.1682 vocab=0.7001 token=1.4309 | lr=9.97e-05 | 6.16 steps/s | VRAM=32.76 Go [2026-07-18 09:43:48] step 737600/1000000 | loss=2.1167 vocab=0.9454 token=1.4357 | lr=9.97e-05 | 6.13 steps/s | VRAM=28.86 Go [2026-07-18 09:43:57] step 737650/1000000 | loss=2.1502 vocab=1.2579 token=1.1470 | lr=9.97e-05 | 6.12 steps/s | VRAM=33.00 Go [2026-07-18 09:44:05] step 737700/1000000 | loss=2.2071 vocab=0.7783 token=1.4324 | lr=9.97e-05 | 6.15 steps/s | VRAM=31.59 Go [2026-07-18 09:44:13] step 737750/1000000 | loss=2.0119 vocab=0.4701 token=0.9367 | lr=9.97e-05 | 6.14 steps/s | VRAM=31.73 Go [2026-07-18 09:44:21] step 737800/1000000 | loss=2.2613 vocab=0.6270 token=1.4711 | lr=9.97e-05 | 6.15 steps/s | VRAM=32.31 Go [2026-07-18 09:44:29] step 737850/1000000 | loss=2.2354 vocab=1.0315 token=1.4872 | lr=9.96e-05 | 6.13 steps/s | VRAM=31.66 Go [2026-07-18 09:44:37] step 737900/1000000 | loss=2.2731 vocab=0.8632 token=1.5382 | lr=9.96e-05 | 6.12 steps/s | VRAM=29.37 Go [2026-07-18 09:44:45] step 737950/1000000 | loss=2.2716 vocab=0.9264 token=1.5121 | lr=9.96e-05 | 6.15 steps/s | VRAM=29.00 Go [2026-07-18 09:44:54] step 738000/1000000 | loss=2.1376 vocab=0.6948 token=0.9888 | lr=9.96e-05 | 6.14 steps/s | VRAM=28.61 Go [2026-07-18 09:45:02] step 738050/1000000 | loss=2.1564 vocab=0.8251 token=0.9779 | lr=9.96e-05 | 6.14 steps/s | VRAM=25.65 Go [2026-07-18 09:45:10] step 738100/1000000 | loss=2.2668 vocab=1.1298 token=1.0818 | lr=9.96e-05 | 6.17 steps/s | VRAM=31.96 Go [2026-07-18 09:45:18] step 738150/1000000 | loss=2.1446 vocab=0.7314 token=0.9682 | lr=9.96e-05 | 6.13 steps/s | VRAM=28.86 Go [2026-07-18 09:45:26] step 738200/1000000 | loss=2.1888 vocab=0.8052 token=1.6292 | lr=9.96e-05 | 6.10 steps/s | VRAM=26.94 Go [2026-07-18 09:45:34] step 738250/1000000 | loss=2.1318 vocab=0.8138 token=1.4073 | lr=9.96e-05 | 6.13 steps/s | VRAM=26.85 Go [2026-07-18 09:45:42] step 738300/1000000 | loss=2.0789 vocab=0.9377 token=1.4665 | lr=9.96e-05 | 6.14 steps/s | VRAM=31.82 Go [2026-07-18 09:45:51] step 738350/1000000 | loss=2.2012 vocab=0.9418 token=1.2469 | lr=9.96e-05 | 6.11 steps/s | VRAM=27.15 Go [2026-07-18 09:45:59] step 738400/1000000 | loss=2.3629 vocab=1.2772 token=1.5642 | lr=9.96e-05 | 6.14 steps/s | VRAM=28.11 Go [2026-07-18 09:46:07] step 738450/1000000 | loss=2.2127 vocab=1.0293 token=1.6441 | lr=9.96e-05 | 6.16 steps/s | VRAM=28.33 Go [2026-07-18 09:46:15] step 738500/1000000 | loss=2.1844 vocab=1.1329 token=1.7757 | lr=9.96e-05 | 6.13 steps/s | VRAM=29.14 Go [2026-07-18 09:46:23] step 738550/1000000 | loss=2.1443 vocab=0.9575 token=1.2644 | lr=9.96e-05 | 6.12 steps/s | VRAM=29.28 Go [2026-07-18 09:46:31] step 738600/1000000 | loss=2.1970 vocab=0.7150 token=1.4172 | lr=9.96e-05 | 6.17 steps/s | VRAM=32.00 Go [2026-07-18 09:46:39] step 738650/1000000 | loss=2.2094 vocab=0.8431 token=1.2081 | lr=9.96e-05 | 6.13 steps/s | VRAM=32.79 Go [2026-07-18 09:46:48] step 738700/1000000 | loss=2.1539 vocab=0.4487 token=1.0945 | lr=9.96e-05 | 6.14 steps/s | VRAM=29.41 Go [2026-07-18 09:46:56] step 738750/1000000 | loss=2.2391 vocab=0.6694 token=1.5375 | lr=9.96e-05 | 6.11 steps/s | VRAM=26.30 Go [2026-07-18 09:47:04] step 738800/1000000 | loss=2.1788 vocab=0.8892 token=1.2279 | lr=9.96e-05 | 6.17 steps/s | VRAM=28.13 Go [2026-07-18 09:47:12] step 738850/1000000 | loss=2.2266 vocab=0.5719 token=0.9941 | lr=9.96e-05 | 6.15 steps/s | VRAM=29.48 Go [2026-07-18 09:47:20] step 738900/1000000 | loss=2.1181 vocab=0.6610 token=1.2178 | lr=9.96e-05 | 6.11 steps/s | VRAM=29.08 Go [2026-07-18 09:47:28] step 738950/1000000 | loss=2.2501 vocab=1.1146 token=1.4635 | lr=9.96e-05 | 6.14 steps/s | VRAM=28.96 Go [2026-07-18 09:47:36] step 739000/1000000 | loss=2.1592 vocab=1.2624 token=1.5950 | lr=9.96e-05 | 6.15 steps/s | VRAM=31.67 Go [2026-07-18 09:47:45] step 739050/1000000 | loss=2.1808 vocab=0.9410 token=1.0310 | lr=9.96e-05 | 6.11 steps/s | VRAM=27.03 Go [2026-07-18 09:47:53] step 739100/1000000 | loss=2.2311 vocab=1.2876 token=1.5691 | lr=9.96e-05 | 6.18 steps/s | VRAM=29.59 Go [2026-07-18 09:48:01] step 739150/1000000 | loss=2.2807 vocab=0.8282 token=1.5059 | lr=9.96e-05 | 6.19 steps/s | VRAM=23.87 Go [2026-07-18 09:48:09] step 739200/1000000 | loss=2.2384 vocab=0.5567 token=1.2282 | lr=9.96e-05 | 6.17 steps/s | VRAM=29.62 Go [2026-07-18 09:48:17] step 739250/1000000 | loss=2.2624 vocab=1.1547 token=1.6835 | lr=9.95e-05 | 6.16 steps/s | VRAM=28.63 Go [2026-07-18 09:48:25] step 739300/1000000 | loss=2.0840 vocab=1.0271 token=1.6198 | lr=9.95e-05 | 6.17 steps/s | VRAM=29.65 Go [2026-07-18 09:48:33] step 739350/1000000 | loss=2.2420 vocab=0.5583 token=1.2226 | lr=9.95e-05 | 6.16 steps/s | VRAM=29.31 Go [2026-07-18 09:48:41] step 739400/1000000 | loss=2.2042 vocab=0.8010 token=1.3573 | lr=9.95e-05 | 6.20 steps/s | VRAM=31.65 Go [2026-07-18 09:48:49] step 739450/1000000 | loss=2.2189 vocab=1.2851 token=1.4392 | lr=9.95e-05 | 6.20 steps/s | VRAM=25.27 Go [2026-07-18 09:48:58] step 739500/1000000 | loss=2.1596 vocab=0.3809 token=1.0829 | lr=9.95e-05 | 6.17 steps/s | VRAM=28.49 Go [2026-07-18 09:49:06] step 739550/1000000 | loss=2.2363 vocab=0.5519 token=1.5036 | lr=9.95e-05 | 6.15 steps/s | VRAM=27.61 Go [2026-07-18 09:49:14] step 739600/1000000 | loss=2.1498 vocab=0.5396 token=1.0404 | lr=9.95e-05 | 6.12 steps/s | VRAM=26.03 Go [2026-07-18 09:49:22] step 739650/1000000 | loss=2.1854 vocab=0.7543 token=0.9081 | lr=9.95e-05 | 6.14 steps/s | VRAM=27.84 Go [2026-07-18 09:49:30] step 739700/1000000 | loss=2.2124 vocab=0.9882 token=1.3426 | lr=9.95e-05 | 6.12 steps/s | VRAM=28.82 Go [2026-07-18 09:49:38] step 739750/1000000 | loss=2.2073 vocab=0.3752 token=1.0259 | lr=9.95e-05 | 6.15 steps/s | VRAM=31.07 Go [2026-07-18 09:49:46] step 739800/1000000 | loss=2.1678 vocab=1.2422 token=1.4550 | lr=9.95e-05 | 6.15 steps/s | VRAM=31.55 Go [2026-07-18 09:49:55] step 739850/1000000 | loss=2.2789 vocab=0.8600 token=1.5365 | lr=9.95e-05 | 6.12 steps/s | VRAM=28.86 Go [2026-07-18 09:50:03] step 739900/1000000 | loss=2.1479 vocab=1.0609 token=1.9602 | lr=9.95e-05 | 6.16 steps/s | VRAM=32.09 Go [2026-07-18 09:50:11] step 739950/1000000 | loss=2.2304 vocab=1.2917 token=1.2999 | lr=9.95e-05 | 6.13 steps/s | VRAM=28.90 Go [2026-07-18 09:50:19] step 740000/1000000 | loss=2.1697 vocab=0.7904 token=1.0201 | lr=9.95e-05 | 6.13 steps/s | VRAM=24.88 Go [2026-07-18 09:50:27] step 740050/1000000 | loss=2.2811 vocab=1.6245 token=1.6498 | lr=9.95e-05 | 6.13 steps/s | VRAM=32.85 Go [2026-07-18 09:50:35] step 740100/1000000 | loss=2.1218 vocab=0.7728 token=1.2099 | lr=9.95e-05 | 6.11 steps/s | VRAM=30.20 Go [2026-07-18 09:50:44] step 740150/1000000 | loss=2.2030 vocab=0.6736 token=1.1961 | lr=9.95e-05 | 6.11 steps/s | VRAM=27.34 Go [2026-07-18 09:50:52] step 740200/1000000 | loss=2.1320 vocab=0.8040 token=1.4658 | lr=9.95e-05 | 6.19 steps/s | VRAM=32.81 Go [2026-07-18 09:51:00] step 740250/1000000 | loss=2.2857 vocab=1.2700 token=1.3152 | lr=9.95e-05 | 6.15 steps/s | VRAM=29.37 Go [2026-07-18 09:51:08] step 740300/1000000 | loss=2.2233 vocab=0.4466 token=0.5567 | lr=9.95e-05 | 6.16 steps/s | VRAM=32.14 Go [2026-07-18 09:51:16] step 740350/1000000 | loss=2.2010 vocab=1.0556 token=1.6468 | lr=9.95e-05 | 6.13 steps/s | VRAM=29.54 Go [2026-07-18 09:51:24] step 740400/1000000 | loss=2.2596 vocab=0.7977 token=1.3914 | lr=9.95e-05 | 6.15 steps/s | VRAM=29.65 Go [2026-07-18 09:51:32] step 740450/1000000 | loss=2.1087 vocab=1.0159 token=1.3367 | lr=9.95e-05 | 6.15 steps/s | VRAM=32.02 Go [2026-07-18 09:51:40] step 740500/1000000 | loss=2.2240 vocab=0.5685 token=1.1184 | lr=9.94e-05 | 6.17 steps/s | VRAM=29.05 Go [2026-07-18 09:51:48] step 740550/1000000 | loss=2.1174 vocab=1.1348 token=1.0969 | lr=9.94e-05 | 6.17 steps/s | VRAM=32.21 Go [2026-07-18 09:51:57] step 740600/1000000 | loss=2.1175 vocab=1.1170 token=1.3266 | lr=9.94e-05 | 6.12 steps/s | VRAM=28.52 Go [2026-07-18 09:52:05] step 740650/1000000 | loss=2.2094 vocab=1.2822 token=1.7027 | lr=9.94e-05 | 6.13 steps/s | VRAM=29.32 Go [2026-07-18 09:52:13] step 740700/1000000 | loss=2.2124 vocab=1.1972 token=1.2074 | lr=9.94e-05 | 6.16 steps/s | VRAM=28.05 Go [2026-07-18 09:52:21] step 740750/1000000 | loss=2.0666 vocab=0.7503 token=1.3753 | lr=9.94e-05 | 6.16 steps/s | VRAM=28.28 Go [2026-07-18 09:52:29] step 740800/1000000 | loss=2.0333 vocab=0.6800 token=1.1787 | lr=9.94e-05 | 6.17 steps/s | VRAM=31.98 Go [2026-07-18 09:52:37] step 740850/1000000 | loss=2.2405 vocab=1.3278 token=1.3732 | lr=9.94e-05 | 6.17 steps/s | VRAM=31.48 Go [2026-07-18 09:52:45] step 740900/1000000 | loss=2.2435 vocab=0.6289 token=1.3228 | lr=9.94e-05 | 6.17 steps/s | VRAM=31.97 Go [2026-07-18 09:52:53] step 740950/1000000 | loss=2.1615 vocab=0.5272 token=1.2982 | lr=9.94e-05 | 6.17 steps/s | VRAM=31.71 Go [2026-07-18 09:53:02] step 741000/1000000 | loss=2.2598 vocab=1.1184 token=1.4484 | lr=9.94e-05 | 6.14 steps/s | VRAM=28.96 Go [2026-07-18 09:53:10] step 741050/1000000 | loss=2.1663 vocab=0.5202 token=1.1569 | lr=9.94e-05 | 6.13 steps/s | VRAM=32.94 Go [2026-07-18 09:53:18] step 741100/1000000 | loss=2.0803 vocab=1.3867 token=1.2538 | lr=9.94e-05 | 6.17 steps/s | VRAM=31.67 Go [2026-07-18 09:53:26] step 741150/1000000 | loss=2.1865 vocab=1.0220 token=1.3726 | lr=9.94e-05 | 6.14 steps/s | VRAM=32.15 Go [2026-07-18 09:53:34] step 741200/1000000 | loss=2.2393 vocab=0.6894 token=1.0691 | lr=9.94e-05 | 6.16 steps/s | VRAM=30.95 Go [2026-07-18 09:53:42] step 741250/1000000 | loss=2.2375 vocab=0.7466 token=1.1645 | lr=9.94e-05 | 6.16 steps/s | VRAM=26.34 Go [2026-07-18 09:53:50] step 741300/1000000 | loss=2.1608 vocab=0.6595 token=1.3847 | lr=9.94e-05 | 6.15 steps/s | VRAM=26.17 Go [2026-07-18 09:53:58] step 741350/1000000 | loss=2.2105 vocab=0.5800 token=1.1175 | lr=9.94e-05 | 6.17 steps/s | VRAM=31.94 Go [2026-07-18 09:54:07] step 741400/1000000 | loss=2.1916 vocab=0.8609 token=1.4314 | lr=9.94e-05 | 6.17 steps/s | VRAM=29.28 Go [2026-07-18 09:54:15] step 741450/1000000 | loss=2.1682 vocab=0.5523 token=0.9796 | lr=9.94e-05 | 6.11 steps/s | VRAM=30.01 Go [2026-07-18 09:54:23] step 741500/1000000 | loss=2.1429 vocab=0.9860 token=1.7868 | lr=9.94e-05 | 6.16 steps/s | VRAM=29.83 Go [2026-07-18 09:54:31] step 741550/1000000 | loss=2.2003 vocab=0.9296 token=1.1934 | lr=9.94e-05 | 6.14 steps/s | VRAM=25.47 Go [2026-07-18 09:54:39] step 741600/1000000 | loss=2.1969 vocab=0.6529 token=1.4411 | lr=9.93e-05 | 6.15 steps/s | VRAM=28.48 Go [2026-07-18 09:54:47] step 741650/1000000 | loss=2.1869 vocab=0.6814 token=0.8539 | lr=9.93e-05 | 6.15 steps/s | VRAM=29.25 Go [2026-07-18 09:54:55] step 741700/1000000 | loss=2.1934 vocab=0.6051 token=1.1276 | lr=9.93e-05 | 6.17 steps/s | VRAM=32.14 Go [2026-07-18 09:55:04] step 741750/1000000 | loss=2.0871 vocab=1.1490 token=1.2544 | lr=9.93e-05 | 6.11 steps/s | VRAM=26.72 Go [2026-07-18 09:55:12] step 741800/1000000 | loss=2.2032 vocab=0.8296 token=1.0418 | lr=9.93e-05 | 6.14 steps/s | VRAM=26.19 Go [2026-07-18 09:55:20] step 741850/1000000 | loss=2.2045 vocab=0.7150 token=1.0687 | lr=9.93e-05 | 6.14 steps/s | VRAM=26.12 Go [2026-07-18 09:55:28] step 741900/1000000 | loss=2.1559 vocab=0.7044 token=1.4919 | lr=9.93e-05 | 6.16 steps/s | VRAM=31.57 Go [2026-07-18 09:55:36] step 741950/1000000 | loss=2.1867 vocab=0.7962 token=1.0275 | lr=9.93e-05 | 6.12 steps/s | VRAM=28.35 Go [2026-07-18 09:55:44] step 742000/1000000 | loss=2.1702 vocab=0.8878 token=1.4805 | lr=9.93e-05 | 6.15 steps/s | VRAM=32.30 Go [2026-07-18 09:55:52] step 742050/1000000 | loss=2.1774 vocab=0.8293 token=1.0012 | lr=9.93e-05 | 6.16 steps/s | VRAM=28.56 Go [2026-07-18 09:56:01] step 742100/1000000 | loss=2.1647 vocab=0.4282 token=1.0146 | lr=9.93e-05 | 6.14 steps/s | VRAM=26.58 Go [2026-07-18 09:56:09] step 742150/1000000 | loss=2.1153 vocab=0.6560 token=1.3125 | lr=9.93e-05 | 6.12 steps/s | VRAM=28.74 Go [2026-07-18 09:56:17] step 742200/1000000 | loss=2.1391 vocab=0.6467 token=0.6569 | lr=9.93e-05 | 6.12 steps/s | VRAM=28.86 Go [2026-07-18 09:56:25] step 742250/1000000 | loss=2.2438 vocab=1.3351 token=1.2824 | lr=9.93e-05 | 6.11 steps/s | VRAM=29.36 Go [2026-07-18 09:56:33] step 742300/1000000 | loss=2.1398 vocab=0.7250 token=1.2651 | lr=9.93e-05 | 6.14 steps/s | VRAM=26.74 Go [2026-07-18 09:56:41] step 742350/1000000 | loss=2.1492 vocab=1.0022 token=1.0184 | lr=9.93e-05 | 6.15 steps/s | VRAM=28.06 Go [2026-07-18 09:56:50] step 742400/1000000 | loss=2.1438 vocab=0.8737 token=1.0874 | lr=9.93e-05 | 6.11 steps/s | VRAM=26.14 Go [2026-07-18 09:56:58] step 742450/1000000 | loss=2.2515 vocab=0.9422 token=1.2195 | lr=9.93e-05 | 6.16 steps/s | VRAM=24.64 Go [2026-07-18 09:57:06] step 742500/1000000 | loss=2.1722 vocab=1.4748 token=1.5040 | lr=9.93e-05 | 6.11 steps/s | VRAM=28.45 Go [2026-07-18 09:57:14] step 742550/1000000 | loss=2.1931 vocab=1.0771 token=1.2471 | lr=9.93e-05 | 6.16 steps/s | VRAM=27.78 Go [2026-07-18 09:57:22] step 742600/1000000 | loss=2.1641 vocab=0.6834 token=1.2144 | lr=9.93e-05 | 6.14 steps/s | VRAM=26.05 Go [2026-07-18 09:57:30] step 742650/1000000 | loss=2.2054 vocab=1.0266 token=1.3177 | lr=9.92e-05 | 6.15 steps/s | VRAM=32.80 Go [2026-07-18 09:57:38] step 742700/1000000 | loss=2.2435 vocab=1.0710 token=1.1014 | lr=9.92e-05 | 6.16 steps/s | VRAM=28.87 Go [2026-07-18 09:57:46] step 742750/1000000 | loss=2.2263 vocab=1.1802 token=1.4559 | lr=9.92e-05 | 6.14 steps/s | VRAM=28.49 Go [2026-07-18 09:57:55] step 742800/1000000 | loss=2.1675 vocab=1.4914 token=1.2126 | lr=9.92e-05 | 6.16 steps/s | VRAM=28.49 Go [2026-07-18 09:58:03] step 742850/1000000 | loss=2.2656 vocab=1.0661 token=1.4850 | lr=9.92e-05 | 6.15 steps/s | VRAM=28.21 Go [2026-07-18 09:58:11] step 742900/1000000 | loss=2.1362 vocab=1.3125 token=1.3476 | lr=9.92e-05 | 6.16 steps/s | VRAM=31.62 Go [2026-07-18 09:58:19] step 742950/1000000 | loss=2.1873 vocab=1.0244 token=1.6022 | lr=9.92e-05 | 6.15 steps/s | VRAM=28.92 Go [2026-07-18 09:58:27] step 743000/1000000 | loss=2.1396 vocab=0.5930 token=1.1169 | lr=9.92e-05 | 6.11 steps/s | VRAM=25.00 Go [2026-07-18 09:58:35] step 743050/1000000 | loss=2.2545 vocab=0.6696 token=1.4480 | lr=9.92e-05 | 6.14 steps/s | VRAM=32.83 Go [2026-07-18 09:58:43] step 743100/1000000 | loss=2.2203 vocab=0.4936 token=0.9338 | lr=9.92e-05 | 6.10 steps/s | VRAM=29.24 Go [2026-07-18 09:58:52] step 743150/1000000 | loss=2.2527 vocab=0.8234 token=0.8169 | lr=9.92e-05 | 6.15 steps/s | VRAM=29.44 Go [2026-07-18 09:59:00] step 743200/1000000 | loss=2.1568 vocab=1.1254 token=1.2930 | lr=9.92e-05 | 6.16 steps/s | VRAM=29.49 Go [2026-07-18 09:59:08] step 743250/1000000 | loss=2.0953 vocab=0.6794 token=0.9613 | lr=9.92e-05 | 6.13 steps/s | VRAM=29.53 Go [2026-07-18 09:59:16] step 743300/1000000 | loss=2.1585 vocab=0.9112 token=1.1267 | lr=9.92e-05 | 6.17 steps/s | VRAM=28.15 Go [2026-07-18 09:59:24] step 743350/1000000 | loss=2.2414 vocab=0.8724 token=1.3581 | lr=9.92e-05 | 6.13 steps/s | VRAM=27.04 Go [2026-07-18 09:59:32] step 743400/1000000 | loss=2.1328 vocab=0.9078 token=1.2604 | lr=9.92e-05 | 6.15 steps/s | VRAM=28.10 Go [2026-07-18 09:59:40] step 743450/1000000 | loss=2.1838 vocab=1.3309 token=1.2336 | lr=9.92e-05 | 6.09 steps/s | VRAM=26.73 Go [2026-07-18 09:59:49] step 743500/1000000 | loss=2.2766 vocab=0.8011 token=1.3511 | lr=9.92e-05 | 6.12 steps/s | VRAM=27.94 Go [2026-07-18 09:59:57] step 743550/1000000 | loss=2.2240 vocab=0.8210 token=1.5231 | lr=9.92e-05 | 6.14 steps/s | VRAM=25.68 Go [2026-07-18 10:00:05] step 743600/1000000 | loss=2.0606 vocab=0.9615 token=1.5787 | lr=9.92e-05 | 6.16 steps/s | VRAM=32.14 Go [2026-07-18 10:00:13] step 743650/1000000 | loss=2.1075 vocab=1.0396 token=1.2932 | lr=9.91e-05 | 6.14 steps/s | VRAM=25.05 Go [2026-07-18 10:00:21] step 743700/1000000 | loss=2.1291 vocab=0.6199 token=1.1803 | lr=9.91e-05 | 6.15 steps/s | VRAM=27.98 Go [2026-07-18 10:00:29] step 743750/1000000 | loss=2.2107 vocab=1.0927 token=1.3455 | lr=9.91e-05 | 6.16 steps/s | VRAM=28.20 Go [2026-07-18 10:00:37] step 743800/1000000 | loss=2.1720 vocab=0.8990 token=1.2363 | lr=9.91e-05 | 6.13 steps/s | VRAM=26.39 Go [2026-07-18 10:00:46] step 743850/1000000 | loss=2.1368 vocab=0.9236 token=0.9265 | lr=9.91e-05 | 6.14 steps/s | VRAM=29.99 Go [2026-07-18 10:00:54] step 743900/1000000 | loss=2.2232 vocab=0.6238 token=1.1216 | lr=9.91e-05 | 6.15 steps/s | VRAM=25.18 Go [2026-07-18 10:01:02] step 743950/1000000 | loss=2.1698 vocab=0.8227 token=1.0513 | lr=9.91e-05 | 6.18 steps/s | VRAM=27.86 Go [2026-07-18 10:01:10] step 744000/1000000 | loss=2.2132 vocab=0.9468 token=1.4013 | lr=9.91e-05 | 6.13 steps/s | VRAM=27.10 Go [2026-07-18 10:01:18] step 744050/1000000 | loss=2.2608 vocab=0.5861 token=1.1684 | lr=9.91e-05 | 6.16 steps/s | VRAM=24.98 Go [2026-07-18 10:01:26] step 744100/1000000 | loss=2.1981 vocab=0.6742 token=1.0876 | lr=9.91e-05 | 6.11 steps/s | VRAM=28.47 Go [2026-07-18 10:01:34] step 744150/1000000 | loss=2.2356 vocab=0.6025 token=1.0755 | lr=9.91e-05 | 6.16 steps/s | VRAM=31.37 Go [2026-07-18 10:01:43] step 744200/1000000 | loss=2.2061 vocab=0.6555 token=0.9528 | lr=9.91e-05 | 6.15 steps/s | VRAM=30.04 Go [2026-07-18 10:01:51] step 744250/1000000 | loss=2.1265 vocab=1.0767 token=1.4601 | lr=9.91e-05 | 6.12 steps/s | VRAM=28.18 Go [2026-07-18 10:01:59] step 744300/1000000 | loss=2.2192 vocab=0.8223 token=1.3122 | lr=9.91e-05 | 6.11 steps/s | VRAM=26.90 Go [2026-07-18 10:02:07] step 744350/1000000 | loss=2.2251 vocab=0.8763 token=1.3650 | lr=9.91e-05 | 6.14 steps/s | VRAM=29.53 Go [2026-07-18 10:02:15] step 744400/1000000 | loss=2.2176 vocab=0.9241 token=1.6464 | lr=9.91e-05 | 6.17 steps/s | VRAM=33.02 Go [2026-07-18 10:02:23] step 744450/1000000 | loss=2.1459 vocab=0.9160 token=1.3538 | lr=9.91e-05 | 6.15 steps/s | VRAM=26.31 Go [2026-07-18 10:02:31] step 744500/1000000 | loss=2.2220 vocab=1.0316 token=1.3815 | lr=9.91e-05 | 6.15 steps/s | VRAM=31.85 Go [2026-07-18 10:02:40] step 744550/1000000 | loss=2.2326 vocab=0.9155 token=0.8996 | lr=9.90e-05 | 6.17 steps/s | VRAM=24.80 Go [2026-07-18 10:02:48] step 744600/1000000 | loss=2.1423 vocab=0.6823 token=0.6379 | lr=9.90e-05 | 6.17 steps/s | VRAM=28.83 Go [2026-07-18 10:02:56] step 744650/1000000 | loss=2.1177 vocab=0.8934 token=0.8934 | lr=9.90e-05 | 6.15 steps/s | VRAM=28.35 Go [2026-07-18 10:03:04] step 744700/1000000 | loss=2.1639 vocab=0.8994 token=1.1095 | lr=9.90e-05 | 6.17 steps/s | VRAM=31.12 Go [2026-07-18 10:03:12] step 744750/1000000 | loss=2.2452 vocab=1.1991 token=1.3327 | lr=9.90e-05 | 6.12 steps/s | VRAM=26.65 Go [2026-07-18 10:03:20] step 744800/1000000 | loss=2.1386 vocab=0.6830 token=1.2514 | lr=9.90e-05 | 6.17 steps/s | VRAM=26.00 Go [2026-07-18 10:03:28] step 744850/1000000 | loss=2.1823 vocab=1.1127 token=1.4431 | lr=9.90e-05 | 6.14 steps/s | VRAM=28.84 Go [2026-07-18 10:03:36] step 744900/1000000 | loss=2.1933 vocab=0.5722 token=1.2363 | lr=9.90e-05 | 6.15 steps/s | VRAM=28.11 Go [2026-07-18 10:03:45] step 744950/1000000 | loss=2.1144 vocab=0.7101 token=1.4788 | lr=9.90e-05 | 6.13 steps/s | VRAM=28.41 Go [2026-07-18 10:03:53] step 745000/1000000 | loss=2.1777 vocab=1.0481 token=1.1314 | lr=9.90e-05 | 6.15 steps/s | VRAM=32.39 Go [2026-07-18 10:04:01] step 745050/1000000 | loss=2.2877 vocab=0.7369 token=1.0946 | lr=9.90e-05 | 6.15 steps/s | VRAM=31.61 Go [2026-07-18 10:04:09] step 745100/1000000 | loss=2.1505 vocab=0.7242 token=1.1030 | lr=9.90e-05 | 6.13 steps/s | VRAM=32.94 Go [2026-07-18 10:04:17] step 745150/1000000 | loss=2.2024 vocab=0.6477 token=1.4280 | lr=9.90e-05 | 6.12 steps/s | VRAM=32.45 Go [2026-07-18 10:04:25] step 745200/1000000 | loss=2.1092 vocab=1.0841 token=1.3335 | lr=9.90e-05 | 6.14 steps/s | VRAM=26.10 Go [2026-07-18 10:04:33] step 745250/1000000 | loss=2.1316 vocab=1.1610 token=1.4139 | lr=9.90e-05 | 6.15 steps/s | VRAM=28.07 Go [2026-07-18 10:04:42] step 745300/1000000 | loss=2.2434 vocab=0.7029 token=1.1446 | lr=9.90e-05 | 6.12 steps/s | VRAM=28.80 Go [2026-07-18 10:04:50] step 745350/1000000 | loss=2.2699 vocab=0.6359 token=0.9645 | lr=9.90e-05 | 6.14 steps/s | VRAM=29.63 Go [2026-07-18 10:04:58] step 745400/1000000 | loss=2.2859 vocab=0.6583 token=1.1686 | lr=9.90e-05 | 6.13 steps/s | VRAM=29.66 Go [2026-07-18 10:05:06] step 745450/1000000 | loss=2.2409 vocab=0.9618 token=1.5912 | lr=9.89e-05 | 6.14 steps/s | VRAM=28.97 Go [2026-07-18 10:05:14] step 745500/1000000 | loss=2.3518 vocab=1.1728 token=1.6622 | lr=9.89e-05 | 6.20 steps/s | VRAM=24.60 Go [2026-07-18 10:05:22] step 745550/1000000 | loss=2.1925 vocab=0.9191 token=1.3908 | lr=9.89e-05 | 6.12 steps/s | VRAM=28.54 Go [2026-07-18 10:05:30] step 745600/1000000 | loss=2.1840 vocab=1.1534 token=1.2172 | lr=9.89e-05 | 6.16 steps/s | VRAM=32.79 Go [2026-07-18 10:05:38] step 745650/1000000 | loss=2.3942 vocab=1.0075 token=1.6733 | lr=9.89e-05 | 6.15 steps/s | VRAM=27.76 Go [2026-07-18 10:05:47] step 745700/1000000 | loss=2.1689 vocab=0.6485 token=1.2188 | lr=9.89e-05 | 6.14 steps/s | VRAM=26.92 Go [2026-07-18 10:05:55] step 745750/1000000 | loss=2.2291 vocab=0.4784 token=1.2332 | lr=9.89e-05 | 6.16 steps/s | VRAM=26.24 Go [2026-07-18 10:06:03] step 745800/1000000 | loss=2.2320 vocab=0.7426 token=1.8897 | lr=9.89e-05 | 6.13 steps/s | VRAM=28.96 Go [2026-07-18 10:06:11] step 745850/1000000 | loss=2.1741 vocab=0.6926 token=1.3368 | lr=9.89e-05 | 6.16 steps/s | VRAM=28.81 Go [2026-07-18 10:06:19] step 745900/1000000 | loss=2.1888 vocab=1.1330 token=1.2586 | lr=9.89e-05 | 6.13 steps/s | VRAM=28.25 Go [2026-07-18 10:06:27] step 745950/1000000 | loss=2.1754 vocab=0.7594 token=0.9949 | lr=9.89e-05 | 6.16 steps/s | VRAM=32.77 Go [2026-07-18 10:06:35] step 746000/1000000 | loss=2.2458 vocab=0.7800 token=1.1673 | lr=9.89e-05 | 6.11 steps/s | VRAM=29.72 Go [2026-07-18 10:06:44] step 746050/1000000 | loss=2.2623 vocab=0.8719 token=1.8875 | lr=9.89e-05 | 6.17 steps/s | VRAM=28.84 Go [2026-07-18 10:06:52] step 746100/1000000 | loss=2.3377 vocab=0.9141 token=1.3408 | lr=9.89e-05 | 6.12 steps/s | VRAM=26.18 Go [2026-07-18 10:07:00] step 746150/1000000 | loss=2.1753 vocab=0.7612 token=1.3941 | lr=9.89e-05 | 6.17 steps/s | VRAM=31.34 Go [2026-07-18 10:07:08] step 746200/1000000 | loss=2.2401 vocab=0.9217 token=0.8628 | lr=9.89e-05 | 6.17 steps/s | VRAM=28.43 Go [2026-07-18 10:07:16] step 746250/1000000 | loss=2.1708 vocab=0.6318 token=0.9037 | lr=9.88e-05 | 6.18 steps/s | VRAM=30.28 Go [2026-07-18 10:07:24] step 746300/1000000 | loss=2.2234 vocab=0.6584 token=0.9679 | lr=9.88e-05 | 6.14 steps/s | VRAM=29.15 Go [2026-07-18 10:07:32] step 746350/1000000 | loss=2.0428 vocab=0.3458 token=0.7206 | lr=9.88e-05 | 6.14 steps/s | VRAM=27.09 Go [2026-07-18 10:07:40] step 746400/1000000 | loss=2.1496 vocab=0.8961 token=1.1664 | lr=9.88e-05 | 6.18 steps/s | VRAM=25.55 Go [2026-07-18 10:07:49] step 746450/1000000 | loss=2.1861 vocab=0.5595 token=1.2549 | lr=9.88e-05 | 6.16 steps/s | VRAM=28.88 Go [2026-07-18 10:07:57] step 746500/1000000 | loss=2.2733 vocab=0.5623 token=1.1357 | lr=9.88e-05 | 6.14 steps/s | VRAM=29.24 Go [2026-07-18 10:08:05] step 746550/1000000 | loss=2.1207 vocab=1.0016 token=1.4010 | lr=9.88e-05 | 6.18 steps/s | VRAM=28.39 Go [2026-07-18 10:08:13] step 746600/1000000 | loss=2.1694 vocab=0.7214 token=1.2913 | lr=9.88e-05 | 6.12 steps/s | VRAM=28.94 Go [2026-07-18 10:08:21] step 746650/1000000 | loss=2.2259 vocab=1.2156 token=1.9370 | lr=9.88e-05 | 6.18 steps/s | VRAM=31.91 Go [2026-07-18 10:08:29] step 746700/1000000 | loss=2.2783 vocab=0.9542 token=0.9619 | lr=9.88e-05 | 6.17 steps/s | VRAM=27.77 Go [2026-07-18 10:08:37] step 746750/1000000 | loss=2.1470 vocab=1.0140 token=1.3857 | lr=9.88e-05 | 6.14 steps/s | VRAM=32.19 Go [2026-07-18 10:08:45] step 746800/1000000 | loss=2.1135 vocab=1.5598 token=1.1647 | lr=9.88e-05 | 6.15 steps/s | VRAM=25.70 Go [2026-07-18 10:08:54] step 746850/1000000 | loss=2.1149 vocab=0.6595 token=1.2356 | lr=9.88e-05 | 6.17 steps/s | VRAM=31.37 Go [2026-07-18 10:09:02] step 746900/1000000 | loss=2.2540 vocab=0.7104 token=1.2348 | lr=9.88e-05 | 6.17 steps/s | VRAM=29.00 Go [2026-07-18 10:09:10] step 746950/1000000 | loss=2.1627 vocab=0.6934 token=1.5100 | lr=9.88e-05 | 6.15 steps/s | VRAM=26.11 Go [2026-07-18 10:09:18] step 747000/1000000 | loss=2.2249 vocab=0.7645 token=1.3260 | lr=9.88e-05 | 6.17 steps/s | VRAM=25.55 Go [2026-07-18 10:09:26] step 747050/1000000 | loss=2.1021 vocab=0.5131 token=0.8544 | lr=9.87e-05 | 6.16 steps/s | VRAM=28.19 Go [2026-07-18 10:09:34] step 747100/1000000 | loss=2.1838 vocab=1.0165 token=1.4838 | lr=9.87e-05 | 6.17 steps/s | VRAM=27.29 Go [2026-07-18 10:09:42] step 747150/1000000 | loss=2.2081 vocab=0.7304 token=0.7363 | lr=9.87e-05 | 6.15 steps/s | VRAM=29.80 Go [2026-07-18 10:09:50] step 747200/1000000 | loss=2.0572 vocab=1.2279 token=1.6658 | lr=9.87e-05 | 6.16 steps/s | VRAM=32.31 Go [2026-07-18 10:09:58] step 747250/1000000 | loss=2.1566 vocab=0.9271 token=1.4679 | lr=9.87e-05 | 6.17 steps/s | VRAM=28.06 Go [2026-07-18 10:10:07] step 747300/1000000 | loss=2.2118 vocab=0.7479 token=1.2675 | lr=9.87e-05 | 6.17 steps/s | VRAM=28.10 Go [2026-07-18 10:10:15] step 747350/1000000 | loss=2.1762 vocab=0.6956 token=1.3636 | lr=9.87e-05 | 6.14 steps/s | VRAM=27.85 Go [2026-07-18 10:10:23] step 747400/1000000 | loss=2.0188 vocab=0.9230 token=1.5192 | lr=9.87e-05 | 6.17 steps/s | VRAM=25.11 Go [2026-07-18 10:10:31] step 747450/1000000 | loss=2.2071 vocab=0.5565 token=1.0422 | lr=9.87e-05 | 6.16 steps/s | VRAM=31.62 Go [2026-07-18 10:10:39] step 747500/1000000 | loss=2.2265 vocab=0.8790 token=1.5332 | lr=9.87e-05 | 6.15 steps/s | VRAM=25.25 Go [2026-07-18 10:10:47] step 747550/1000000 | loss=2.2025 vocab=0.5944 token=0.9894 | lr=9.87e-05 | 6.17 steps/s | VRAM=25.26 Go [2026-07-18 10:10:55] step 747600/1000000 | loss=2.2989 vocab=1.1009 token=1.0985 | lr=9.87e-05 | 6.16 steps/s | VRAM=28.46 Go [2026-07-18 10:11:03] step 747650/1000000 | loss=2.1892 vocab=0.8167 token=1.2293 | lr=9.87e-05 | 6.16 steps/s | VRAM=28.85 Go [2026-07-18 10:11:12] step 747700/1000000 | loss=2.2403 vocab=0.8950 token=1.2288 | lr=9.87e-05 | 6.12 steps/s | VRAM=26.37 Go [2026-07-18 10:11:20] step 747750/1000000 | loss=2.1718 vocab=1.1130 token=1.3374 | lr=9.87e-05 | 6.14 steps/s | VRAM=28.96 Go [2026-07-18 10:11:28] step 747800/1000000 | loss=2.2212 vocab=0.9437 token=1.3761 | lr=9.87e-05 | 6.15 steps/s | VRAM=27.85 Go [2026-07-18 10:11:36] step 747850/1000000 | loss=2.2203 vocab=1.0709 token=1.3842 | lr=9.86e-05 | 6.15 steps/s | VRAM=29.39 Go [2026-07-18 10:11:44] step 747900/1000000 | loss=2.2196 vocab=0.7941 token=1.3703 | lr=9.86e-05 | 6.12 steps/s | VRAM=29.38 Go [2026-07-18 10:11:52] step 747950/1000000 | loss=2.2063 vocab=0.8951 token=1.1815 | lr=9.86e-05 | 6.13 steps/s | VRAM=23.32 Go [2026-07-18 10:12:00] step 748000/1000000 | loss=2.1803 vocab=1.1033 token=1.0910 | lr=9.86e-05 | 6.16 steps/s | VRAM=32.90 Go [2026-07-18 10:12:09] step 748050/1000000 | loss=2.1539 vocab=1.0024 token=1.3280 | lr=9.86e-05 | 6.14 steps/s | VRAM=26.30 Go [2026-07-18 10:12:17] step 748100/1000000 | loss=2.2517 vocab=0.9493 token=1.4766 | lr=9.86e-05 | 6.14 steps/s | VRAM=29.24 Go [2026-07-18 10:12:25] step 748150/1000000 | loss=2.2126 vocab=0.5806 token=1.1562 | lr=9.86e-05 | 6.16 steps/s | VRAM=29.19 Go [2026-07-18 10:12:33] step 748200/1000000 | loss=2.0832 vocab=0.8056 token=1.3519 | lr=9.86e-05 | 6.15 steps/s | VRAM=31.84 Go [2026-07-18 10:12:41] step 748250/1000000 | loss=2.2106 vocab=0.8708 token=1.1312 | lr=9.86e-05 | 6.15 steps/s | VRAM=32.70 Go [2026-07-18 10:12:49] step 748300/1000000 | loss=2.2490 vocab=0.6382 token=1.0983 | lr=9.86e-05 | 6.14 steps/s | VRAM=26.31 Go [2026-07-18 10:12:57] step 748350/1000000 | loss=2.1449 vocab=0.9039 token=1.0726 | lr=9.86e-05 | 6.10 steps/s | VRAM=27.85 Go [2026-07-18 10:13:06] step 748400/1000000 | loss=2.0556 vocab=1.0002 token=1.0902 | lr=9.86e-05 | 6.14 steps/s | VRAM=30.40 Go [2026-07-18 10:13:14] step 748450/1000000 | loss=2.2148 vocab=0.6354 token=1.0365 | lr=9.86e-05 | 6.10 steps/s | VRAM=30.07 Go [2026-07-18 10:13:22] step 748500/1000000 | loss=2.1273 vocab=1.5569 token=1.3616 | lr=9.86e-05 | 6.15 steps/s | VRAM=28.64 Go [2026-07-18 10:13:30] step 748550/1000000 | loss=2.1669 vocab=0.6308 token=1.1930 | lr=9.86e-05 | 6.14 steps/s | VRAM=29.21 Go [2026-07-18 10:13:38] step 748600/1000000 | loss=2.1496 vocab=0.6536 token=1.0377 | lr=9.85e-05 | 6.14 steps/s | VRAM=32.50 Go [2026-07-18 10:13:46] step 748650/1000000 | loss=2.1627 vocab=0.5653 token=1.0662 | lr=9.85e-05 | 6.15 steps/s | VRAM=25.93 Go [2026-07-18 10:13:54] step 748700/1000000 | loss=2.2593 vocab=0.6587 token=0.8003 | lr=9.85e-05 | 6.14 steps/s | VRAM=28.99 Go [2026-07-18 10:14:03] step 748750/1000000 | loss=2.1790 vocab=1.0765 token=1.5058 | lr=9.85e-05 | 6.14 steps/s | VRAM=32.86 Go [2026-07-18 10:14:11] step 748800/1000000 | loss=2.1949 vocab=0.7186 token=1.1087 | lr=9.85e-05 | 6.12 steps/s | VRAM=25.70 Go [2026-07-18 10:14:19] step 748850/1000000 | loss=2.0684 vocab=0.7166 token=1.4938 | lr=9.85e-05 | 6.17 steps/s | VRAM=31.95 Go [2026-07-18 10:14:27] step 748900/1000000 | loss=2.1524 vocab=0.7562 token=1.2071 | lr=9.85e-05 | 6.15 steps/s | VRAM=32.30 Go [2026-07-18 10:14:35] step 748950/1000000 | loss=2.1980 vocab=0.8676 token=1.1733 | lr=9.85e-05 | 6.16 steps/s | VRAM=31.48 Go [2026-07-18 10:14:43] step 749000/1000000 | loss=2.0974 vocab=1.4129 token=1.7027 | lr=9.85e-05 | 6.16 steps/s | VRAM=31.74 Go [2026-07-18 10:14:51] step 749050/1000000 | loss=2.2909 vocab=0.8395 token=1.0986 | lr=9.85e-05 | 6.13 steps/s | VRAM=32.27 Go [2026-07-18 10:15:00] step 749100/1000000 | loss=2.1591 vocab=0.8812 token=1.3563 | lr=9.85e-05 | 6.17 steps/s | VRAM=28.56 Go [2026-07-18 10:15:08] step 749150/1000000 | loss=2.1591 vocab=0.9362 token=1.4043 | lr=9.85e-05 | 6.15 steps/s | VRAM=25.87 Go [2026-07-18 10:15:16] step 749200/1000000 | loss=2.1344 vocab=0.6716 token=1.0965 | lr=9.85e-05 | 6.14 steps/s | VRAM=27.24 Go [2026-07-18 10:15:24] step 749250/1000000 | loss=2.1501 vocab=0.9879 token=1.6071 | lr=9.85e-05 | 6.13 steps/s | VRAM=32.10 Go [2026-07-18 10:15:32] step 749300/1000000 | loss=2.2097 vocab=0.7979 token=1.0990 | lr=9.84e-05 | 6.11 steps/s | VRAM=26.49 Go [2026-07-18 10:15:40] step 749350/1000000 | loss=2.1641 vocab=0.9520 token=1.5818 | lr=9.84e-05 | 6.15 steps/s | VRAM=29.16 Go [2026-07-18 10:15:48] step 749400/1000000 | loss=2.2029 vocab=0.9130 token=1.3961 | lr=9.84e-05 | 6.17 steps/s | VRAM=28.80 Go [2026-07-18 10:15:57] step 749450/1000000 | loss=2.2259 vocab=1.7026 token=1.1057 | lr=9.84e-05 | 6.14 steps/s | VRAM=29.52 Go [2026-07-18 10:16:05] step 749500/1000000 | loss=2.1880 vocab=0.4876 token=1.2544 | lr=9.84e-05 | 6.12 steps/s | VRAM=24.65 Go [2026-07-18 10:16:13] step 749550/1000000 | loss=2.2161 vocab=1.0223 token=1.5858 | lr=9.84e-05 | 6.16 steps/s | VRAM=28.34 Go [2026-07-18 10:16:21] step 749600/1000000 | loss=2.2445 vocab=1.4231 token=1.6955 | lr=9.84e-05 | 6.11 steps/s | VRAM=29.91 Go [2026-07-18 10:16:29] step 749650/1000000 | loss=2.2733 vocab=1.2692 token=1.6001 | lr=9.84e-05 | 6.14 steps/s | VRAM=32.77 Go [2026-07-18 10:16:37] step 749700/1000000 | loss=2.1167 vocab=0.4848 token=0.8429 | lr=9.84e-05 | 6.16 steps/s | VRAM=30.66 Go [2026-07-18 10:16:45] step 749750/1000000 | loss=2.1679 vocab=1.0090 token=1.4679 | lr=9.84e-05 | 6.13 steps/s | VRAM=29.02 Go [2026-07-18 10:16:54] step 749800/1000000 | loss=2.2965 vocab=0.6856 token=1.3771 | lr=9.84e-05 | 6.16 steps/s | VRAM=29.00 Go [2026-07-18 10:17:02] step 749850/1000000 | loss=2.2330 vocab=1.4199 token=1.2464 | lr=9.84e-05 | 6.14 steps/s | VRAM=29.42 Go [2026-07-18 10:17:10] step 749900/1000000 | loss=2.1334 vocab=0.7449 token=1.1993 | lr=9.84e-05 | 6.12 steps/s | VRAM=28.42 Go [2026-07-18 10:17:18] step 749950/1000000 | loss=2.1973 vocab=0.5061 token=1.5617 | lr=9.84e-05 | 6.14 steps/s | VRAM=32.28 Go [2026-07-18 10:17:26] step 750000/1000000 | loss=2.1431 vocab=0.8324 token=1.3249 | lr=9.83e-05 | 6.10 steps/s | VRAM=26.61 Go [accuracy] step 750000 : 63.42% (log: /workspace/plbert_train/precision_log.csv) [push] step 750000 OK (loss_token=1.3249, accuracy=63.42%) [2026-07-18 10:19:00] step 750050/1000000 | loss=2.2785 vocab=1.0543 token=1.5948 | lr=9.83e-05 | 0.53 steps/s | VRAM=31.42 Go [2026-07-18 10:19:08] step 750100/1000000 | loss=2.0945 vocab=0.6641 token=1.1273 | lr=9.83e-05 | 6.25 steps/s | VRAM=27.82 Go [2026-07-18 10:19:16] step 750150/1000000 | loss=2.2847 vocab=0.8596 token=1.3718 | lr=9.83e-05 | 6.22 steps/s | VRAM=31.07 Go [2026-07-18 10:19:24] step 750200/1000000 | loss=2.1339 vocab=0.8577 token=1.4910 | lr=9.83e-05 | 6.22 steps/s | VRAM=31.39 Go [2026-07-18 10:19:32] step 750250/1000000 | loss=2.1432 vocab=0.8727 token=0.8958 | lr=9.83e-05 | 6.19 steps/s | VRAM=25.14 Go [2026-07-18 10:19:40] step 750300/1000000 | loss=2.2331 vocab=0.6621 token=1.2264 | lr=9.83e-05 | 6.18 steps/s | VRAM=25.62 Go [2026-07-18 10:19:49] step 750350/1000000 | loss=2.2090 vocab=0.6813 token=1.1054 | lr=9.83e-05 | 6.16 steps/s | VRAM=26.67 Go [2026-07-18 10:19:57] step 750400/1000000 | loss=2.2707 vocab=1.0271 token=1.3365 | lr=9.83e-05 | 6.16 steps/s | VRAM=32.30 Go [2026-07-18 10:20:05] step 750450/1000000 | loss=2.1084 vocab=0.9820 token=1.6655 | lr=9.83e-05 | 6.15 steps/s | VRAM=25.34 Go [2026-07-18 10:20:13] step 750500/1000000 | loss=2.2002 vocab=1.0087 token=1.0678 | lr=9.83e-05 | 6.16 steps/s | VRAM=28.84 Go [2026-07-18 10:20:21] step 750550/1000000 | loss=2.1626 vocab=0.8067 token=1.4381 | lr=9.83e-05 | 6.15 steps/s | VRAM=28.24 Go [2026-07-18 10:20:29] step 750600/1000000 | loss=2.2662 vocab=1.1356 token=1.2930 | lr=9.83e-05 | 6.16 steps/s | VRAM=25.81 Go [2026-07-18 10:20:37] step 750650/1000000 | loss=2.2307 vocab=0.9708 token=1.5886 | lr=9.82e-05 | 6.18 steps/s | VRAM=26.57 Go [2026-07-18 10:20:45] step 750700/1000000 | loss=2.2428 vocab=1.0578 token=1.3947 | lr=9.82e-05 | 6.15 steps/s | VRAM=28.74 Go [2026-07-18 10:20:54] step 750750/1000000 | loss=2.0730 vocab=0.9633 token=1.5823 | lr=9.82e-05 | 6.10 steps/s | VRAM=27.65 Go [2026-07-18 10:21:02] step 750800/1000000 | loss=2.1947 vocab=1.2991 token=1.7491 | lr=9.82e-05 | 6.14 steps/s | VRAM=27.78 Go [2026-07-18 10:21:10] step 750850/1000000 | loss=2.2607 vocab=0.7637 token=1.6450 | lr=9.82e-05 | 6.12 steps/s | VRAM=26.92 Go [2026-07-18 10:21:18] step 750900/1000000 | loss=2.2033 vocab=0.4184 token=1.3049 | lr=9.82e-05 | 6.13 steps/s | VRAM=27.51 Go [2026-07-18 10:21:26] step 750950/1000000 | loss=2.2062 vocab=0.9045 token=1.1836 | lr=9.82e-05 | 6.14 steps/s | VRAM=28.05 Go [2026-07-18 10:21:34] step 751000/1000000 | loss=2.1592 vocab=1.4166 token=1.7868 | lr=9.82e-05 | 6.12 steps/s | VRAM=29.24 Go [2026-07-18 10:21:43] step 751050/1000000 | loss=2.2199 vocab=0.7871 token=1.3442 | lr=9.82e-05 | 6.11 steps/s | VRAM=30.20 Go [2026-07-18 10:21:51] step 751100/1000000 | loss=2.1765 vocab=0.7827 token=1.7325 | lr=9.82e-05 | 6.14 steps/s | VRAM=29.11 Go [2026-07-18 10:21:59] step 751150/1000000 | loss=2.1201 vocab=0.8524 token=1.2391 | lr=9.82e-05 | 6.15 steps/s | VRAM=32.02 Go [2026-07-18 10:22:07] step 751200/1000000 | loss=2.1235 vocab=0.6997 token=1.0542 | lr=9.82e-05 | 6.14 steps/s | VRAM=27.47 Go [2026-07-18 10:22:15] step 751250/1000000 | loss=2.2142 vocab=0.6197 token=1.2388 | lr=9.82e-05 | 6.14 steps/s | VRAM=31.30 Go [2026-07-18 10:22:23] step 751300/1000000 | loss=2.1550 vocab=1.2392 token=1.3007 | lr=9.82e-05 | 6.12 steps/s | VRAM=29.49 Go [2026-07-18 10:22:31] step 751350/1000000 | loss=2.4126 vocab=0.6856 token=0.9954 | lr=9.81e-05 | 6.15 steps/s | VRAM=25.06 Go [2026-07-18 10:22:40] step 751400/1000000 | loss=2.2119 vocab=1.0308 token=1.0042 | lr=9.81e-05 | 6.10 steps/s | VRAM=25.99 Go [2026-07-18 10:22:48] step 751450/1000000 | loss=2.1654 vocab=0.8457 token=1.0385 | lr=9.81e-05 | 6.14 steps/s | VRAM=26.47 Go [2026-07-18 10:22:56] step 751500/1000000 | loss=2.2400 vocab=0.6648 token=0.9258 | lr=9.81e-05 | 6.14 steps/s | VRAM=27.56 Go [2026-07-18 10:23:04] step 751550/1000000 | loss=2.1716 vocab=0.7788 token=1.3938 | lr=9.81e-05 | 6.10 steps/s | VRAM=28.20 Go [2026-07-18 10:23:12] step 751600/1000000 | loss=2.1625 vocab=0.7155 token=1.6564 | lr=9.81e-05 | 6.11 steps/s | VRAM=23.67 Go [2026-07-18 10:23:20] step 751650/1000000 | loss=2.1711 vocab=0.6382 token=1.6236 | lr=9.81e-05 | 6.17 steps/s | VRAM=28.08 Go [2026-07-18 10:23:29] step 751700/1000000 | loss=2.2195 vocab=0.7906 token=1.4094 | lr=9.81e-05 | 6.14 steps/s | VRAM=29.58 Go [2026-07-18 10:23:37] step 751750/1000000 | loss=2.2218 vocab=0.9318 token=1.7276 | lr=9.81e-05 | 6.18 steps/s | VRAM=29.53 Go [2026-07-18 10:23:45] step 751800/1000000 | loss=2.1778 vocab=0.6415 token=1.6686 | lr=9.81e-05 | 6.14 steps/s | VRAM=29.48 Go [2026-07-18 10:23:53] step 751850/1000000 | loss=2.3488 vocab=1.4781 token=1.3344 | lr=9.81e-05 | 6.17 steps/s | VRAM=28.22 Go [2026-07-18 10:24:01] step 751900/1000000 | loss=2.1777 vocab=0.6499 token=1.5349 | lr=9.81e-05 | 6.17 steps/s | VRAM=28.09 Go [2026-07-18 10:24:09] step 751950/1000000 | loss=2.1551 vocab=1.0851 token=0.6215 | lr=9.80e-05 | 6.17 steps/s | VRAM=27.93 Go [2026-07-18 10:24:17] step 752000/1000000 | loss=2.0309 vocab=1.1720 token=1.2903 | lr=9.80e-05 | 6.19 steps/s | VRAM=30.84 Go [2026-07-18 10:24:25] step 752050/1000000 | loss=2.0713 vocab=0.8035 token=1.1520 | lr=9.80e-05 | 6.15 steps/s | VRAM=30.18 Go [2026-07-18 10:24:33] step 752100/1000000 | loss=2.1594 vocab=0.8488 token=1.5299 | lr=9.80e-05 | 6.17 steps/s | VRAM=28.00 Go [2026-07-18 10:24:41] step 752150/1000000 | loss=2.3018 vocab=0.9576 token=1.0284 | lr=9.80e-05 | 6.19 steps/s | VRAM=32.98 Go [2026-07-18 10:24:50] step 752200/1000000 | loss=2.1797 vocab=0.8830 token=1.3697 | lr=9.80e-05 | 6.14 steps/s | VRAM=27.89 Go [2026-07-18 10:24:58] step 752250/1000000 | loss=2.2550 vocab=0.7518 token=0.7853 | lr=9.80e-05 | 6.15 steps/s | VRAM=28.62 Go [2026-07-18 10:25:06] step 752300/1000000 | loss=2.2304 vocab=0.7491 token=1.6121 | lr=9.80e-05 | 6.15 steps/s | VRAM=25.87 Go [2026-07-18 10:25:14] step 752350/1000000 | loss=2.2243 vocab=0.7418 token=1.0613 | lr=9.80e-05 | 6.19 steps/s | VRAM=30.83 Go [2026-07-18 10:25:22] step 752400/1000000 | loss=2.1827 vocab=1.0426 token=1.4195 | lr=9.80e-05 | 6.12 steps/s | VRAM=28.76 Go [2026-07-18 10:25:30] step 752450/1000000 | loss=2.2015 vocab=1.2020 token=1.3964 | lr=9.80e-05 | 6.17 steps/s | VRAM=31.85 Go [2026-07-18 10:25:38] step 752500/1000000 | loss=2.0448 vocab=0.9450 token=1.1981 | lr=9.80e-05 | 6.15 steps/s | VRAM=28.87 Go [2026-07-18 10:25:46] step 752550/1000000 | loss=2.2124 vocab=0.6907 token=1.2758 | lr=9.80e-05 | 6.15 steps/s | VRAM=25.97 Go [2026-07-18 10:25:55] step 752600/1000000 | loss=2.1469 vocab=0.8774 token=0.9926 | lr=9.79e-05 | 6.16 steps/s | VRAM=25.42 Go [2026-07-18 10:26:03] step 752650/1000000 | loss=2.2148 vocab=0.9737 token=1.4225 | lr=9.79e-05 | 6.18 steps/s | VRAM=25.86 Go [2026-07-18 10:26:11] step 752700/1000000 | loss=2.1850 vocab=0.5581 token=0.9426 | lr=9.79e-05 | 6.17 steps/s | VRAM=27.30 Go [2026-07-18 10:26:19] step 752750/1000000 | loss=2.2110 vocab=1.1333 token=1.2838 | lr=9.79e-05 | 6.16 steps/s | VRAM=31.55 Go [2026-07-18 10:26:27] step 752800/1000000 | loss=2.1259 vocab=1.2546 token=1.2814 | lr=9.79e-05 | 6.18 steps/s | VRAM=31.54 Go [2026-07-18 10:26:35] step 752850/1000000 | loss=2.2252 vocab=0.3367 token=0.8069 | lr=9.79e-05 | 6.11 steps/s | VRAM=25.84 Go [2026-07-18 10:26:43] step 752900/1000000 | loss=2.0648 vocab=1.0120 token=1.2849 | lr=9.79e-05 | 6.18 steps/s | VRAM=31.98 Go [2026-07-18 10:26:51] step 752950/1000000 | loss=2.2222 vocab=0.9579 token=1.1483 | lr=9.79e-05 | 6.16 steps/s | VRAM=32.06 Go [2026-07-18 10:27:00] step 753000/1000000 | loss=2.2203 vocab=0.4401 token=0.8995 | lr=9.79e-05 | 6.14 steps/s | VRAM=28.22 Go [2026-07-18 10:27:08] step 753050/1000000 | loss=2.2000 vocab=1.0430 token=1.3487 | lr=9.79e-05 | 6.16 steps/s | VRAM=31.20 Go [2026-07-18 10:27:16] step 753100/1000000 | loss=2.1915 vocab=0.7727 token=1.6402 | lr=9.79e-05 | 6.15 steps/s | VRAM=29.80 Go [2026-07-18 10:27:24] step 753150/1000000 | loss=2.2025 vocab=0.5706 token=1.3909 | lr=9.79e-05 | 6.16 steps/s | VRAM=31.89 Go [2026-07-18 10:27:32] step 753200/1000000 | loss=2.1336 vocab=0.8970 token=1.3083 | lr=9.78e-05 | 6.11 steps/s | VRAM=26.99 Go [2026-07-18 10:27:40] step 753250/1000000 | loss=2.2420 vocab=1.0399 token=1.4158 | lr=9.78e-05 | 6.20 steps/s | VRAM=31.92 Go [2026-07-18 10:27:48] step 753300/1000000 | loss=2.1975 vocab=0.9937 token=1.5109 | lr=9.78e-05 | 6.17 steps/s | VRAM=32.68 Go [2026-07-18 10:27:56] step 753350/1000000 | loss=2.1874 vocab=0.8096 token=1.5476 | lr=9.78e-05 | 6.17 steps/s | VRAM=28.60 Go [2026-07-18 10:28:05] step 753400/1000000 | loss=2.1787 vocab=0.5383 token=1.3973 | lr=9.78e-05 | 6.15 steps/s | VRAM=28.97 Go [2026-07-18 10:28:13] step 753450/1000000 | loss=2.1369 vocab=0.8601 token=1.5428 | lr=9.78e-05 | 6.14 steps/s | VRAM=26.11 Go [2026-07-18 10:28:21] step 753500/1000000 | loss=2.1512 vocab=0.9537 token=1.2679 | lr=9.78e-05 | 6.14 steps/s | VRAM=29.30 Go [2026-07-18 10:28:29] step 753550/1000000 | loss=2.2002 vocab=1.0874 token=1.4331 | lr=9.78e-05 | 6.17 steps/s | VRAM=25.44 Go [2026-07-18 10:28:37] step 753600/1000000 | loss=2.1891 vocab=0.7388 token=1.4246 | lr=9.78e-05 | 6.16 steps/s | VRAM=26.00 Go [2026-07-18 10:28:45] step 753650/1000000 | loss=2.0963 vocab=0.7294 token=1.3274 | lr=9.78e-05 | 6.16 steps/s | VRAM=26.46 Go [2026-07-18 10:28:53] step 753700/1000000 | loss=2.1905 vocab=0.7532 token=1.1434 | lr=9.78e-05 | 6.16 steps/s | VRAM=32.27 Go [2026-07-18 10:29:01] step 753750/1000000 | loss=2.0991 vocab=1.7964 token=1.8550 | lr=9.78e-05 | 6.15 steps/s | VRAM=31.83 Go [2026-07-18 10:29:09] step 753800/1000000 | loss=2.1587 vocab=0.9982 token=1.1301 | lr=9.77e-05 | 6.16 steps/s | VRAM=32.64 Go [2026-07-18 10:29:18] step 753850/1000000 | loss=2.1356 vocab=0.7066 token=1.6282 | lr=9.77e-05 | 6.12 steps/s | VRAM=29.16 Go [2026-07-18 10:29:26] step 753900/1000000 | loss=2.3627 vocab=0.7872 token=1.2016 | lr=9.77e-05 | 6.12 steps/s | VRAM=29.19 Go [2026-07-18 10:29:34] step 753950/1000000 | loss=2.2616 vocab=1.0617 token=1.2834 | lr=9.77e-05 | 6.16 steps/s | VRAM=28.54 Go [2026-07-18 10:29:42] step 754000/1000000 | loss=2.1775 vocab=0.9675 token=1.1979 | lr=9.77e-05 | 6.16 steps/s | VRAM=25.65 Go [2026-07-18 10:29:50] step 754050/1000000 | loss=2.2077 vocab=1.0882 token=1.3839 | lr=9.77e-05 | 6.14 steps/s | VRAM=26.15 Go [2026-07-18 10:29:58] step 754100/1000000 | loss=2.1422 vocab=1.3017 token=1.4022 | lr=9.77e-05 | 6.14 steps/s | VRAM=29.80 Go [2026-07-18 10:30:06] step 754150/1000000 | loss=2.1509 vocab=1.3197 token=1.9402 | lr=9.77e-05 | 6.16 steps/s | VRAM=26.60 Go [2026-07-18 10:30:15] step 754200/1000000 | loss=2.2508 vocab=0.7824 token=1.3431 | lr=9.77e-05 | 6.16 steps/s | VRAM=25.98 Go [2026-07-18 10:30:23] step 754250/1000000 | loss=2.1725 vocab=0.9585 token=1.4716 | lr=9.77e-05 | 6.18 steps/s | VRAM=32.35 Go [2026-07-18 10:30:31] step 754300/1000000 | loss=2.1042 vocab=0.8178 token=0.9160 | lr=9.77e-05 | 6.16 steps/s | VRAM=29.36 Go [2026-07-18 10:30:39] step 754350/1000000 | loss=2.2637 vocab=1.1264 token=1.5652 | lr=9.76e-05 | 6.19 steps/s | VRAM=31.56 Go [2026-07-18 10:30:47] step 754400/1000000 | loss=2.1250 vocab=0.7646 token=1.2428 | lr=9.76e-05 | 6.16 steps/s | VRAM=24.30 Go [2026-07-18 10:30:55] step 754450/1000000 | loss=2.2365 vocab=0.7735 token=1.0837 | lr=9.76e-05 | 6.17 steps/s | VRAM=24.64 Go [2026-07-18 10:31:03] step 754500/1000000 | loss=2.1690 vocab=0.6291 token=1.2472 | lr=9.76e-05 | 6.18 steps/s | VRAM=32.46 Go [2026-07-18 10:31:11] step 754550/1000000 | loss=2.1788 vocab=0.7045 token=1.5693 | lr=9.76e-05 | 6.15 steps/s | VRAM=28.75 Go [2026-07-18 10:31:19] step 754600/1000000 | loss=2.1144 vocab=0.6346 token=1.2350 | lr=9.76e-05 | 6.15 steps/s | VRAM=28.36 Go [2026-07-18 10:31:28] step 754650/1000000 | loss=2.1233 vocab=1.1587 token=1.2478 | lr=9.76e-05 | 6.14 steps/s | VRAM=25.52 Go [2026-07-18 10:31:36] step 754700/1000000 | loss=2.1725 vocab=0.6967 token=1.4874 | lr=9.76e-05 | 6.15 steps/s | VRAM=32.23 Go [2026-07-18 10:31:44] step 754750/1000000 | loss=2.2167 vocab=1.1739 token=1.2556 | lr=9.76e-05 | 6.16 steps/s | VRAM=25.82 Go [2026-07-18 10:31:52] step 754800/1000000 | loss=2.1723 vocab=0.7378 token=1.3123 | lr=9.76e-05 | 6.15 steps/s | VRAM=32.94 Go [2026-07-18 10:32:00] step 754850/1000000 | loss=2.2145 vocab=0.8909 token=1.5031 | lr=9.76e-05 | 6.17 steps/s | VRAM=25.44 Go [2026-07-18 10:32:08] step 754900/1000000 | loss=2.1228 vocab=1.0554 token=1.3560 | lr=9.76e-05 | 6.18 steps/s | VRAM=31.17 Go [2026-07-18 10:32:16] step 754950/1000000 | loss=2.1741 vocab=0.8682 token=1.3418 | lr=9.75e-05 | 6.13 steps/s | VRAM=29.53 Go [2026-07-18 10:32:24] step 755000/1000000 | loss=2.2479 vocab=1.2226 token=1.5629 | lr=9.75e-05 | 6.15 steps/s | VRAM=29.17 Go [2026-07-18 10:32:33] step 755050/1000000 | loss=2.1968 vocab=1.0886 token=1.3248 | lr=9.75e-05 | 6.15 steps/s | VRAM=26.76 Go [2026-07-18 10:32:41] step 755100/1000000 | loss=2.2114 vocab=0.8944 token=1.3093 | lr=9.75e-05 | 6.17 steps/s | VRAM=31.83 Go [2026-07-18 10:32:49] step 755150/1000000 | loss=2.1726 vocab=0.8668 token=1.2933 | lr=9.75e-05 | 6.14 steps/s | VRAM=28.06 Go [2026-07-18 10:32:57] step 755200/1000000 | loss=2.2317 vocab=0.6322 token=1.2496 | lr=9.75e-05 | 6.16 steps/s | VRAM=28.72 Go [2026-07-18 10:33:05] step 755250/1000000 | loss=2.1869 vocab=0.5169 token=1.5099 | lr=9.75e-05 | 6.16 steps/s | VRAM=31.17 Go [2026-07-18 10:33:13] step 755300/1000000 | loss=2.1205 vocab=1.0075 token=1.3783 | lr=9.75e-05 | 6.12 steps/s | VRAM=29.00 Go [2026-07-18 10:33:21] step 755350/1000000 | loss=2.2869 vocab=1.0129 token=1.5470 | lr=9.75e-05 | 6.16 steps/s | VRAM=31.09 Go [2026-07-18 10:33:30] step 755400/1000000 | loss=2.1157 vocab=1.1514 token=1.4344 | lr=9.75e-05 | 6.11 steps/s | VRAM=32.75 Go [2026-07-18 10:33:38] step 755450/1000000 | loss=2.1712 vocab=1.0696 token=0.9714 | lr=9.75e-05 | 6.11 steps/s | VRAM=28.70 Go [2026-07-18 10:33:46] step 755500/1000000 | loss=2.1801 vocab=0.3294 token=1.2525 | lr=9.74e-05 | 6.12 steps/s | VRAM=28.67 Go [2026-07-18 10:33:54] step 755550/1000000 | loss=2.0826 vocab=0.5734 token=1.2731 | lr=9.74e-05 | 6.13 steps/s | VRAM=28.35 Go [2026-07-18 10:34:02] step 755600/1000000 | loss=2.2028 vocab=0.8942 token=1.3741 | lr=9.74e-05 | 6.17 steps/s | VRAM=31.76 Go [2026-07-18 10:34:10] step 755650/1000000 | loss=2.2246 vocab=0.9800 token=1.3684 | lr=9.74e-05 | 6.14 steps/s | VRAM=25.30 Go [2026-07-18 10:34:18] step 755700/1000000 | loss=2.2089 vocab=1.0890 token=1.4481 | lr=9.74e-05 | 6.16 steps/s | VRAM=31.76 Go [2026-07-18 10:34:27] step 755750/1000000 | loss=2.2153 vocab=0.6916 token=1.5090 | lr=9.74e-05 | 6.12 steps/s | VRAM=26.06 Go [2026-07-18 10:34:35] step 755800/1000000 | loss=2.1926 vocab=0.8531 token=1.5027 | lr=9.74e-05 | 6.11 steps/s | VRAM=26.54 Go [2026-07-18 10:34:43] step 755850/1000000 | loss=2.1807 vocab=0.8907 token=1.1738 | lr=9.74e-05 | 6.12 steps/s | VRAM=29.20 Go [2026-07-18 10:34:51] step 755900/1000000 | loss=2.1348 vocab=1.1793 token=1.0887 | lr=9.74e-05 | 6.13 steps/s | VRAM=26.30 Go [2026-07-18 10:34:59] step 755950/1000000 | loss=2.2223 vocab=0.9519 token=1.0995 | lr=9.74e-05 | 6.13 steps/s | VRAM=28.61 Go [2026-07-18 10:35:07] step 756000/1000000 | loss=2.0923 vocab=0.7673 token=1.1901 | lr=9.74e-05 | 6.13 steps/s | VRAM=28.00 Go [2026-07-18 10:35:16] step 756050/1000000 | loss=2.2123 vocab=0.8369 token=1.1025 | lr=9.73e-05 | 6.15 steps/s | VRAM=32.30 Go [2026-07-18 10:35:24] step 756100/1000000 | loss=2.2201 vocab=0.6022 token=1.1710 | lr=9.73e-05 | 6.11 steps/s | VRAM=32.62 Go [2026-07-18 10:35:32] step 756150/1000000 | loss=2.0989 vocab=1.2468 token=0.9794 | lr=9.73e-05 | 6.14 steps/s | VRAM=28.06 Go [2026-07-18 10:35:40] step 756200/1000000 | loss=2.1931 vocab=0.4351 token=1.0912 | lr=9.73e-05 | 6.11 steps/s | VRAM=32.45 Go [2026-07-18 10:35:48] step 756250/1000000 | loss=2.1455 vocab=0.7366 token=1.0884 | lr=9.73e-05 | 6.11 steps/s | VRAM=29.51 Go [2026-07-18 10:35:56] step 756300/1000000 | loss=2.2179 vocab=0.8579 token=1.3769 | lr=9.73e-05 | 6.17 steps/s | VRAM=29.73 Go [2026-07-18 10:36:04] step 756350/1000000 | loss=2.2011 vocab=1.0927 token=1.1459 | lr=9.73e-05 | 6.13 steps/s | VRAM=31.31 Go [2026-07-18 10:36:13] step 756400/1000000 | loss=2.1971 vocab=0.7959 token=1.1365 | lr=9.73e-05 | 6.15 steps/s | VRAM=28.79 Go [2026-07-18 10:36:21] step 756450/1000000 | loss=2.1435 vocab=0.8124 token=1.0809 | lr=9.73e-05 | 6.13 steps/s | VRAM=32.58 Go [2026-07-18 10:36:29] step 756500/1000000 | loss=2.1054 vocab=0.4283 token=0.8974 | lr=9.73e-05 | 6.12 steps/s | VRAM=32.11 Go [2026-07-18 10:36:37] step 756550/1000000 | loss=2.2193 vocab=0.8140 token=1.5418 | lr=9.73e-05 | 6.15 steps/s | VRAM=31.38 Go [2026-07-18 10:36:45] step 756600/1000000 | loss=2.2098 vocab=1.1412 token=1.4346 | lr=9.72e-05 | 6.15 steps/s | VRAM=24.59 Go [2026-07-18 10:36:53] step 756650/1000000 | loss=2.1289 vocab=1.3036 token=1.1200 | lr=9.72e-05 | 6.12 steps/s | VRAM=28.46 Go [2026-07-18 10:37:02] step 756700/1000000 | loss=2.1790 vocab=0.9937 token=1.3937 | lr=9.72e-05 | 6.14 steps/s | VRAM=29.01 Go [2026-07-18 10:37:10] step 756750/1000000 | loss=2.2697 vocab=0.7901 token=1.3392 | lr=9.72e-05 | 6.08 steps/s | VRAM=26.10 Go [2026-07-18 10:37:18] step 756800/1000000 | loss=2.1376 vocab=0.9511 token=0.8331 | lr=9.72e-05 | 6.14 steps/s | VRAM=32.65 Go [2026-07-18 10:37:26] step 756850/1000000 | loss=2.2237 vocab=1.1862 token=1.4429 | lr=9.72e-05 | 6.13 steps/s | VRAM=28.17 Go [2026-07-18 10:37:34] step 756900/1000000 | loss=2.1960 vocab=1.2983 token=1.6541 | lr=9.72e-05 | 6.13 steps/s | VRAM=26.50 Go [2026-07-18 10:37:42] step 756950/1000000 | loss=2.1712 vocab=1.1382 token=1.4954 | lr=9.72e-05 | 6.15 steps/s | VRAM=28.63 Go [2026-07-18 10:37:51] step 757000/1000000 | loss=2.2402 vocab=0.9766 token=1.6358 | lr=9.72e-05 | 6.11 steps/s | VRAM=28.87 Go [2026-07-18 10:37:59] step 757050/1000000 | loss=2.2056 vocab=1.4115 token=1.7441 | lr=9.72e-05 | 6.11 steps/s | VRAM=26.38 Go [2026-07-18 10:38:07] step 757100/1000000 | loss=2.2150 vocab=1.0414 token=1.1635 | lr=9.71e-05 | 6.17 steps/s | VRAM=31.29 Go [2026-07-18 10:38:15] step 757150/1000000 | loss=2.1456 vocab=0.7502 token=1.2529 | lr=9.71e-05 | 6.11 steps/s | VRAM=25.61 Go [2026-07-18 10:38:23] step 757200/1000000 | loss=2.1040 vocab=0.5782 token=1.2688 | lr=9.71e-05 | 6.10 steps/s | VRAM=26.35 Go [2026-07-18 10:38:31] step 757250/1000000 | loss=2.1790 vocab=0.7322 token=0.7332 | lr=9.71e-05 | 6.14 steps/s | VRAM=25.54 Go [2026-07-18 10:38:40] step 757300/1000000 | loss=2.2375 vocab=0.6795 token=1.1906 | lr=9.71e-05 | 6.14 steps/s | VRAM=25.37 Go [2026-07-18 10:38:48] step 757350/1000000 | loss=2.2664 vocab=1.0815 token=1.3507 | lr=9.71e-05 | 6.15 steps/s | VRAM=28.16 Go [2026-07-18 10:38:56] step 757400/1000000 | loss=2.2147 vocab=1.2621 token=1.4338 | lr=9.71e-05 | 6.12 steps/s | VRAM=25.96 Go [2026-07-18 10:39:04] step 757450/1000000 | loss=2.1913 vocab=0.5272 token=1.1368 | lr=9.71e-05 | 6.16 steps/s | VRAM=32.29 Go [2026-07-18 10:39:12] step 757500/1000000 | loss=2.1403 vocab=0.5255 token=0.8065 | lr=9.71e-05 | 6.12 steps/s | VRAM=32.16 Go [2026-07-18 10:39:20] step 757550/1000000 | loss=2.1806 vocab=0.8664 token=0.9425 | lr=9.71e-05 | 6.15 steps/s | VRAM=31.49 Go [2026-07-18 10:39:28] step 757600/1000000 | loss=2.1544 vocab=1.1186 token=1.1524 | lr=9.71e-05 | 6.12 steps/s | VRAM=28.98 Go [2026-07-18 10:39:37] step 757650/1000000 | loss=2.2613 vocab=0.5146 token=1.3947 | lr=9.70e-05 | 6.14 steps/s | VRAM=30.10 Go [2026-07-18 10:39:45] step 757700/1000000 | loss=2.1228 vocab=1.2709 token=1.5716 | lr=9.70e-05 | 6.13 steps/s | VRAM=31.80 Go [2026-07-18 10:39:53] step 757750/1000000 | loss=2.1466 vocab=0.9666 token=0.8897 | lr=9.70e-05 | 6.13 steps/s | VRAM=28.83 Go [2026-07-18 10:40:01] step 757800/1000000 | loss=2.2262 vocab=0.5918 token=1.1339 | lr=9.70e-05 | 6.13 steps/s | VRAM=27.89 Go [2026-07-18 10:40:09] step 757850/1000000 | loss=2.2071 vocab=0.8407 token=1.2180 | lr=9.70e-05 | 6.14 steps/s | VRAM=32.24 Go [2026-07-18 10:40:17] step 757900/1000000 | loss=2.2210 vocab=1.2170 token=1.4613 | lr=9.70e-05 | 6.12 steps/s | VRAM=28.40 Go [2026-07-18 10:40:26] step 757950/1000000 | loss=2.2274 vocab=0.8311 token=1.1511 | lr=9.70e-05 | 6.11 steps/s | VRAM=32.66 Go [2026-07-18 10:40:34] step 758000/1000000 | loss=2.2237 vocab=1.1692 token=1.6564 | lr=9.70e-05 | 6.09 steps/s | VRAM=25.88 Go [2026-07-18 10:40:42] step 758050/1000000 | loss=2.1412 vocab=1.0471 token=1.8972 | lr=9.70e-05 | 6.11 steps/s | VRAM=26.42 Go [2026-07-18 10:40:50] step 758100/1000000 | loss=2.2182 vocab=1.0376 token=1.5804 | lr=9.70e-05 | 6.13 steps/s | VRAM=32.78 Go [2026-07-18 10:40:58] step 758150/1000000 | loss=2.1150 vocab=1.1359 token=1.2557 | lr=9.69e-05 | 6.11 steps/s | VRAM=26.52 Go [2026-07-18 10:41:06] step 758200/1000000 | loss=2.1216 vocab=0.6315 token=1.4467 | lr=9.69e-05 | 6.16 steps/s | VRAM=31.10 Go [2026-07-18 10:41:15] step 758250/1000000 | loss=2.1575 vocab=0.6991 token=1.1674 | lr=9.69e-05 | 6.13 steps/s | VRAM=26.24 Go [2026-07-18 10:41:23] step 758300/1000000 | loss=2.1612 vocab=0.8004 token=0.8756 | lr=9.69e-05 | 6.17 steps/s | VRAM=28.02 Go [2026-07-18 10:41:31] step 758350/1000000 | loss=2.1250 vocab=0.4708 token=0.8716 | lr=9.69e-05 | 6.14 steps/s | VRAM=28.43 Go [2026-07-18 10:41:39] step 758400/1000000 | loss=2.2594 vocab=1.0215 token=1.0779 | lr=9.69e-05 | 6.14 steps/s | VRAM=29.00 Go [2026-07-18 10:41:47] step 758450/1000000 | loss=2.2222 vocab=1.3898 token=1.8128 | lr=9.69e-05 | 6.13 steps/s | VRAM=29.40 Go [2026-07-18 10:41:55] step 758500/1000000 | loss=2.2586 vocab=0.8009 token=0.9545 | lr=9.69e-05 | 6.11 steps/s | VRAM=26.80 Go [2026-07-18 10:42:03] step 758550/1000000 | loss=2.2160 vocab=1.1124 token=1.7156 | lr=9.69e-05 | 6.14 steps/s | VRAM=28.31 Go [2026-07-18 10:42:12] step 758600/1000000 | loss=2.1722 vocab=0.5086 token=1.4499 | lr=9.69e-05 | 6.10 steps/s | VRAM=26.36 Go [2026-07-18 10:42:20] step 758650/1000000 | loss=2.1235 vocab=0.8187 token=1.0892 | lr=9.68e-05 | 6.16 steps/s | VRAM=25.05 Go [2026-07-18 10:42:28] step 758700/1000000 | loss=2.1768 vocab=1.0169 token=1.3181 | lr=9.68e-05 | 6.12 steps/s | VRAM=32.59 Go [2026-07-18 10:42:36] step 758750/1000000 | loss=2.2872 vocab=0.9317 token=1.3929 | lr=9.68e-05 | 6.14 steps/s | VRAM=28.98 Go [2026-07-18 10:42:44] step 758800/1000000 | loss=2.2450 vocab=1.0827 token=1.3895 | lr=9.68e-05 | 6.15 steps/s | VRAM=32.69 Go [2026-07-18 10:42:52] step 758850/1000000 | loss=2.1799 vocab=1.4613 token=1.2906 | lr=9.68e-05 | 6.15 steps/s | VRAM=26.09 Go [2026-07-18 10:43:00] step 758900/1000000 | loss=2.1902 vocab=0.8496 token=1.1692 | lr=9.68e-05 | 6.15 steps/s | VRAM=28.16 Go [2026-07-18 10:43:09] step 758950/1000000 | loss=2.2186 vocab=0.9253 token=1.6383 | lr=9.68e-05 | 6.12 steps/s | VRAM=29.54 Go [2026-07-18 10:43:17] step 759000/1000000 | loss=2.1828 vocab=0.8731 token=1.6394 | lr=9.68e-05 | 6.12 steps/s | VRAM=28.48 Go [2026-07-18 10:43:25] step 759050/1000000 | loss=2.0851 vocab=1.1230 token=1.4719 | lr=9.68e-05 | 6.18 steps/s | VRAM=31.14 Go [2026-07-18 10:43:33] step 759100/1000000 | loss=2.1094 vocab=0.9725 token=1.2514 | lr=9.68e-05 | 6.14 steps/s | VRAM=28.17 Go [2026-07-18 10:43:41] step 759150/1000000 | loss=2.2676 vocab=1.0529 token=0.9850 | lr=9.67e-05 | 6.13 steps/s | VRAM=32.68 Go [2026-07-18 10:43:49] step 759200/1000000 | loss=2.1710 vocab=0.7384 token=1.2491 | lr=9.67e-05 | 6.13 steps/s | VRAM=32.02 Go [2026-07-18 10:43:57] step 759250/1000000 | loss=2.2046 vocab=0.8586 token=1.4940 | lr=9.67e-05 | 6.13 steps/s | VRAM=28.00 Go [2026-07-18 10:44:06] step 759300/1000000 | loss=2.2499 vocab=0.7582 token=1.2688 | lr=9.67e-05 | 6.11 steps/s | VRAM=29.70 Go [2026-07-18 10:44:14] step 759350/1000000 | loss=2.2238 vocab=1.5182 token=1.3848 | lr=9.67e-05 | 6.15 steps/s | VRAM=26.58 Go [2026-07-18 10:44:22] step 759400/1000000 | loss=2.2156 vocab=0.9313 token=1.2946 | lr=9.67e-05 | 6.17 steps/s | VRAM=31.45 Go [2026-07-18 10:44:30] step 759450/1000000 | loss=2.2148 vocab=1.0548 token=1.3646 | lr=9.67e-05 | 6.13 steps/s | VRAM=25.83 Go [2026-07-18 10:44:38] step 759500/1000000 | loss=2.0914 vocab=0.9933 token=1.0433 | lr=9.67e-05 | 6.13 steps/s | VRAM=28.40 Go [2026-07-18 10:44:46] step 759550/1000000 | loss=2.1304 vocab=0.7943 token=0.9649 | lr=9.67e-05 | 6.13 steps/s | VRAM=29.59 Go [2026-07-18 10:44:55] step 759600/1000000 | loss=2.1460 vocab=0.8722 token=1.2441 | lr=9.67e-05 | 6.12 steps/s | VRAM=28.84 Go [2026-07-18 10:45:03] step 759650/1000000 | loss=2.1189 vocab=0.6280 token=0.9685 | lr=9.66e-05 | 6.11 steps/s | VRAM=25.48 Go [2026-07-18 10:45:11] step 759700/1000000 | loss=2.1493 vocab=0.7193 token=1.3546 | lr=9.66e-05 | 6.14 steps/s | VRAM=24.92 Go [2026-07-18 10:45:19] step 759750/1000000 | loss=2.1629 vocab=1.0426 token=1.3220 | lr=9.66e-05 | 6.15 steps/s | VRAM=32.26 Go [2026-07-18 10:45:27] step 759800/1000000 | loss=2.2865 vocab=1.3445 token=1.3808 | lr=9.66e-05 | 6.10 steps/s | VRAM=25.78 Go [2026-07-18 10:45:35] step 759850/1000000 | loss=2.2692 vocab=1.1204 token=1.3492 | lr=9.66e-05 | 6.14 steps/s | VRAM=28.47 Go [2026-07-18 10:45:43] step 759900/1000000 | loss=2.2608 vocab=0.7309 token=0.9683 | lr=9.66e-05 | 6.12 steps/s | VRAM=31.68 Go [2026-07-18 10:45:52] step 759950/1000000 | loss=2.1911 vocab=1.2918 token=1.5518 | lr=9.66e-05 | 6.12 steps/s | VRAM=30.79 Go [2026-07-18 10:46:00] step 760000/1000000 | loss=2.1946 vocab=1.4241 token=1.6662 | lr=9.66e-05 | 6.12 steps/s | VRAM=28.98 Go [2026-07-18 10:46:08] step 760050/1000000 | loss=2.1434 vocab=0.7205 token=1.0640 | lr=9.66e-05 | 6.13 steps/s | VRAM=31.89 Go [2026-07-18 10:46:16] step 760100/1000000 | loss=2.0871 vocab=0.5676 token=1.0597 | lr=9.65e-05 | 6.14 steps/s | VRAM=31.96 Go [2026-07-18 10:46:24] step 760150/1000000 | loss=2.1195 vocab=0.5619 token=1.2953 | lr=9.65e-05 | 6.13 steps/s | VRAM=26.61 Go [2026-07-18 10:46:32] step 760200/1000000 | loss=2.2869 vocab=1.1876 token=1.3798 | lr=9.65e-05 | 6.15 steps/s | VRAM=28.34 Go [2026-07-18 10:46:41] step 760250/1000000 | loss=2.1615 vocab=0.8541 token=1.2200 | lr=9.65e-05 | 6.13 steps/s | VRAM=27.94 Go [2026-07-18 10:46:49] step 760300/1000000 | loss=2.0739 vocab=0.4957 token=1.1565 | lr=9.65e-05 | 6.12 steps/s | VRAM=29.20 Go [2026-07-18 10:46:57] step 760350/1000000 | loss=2.1966 vocab=0.5959 token=1.1196 | lr=9.65e-05 | 6.14 steps/s | VRAM=29.26 Go [2026-07-18 10:47:05] step 760400/1000000 | loss=2.2281 vocab=0.7779 token=1.1174 | lr=9.65e-05 | 6.16 steps/s | VRAM=31.51 Go [2026-07-18 10:47:13] step 760450/1000000 | loss=2.1373 vocab=1.0147 token=0.8395 | lr=9.65e-05 | 6.14 steps/s | VRAM=32.32 Go [2026-07-18 10:47:21] step 760500/1000000 | loss=2.2926 vocab=0.9912 token=1.4208 | lr=9.65e-05 | 6.14 steps/s | VRAM=29.61 Go [2026-07-18 10:47:29] step 760550/1000000 | loss=2.1867 vocab=0.6609 token=1.1929 | lr=9.65e-05 | 6.12 steps/s | VRAM=28.21 Go [2026-07-18 10:47:38] step 760600/1000000 | loss=2.1353 vocab=0.7585 token=0.8394 | lr=9.64e-05 | 6.17 steps/s | VRAM=28.57 Go [2026-07-18 10:47:46] step 760650/1000000 | loss=2.3237 vocab=0.5583 token=1.0445 | lr=9.64e-05 | 6.09 steps/s | VRAM=28.37 Go [2026-07-18 10:47:54] step 760700/1000000 | loss=2.0786 vocab=0.3929 token=0.7767 | lr=9.64e-05 | 6.13 steps/s | VRAM=31.00 Go [2026-07-18 10:48:02] step 760750/1000000 | loss=2.2300 vocab=0.9626 token=1.4883 | lr=9.64e-05 | 6.10 steps/s | VRAM=27.73 Go [2026-07-18 10:48:10] step 760800/1000000 | loss=2.1675 vocab=0.6616 token=1.2286 | lr=9.64e-05 | 6.11 steps/s | VRAM=27.53 Go [2026-07-18 10:48:18] step 760850/1000000 | loss=2.0994 vocab=0.8808 token=1.2924 | lr=9.64e-05 | 6.12 steps/s | VRAM=31.89 Go [2026-07-18 10:48:27] step 760900/1000000 | loss=2.2354 vocab=1.1522 token=1.3895 | lr=9.64e-05 | 6.14 steps/s | VRAM=28.67 Go [2026-07-18 10:48:35] step 760950/1000000 | loss=2.1350 vocab=0.8300 token=1.2955 | lr=9.64e-05 | 6.15 steps/s | VRAM=24.84 Go [2026-07-18 10:48:43] step 761000/1000000 | loss=2.1202 vocab=0.8244 token=1.3674 | lr=9.64e-05 | 6.17 steps/s | VRAM=27.91 Go [2026-07-18 10:48:51] step 761050/1000000 | loss=2.0729 vocab=0.6243 token=1.1155 | lr=9.63e-05 | 6.15 steps/s | VRAM=31.32 Go [2026-07-18 10:48:59] step 761100/1000000 | loss=2.1015 vocab=0.6440 token=0.9236 | lr=9.63e-05 | 6.13 steps/s | VRAM=29.53 Go [2026-07-18 10:49:07] step 761150/1000000 | loss=2.1641 vocab=1.2355 token=1.5393 | lr=9.63e-05 | 6.16 steps/s | VRAM=31.76 Go [2026-07-18 10:49:15] step 761200/1000000 | loss=2.2491 vocab=0.8440 token=1.2604 | lr=9.63e-05 | 6.12 steps/s | VRAM=28.61 Go [2026-07-18 10:49:24] step 761250/1000000 | loss=2.2656 vocab=0.8223 token=1.3353 | lr=9.63e-05 | 6.12 steps/s | VRAM=28.33 Go [2026-07-18 10:49:32] step 761300/1000000 | loss=2.2959 vocab=0.8565 token=1.3198 | lr=9.63e-05 | 6.15 steps/s | VRAM=30.45 Go [2026-07-18 10:49:40] step 761350/1000000 | loss=2.2536 vocab=0.7365 token=1.1745 | lr=9.63e-05 | 6.18 steps/s | VRAM=31.52 Go [2026-07-18 10:49:48] step 761400/1000000 | loss=2.1033 vocab=0.4969 token=0.5856 | lr=9.63e-05 | 6.16 steps/s | VRAM=28.41 Go [2026-07-18 10:49:56] step 761450/1000000 | loss=2.1491 vocab=0.6566 token=0.8216 | lr=9.63e-05 | 6.14 steps/s | VRAM=29.18 Go [2026-07-18 10:50:04] step 761500/1000000 | loss=2.1207 vocab=0.5534 token=1.4206 | lr=9.62e-05 | 6.17 steps/s | VRAM=29.53 Go [2026-07-18 10:50:12] step 761550/1000000 | loss=2.2010 vocab=1.1822 token=1.6726 | lr=9.62e-05 | 6.13 steps/s | VRAM=28.86 Go [2026-07-18 10:50:20] step 761600/1000000 | loss=2.1559 vocab=1.0570 token=1.7343 | lr=9.62e-05 | 6.15 steps/s | VRAM=29.37 Go [2026-07-18 10:50:29] step 761650/1000000 | loss=2.2584 vocab=1.0249 token=1.3600 | lr=9.62e-05 | 6.12 steps/s | VRAM=28.96 Go [2026-07-18 10:50:37] step 761700/1000000 | loss=2.1736 vocab=1.3899 token=1.4473 | lr=9.62e-05 | 6.18 steps/s | VRAM=24.30 Go [2026-07-18 10:50:45] step 761750/1000000 | loss=2.1840 vocab=0.7558 token=0.8747 | lr=9.62e-05 | 6.14 steps/s | VRAM=31.87 Go [2026-07-18 10:50:53] step 761800/1000000 | loss=2.1008 vocab=0.8569 token=1.2781 | lr=9.62e-05 | 6.13 steps/s | VRAM=25.99 Go [2026-07-18 10:51:01] step 761850/1000000 | loss=2.2163 vocab=0.6637 token=0.9623 | lr=9.62e-05 | 6.14 steps/s | VRAM=26.50 Go [2026-07-18 10:51:09] step 761900/1000000 | loss=2.2001 vocab=1.0335 token=1.4501 | lr=9.62e-05 | 6.13 steps/s | VRAM=28.82 Go [2026-07-18 10:51:17] step 761950/1000000 | loss=2.1345 vocab=0.5995 token=0.7896 | lr=9.61e-05 | 6.14 steps/s | VRAM=29.45 Go [2026-07-18 10:51:26] step 762000/1000000 | loss=2.0741 vocab=0.6291 token=1.2712 | lr=9.61e-05 | 6.16 steps/s | VRAM=28.56 Go [2026-07-18 10:51:34] step 762050/1000000 | loss=2.1404 vocab=0.7309 token=1.1139 | lr=9.61e-05 | 6.13 steps/s | VRAM=29.12 Go [2026-07-18 10:51:42] step 762100/1000000 | loss=2.2796 vocab=1.2766 token=1.6737 | lr=9.61e-05 | 6.12 steps/s | VRAM=32.56 Go [2026-07-18 10:51:50] step 762150/1000000 | loss=2.1135 vocab=1.1237 token=1.7310 | lr=9.61e-05 | 6.13 steps/s | VRAM=28.31 Go [2026-07-18 10:51:58] step 762200/1000000 | loss=2.1406 vocab=1.0330 token=1.5552 | lr=9.61e-05 | 6.13 steps/s | VRAM=27.97 Go [2026-07-18 10:52:06] step 762250/1000000 | loss=2.1628 vocab=0.8516 token=1.1971 | lr=9.61e-05 | 6.12 steps/s | VRAM=29.13 Go [2026-07-18 10:52:15] step 762300/1000000 | loss=2.1553 vocab=0.6804 token=1.3935 | lr=9.61e-05 | 6.16 steps/s | VRAM=32.09 Go [2026-07-18 10:52:23] step 762350/1000000 | loss=2.1093 vocab=1.0358 token=1.3894 | lr=9.61e-05 | 6.12 steps/s | VRAM=27.06 Go [2026-07-18 10:52:31] step 762400/1000000 | loss=2.2400 vocab=1.1466 token=1.1722 | lr=9.60e-05 | 6.13 steps/s | VRAM=28.34 Go [2026-07-18 10:52:39] step 762450/1000000 | loss=2.2415 vocab=0.9661 token=1.3054 | lr=9.60e-05 | 6.13 steps/s | VRAM=25.94 Go [2026-07-18 10:52:47] step 762500/1000000 | loss=2.1340 vocab=0.7201 token=1.0668 | lr=9.60e-05 | 6.12 steps/s | VRAM=28.63 Go [2026-07-18 10:52:55] step 762550/1000000 | loss=2.2600 vocab=0.6031 token=1.0475 | lr=9.60e-05 | 6.10 steps/s | VRAM=26.54 Go [2026-07-18 10:53:04] step 762600/1000000 | loss=2.1362 vocab=0.8904 token=1.2793 | lr=9.60e-05 | 6.13 steps/s | VRAM=28.45 Go [2026-07-18 10:53:12] step 762650/1000000 | loss=2.1644 vocab=0.5232 token=1.3470 | lr=9.60e-05 | 6.13 steps/s | VRAM=31.93 Go [2026-07-18 10:53:20] step 762700/1000000 | loss=2.2193 vocab=0.5697 token=1.1790 | lr=9.60e-05 | 6.14 steps/s | VRAM=28.65 Go [2026-07-18 10:53:28] step 762750/1000000 | loss=2.2260 vocab=0.8548 token=1.2890 | lr=9.60e-05 | 6.12 steps/s | VRAM=28.47 Go [2026-07-18 10:53:36] step 762800/1000000 | loss=2.2688 vocab=1.3380 token=1.3854 | lr=9.60e-05 | 6.10 steps/s | VRAM=28.99 Go [2026-07-18 10:53:44] step 762850/1000000 | loss=2.3765 vocab=0.9408 token=1.4271 | lr=9.59e-05 | 6.14 steps/s | VRAM=25.96 Go [2026-07-18 10:53:52] step 762900/1000000 | loss=2.1911 vocab=1.2421 token=1.2177 | lr=9.59e-05 | 6.17 steps/s | VRAM=31.82 Go [2026-07-18 10:54:01] step 762950/1000000 | loss=2.1487 vocab=0.4448 token=1.0683 | lr=9.59e-05 | 6.14 steps/s | VRAM=26.53 Go [2026-07-18 10:54:09] step 763000/1000000 | loss=2.1394 vocab=1.1441 token=1.3284 | lr=9.59e-05 | 6.16 steps/s | VRAM=24.75 Go [2026-07-18 10:54:17] step 763050/1000000 | loss=2.1675 vocab=0.6912 token=1.0635 | lr=9.59e-05 | 6.16 steps/s | VRAM=31.04 Go [2026-07-18 10:54:25] step 763100/1000000 | loss=2.0913 vocab=0.6278 token=1.2528 | lr=9.59e-05 | 6.11 steps/s | VRAM=32.19 Go [2026-07-18 10:54:33] step 763150/1000000 | loss=2.1128 vocab=0.7457 token=1.2995 | lr=9.59e-05 | 6.09 steps/s | VRAM=28.81 Go [2026-07-18 10:54:41] step 763200/1000000 | loss=2.2212 vocab=0.7764 token=1.4056 | lr=9.59e-05 | 6.09 steps/s | VRAM=28.96 Go [2026-07-18 10:54:50] step 763250/1000000 | loss=2.1188 vocab=0.8661 token=1.5256 | lr=9.59e-05 | 6.16 steps/s | VRAM=32.21 Go [2026-07-18 10:54:58] step 763300/1000000 | loss=2.2401 vocab=0.6731 token=1.3829 | lr=9.58e-05 | 6.10 steps/s | VRAM=27.92 Go [2026-07-18 10:55:06] step 763350/1000000 | loss=2.1750 vocab=1.4204 token=1.3814 | lr=9.58e-05 | 6.14 steps/s | VRAM=25.90 Go [2026-07-18 10:55:14] step 763400/1000000 | loss=2.1819 vocab=0.7375 token=1.0827 | lr=9.58e-05 | 6.13 steps/s | VRAM=26.73 Go [2026-07-18 10:55:22] step 763450/1000000 | loss=2.2173 vocab=1.2226 token=1.6570 | lr=9.58e-05 | 6.14 steps/s | VRAM=24.97 Go [2026-07-18 10:55:30] step 763500/1000000 | loss=2.1379 vocab=1.4281 token=1.5589 | lr=9.58e-05 | 6.15 steps/s | VRAM=31.47 Go [2026-07-18 10:55:38] step 763550/1000000 | loss=2.2121 vocab=0.8656 token=1.3873 | lr=9.58e-05 | 6.13 steps/s | VRAM=31.94 Go [2026-07-18 10:55:47] step 763600/1000000 | loss=2.2253 vocab=1.1797 token=1.3559 | lr=9.58e-05 | 6.12 steps/s | VRAM=29.31 Go [2026-07-18 10:55:55] step 763650/1000000 | loss=2.1756 vocab=0.6422 token=1.1048 | lr=9.58e-05 | 6.11 steps/s | VRAM=32.21 Go [2026-07-18 10:56:03] step 763700/1000000 | loss=2.2046 vocab=0.6811 token=1.1384 | lr=9.58e-05 | 6.11 steps/s | VRAM=29.02 Go [2026-07-18 10:56:11] step 763750/1000000 | loss=2.3597 vocab=1.0263 token=1.3967 | lr=9.57e-05 | 6.14 steps/s | VRAM=29.21 Go [2026-07-18 10:56:19] step 763800/1000000 | loss=2.3008 vocab=0.5049 token=1.1133 | lr=9.57e-05 | 6.15 steps/s | VRAM=28.46 Go [2026-07-18 10:56:27] step 763850/1000000 | loss=2.1564 vocab=0.9112 token=1.3459 | lr=9.57e-05 | 6.16 steps/s | VRAM=31.37 Go [2026-07-18 10:56:36] step 763900/1000000 | loss=2.2556 vocab=0.9846 token=1.4150 | lr=9.57e-05 | 6.14 steps/s | VRAM=28.80 Go [2026-07-18 10:56:44] step 763950/1000000 | loss=2.1949 vocab=0.6385 token=0.9139 | lr=9.57e-05 | 6.12 steps/s | VRAM=32.85 Go [2026-07-18 10:56:52] step 764000/1000000 | loss=2.1006 vocab=1.2213 token=1.5396 | lr=9.57e-05 | 6.11 steps/s | VRAM=30.02 Go [2026-07-18 10:57:00] step 764050/1000000 | loss=2.1288 vocab=1.3660 token=1.3257 | lr=9.57e-05 | 6.16 steps/s | VRAM=30.52 Go [2026-07-18 10:57:08] step 764100/1000000 | loss=2.2472 vocab=0.8323 token=1.4506 | lr=9.57e-05 | 6.12 steps/s | VRAM=26.66 Go [2026-07-18 10:57:16] step 764150/1000000 | loss=2.4300 vocab=0.7808 token=1.4440 | lr=9.56e-05 | 6.14 steps/s | VRAM=29.20 Go [2026-07-18 10:57:24] step 764200/1000000 | loss=2.1525 vocab=1.0343 token=0.9921 | lr=9.56e-05 | 6.14 steps/s | VRAM=26.36 Go [2026-07-18 10:57:33] step 764250/1000000 | loss=2.1320 vocab=1.1496 token=1.3702 | lr=9.56e-05 | 6.17 steps/s | VRAM=28.73 Go [2026-07-18 10:57:41] step 764300/1000000 | loss=2.1583 vocab=1.0697 token=1.3590 | lr=9.56e-05 | 6.13 steps/s | VRAM=28.18 Go [2026-07-18 10:57:49] step 764350/1000000 | loss=2.1419 vocab=0.5480 token=0.9448 | lr=9.56e-05 | 6.15 steps/s | VRAM=29.18 Go [2026-07-18 10:57:57] step 764400/1000000 | loss=2.0637 vocab=1.0553 token=1.3150 | lr=9.56e-05 | 6.15 steps/s | VRAM=28.27 Go [2026-07-18 10:58:05] step 764450/1000000 | loss=2.1512 vocab=1.2036 token=1.4073 | lr=9.56e-05 | 6.16 steps/s | VRAM=32.43 Go [2026-07-18 10:58:13] step 764500/1000000 | loss=2.1542 vocab=0.6509 token=1.0198 | lr=9.56e-05 | 6.15 steps/s | VRAM=28.44 Go [2026-07-18 10:58:21] step 764550/1000000 | loss=2.1862 vocab=1.1880 token=1.4268 | lr=9.56e-05 | 6.17 steps/s | VRAM=32.18 Go [2026-07-18 10:58:29] step 764600/1000000 | loss=2.1610 vocab=0.8903 token=1.2813 | lr=9.55e-05 | 6.14 steps/s | VRAM=29.85 Go [2026-07-18 10:58:38] step 764650/1000000 | loss=2.1832 vocab=0.8543 token=1.7131 | lr=9.55e-05 | 6.14 steps/s | VRAM=25.91 Go [2026-07-18 10:58:46] step 764700/1000000 | loss=2.1761 vocab=0.9686 token=1.0851 | lr=9.55e-05 | 6.17 steps/s | VRAM=31.57 Go [2026-07-18 10:58:54] step 764750/1000000 | loss=2.1316 vocab=0.8376 token=1.2762 | lr=9.55e-05 | 6.14 steps/s | VRAM=28.41 Go [2026-07-18 10:59:02] step 764800/1000000 | loss=2.2446 vocab=0.4955 token=1.1526 | lr=9.55e-05 | 6.13 steps/s | VRAM=25.96 Go [2026-07-18 10:59:10] step 764850/1000000 | loss=2.2809 vocab=0.8828 token=1.3823 | lr=9.55e-05 | 6.15 steps/s | VRAM=28.97 Go [2026-07-18 10:59:18] step 764900/1000000 | loss=2.2512 vocab=0.9089 token=1.2020 | lr=9.55e-05 | 6.13 steps/s | VRAM=32.12 Go [2026-07-18 10:59:26] step 764950/1000000 | loss=2.1421 vocab=0.4373 token=0.9436 | lr=9.55e-05 | 6.16 steps/s | VRAM=29.40 Go [2026-07-18 10:59:35] step 765000/1000000 | loss=2.1859 vocab=1.7043 token=1.1207 | lr=9.54e-05 | 6.14 steps/s | VRAM=28.44 Go [2026-07-18 10:59:43] step 765050/1000000 | loss=2.2093 vocab=0.6157 token=0.9511 | lr=9.54e-05 | 6.13 steps/s | VRAM=26.11 Go [2026-07-18 10:59:51] step 765100/1000000 | loss=2.1764 vocab=0.5725 token=1.2799 | lr=9.54e-05 | 6.17 steps/s | VRAM=29.16 Go [2026-07-18 10:59:59] step 765150/1000000 | loss=2.2275 vocab=0.9200 token=1.5243 | lr=9.54e-05 | 6.11 steps/s | VRAM=28.54 Go [2026-07-18 11:00:07] step 765200/1000000 | loss=2.1906 vocab=1.3582 token=1.0299 | lr=9.54e-05 | 6.15 steps/s | VRAM=32.68 Go [2026-07-18 11:00:15] step 765250/1000000 | loss=2.1936 vocab=1.1151 token=1.1397 | lr=9.54e-05 | 6.17 steps/s | VRAM=32.07 Go [2026-07-18 11:00:23] step 765300/1000000 | loss=2.2799 vocab=0.5677 token=1.2109 | lr=9.54e-05 | 6.15 steps/s | VRAM=28.37 Go [2026-07-18 11:00:31] step 765350/1000000 | loss=2.1153 vocab=0.6466 token=1.1937 | lr=9.54e-05 | 6.18 steps/s | VRAM=32.02 Go [2026-07-18 11:00:40] step 765400/1000000 | loss=2.1629 vocab=1.0382 token=1.4520 | lr=9.54e-05 | 6.14 steps/s | VRAM=28.85 Go [2026-07-18 11:00:48] step 765450/1000000 | loss=2.0524 vocab=0.6830 token=1.2949 | lr=9.53e-05 | 6.15 steps/s | VRAM=31.90 Go [2026-07-18 11:00:56] step 765500/1000000 | loss=2.1989 vocab=0.7369 token=1.2742 | lr=9.53e-05 | 6.16 steps/s | VRAM=32.03 Go [2026-07-18 11:01:04] step 765550/1000000 | loss=2.1776 vocab=0.9252 token=1.0599 | lr=9.53e-05 | 6.16 steps/s | VRAM=28.63 Go [2026-07-18 11:01:12] step 765600/1000000 | loss=2.1650 vocab=0.8671 token=1.5494 | lr=9.53e-05 | 6.15 steps/s | VRAM=29.14 Go [2026-07-18 11:01:20] step 765650/1000000 | loss=2.1441 vocab=1.4457 token=1.2159 | lr=9.53e-05 | 6.17 steps/s | VRAM=29.00 Go [2026-07-18 11:01:28] step 765700/1000000 | loss=2.1768 vocab=0.8837 token=1.1496 | lr=9.53e-05 | 6.12 steps/s | VRAM=26.46 Go [2026-07-18 11:01:36] step 765750/1000000 | loss=2.3325 vocab=1.5432 token=1.4817 | lr=9.53e-05 | 6.18 steps/s | VRAM=29.30 Go [2026-07-18 11:01:45] step 765800/1000000 | loss=2.1988 vocab=1.0606 token=1.2089 | lr=9.53e-05 | 6.18 steps/s | VRAM=29.24 Go [2026-07-18 11:01:53] step 765850/1000000 | loss=2.1968 vocab=1.1761 token=1.0733 | lr=9.52e-05 | 6.16 steps/s | VRAM=28.54 Go [2026-07-18 11:02:01] step 765900/1000000 | loss=2.1485 vocab=0.6613 token=1.3353 | lr=9.52e-05 | 6.15 steps/s | VRAM=29.34 Go [2026-07-18 11:02:09] step 765950/1000000 | loss=2.1832 vocab=1.0046 token=1.4081 | lr=9.52e-05 | 6.17 steps/s | VRAM=29.03 Go [2026-07-18 11:02:17] step 766000/1000000 | loss=2.2146 vocab=1.0753 token=1.1777 | lr=9.52e-05 | 6.15 steps/s | VRAM=28.07 Go [2026-07-18 11:02:25] step 766050/1000000 | loss=2.1768 vocab=1.4149 token=1.2738 | lr=9.52e-05 | 6.15 steps/s | VRAM=24.41 Go [2026-07-18 11:02:33] step 766100/1000000 | loss=2.1623 vocab=0.7751 token=1.3869 | lr=9.52e-05 | 6.18 steps/s | VRAM=27.47 Go [2026-07-18 11:02:41] step 766150/1000000 | loss=2.1804 vocab=1.0531 token=1.1829 | lr=9.52e-05 | 6.19 steps/s | VRAM=27.77 Go [2026-07-18 11:02:49] step 766200/1000000 | loss=2.1049 vocab=0.8528 token=1.2980 | lr=9.52e-05 | 6.18 steps/s | VRAM=26.18 Go [2026-07-18 11:02:57] step 766250/1000000 | loss=2.0860 vocab=1.1902 token=1.3088 | lr=9.51e-05 | 6.19 steps/s | VRAM=29.23 Go [2026-07-18 11:03:06] step 766300/1000000 | loss=2.0873 vocab=0.7096 token=1.2795 | lr=9.51e-05 | 6.13 steps/s | VRAM=29.86 Go [2026-07-18 11:03:14] step 766350/1000000 | loss=2.3191 vocab=1.4028 token=1.2261 | lr=9.51e-05 | 6.17 steps/s | VRAM=31.34 Go [2026-07-18 11:03:22] step 766400/1000000 | loss=2.0971 vocab=1.1793 token=1.5081 | lr=9.51e-05 | 6.19 steps/s | VRAM=32.62 Go [2026-07-18 11:03:30] step 766450/1000000 | loss=2.1433 vocab=0.9832 token=1.0494 | lr=9.51e-05 | 6.17 steps/s | VRAM=28.16 Go [2026-07-18 11:03:38] step 766500/1000000 | loss=2.1436 vocab=0.6671 token=1.2678 | lr=9.51e-05 | 6.18 steps/s | VRAM=28.24 Go [2026-07-18 11:03:46] step 766550/1000000 | loss=2.2481 vocab=0.7203 token=1.1480 | lr=9.51e-05 | 6.12 steps/s | VRAM=26.23 Go [2026-07-18 11:03:54] step 766600/1000000 | loss=2.1742 vocab=0.6794 token=1.1418 | lr=9.51e-05 | 6.14 steps/s | VRAM=31.61 Go [2026-07-18 11:04:02] step 766650/1000000 | loss=2.0616 vocab=1.0116 token=1.0808 | lr=9.50e-05 | 6.16 steps/s | VRAM=31.57 Go [2026-07-18 11:04:11] step 766700/1000000 | loss=2.1537 vocab=0.3335 token=0.9274 | lr=9.50e-05 | 6.11 steps/s | VRAM=29.63 Go [2026-07-18 11:04:19] step 766750/1000000 | loss=2.0867 vocab=0.6151 token=1.0119 | lr=9.50e-05 | 6.15 steps/s | VRAM=28.04 Go [2026-07-18 11:04:27] step 766800/1000000 | loss=2.2747 vocab=0.7220 token=1.1291 | lr=9.50e-05 | 6.17 steps/s | VRAM=28.95 Go [2026-07-18 11:04:35] step 766850/1000000 | loss=2.1690 vocab=1.1937 token=1.5917 | lr=9.50e-05 | 6.14 steps/s | VRAM=26.41 Go [2026-07-18 11:04:43] step 766900/1000000 | loss=2.2528 vocab=1.5049 token=1.7085 | lr=9.50e-05 | 6.16 steps/s | VRAM=28.40 Go [2026-07-18 11:04:51] step 766950/1000000 | loss=2.2062 vocab=0.8493 token=1.3338 | lr=9.50e-05 | 6.15 steps/s | VRAM=26.49 Go [2026-07-18 11:04:59] step 767000/1000000 | loss=2.1713 vocab=0.7459 token=1.1827 | lr=9.50e-05 | 6.17 steps/s | VRAM=28.09 Go [2026-07-18 11:05:08] step 767050/1000000 | loss=2.1495 vocab=0.8256 token=1.2784 | lr=9.49e-05 | 6.16 steps/s | VRAM=32.12 Go [2026-07-18 11:05:16] step 767100/1000000 | loss=2.1130 vocab=0.7350 token=1.0414 | lr=9.49e-05 | 6.16 steps/s | VRAM=28.75 Go [2026-07-18 11:05:24] step 767150/1000000 | loss=2.1773 vocab=1.0163 token=1.5792 | lr=9.49e-05 | 6.14 steps/s | VRAM=31.24 Go [2026-07-18 11:05:32] step 767200/1000000 | loss=2.1332 vocab=0.9291 token=1.3513 | lr=9.49e-05 | 6.18 steps/s | VRAM=28.95 Go [2026-07-18 11:05:40] step 767250/1000000 | loss=2.1778 vocab=1.0444 token=1.2854 | lr=9.49e-05 | 6.17 steps/s | VRAM=28.12 Go [2026-07-18 11:05:48] step 767300/1000000 | loss=2.1688 vocab=0.8105 token=1.4303 | lr=9.49e-05 | 6.16 steps/s | VRAM=32.32 Go [2026-07-18 11:05:56] step 767350/1000000 | loss=2.2925 vocab=0.9522 token=1.2702 | lr=9.49e-05 | 6.14 steps/s | VRAM=32.65 Go [2026-07-18 11:06:04] step 767400/1000000 | loss=2.2363 vocab=1.2090 token=0.9882 | lr=9.49e-05 | 6.14 steps/s | VRAM=29.26 Go [2026-07-18 11:06:13] step 767450/1000000 | loss=2.1946 vocab=1.1353 token=1.2558 | lr=9.48e-05 | 6.12 steps/s | VRAM=26.53 Go [2026-07-18 11:06:21] step 767500/1000000 | loss=2.2472 vocab=0.8844 token=1.3741 | lr=9.48e-05 | 6.15 steps/s | VRAM=31.45 Go [2026-07-18 11:06:29] step 767550/1000000 | loss=2.1907 vocab=0.9498 token=1.5782 | lr=9.48e-05 | 6.17 steps/s | VRAM=29.15 Go [2026-07-18 11:06:37] step 767600/1000000 | loss=2.1542 vocab=0.3832 token=0.6833 | lr=9.48e-05 | 6.13 steps/s | VRAM=29.91 Go [2026-07-18 11:06:45] step 767650/1000000 | loss=2.1628 vocab=0.9420 token=1.4547 | lr=9.48e-05 | 6.17 steps/s | VRAM=31.74 Go [2026-07-18 11:06:53] step 767700/1000000 | loss=2.2834 vocab=0.7524 token=1.2343 | lr=9.48e-05 | 6.14 steps/s | VRAM=32.76 Go [2026-07-18 11:07:01] step 767750/1000000 | loss=2.1259 vocab=0.8691 token=1.1256 | lr=9.48e-05 | 6.13 steps/s | VRAM=28.80 Go [2026-07-18 11:07:09] step 767800/1000000 | loss=2.1223 vocab=1.4749 token=1.2139 | lr=9.48e-05 | 6.17 steps/s | VRAM=31.39 Go [2026-07-18 11:07:18] step 767850/1000000 | loss=2.3464 vocab=1.3417 token=1.3102 | lr=9.47e-05 | 6.10 steps/s | VRAM=30.44 Go [2026-07-18 11:07:26] step 767900/1000000 | loss=2.1893 vocab=1.1158 token=1.4648 | lr=9.47e-05 | 6.13 steps/s | VRAM=29.04 Go [2026-07-18 11:07:34] step 767950/1000000 | loss=2.0941 vocab=0.6705 token=1.2193 | lr=9.47e-05 | 6.13 steps/s | VRAM=28.17 Go [2026-07-18 11:07:42] step 768000/1000000 | loss=2.0986 vocab=0.8173 token=0.9907 | lr=9.47e-05 | 6.14 steps/s | VRAM=26.41 Go [2026-07-18 11:07:50] step 768050/1000000 | loss=2.0433 vocab=0.8459 token=1.2304 | lr=9.47e-05 | 6.14 steps/s | VRAM=28.08 Go [2026-07-18 11:07:58] step 768100/1000000 | loss=2.2710 vocab=0.8071 token=1.4133 | lr=9.47e-05 | 6.17 steps/s | VRAM=29.83 Go [2026-07-18 11:08:06] step 768150/1000000 | loss=2.2910 vocab=0.7413 token=1.3682 | lr=9.47e-05 | 6.16 steps/s | VRAM=28.07 Go [2026-07-18 11:08:15] step 768200/1000000 | loss=2.1473 vocab=0.9646 token=1.5524 | lr=9.47e-05 | 6.13 steps/s | VRAM=28.79 Go [2026-07-18 11:08:23] step 768250/1000000 | loss=2.2579 vocab=0.9634 token=1.1639 | lr=9.46e-05 | 6.12 steps/s | VRAM=31.64 Go [2026-07-18 11:08:31] step 768300/1000000 | loss=2.0665 vocab=0.6994 token=1.1745 | lr=9.46e-05 | 6.15 steps/s | VRAM=26.69 Go [2026-07-18 11:08:39] step 768350/1000000 | loss=2.2621 vocab=0.8597 token=1.3615 | lr=9.46e-05 | 6.14 steps/s | VRAM=28.70 Go [2026-07-18 11:08:47] step 768400/1000000 | loss=2.1110 vocab=1.1569 token=0.9613 | lr=9.46e-05 | 6.13 steps/s | VRAM=26.25 Go [2026-07-18 11:08:55] step 768450/1000000 | loss=2.1934 vocab=0.8809 token=1.4827 | lr=9.46e-05 | 6.14 steps/s | VRAM=28.50 Go [2026-07-18 11:09:04] step 768500/1000000 | loss=2.2103 vocab=1.3974 token=1.1814 | lr=9.46e-05 | 6.14 steps/s | VRAM=32.25 Go [2026-07-18 11:09:12] step 768550/1000000 | loss=2.1486 vocab=0.4573 token=1.1001 | lr=9.46e-05 | 6.15 steps/s | VRAM=32.68 Go [2026-07-18 11:09:20] step 768600/1000000 | loss=2.1159 vocab=0.9531 token=1.4860 | lr=9.45e-05 | 6.13 steps/s | VRAM=25.03 Go [2026-07-18 11:09:28] step 768650/1000000 | loss=2.1054 vocab=0.7184 token=1.2102 | lr=9.45e-05 | 6.14 steps/s | VRAM=32.93 Go [2026-07-18 11:09:36] step 768700/1000000 | loss=2.1045 vocab=0.7941 token=1.6004 | lr=9.45e-05 | 6.13 steps/s | VRAM=28.31 Go [2026-07-18 11:09:44] step 768750/1000000 | loss=2.0874 vocab=0.7383 token=1.4783 | lr=9.45e-05 | 6.14 steps/s | VRAM=28.75 Go [2026-07-18 11:09:52] step 768800/1000000 | loss=2.2259 vocab=1.0512 token=1.2878 | lr=9.45e-05 | 6.10 steps/s | VRAM=25.33 Go [2026-07-18 11:10:01] step 768850/1000000 | loss=2.1142 vocab=0.8899 token=1.1058 | lr=9.45e-05 | 6.16 steps/s | VRAM=24.81 Go [2026-07-18 11:10:09] step 768900/1000000 | loss=2.0271 vocab=0.8232 token=1.4525 | lr=9.45e-05 | 6.13 steps/s | VRAM=26.31 Go [2026-07-18 11:10:17] step 768950/1000000 | loss=2.0950 vocab=1.4031 token=1.0810 | lr=9.45e-05 | 6.11 steps/s | VRAM=30.10 Go [2026-07-18 11:10:25] step 769000/1000000 | loss=2.2238 vocab=0.8043 token=1.3633 | lr=9.44e-05 | 6.09 steps/s | VRAM=29.04 Go [2026-07-18 11:10:33] step 769050/1000000 | loss=2.1333 vocab=1.0306 token=1.6753 | lr=9.44e-05 | 6.09 steps/s | VRAM=26.55 Go [2026-07-18 11:10:41] step 769100/1000000 | loss=2.1695 vocab=0.6205 token=1.2858 | lr=9.44e-05 | 6.14 steps/s | VRAM=25.60 Go [2026-07-18 11:10:50] step 769150/1000000 | loss=2.1409 vocab=1.3577 token=1.4563 | lr=9.44e-05 | 6.12 steps/s | VRAM=30.97 Go [2026-07-18 11:10:58] step 769200/1000000 | loss=2.1085 vocab=0.6833 token=1.4584 | lr=9.44e-05 | 6.12 steps/s | VRAM=32.53 Go [2026-07-18 11:11:06] step 769250/1000000 | loss=2.1863 vocab=0.5094 token=0.8716 | lr=9.44e-05 | 6.13 steps/s | VRAM=26.73 Go [2026-07-18 11:11:14] step 769300/1000000 | loss=2.1953 vocab=0.6669 token=1.3879 | lr=9.44e-05 | 6.13 steps/s | VRAM=27.99 Go [2026-07-18 11:11:22] step 769350/1000000 | loss=2.1616 vocab=0.9689 token=1.1930 | lr=9.44e-05 | 6.13 steps/s | VRAM=28.58 Go [2026-07-18 11:11:30] step 769400/1000000 | loss=2.1639 vocab=0.5706 token=1.1791 | lr=9.43e-05 | 6.14 steps/s | VRAM=29.75 Go [2026-07-18 11:11:39] step 769450/1000000 | loss=2.1898 vocab=0.7825 token=1.4293 | lr=9.43e-05 | 6.15 steps/s | VRAM=27.68 Go [2026-07-18 11:11:47] step 769500/1000000 | loss=2.0709 vocab=0.7666 token=1.1207 | lr=9.43e-05 | 6.13 steps/s | VRAM=28.09 Go [2026-07-18 11:11:55] step 769550/1000000 | loss=2.1237 vocab=1.1380 token=1.0305 | lr=9.43e-05 | 6.13 steps/s | VRAM=28.51 Go [2026-07-18 11:12:03] step 769600/1000000 | loss=2.1588 vocab=0.7115 token=1.2795 | lr=9.43e-05 | 6.10 steps/s | VRAM=28.04 Go [2026-07-18 11:12:11] step 769650/1000000 | loss=2.1779 vocab=1.0320 token=1.7290 | lr=9.43e-05 | 6.13 steps/s | VRAM=25.93 Go [2026-07-18 11:12:19] step 769700/1000000 | loss=2.0898 vocab=0.9434 token=1.2225 | lr=9.43e-05 | 6.15 steps/s | VRAM=32.16 Go [2026-07-18 11:12:28] step 769750/1000000 | loss=2.2452 vocab=0.6511 token=0.6911 | lr=9.42e-05 | 6.12 steps/s | VRAM=26.46 Go [2026-07-18 11:12:36] step 769800/1000000 | loss=2.1319 vocab=1.0033 token=1.2478 | lr=9.42e-05 | 6.14 steps/s | VRAM=29.16 Go [2026-07-18 11:12:44] step 769850/1000000 | loss=2.0963 vocab=0.8241 token=1.3311 | lr=9.42e-05 | 6.14 steps/s | VRAM=31.67 Go [2026-07-18 11:12:52] step 769900/1000000 | loss=2.0800 vocab=1.0363 token=1.6405 | lr=9.42e-05 | 6.09 steps/s | VRAM=32.55 Go [2026-07-18 11:13:00] step 769950/1000000 | loss=2.1196 vocab=0.6301 token=1.1784 | lr=9.42e-05 | 6.15 steps/s | VRAM=30.64 Go [2026-07-18 11:13:08] step 770000/1000000 | loss=2.2162 vocab=0.6236 token=1.3090 | lr=9.42e-05 | 6.12 steps/s | VRAM=32.50 Go [2026-07-18 11:13:16] step 770050/1000000 | loss=2.2047 vocab=0.6139 token=0.8932 | lr=9.42e-05 | 6.12 steps/s | VRAM=30.24 Go [2026-07-18 11:13:25] step 770100/1000000 | loss=2.2323 vocab=0.7362 token=1.9100 | lr=9.42e-05 | 6.15 steps/s | VRAM=27.99 Go [2026-07-18 11:13:33] step 770150/1000000 | loss=2.1189 vocab=1.0649 token=1.6457 | lr=9.41e-05 | 6.14 steps/s | VRAM=32.05 Go [2026-07-18 11:13:41] step 770200/1000000 | loss=2.2097 vocab=1.0675 token=1.3510 | lr=9.41e-05 | 6.12 steps/s | VRAM=29.18 Go [2026-07-18 11:13:49] step 770250/1000000 | loss=2.1467 vocab=0.6439 token=1.2046 | lr=9.41e-05 | 6.15 steps/s | VRAM=31.62 Go [2026-07-18 11:13:57] step 770300/1000000 | loss=2.1013 vocab=0.8723 token=1.5522 | lr=9.41e-05 | 6.13 steps/s | VRAM=28.73 Go [2026-07-18 11:14:05] step 770350/1000000 | loss=2.1574 vocab=0.7153 token=1.2908 | lr=9.41e-05 | 6.15 steps/s | VRAM=32.05 Go [2026-07-18 11:14:13] step 770400/1000000 | loss=2.2525 vocab=1.0152 token=1.2057 | lr=9.41e-05 | 6.17 steps/s | VRAM=29.19 Go [2026-07-18 11:14:22] step 770450/1000000 | loss=2.1457 vocab=0.7057 token=1.0841 | lr=9.41e-05 | 6.16 steps/s | VRAM=31.29 Go [2026-07-18 11:14:30] step 770500/1000000 | loss=2.1084 vocab=0.8197 token=0.9353 | lr=9.40e-05 | 6.12 steps/s | VRAM=29.10 Go [2026-07-18 11:14:38] step 770550/1000000 | loss=2.2671 vocab=0.8456 token=1.3647 | lr=9.40e-05 | 6.12 steps/s | VRAM=28.27 Go [2026-07-18 11:14:46] step 770600/1000000 | loss=2.1562 vocab=0.9912 token=1.0149 | lr=9.40e-05 | 6.13 steps/s | VRAM=25.59 Go [2026-07-18 11:14:54] step 770650/1000000 | loss=2.2355 vocab=0.8559 token=1.4475 | lr=9.40e-05 | 6.14 steps/s | VRAM=25.84 Go [2026-07-18 11:15:02] step 770700/1000000 | loss=2.1231 vocab=0.9386 token=1.2883 | lr=9.40e-05 | 6.12 steps/s | VRAM=32.64 Go [2026-07-18 11:15:11] step 770750/1000000 | loss=2.2302 vocab=0.6514 token=0.9315 | lr=9.40e-05 | 6.12 steps/s | VRAM=28.98 Go [2026-07-18 11:15:19] step 770800/1000000 | loss=2.1299 vocab=0.6891 token=0.9700 | lr=9.40e-05 | 6.14 steps/s | VRAM=28.94 Go [2026-07-18 11:15:27] step 770850/1000000 | loss=2.2175 vocab=2.1403 token=1.8480 | lr=9.39e-05 | 6.16 steps/s | VRAM=25.08 Go [2026-07-18 11:15:35] step 770900/1000000 | loss=2.1549 vocab=0.7962 token=1.1229 | lr=9.39e-05 | 6.13 steps/s | VRAM=29.71 Go [2026-07-18 11:15:43] step 770950/1000000 | loss=2.1293 vocab=0.4004 token=1.0652 | lr=9.39e-05 | 6.14 steps/s | VRAM=26.76 Go [2026-07-18 11:15:51] step 771000/1000000 | loss=2.2105 vocab=1.1789 token=1.3021 | lr=9.39e-05 | 6.12 steps/s | VRAM=28.25 Go [2026-07-18 11:15:59] step 771050/1000000 | loss=2.1498 vocab=0.7157 token=1.2983 | lr=9.39e-05 | 6.15 steps/s | VRAM=26.10 Go [2026-07-18 11:16:08] step 771100/1000000 | loss=2.1510 vocab=0.9437 token=1.3225 | lr=9.39e-05 | 6.14 steps/s | VRAM=32.26 Go [2026-07-18 11:16:16] step 771150/1000000 | loss=2.1066 vocab=0.9128 token=1.3648 | lr=9.39e-05 | 6.15 steps/s | VRAM=31.85 Go [2026-07-18 11:16:24] step 771200/1000000 | loss=2.1804 vocab=0.9681 token=1.3409 | lr=9.39e-05 | 6.12 steps/s | VRAM=29.00 Go [2026-07-18 11:16:32] step 771250/1000000 | loss=2.2276 vocab=1.3035 token=1.6077 | lr=9.38e-05 | 6.12 steps/s | VRAM=25.86 Go [2026-07-18 11:16:40] step 771300/1000000 | loss=2.0659 vocab=1.7263 token=1.3402 | lr=9.38e-05 | 6.13 steps/s | VRAM=26.17 Go [2026-07-18 11:16:48] step 771350/1000000 | loss=2.1165 vocab=0.5681 token=1.2296 | lr=9.38e-05 | 6.14 steps/s | VRAM=32.18 Go [2026-07-18 11:16:56] step 771400/1000000 | loss=2.2076 vocab=0.8480 token=1.5451 | lr=9.38e-05 | 6.13 steps/s | VRAM=29.68 Go [2026-07-18 11:17:05] step 771450/1000000 | loss=2.1812 vocab=1.0579 token=1.2479 | lr=9.38e-05 | 6.14 steps/s | VRAM=31.32 Go [2026-07-18 11:17:13] step 771500/1000000 | loss=2.0590 vocab=1.0541 token=1.2250 | lr=9.38e-05 | 6.10 steps/s | VRAM=27.23 Go [2026-07-18 11:17:21] step 771550/1000000 | loss=2.1390 vocab=0.8278 token=1.2493 | lr=9.38e-05 | 6.12 steps/s | VRAM=27.49 Go [2026-07-18 11:17:29] step 771600/1000000 | loss=2.2608 vocab=1.5291 token=1.2845 | lr=9.37e-05 | 6.14 steps/s | VRAM=26.20 Go [2026-07-18 11:17:37] step 771650/1000000 | loss=2.1287 vocab=0.8629 token=1.0435 | lr=9.37e-05 | 6.15 steps/s | VRAM=31.72 Go [2026-07-18 11:17:45] step 771700/1000000 | loss=2.2818 vocab=1.0144 token=1.5386 | lr=9.37e-05 | 6.13 steps/s | VRAM=32.30 Go [2026-07-18 11:17:54] step 771750/1000000 | loss=2.2684 vocab=0.9071 token=1.1559 | lr=9.37e-05 | 6.13 steps/s | VRAM=32.08 Go [2026-07-18 11:18:02] step 771800/1000000 | loss=2.1784 vocab=0.6228 token=0.8560 | lr=9.37e-05 | 6.14 steps/s | VRAM=32.18 Go [2026-07-18 11:18:10] step 771850/1000000 | loss=2.2427 vocab=1.1876 token=1.4111 | lr=9.37e-05 | 6.13 steps/s | VRAM=28.83 Go [2026-07-18 11:18:18] step 771900/1000000 | loss=2.2699 vocab=1.1186 token=1.0667 | lr=9.37e-05 | 6.14 steps/s | VRAM=29.45 Go [2026-07-18 11:18:26] step 771950/1000000 | loss=2.2341 vocab=1.3044 token=1.4609 | lr=9.36e-05 | 6.15 steps/s | VRAM=28.05 Go [2026-07-18 11:18:34] step 772000/1000000 | loss=2.1187 vocab=0.8078 token=1.1467 | lr=9.36e-05 | 6.13 steps/s | VRAM=28.99 Go [2026-07-18 11:18:42] step 772050/1000000 | loss=2.2038 vocab=0.7665 token=1.3337 | lr=9.36e-05 | 6.15 steps/s | VRAM=28.46 Go [2026-07-18 11:18:51] step 772100/1000000 | loss=2.1984 vocab=0.7876 token=1.1924 | lr=9.36e-05 | 6.15 steps/s | VRAM=31.94 Go [2026-07-18 11:18:59] step 772150/1000000 | loss=2.1740 vocab=0.7093 token=1.3748 | lr=9.36e-05 | 6.15 steps/s | VRAM=31.38 Go [2026-07-18 11:19:07] step 772200/1000000 | loss=2.2037 vocab=0.7727 token=1.4097 | lr=9.36e-05 | 6.10 steps/s | VRAM=25.79 Go [2026-07-18 11:19:15] step 772250/1000000 | loss=2.2969 vocab=1.2457 token=1.2178 | lr=9.36e-05 | 6.13 steps/s | VRAM=32.36 Go [2026-07-18 11:19:23] step 772300/1000000 | loss=2.1972 vocab=0.5580 token=0.9525 | lr=9.35e-05 | 6.14 steps/s | VRAM=28.94 Go [2026-07-18 11:19:31] step 772350/1000000 | loss=2.2725 vocab=1.4635 token=0.9137 | lr=9.35e-05 | 6.12 steps/s | VRAM=25.86 Go [2026-07-18 11:19:40] step 772400/1000000 | loss=2.0713 vocab=0.7017 token=1.5436 | lr=9.35e-05 | 6.12 steps/s | VRAM=28.51 Go [2026-07-18 11:19:48] step 772450/1000000 | loss=2.1609 vocab=0.5855 token=1.1826 | lr=9.35e-05 | 6.12 steps/s | VRAM=28.05 Go [2026-07-18 11:19:56] step 772500/1000000 | loss=2.2981 vocab=0.6429 token=1.1701 | lr=9.35e-05 | 6.15 steps/s | VRAM=31.81 Go [2026-07-18 11:20:04] step 772550/1000000 | loss=2.1071 vocab=1.0699 token=1.3868 | lr=9.35e-05 | 6.14 steps/s | VRAM=28.26 Go [2026-07-18 11:20:12] step 772600/1000000 | loss=2.1401 vocab=0.9931 token=1.7250 | lr=9.35e-05 | 6.15 steps/s | VRAM=28.30 Go [2026-07-18 11:20:20] step 772650/1000000 | loss=2.1782 vocab=1.0204 token=1.0982 | lr=9.34e-05 | 6.12 steps/s | VRAM=30.04 Go [2026-07-18 11:20:28] step 772700/1000000 | loss=2.2011 vocab=0.6947 token=0.9442 | lr=9.34e-05 | 6.11 steps/s | VRAM=28.88 Go [2026-07-18 11:20:37] step 772750/1000000 | loss=2.1885 vocab=0.7192 token=1.2347 | lr=9.34e-05 | 6.12 steps/s | VRAM=28.49 Go [2026-07-18 11:20:45] step 772800/1000000 | loss=2.2046 vocab=0.9526 token=1.4149 | lr=9.34e-05 | 6.14 steps/s | VRAM=29.02 Go [2026-07-18 11:20:53] step 772850/1000000 | loss=2.1450 vocab=0.5423 token=1.2369 | lr=9.34e-05 | 6.12 steps/s | VRAM=28.32 Go [2026-07-18 11:21:01] step 772900/1000000 | loss=2.1678 vocab=0.6245 token=0.9666 | lr=9.34e-05 | 6.07 steps/s | VRAM=26.37 Go [2026-07-18 11:21:09] step 772950/1000000 | loss=2.2012 vocab=0.7610 token=1.1517 | lr=9.34e-05 | 6.15 steps/s | VRAM=28.01 Go [2026-07-18 11:21:17] step 773000/1000000 | loss=2.1862 vocab=0.6934 token=1.1183 | lr=9.33e-05 | 6.10 steps/s | VRAM=26.12 Go [2026-07-18 11:21:26] step 773050/1000000 | loss=2.2204 vocab=0.8954 token=1.4685 | lr=9.33e-05 | 6.14 steps/s | VRAM=28.42 Go [2026-07-18 11:21:34] step 773100/1000000 | loss=2.2545 vocab=0.9108 token=1.1219 | lr=9.33e-05 | 6.15 steps/s | VRAM=32.43 Go [2026-07-18 11:21:42] step 773150/1000000 | loss=2.0430 vocab=0.3819 token=1.1585 | lr=9.33e-05 | 6.14 steps/s | VRAM=25.43 Go [2026-07-18 11:21:50] step 773200/1000000 | loss=2.0847 vocab=1.2810 token=1.2823 | lr=9.33e-05 | 6.13 steps/s | VRAM=29.53 Go [2026-07-18 11:21:58] step 773250/1000000 | loss=2.1240 vocab=0.9915 token=1.3642 | lr=9.33e-05 | 6.11 steps/s | VRAM=25.14 Go [2026-07-18 11:22:06] step 773300/1000000 | loss=2.1475 vocab=0.6243 token=0.8059 | lr=9.33e-05 | 6.14 steps/s | VRAM=27.71 Go [2026-07-18 11:22:15] step 773350/1000000 | loss=2.1418 vocab=0.7878 token=1.2299 | lr=9.32e-05 | 6.15 steps/s | VRAM=29.19 Go [2026-07-18 11:22:23] step 773400/1000000 | loss=2.1413 vocab=1.2576 token=1.4264 | lr=9.32e-05 | 6.15 steps/s | VRAM=27.79 Go [2026-07-18 11:22:31] step 773450/1000000 | loss=2.1495 vocab=0.9426 token=1.3806 | lr=9.32e-05 | 6.11 steps/s | VRAM=31.98 Go [2026-07-18 11:22:39] step 773500/1000000 | loss=2.2416 vocab=0.9785 token=1.6220 | lr=9.32e-05 | 6.12 steps/s | VRAM=28.98 Go [2026-07-18 11:22:47] step 773550/1000000 | loss=2.1520 vocab=0.7918 token=0.8936 | lr=9.32e-05 | 6.15 steps/s | VRAM=28.86 Go [2026-07-18 11:22:55] step 773600/1000000 | loss=2.2382 vocab=0.9434 token=1.3744 | lr=9.32e-05 | 6.12 steps/s | VRAM=29.05 Go [2026-07-18 11:23:03] step 773650/1000000 | loss=2.2690 vocab=1.1247 token=1.3336 | lr=9.32e-05 | 6.16 steps/s | VRAM=27.93 Go [2026-07-18 11:23:12] step 773700/1000000 | loss=2.1973 vocab=1.1321 token=1.6717 | lr=9.31e-05 | 6.12 steps/s | VRAM=26.24 Go [2026-07-18 11:23:20] step 773750/1000000 | loss=2.1628 vocab=0.6130 token=1.2998 | lr=9.31e-05 | 6.10 steps/s | VRAM=28.92 Go [2026-07-18 11:23:28] step 773800/1000000 | loss=2.1342 vocab=1.1113 token=1.3692 | lr=9.31e-05 | 6.09 steps/s | VRAM=28.45 Go [2026-07-18 11:23:36] step 773850/1000000 | loss=2.1853 vocab=1.0005 token=1.2774 | lr=9.31e-05 | 6.13 steps/s | VRAM=24.61 Go [2026-07-18 11:23:44] step 773900/1000000 | loss=2.1157 vocab=0.6074 token=1.2760 | lr=9.31e-05 | 6.13 steps/s | VRAM=32.46 Go [2026-07-18 11:23:52] step 773950/1000000 | loss=2.0710 vocab=0.7238 token=1.2373 | lr=9.31e-05 | 6.13 steps/s | VRAM=32.55 Go [2026-07-18 11:24:01] step 774000/1000000 | loss=2.2215 vocab=0.8340 token=1.2737 | lr=9.31e-05 | 6.14 steps/s | VRAM=24.46 Go [2026-07-18 11:24:09] step 774050/1000000 | loss=2.2193 vocab=1.2577 token=1.3074 | lr=9.30e-05 | 6.13 steps/s | VRAM=28.56 Go [2026-07-18 11:24:17] step 774100/1000000 | loss=2.0729 vocab=1.3526 token=1.7961 | lr=9.30e-05 | 6.13 steps/s | VRAM=26.82 Go [2026-07-18 11:24:25] step 774150/1000000 | loss=2.1329 vocab=1.0256 token=1.3606 | lr=9.30e-05 | 6.13 steps/s | VRAM=28.20 Go [2026-07-18 11:24:33] step 774200/1000000 | loss=2.1401 vocab=0.7035 token=1.4628 | lr=9.30e-05 | 6.17 steps/s | VRAM=31.16 Go [2026-07-18 11:24:41] step 774250/1000000 | loss=2.3083 vocab=0.7024 token=0.7598 | lr=9.30e-05 | 6.13 steps/s | VRAM=29.02 Go [2026-07-18 11:24:49] step 774300/1000000 | loss=2.1797 vocab=0.6759 token=1.2803 | lr=9.30e-05 | 6.13 steps/s | VRAM=28.76 Go [2026-07-18 11:24:58] step 774350/1000000 | loss=2.1111 vocab=0.7983 token=1.4828 | lr=9.30e-05 | 6.15 steps/s | VRAM=28.85 Go [2026-07-18 11:25:06] step 774400/1000000 | loss=2.1342 vocab=1.3216 token=1.4613 | lr=9.29e-05 | 6.13 steps/s | VRAM=26.22 Go [2026-07-18 11:25:14] step 774450/1000000 | loss=2.2199 vocab=0.9017 token=1.6631 | lr=9.29e-05 | 6.15 steps/s | VRAM=24.74 Go [2026-07-18 11:25:22] step 774500/1000000 | loss=2.1717 vocab=0.7535 token=0.6228 | lr=9.29e-05 | 6.13 steps/s | VRAM=28.98 Go [2026-07-18 11:25:30] step 774550/1000000 | loss=2.2065 vocab=1.0637 token=1.2204 | lr=9.29e-05 | 6.14 steps/s | VRAM=32.78 Go [2026-07-18 11:25:38] step 774600/1000000 | loss=2.0998 vocab=0.6109 token=1.0589 | lr=9.29e-05 | 6.14 steps/s | VRAM=29.24 Go [2026-07-18 11:25:46] step 774650/1000000 | loss=2.2604 vocab=1.3162 token=1.5669 | lr=9.29e-05 | 6.14 steps/s | VRAM=28.54 Go [2026-07-18 11:25:55] step 774700/1000000 | loss=2.1270 vocab=0.9463 token=1.5116 | lr=9.29e-05 | 6.12 steps/s | VRAM=24.68 Go [2026-07-18 11:26:03] step 774750/1000000 | loss=2.1287 vocab=0.6394 token=1.4889 | lr=9.28e-05 | 6.13 steps/s | VRAM=28.74 Go [2026-07-18 11:26:11] step 774800/1000000 | loss=2.2360 vocab=0.6153 token=1.2563 | lr=9.28e-05 | 6.14 steps/s | VRAM=28.25 Go [2026-07-18 11:26:19] step 774850/1000000 | loss=2.1750 vocab=0.8231 token=1.4088 | lr=9.28e-05 | 6.12 steps/s | VRAM=29.36 Go [2026-07-18 11:26:27] step 774900/1000000 | loss=2.1591 vocab=0.9964 token=1.1850 | lr=9.28e-05 | 6.15 steps/s | VRAM=25.81 Go [2026-07-18 11:26:35] step 774950/1000000 | loss=2.1491 vocab=0.5934 token=0.9269 | lr=9.28e-05 | 6.15 steps/s | VRAM=29.04 Go [2026-07-18 11:26:44] step 775000/1000000 | loss=2.1068 vocab=1.2052 token=1.4525 | lr=9.28e-05 | 6.15 steps/s | VRAM=28.99 Go [2026-07-18 11:26:52] step 775050/1000000 | loss=2.1864 vocab=1.1347 token=1.5809 | lr=9.28e-05 | 6.13 steps/s | VRAM=28.92 Go [2026-07-18 11:27:00] step 775100/1000000 | loss=2.1999 vocab=0.9481 token=1.2373 | lr=9.27e-05 | 6.15 steps/s | VRAM=31.86 Go [2026-07-18 11:27:08] step 775150/1000000 | loss=2.2022 vocab=0.8925 token=1.2476 | lr=9.27e-05 | 6.16 steps/s | VRAM=28.69 Go [2026-07-18 11:27:16] step 775200/1000000 | loss=2.1204 vocab=0.6512 token=0.9551 | lr=9.27e-05 | 6.12 steps/s | VRAM=29.02 Go [2026-07-18 11:27:24] step 775250/1000000 | loss=2.1658 vocab=1.1200 token=1.4264 | lr=9.27e-05 | 6.15 steps/s | VRAM=29.42 Go [2026-07-18 11:27:32] step 775300/1000000 | loss=2.1498 vocab=1.2181 token=1.5142 | lr=9.27e-05 | 6.13 steps/s | VRAM=29.02 Go [2026-07-18 11:27:41] step 775350/1000000 | loss=2.1357 vocab=0.9504 token=1.5903 | lr=9.27e-05 | 6.15 steps/s | VRAM=29.66 Go [2026-07-18 11:27:49] step 775400/1000000 | loss=2.2128 vocab=0.6842 token=1.2507 | lr=9.26e-05 | 6.16 steps/s | VRAM=28.97 Go [2026-07-18 11:27:57] step 775450/1000000 | loss=2.1733 vocab=0.7416 token=1.4011 | lr=9.26e-05 | 6.14 steps/s | VRAM=28.62 Go [2026-07-18 11:28:05] step 775500/1000000 | loss=2.1463 vocab=0.9986 token=1.2679 | lr=9.26e-05 | 6.12 steps/s | VRAM=26.52 Go [2026-07-18 11:28:13] step 775550/1000000 | loss=2.1264 vocab=1.5168 token=2.1421 | lr=9.26e-05 | 6.15 steps/s | VRAM=32.82 Go [2026-07-18 11:28:21] step 775600/1000000 | loss=2.1816 vocab=0.9464 token=1.2607 | lr=9.26e-05 | 6.17 steps/s | VRAM=25.10 Go [2026-07-18 11:28:29] step 775650/1000000 | loss=2.1474 vocab=0.9729 token=0.9323 | lr=9.26e-05 | 6.14 steps/s | VRAM=28.28 Go [2026-07-18 11:28:38] step 775700/1000000 | loss=2.1008 vocab=0.4855 token=1.2051 | lr=9.26e-05 | 6.08 steps/s | VRAM=26.27 Go [2026-07-18 11:28:46] step 775750/1000000 | loss=2.1790 vocab=0.5864 token=1.5875 | lr=9.25e-05 | 6.12 steps/s | VRAM=29.04 Go [2026-07-18 11:28:54] step 775800/1000000 | loss=2.2060 vocab=0.4512 token=1.3237 | lr=9.25e-05 | 6.10 steps/s | VRAM=26.52 Go [2026-07-18 11:29:02] step 775850/1000000 | loss=2.1868 vocab=0.9933 token=1.4255 | lr=9.25e-05 | 6.15 steps/s | VRAM=31.57 Go [2026-07-18 11:29:10] step 775900/1000000 | loss=2.2122 vocab=0.7865 token=1.5544 | lr=9.25e-05 | 6.13 steps/s | VRAM=32.50 Go [2026-07-18 11:29:18] step 775950/1000000 | loss=2.1139 vocab=0.9354 token=1.1313 | lr=9.25e-05 | 6.12 steps/s | VRAM=32.41 Go [2026-07-18 11:29:27] step 776000/1000000 | loss=2.1427 vocab=0.7733 token=1.1498 | lr=9.25e-05 | 6.11 steps/s | VRAM=25.60 Go [2026-07-18 11:29:35] step 776050/1000000 | loss=2.2760 vocab=1.2207 token=1.0304 | lr=9.24e-05 | 6.15 steps/s | VRAM=32.76 Go [2026-07-18 11:29:43] step 776100/1000000 | loss=2.2466 vocab=0.4712 token=0.9627 | lr=9.24e-05 | 6.11 steps/s | VRAM=28.76 Go [2026-07-18 11:29:51] step 776150/1000000 | loss=2.2997 vocab=0.5156 token=1.3246 | lr=9.24e-05 | 6.14 steps/s | VRAM=31.52 Go [2026-07-18 11:29:59] step 776200/1000000 | loss=2.1685 vocab=0.7785 token=1.1287 | lr=9.24e-05 | 6.12 steps/s | VRAM=28.64 Go [2026-07-18 11:30:07] step 776250/1000000 | loss=2.1245 vocab=0.6037 token=0.9803 | lr=9.24e-05 | 6.14 steps/s | VRAM=28.22 Go [2026-07-18 11:30:15] step 776300/1000000 | loss=2.1630 vocab=0.8347 token=1.3600 | lr=9.24e-05 | 6.13 steps/s | VRAM=32.53 Go [2026-07-18 11:30:24] step 776350/1000000 | loss=2.2582 vocab=0.7856 token=1.2814 | lr=9.24e-05 | 6.10 steps/s | VRAM=25.36 Go [2026-07-18 11:30:32] step 776400/1000000 | loss=2.1970 vocab=0.4239 token=1.2001 | lr=9.23e-05 | 6.11 steps/s | VRAM=29.13 Go [2026-07-18 11:30:40] step 776450/1000000 | loss=2.3200 vocab=1.3431 token=1.7498 | lr=9.23e-05 | 6.13 steps/s | VRAM=25.40 Go [2026-07-18 11:30:48] step 776500/1000000 | loss=2.1475 vocab=0.8541 token=1.4885 | lr=9.23e-05 | 6.11 steps/s | VRAM=27.94 Go [2026-07-18 11:30:56] step 776550/1000000 | loss=2.1299 vocab=1.1916 token=1.3859 | lr=9.23e-05 | 6.11 steps/s | VRAM=32.57 Go [2026-07-18 11:31:05] step 776600/1000000 | loss=2.1901 vocab=0.6439 token=1.2357 | lr=9.23e-05 | 6.13 steps/s | VRAM=32.22 Go [2026-07-18 11:31:13] step 776650/1000000 | loss=2.1654 vocab=0.6851 token=0.7923 | lr=9.23e-05 | 6.12 steps/s | VRAM=29.12 Go [2026-07-18 11:31:21] step 776700/1000000 | loss=2.1378 vocab=0.8510 token=1.2844 | lr=9.23e-05 | 6.11 steps/s | VRAM=29.37 Go [2026-07-18 11:31:29] step 776750/1000000 | loss=2.2161 vocab=1.0393 token=1.2058 | lr=9.22e-05 | 6.14 steps/s | VRAM=29.25 Go [2026-07-18 11:31:37] step 776800/1000000 | loss=2.1049 vocab=0.7711 token=1.2448 | lr=9.22e-05 | 6.13 steps/s | VRAM=29.34 Go [2026-07-18 11:31:45] step 776850/1000000 | loss=2.2272 vocab=0.8723 token=1.1091 | lr=9.22e-05 | 6.12 steps/s | VRAM=26.26 Go [2026-07-18 11:31:54] step 776900/1000000 | loss=2.0400 vocab=0.5783 token=0.9658 | lr=9.22e-05 | 6.12 steps/s | VRAM=27.99 Go [2026-07-18 11:32:02] step 776950/1000000 | loss=2.2175 vocab=1.4970 token=1.6896 | lr=9.22e-05 | 6.12 steps/s | VRAM=28.09 Go [2026-07-18 11:32:10] step 777000/1000000 | loss=2.1573 vocab=0.8508 token=1.2001 | lr=9.22e-05 | 6.13 steps/s | VRAM=27.38 Go [2026-07-18 11:32:18] step 777050/1000000 | loss=2.0942 vocab=0.7841 token=1.0712 | lr=9.21e-05 | 6.15 steps/s | VRAM=32.59 Go [2026-07-18 11:32:26] step 777100/1000000 | loss=2.2366 vocab=0.5601 token=1.1707 | lr=9.21e-05 | 6.12 steps/s | VRAM=28.62 Go [2026-07-18 11:32:34] step 777150/1000000 | loss=2.1530 vocab=0.9358 token=1.0373 | lr=9.21e-05 | 6.12 steps/s | VRAM=26.49 Go [2026-07-18 11:32:42] step 777200/1000000 | loss=2.1153 vocab=0.9330 token=1.2526 | lr=9.21e-05 | 6.13 steps/s | VRAM=31.40 Go [2026-07-18 11:32:51] step 777250/1000000 | loss=2.1591 vocab=1.3215 token=1.5295 | lr=9.21e-05 | 6.10 steps/s | VRAM=28.64 Go [2026-07-18 11:32:59] step 777300/1000000 | loss=2.1188 vocab=1.2398 token=1.6908 | lr=9.21e-05 | 6.15 steps/s | VRAM=25.03 Go [2026-07-18 11:33:07] step 777350/1000000 | loss=2.1863 vocab=0.7785 token=1.3457 | lr=9.21e-05 | 6.11 steps/s | VRAM=28.92 Go [2026-07-18 11:33:15] step 777400/1000000 | loss=2.2652 vocab=1.1200 token=1.5043 | lr=9.20e-05 | 6.12 steps/s | VRAM=32.49 Go [2026-07-18 11:33:23] step 777450/1000000 | loss=2.2548 vocab=0.8036 token=1.0314 | lr=9.20e-05 | 6.13 steps/s | VRAM=31.72 Go [2026-07-18 11:33:31] step 777500/1000000 | loss=2.2004 vocab=0.8388 token=1.4814 | lr=9.20e-05 | 6.15 steps/s | VRAM=28.35 Go [2026-07-18 11:33:40] step 777550/1000000 | loss=2.1819 vocab=1.0907 token=1.8362 | lr=9.20e-05 | 6.12 steps/s | VRAM=26.46 Go [2026-07-18 11:33:48] step 777600/1000000 | loss=2.1821 vocab=0.6589 token=1.4121 | lr=9.20e-05 | 6.16 steps/s | VRAM=28.12 Go [2026-07-18 11:33:56] step 777650/1000000 | loss=2.1858 vocab=0.6995 token=1.1874 | lr=9.20e-05 | 6.11 steps/s | VRAM=28.78 Go [2026-07-18 11:34:04] step 777700/1000000 | loss=2.2560 vocab=1.2842 token=1.3133 | lr=9.19e-05 | 6.14 steps/s | VRAM=31.47 Go [2026-07-18 11:34:12] step 777750/1000000 | loss=2.1353 vocab=0.6869 token=1.1038 | lr=9.19e-05 | 6.11 steps/s | VRAM=25.69 Go [2026-07-18 11:34:20] step 777800/1000000 | loss=2.0904 vocab=0.5976 token=1.2016 | lr=9.19e-05 | 6.15 steps/s | VRAM=25.90 Go [2026-07-18 11:34:28] step 777850/1000000 | loss=2.1202 vocab=0.9039 token=1.6722 | lr=9.19e-05 | 6.15 steps/s | VRAM=22.52 Go [2026-07-18 11:34:37] step 777900/1000000 | loss=2.1419 vocab=1.4224 token=1.4128 | lr=9.19e-05 | 6.13 steps/s | VRAM=32.56 Go [2026-07-18 11:34:45] step 777950/1000000 | loss=2.0951 vocab=0.8351 token=1.3825 | lr=9.19e-05 | 6.15 steps/s | VRAM=29.33 Go [2026-07-18 11:34:53] step 778000/1000000 | loss=2.1511 vocab=0.6711 token=1.4419 | lr=9.18e-05 | 6.14 steps/s | VRAM=29.03 Go [2026-07-18 11:35:01] step 778050/1000000 | loss=2.1208 vocab=0.6789 token=1.1181 | lr=9.18e-05 | 6.15 steps/s | VRAM=31.43 Go [2026-07-18 11:35:09] step 778100/1000000 | loss=2.2037 vocab=0.8040 token=0.9975 | lr=9.18e-05 | 6.12 steps/s | VRAM=32.89 Go [2026-07-18 11:35:17] step 778150/1000000 | loss=2.2488 vocab=0.4524 token=0.9116 | lr=9.18e-05 | 6.11 steps/s | VRAM=26.72 Go [2026-07-18 11:35:26] step 778200/1000000 | loss=2.2312 vocab=1.2788 token=1.1577 | lr=9.18e-05 | 6.14 steps/s | VRAM=24.87 Go [2026-07-18 11:35:34] step 778250/1000000 | loss=2.1545 vocab=0.4701 token=1.0048 | lr=9.18e-05 | 6.16 steps/s | VRAM=29.24 Go [2026-07-18 11:35:42] step 778300/1000000 | loss=2.2244 vocab=0.8620 token=1.6652 | lr=9.18e-05 | 6.12 steps/s | VRAM=26.81 Go [2026-07-18 11:35:50] step 778350/1000000 | loss=2.1964 vocab=0.6691 token=1.6571 | lr=9.17e-05 | 6.19 steps/s | VRAM=30.68 Go [2026-07-18 11:35:58] step 778400/1000000 | loss=2.2289 vocab=0.7319 token=1.1652 | lr=9.17e-05 | 6.14 steps/s | VRAM=29.14 Go [2026-07-18 11:36:06] step 778450/1000000 | loss=2.3731 vocab=0.9203 token=1.4662 | lr=9.17e-05 | 6.11 steps/s | VRAM=27.38 Go [2026-07-18 11:36:14] step 778500/1000000 | loss=2.1716 vocab=0.6469 token=1.2644 | lr=9.17e-05 | 6.14 steps/s | VRAM=24.70 Go [2026-07-18 11:36:23] step 778550/1000000 | loss=2.1466 vocab=1.0523 token=1.0044 | lr=9.17e-05 | 6.12 steps/s | VRAM=28.29 Go [2026-07-18 11:36:31] step 778600/1000000 | loss=2.2050 vocab=1.1880 token=1.3438 | lr=9.17e-05 | 6.13 steps/s | VRAM=26.59 Go [2026-07-18 11:36:39] step 778650/1000000 | loss=2.1090 vocab=0.9148 token=1.5110 | lr=9.16e-05 | 6.13 steps/s | VRAM=28.44 Go [2026-07-18 11:36:47] step 778700/1000000 | loss=2.1606 vocab=0.9143 token=1.3741 | lr=9.16e-05 | 6.15 steps/s | VRAM=30.19 Go [2026-07-18 11:36:55] step 778750/1000000 | loss=2.1396 vocab=0.8894 token=1.2999 | lr=9.16e-05 | 6.16 steps/s | VRAM=28.55 Go [2026-07-18 11:37:03] step 778800/1000000 | loss=2.0918 vocab=0.3797 token=0.9869 | lr=9.16e-05 | 6.16 steps/s | VRAM=25.50 Go [2026-07-18 11:37:11] step 778850/1000000 | loss=2.2243 vocab=1.3486 token=1.4113 | lr=9.16e-05 | 6.12 steps/s | VRAM=26.54 Go [2026-07-18 11:37:20] step 778900/1000000 | loss=2.2839 vocab=1.0275 token=1.2829 | lr=9.16e-05 | 6.14 steps/s | VRAM=29.39 Go [2026-07-18 11:37:28] step 778950/1000000 | loss=2.2274 vocab=0.6413 token=0.8258 | lr=9.15e-05 | 6.15 steps/s | VRAM=28.88 Go [2026-07-18 11:37:36] step 779000/1000000 | loss=2.1249 vocab=0.3143 token=0.9760 | lr=9.15e-05 | 6.16 steps/s | VRAM=30.96 Go [2026-07-18 11:37:44] step 779050/1000000 | loss=2.1764 vocab=0.9901 token=1.5097 | lr=9.15e-05 | 6.13 steps/s | VRAM=29.65 Go [2026-07-18 11:37:52] step 779100/1000000 | loss=2.1414 vocab=0.7324 token=1.0186 | lr=9.15e-05 | 6.15 steps/s | VRAM=26.17 Go [2026-07-18 11:38:00] step 779150/1000000 | loss=2.1716 vocab=0.9026 token=0.8701 | lr=9.15e-05 | 6.15 steps/s | VRAM=32.47 Go [2026-07-18 11:38:08] step 779200/1000000 | loss=2.2142 vocab=0.5765 token=1.2928 | lr=9.15e-05 | 6.13 steps/s | VRAM=28.44 Go [2026-07-18 11:38:17] step 779250/1000000 | loss=2.1705 vocab=0.2002 token=1.0714 | lr=9.15e-05 | 6.12 steps/s | VRAM=28.67 Go [2026-07-18 11:38:25] step 779300/1000000 | loss=2.1903 vocab=1.2130 token=1.5486 | lr=9.14e-05 | 6.15 steps/s | VRAM=27.97 Go [2026-07-18 11:38:33] step 779350/1000000 | loss=2.0810 vocab=0.5536 token=1.3945 | lr=9.14e-05 | 6.13 steps/s | VRAM=25.99 Go [2026-07-18 11:38:41] step 779400/1000000 | loss=2.2471 vocab=1.3300 token=1.2293 | lr=9.14e-05 | 6.15 steps/s | VRAM=29.12 Go [2026-07-18 11:38:49] step 779450/1000000 | loss=2.1407 vocab=0.6594 token=1.1548 | lr=9.14e-05 | 6.11 steps/s | VRAM=28.02 Go [2026-07-18 11:38:57] step 779500/1000000 | loss=2.1316 vocab=1.4530 token=1.2651 | lr=9.14e-05 | 6.15 steps/s | VRAM=31.47 Go [2026-07-18 11:39:05] step 779550/1000000 | loss=2.1785 vocab=0.7210 token=1.1105 | lr=9.14e-05 | 6.13 steps/s | VRAM=31.83 Go [2026-07-18 11:39:14] step 779600/1000000 | loss=2.0999 vocab=1.0308 token=1.1228 | lr=9.13e-05 | 6.14 steps/s | VRAM=32.49 Go [2026-07-18 11:39:22] step 779650/1000000 | loss=2.1429 vocab=0.5779 token=0.8152 | lr=9.13e-05 | 6.09 steps/s | VRAM=30.41 Go [2026-07-18 11:39:30] step 779700/1000000 | loss=2.1381 vocab=0.4814 token=1.1773 | lr=9.13e-05 | 6.13 steps/s | VRAM=31.33 Go [2026-07-18 11:39:38] step 779750/1000000 | loss=2.2254 vocab=0.3092 token=1.0771 | lr=9.13e-05 | 6.13 steps/s | VRAM=29.69 Go [2026-07-18 11:39:46] step 779800/1000000 | loss=2.1728 vocab=0.7587 token=1.4998 | lr=9.13e-05 | 6.10 steps/s | VRAM=29.73 Go [2026-07-18 11:39:54] step 779850/1000000 | loss=2.1723 vocab=0.8607 token=1.1652 | lr=9.13e-05 | 6.11 steps/s | VRAM=25.92 Go [2026-07-18 11:40:03] step 779900/1000000 | loss=2.1430 vocab=1.0368 token=1.0339 | lr=9.12e-05 | 6.12 steps/s | VRAM=29.36 Go [2026-07-18 11:40:11] step 779950/1000000 | loss=2.2182 vocab=1.6332 token=1.6078 | lr=9.12e-05 | 6.17 steps/s | VRAM=32.58 Go [2026-07-18 11:40:19] step 780000/1000000 | loss=2.1101 vocab=1.0450 token=1.4141 | lr=9.12e-05 | 6.13 steps/s | VRAM=24.65 Go [2026-07-18 11:40:27] step 780050/1000000 | loss=2.1681 vocab=0.4996 token=1.3749 | lr=9.12e-05 | 6.16 steps/s | VRAM=28.31 Go [2026-07-18 11:40:35] step 780100/1000000 | loss=2.1315 vocab=1.0010 token=1.3915 | lr=9.12e-05 | 6.10 steps/s | VRAM=26.80 Go [2026-07-18 11:40:43] step 780150/1000000 | loss=2.1452 vocab=0.8542 token=1.1078 | lr=9.12e-05 | 6.14 steps/s | VRAM=26.21 Go [2026-07-18 11:40:52] step 780200/1000000 | loss=2.0734 vocab=0.8326 token=1.4286 | lr=9.11e-05 | 6.08 steps/s | VRAM=29.35 Go [2026-07-18 11:41:00] step 780250/1000000 | loss=2.1201 vocab=0.8527 token=1.5521 | lr=9.11e-05 | 6.16 steps/s | VRAM=28.41 Go [2026-07-18 11:41:08] step 780300/1000000 | loss=2.1275 vocab=1.4612 token=1.5279 | lr=9.11e-05 | 6.16 steps/s | VRAM=24.71 Go [2026-07-18 11:41:16] step 780350/1000000 | loss=2.2391 vocab=0.9228 token=1.4402 | lr=9.11e-05 | 6.14 steps/s | VRAM=26.02 Go [2026-07-18 11:41:24] step 780400/1000000 | loss=2.1583 vocab=1.1828 token=1.3581 | lr=9.11e-05 | 6.12 steps/s | VRAM=26.14 Go [2026-07-18 11:41:32] step 780450/1000000 | loss=2.3145 vocab=0.8653 token=1.5861 | lr=9.11e-05 | 6.15 steps/s | VRAM=28.33 Go [2026-07-18 11:41:40] step 780500/1000000 | loss=2.2396 vocab=0.6550 token=1.0796 | lr=9.10e-05 | 6.15 steps/s | VRAM=28.23 Go [2026-07-18 11:41:49] step 780550/1000000 | loss=2.1965 vocab=0.7154 token=1.0960 | lr=9.10e-05 | 6.14 steps/s | VRAM=32.29 Go [2026-07-18 11:41:57] step 780600/1000000 | loss=2.0865 vocab=0.5961 token=0.9717 | lr=9.10e-05 | 6.17 steps/s | VRAM=24.89 Go [2026-07-18 11:42:05] step 780650/1000000 | loss=2.2704 vocab=0.7797 token=1.5360 | lr=9.10e-05 | 6.15 steps/s | VRAM=31.62 Go [2026-07-18 11:42:13] step 780700/1000000 | loss=2.1251 vocab=0.9452 token=1.4249 | lr=9.10e-05 | 6.17 steps/s | VRAM=32.46 Go [2026-07-18 11:42:21] step 780750/1000000 | loss=2.1014 vocab=0.3824 token=0.9408 | lr=9.10e-05 | 6.16 steps/s | VRAM=28.55 Go [2026-07-18 11:42:29] step 780800/1000000 | loss=2.1028 vocab=0.6170 token=1.3292 | lr=9.09e-05 | 6.16 steps/s | VRAM=30.84 Go [2026-07-18 11:42:37] step 780850/1000000 | loss=2.3305 vocab=0.9832 token=1.2231 | lr=9.09e-05 | 6.18 steps/s | VRAM=32.01 Go [2026-07-18 11:42:45] step 780900/1000000 | loss=2.2137 vocab=0.8268 token=1.1496 | lr=9.09e-05 | 6.13 steps/s | VRAM=29.70 Go [2026-07-18 11:42:53] step 780950/1000000 | loss=2.1853 vocab=0.6045 token=1.6430 | lr=9.09e-05 | 6.17 steps/s | VRAM=32.23 Go [2026-07-18 11:43:02] step 781000/1000000 | loss=2.2321 vocab=0.9025 token=0.8815 | lr=9.09e-05 | 6.14 steps/s | VRAM=29.25 Go [2026-07-18 11:43:10] step 781050/1000000 | loss=2.1379 vocab=0.6792 token=0.8721 | lr=9.09e-05 | 6.17 steps/s | VRAM=28.88 Go [2026-07-18 11:43:18] step 781100/1000000 | loss=2.2038 vocab=0.9515 token=1.2936 | lr=9.08e-05 | 6.16 steps/s | VRAM=24.60 Go [2026-07-18 11:43:26] step 781150/1000000 | loss=2.1270 vocab=0.9971 token=1.3765 | lr=9.08e-05 | 6.15 steps/s | VRAM=28.13 Go [2026-07-18 11:43:34] step 781200/1000000 | loss=2.1461 vocab=0.5902 token=1.0615 | lr=9.08e-05 | 6.15 steps/s | VRAM=29.09 Go [2026-07-18 11:43:42] step 781250/1000000 | loss=2.1954 vocab=1.0134 token=0.8593 | lr=9.08e-05 | 6.14 steps/s | VRAM=29.87 Go [2026-07-18 11:43:50] step 781300/1000000 | loss=2.1761 vocab=1.0940 token=1.2335 | lr=9.08e-05 | 6.15 steps/s | VRAM=31.54 Go [2026-07-18 11:43:59] step 781350/1000000 | loss=2.0549 vocab=0.4953 token=1.2895 | lr=9.08e-05 | 6.14 steps/s | VRAM=29.20 Go [2026-07-18 11:44:07] step 781400/1000000 | loss=2.2075 vocab=0.7951 token=1.0235 | lr=9.08e-05 | 6.12 steps/s | VRAM=29.52 Go [2026-07-18 11:44:15] step 781450/1000000 | loss=2.0469 vocab=0.6742 token=1.2298 | lr=9.07e-05 | 6.17 steps/s | VRAM=32.70 Go [2026-07-18 11:44:23] step 781500/1000000 | loss=2.2761 vocab=0.6743 token=1.0088 | lr=9.07e-05 | 6.17 steps/s | VRAM=27.98 Go [2026-07-18 11:44:31] step 781550/1000000 | loss=2.0299 vocab=1.2344 token=1.1541 | lr=9.07e-05 | 6.14 steps/s | VRAM=32.80 Go [2026-07-18 11:44:39] step 781600/1000000 | loss=2.1892 vocab=0.5326 token=1.4267 | lr=9.07e-05 | 6.15 steps/s | VRAM=31.66 Go [2026-07-18 11:44:47] step 781650/1000000 | loss=2.1416 vocab=0.6816 token=1.1512 | lr=9.07e-05 | 6.14 steps/s | VRAM=24.71 Go [2026-07-18 11:44:55] step 781700/1000000 | loss=2.1652 vocab=0.7361 token=0.8284 | lr=9.07e-05 | 6.16 steps/s | VRAM=29.01 Go [2026-07-18 11:45:04] step 781750/1000000 | loss=2.1113 vocab=0.6422 token=1.3219 | lr=9.06e-05 | 6.13 steps/s | VRAM=28.67 Go [2026-07-18 11:45:12] step 781800/1000000 | loss=2.3084 vocab=0.9707 token=0.4536 | lr=9.06e-05 | 6.15 steps/s | VRAM=29.13 Go [2026-07-18 11:45:20] step 781850/1000000 | loss=2.1028 vocab=0.6126 token=1.1325 | lr=9.06e-05 | 6.13 steps/s | VRAM=28.16 Go [2026-07-18 11:45:28] step 781900/1000000 | loss=2.2789 vocab=0.5864 token=1.1866 | lr=9.06e-05 | 6.15 steps/s | VRAM=31.99 Go [2026-07-18 11:45:36] step 781950/1000000 | loss=2.1956 vocab=0.8752 token=1.2860 | lr=9.06e-05 | 6.16 steps/s | VRAM=31.74 Go [2026-07-18 11:45:44] step 782000/1000000 | loss=2.2286 vocab=0.7612 token=1.2461 | lr=9.06e-05 | 6.15 steps/s | VRAM=32.95 Go [2026-07-18 11:45:52] step 782050/1000000 | loss=2.1517 vocab=0.7979 token=1.2519 | lr=9.05e-05 | 6.16 steps/s | VRAM=31.29 Go [2026-07-18 11:46:01] step 782100/1000000 | loss=2.1283 vocab=0.6448 token=0.9964 | lr=9.05e-05 | 6.13 steps/s | VRAM=29.41 Go [2026-07-18 11:46:09] step 782150/1000000 | loss=2.3195 vocab=0.8479 token=1.6133 | lr=9.05e-05 | 6.16 steps/s | VRAM=32.24 Go [2026-07-18 11:46:17] step 782200/1000000 | loss=2.1440 vocab=0.7308 token=1.3972 | lr=9.05e-05 | 6.16 steps/s | VRAM=28.50 Go [2026-07-18 11:46:25] step 782250/1000000 | loss=2.0772 vocab=0.9764 token=1.3989 | lr=9.05e-05 | 6.13 steps/s | VRAM=25.65 Go [2026-07-18 11:46:33] step 782300/1000000 | loss=2.2543 vocab=0.9757 token=1.7507 | lr=9.04e-05 | 6.16 steps/s | VRAM=29.52 Go [2026-07-18 11:46:41] step 782350/1000000 | loss=2.1742 vocab=0.8920 token=1.1560 | lr=9.04e-05 | 6.13 steps/s | VRAM=28.45 Go [2026-07-18 11:46:49] step 782400/1000000 | loss=2.1437 vocab=0.8351 token=1.5197 | lr=9.04e-05 | 6.11 steps/s | VRAM=26.18 Go [2026-07-18 11:46:58] step 782450/1000000 | loss=2.1793 vocab=1.1256 token=1.2481 | lr=9.04e-05 | 6.11 steps/s | VRAM=26.46 Go [2026-07-18 11:47:06] step 782500/1000000 | loss=2.2081 vocab=0.7056 token=0.9650 | lr=9.04e-05 | 6.16 steps/s | VRAM=32.27 Go [2026-07-18 11:47:14] step 782550/1000000 | loss=2.2005 vocab=0.6365 token=1.0274 | lr=9.04e-05 | 6.13 steps/s | VRAM=24.66 Go [2026-07-18 11:47:22] step 782600/1000000 | loss=2.1458 vocab=0.9317 token=1.1767 | lr=9.03e-05 | 6.13 steps/s | VRAM=28.08 Go [2026-07-18 11:47:30] step 782650/1000000 | loss=2.1874 vocab=0.9624 token=1.3327 | lr=9.03e-05 | 6.16 steps/s | VRAM=26.38 Go [2026-07-18 11:47:38] step 782700/1000000 | loss=2.1521 vocab=0.8246 token=1.1610 | lr=9.03e-05 | 6.12 steps/s | VRAM=28.17 Go [2026-07-18 11:47:46] step 782750/1000000 | loss=2.2249 vocab=0.9440 token=1.1844 | lr=9.03e-05 | 6.16 steps/s | VRAM=28.20 Go [2026-07-18 11:47:55] step 782800/1000000 | loss=2.1854 vocab=0.5324 token=1.1168 | lr=9.03e-05 | 6.15 steps/s | VRAM=28.64 Go [2026-07-18 11:48:03] step 782850/1000000 | loss=2.1215 vocab=0.7185 token=0.9249 | lr=9.03e-05 | 6.14 steps/s | VRAM=28.91 Go [2026-07-18 11:48:11] step 782900/1000000 | loss=2.1233 vocab=0.6627 token=1.4986 | lr=9.02e-05 | 6.13 steps/s | VRAM=27.27 Go [2026-07-18 11:48:19] step 782950/1000000 | loss=2.1078 vocab=0.9479 token=1.1420 | lr=9.02e-05 | 6.13 steps/s | VRAM=32.46 Go [2026-07-18 11:48:27] step 783000/1000000 | loss=2.1112 vocab=0.8796 token=1.4686 | lr=9.02e-05 | 6.10 steps/s | VRAM=30.44 Go [2026-07-18 11:48:35] step 783050/1000000 | loss=2.1601 vocab=0.4194 token=0.8744 | lr=9.02e-05 | 6.14 steps/s | VRAM=26.43 Go [2026-07-18 11:48:44] step 783100/1000000 | loss=2.1042 vocab=1.3166 token=1.7044 | lr=9.02e-05 | 6.13 steps/s | VRAM=27.90 Go [2026-07-18 11:48:52] step 783150/1000000 | loss=2.1438 vocab=0.6554 token=1.5118 | lr=9.02e-05 | 6.14 steps/s | VRAM=28.75 Go [2026-07-18 11:49:00] step 783200/1000000 | loss=2.1436 vocab=1.1413 token=1.6749 | lr=9.01e-05 | 6.14 steps/s | VRAM=29.23 Go [2026-07-18 11:49:08] step 783250/1000000 | loss=2.1695 vocab=0.7930 token=1.2419 | lr=9.01e-05 | 6.08 steps/s | VRAM=26.47 Go [2026-07-18 11:49:16] step 783300/1000000 | loss=1.9575 vocab=0.6201 token=1.0549 | lr=9.01e-05 | 6.16 steps/s | VRAM=32.81 Go [2026-07-18 11:49:24] step 783350/1000000 | loss=2.1702 vocab=0.6656 token=1.0395 | lr=9.01e-05 | 6.12 steps/s | VRAM=28.41 Go [2026-07-18 11:49:32] step 783400/1000000 | loss=2.1607 vocab=0.8056 token=1.5493 | lr=9.01e-05 | 6.14 steps/s | VRAM=31.79 Go [2026-07-18 11:49:41] step 783450/1000000 | loss=2.1528 vocab=1.2857 token=1.2693 | lr=9.01e-05 | 6.11 steps/s | VRAM=31.77 Go [2026-07-18 11:49:49] step 783500/1000000 | loss=2.0673 vocab=0.9993 token=1.4369 | lr=9.00e-05 | 6.15 steps/s | VRAM=26.10 Go [2026-07-18 11:49:57] step 783550/1000000 | loss=2.2244 vocab=0.3719 token=0.9465 | lr=9.00e-05 | 6.13 steps/s | VRAM=29.21 Go [2026-07-18 11:50:05] step 783600/1000000 | loss=2.0851 vocab=1.1434 token=1.3550 | lr=9.00e-05 | 6.17 steps/s | VRAM=25.28 Go [2026-07-18 11:50:13] step 783650/1000000 | loss=2.0672 vocab=0.3816 token=1.0582 | lr=9.00e-05 | 6.14 steps/s | VRAM=28.14 Go [2026-07-18 11:50:21] step 783700/1000000 | loss=2.2504 vocab=0.8151 token=1.2415 | lr=9.00e-05 | 6.13 steps/s | VRAM=32.47 Go [2026-07-18 11:50:29] step 783750/1000000 | loss=2.2982 vocab=1.1015 token=1.5197 | lr=9.00e-05 | 6.15 steps/s | VRAM=31.53 Go [2026-07-18 11:50:38] step 783800/1000000 | loss=2.0381 vocab=0.4833 token=1.1111 | lr=8.99e-05 | 6.09 steps/s | VRAM=26.43 Go [2026-07-18 11:50:46] step 783850/1000000 | loss=2.2166 vocab=1.4330 token=1.2471 | lr=8.99e-05 | 6.10 steps/s | VRAM=26.59 Go [2026-07-18 11:50:54] step 783900/1000000 | loss=2.2205 vocab=0.8943 token=1.3097 | lr=8.99e-05 | 6.13 steps/s | VRAM=30.41 Go [2026-07-18 11:51:02] step 783950/1000000 | loss=2.3102 vocab=0.7336 token=1.0451 | lr=8.99e-05 | 6.13 steps/s | VRAM=28.92 Go [2026-07-18 11:51:10] step 784000/1000000 | loss=2.1459 vocab=0.7772 token=0.9398 | lr=8.99e-05 | 6.13 steps/s | VRAM=28.56 Go [2026-07-18 11:51:18] step 784050/1000000 | loss=2.2033 vocab=0.7029 token=1.3046 | lr=8.99e-05 | 6.10 steps/s | VRAM=29.15 Go [2026-07-18 11:51:27] step 784100/1000000 | loss=2.2208 vocab=0.6413 token=1.3247 | lr=8.98e-05 | 6.14 steps/s | VRAM=28.34 Go [2026-07-18 11:51:35] step 784150/1000000 | loss=2.1985 vocab=0.9519 token=1.6062 | lr=8.98e-05 | 6.09 steps/s | VRAM=26.90 Go [2026-07-18 11:51:43] step 784200/1000000 | loss=2.2136 vocab=1.1983 token=1.5064 | lr=8.98e-05 | 6.16 steps/s | VRAM=32.14 Go [2026-07-18 11:51:51] step 784250/1000000 | loss=2.2277 vocab=1.0155 token=1.2362 | lr=8.98e-05 | 6.12 steps/s | VRAM=32.03 Go [2026-07-18 11:51:59] step 784300/1000000 | loss=2.1297 vocab=0.4830 token=1.2244 | lr=8.98e-05 | 6.11 steps/s | VRAM=32.47 Go [2026-07-18 11:52:07] step 784350/1000000 | loss=2.1142 vocab=0.4825 token=0.5538 | lr=8.97e-05 | 6.13 steps/s | VRAM=32.58 Go [2026-07-18 11:52:16] step 784400/1000000 | loss=2.1190 vocab=1.2778 token=1.2725 | lr=8.97e-05 | 6.13 steps/s | VRAM=29.54 Go [2026-07-18 11:52:24] step 784450/1000000 | loss=2.2340 vocab=0.9134 token=1.8333 | lr=8.97e-05 | 6.15 steps/s | VRAM=28.39 Go [2026-07-18 11:52:32] step 784500/1000000 | loss=2.2701 vocab=0.8797 token=1.3554 | lr=8.97e-05 | 6.14 steps/s | VRAM=31.37 Go [2026-07-18 11:52:40] step 784550/1000000 | loss=2.2208 vocab=1.0529 token=1.7019 | lr=8.97e-05 | 6.12 steps/s | VRAM=30.22 Go [2026-07-18 11:52:48] step 784600/1000000 | loss=2.1849 vocab=1.0366 token=1.0397 | lr=8.97e-05 | 6.16 steps/s | VRAM=24.86 Go [2026-07-18 11:52:56] step 784650/1000000 | loss=2.1620 vocab=1.0087 token=1.3692 | lr=8.96e-05 | 6.14 steps/s | VRAM=26.41 Go [2026-07-18 11:53:05] step 784700/1000000 | loss=2.1987 vocab=1.3424 token=0.9690 | lr=8.96e-05 | 6.09 steps/s | VRAM=26.82 Go [2026-07-18 11:53:13] step 784750/1000000 | loss=2.1231 vocab=0.7218 token=1.5671 | lr=8.96e-05 | 6.14 steps/s | VRAM=32.49 Go [2026-07-18 11:53:21] step 784800/1000000 | loss=2.1708 vocab=1.1137 token=1.1240 | lr=8.96e-05 | 6.14 steps/s | VRAM=24.68 Go [2026-07-18 11:53:29] step 784850/1000000 | loss=2.1029 vocab=0.7121 token=1.3748 | lr=8.96e-05 | 6.13 steps/s | VRAM=28.95 Go [2026-07-18 11:53:37] step 784900/1000000 | loss=2.1720 vocab=0.6562 token=1.2113 | lr=8.96e-05 | 6.12 steps/s | VRAM=29.28 Go [2026-07-18 11:53:45] step 784950/1000000 | loss=2.2398 vocab=0.6588 token=1.1172 | lr=8.95e-05 | 6.13 steps/s | VRAM=28.71 Go [2026-07-18 11:53:53] step 785000/1000000 | loss=2.1545 vocab=1.1711 token=1.0425 | lr=8.95e-05 | 6.14 steps/s | VRAM=26.83 Go [2026-07-18 11:54:02] step 785050/1000000 | loss=2.1230 vocab=0.6357 token=1.2149 | lr=8.95e-05 | 6.14 steps/s | VRAM=28.65 Go [2026-07-18 11:54:10] step 785100/1000000 | loss=2.1178 vocab=0.7153 token=1.3266 | lr=8.95e-05 | 6.08 steps/s | VRAM=26.84 Go [2026-07-18 11:54:18] step 785150/1000000 | loss=2.2408 vocab=1.5922 token=1.2708 | lr=8.95e-05 | 6.13 steps/s | VRAM=28.24 Go [2026-07-18 11:54:26] step 785200/1000000 | loss=2.2304 vocab=0.8073 token=1.2876 | lr=8.94e-05 | 6.14 steps/s | VRAM=28.60 Go [2026-07-18 11:54:34] step 785250/1000000 | loss=2.2312 vocab=0.5605 token=1.1229 | lr=8.94e-05 | 6.15 steps/s | VRAM=31.86 Go [2026-07-18 11:54:42] step 785300/1000000 | loss=2.0507 vocab=1.1408 token=0.9870 | lr=8.94e-05 | 6.12 steps/s | VRAM=30.31 Go [2026-07-18 11:54:51] step 785350/1000000 | loss=2.1425 vocab=0.8394 token=1.6199 | lr=8.94e-05 | 6.16 steps/s | VRAM=27.77 Go [2026-07-18 11:54:59] step 785400/1000000 | loss=2.2295 vocab=1.4571 token=1.0762 | lr=8.94e-05 | 6.13 steps/s | VRAM=28.79 Go [2026-07-18 11:55:07] step 785450/1000000 | loss=2.1975 vocab=0.9926 token=1.1423 | lr=8.94e-05 | 6.10 steps/s | VRAM=30.01 Go [2026-07-18 11:55:15] step 785500/1000000 | loss=2.1458 vocab=0.7587 token=0.7408 | lr=8.93e-05 | 6.15 steps/s | VRAM=28.82 Go [2026-07-18 11:55:23] step 785550/1000000 | loss=2.1451 vocab=1.1647 token=1.4284 | lr=8.93e-05 | 6.18 steps/s | VRAM=31.45 Go [2026-07-18 11:55:31] step 785600/1000000 | loss=2.1743 vocab=1.4751 token=1.7312 | lr=8.93e-05 | 6.15 steps/s | VRAM=32.53 Go [2026-07-18 11:55:39] step 785650/1000000 | loss=2.1750 vocab=0.9777 token=1.2833 | lr=8.93e-05 | 6.15 steps/s | VRAM=29.19 Go [2026-07-18 11:55:48] step 785700/1000000 | loss=2.1658 vocab=0.9074 token=1.3537 | lr=8.93e-05 | 6.11 steps/s | VRAM=26.04 Go [2026-07-18 11:55:56] step 785750/1000000 | loss=2.2361 vocab=0.7112 token=1.5382 | lr=8.93e-05 | 6.15 steps/s | VRAM=31.49 Go [2026-07-18 11:56:04] step 785800/1000000 | loss=2.1735 vocab=1.2209 token=1.5104 | lr=8.92e-05 | 6.12 steps/s | VRAM=28.98 Go [2026-07-18 11:56:12] step 785850/1000000 | loss=2.1656 vocab=0.8169 token=1.4238 | lr=8.92e-05 | 6.16 steps/s | VRAM=32.47 Go [2026-07-18 11:56:20] step 785900/1000000 | loss=2.2446 vocab=0.7149 token=1.4259 | lr=8.92e-05 | 6.13 steps/s | VRAM=26.42 Go [2026-07-18 11:56:28] step 785950/1000000 | loss=2.2825 vocab=0.7912 token=1.2807 | lr=8.92e-05 | 6.15 steps/s | VRAM=23.72 Go [2026-07-18 11:56:36] step 786000/1000000 | loss=2.2415 vocab=0.8435 token=1.3413 | lr=8.92e-05 | 6.16 steps/s | VRAM=29.03 Go [2026-07-18 11:56:45] step 786050/1000000 | loss=2.2487 vocab=0.9822 token=1.4798 | lr=8.91e-05 | 6.14 steps/s | VRAM=31.29 Go [2026-07-18 11:56:53] step 786100/1000000 | loss=2.0748 vocab=0.7476 token=1.1919 | lr=8.91e-05 | 6.13 steps/s | VRAM=28.05 Go [2026-07-18 11:57:01] step 786150/1000000 | loss=2.1520 vocab=1.0609 token=1.4621 | lr=8.91e-05 | 6.12 steps/s | VRAM=29.70 Go [2026-07-18 11:57:09] step 786200/1000000 | loss=2.0979 vocab=0.8914 token=1.0720 | lr=8.91e-05 | 6.11 steps/s | VRAM=28.76 Go [2026-07-18 11:57:17] step 786250/1000000 | loss=2.1532 vocab=1.1758 token=1.6352 | lr=8.91e-05 | 6.12 steps/s | VRAM=31.25 Go [2026-07-18 11:57:25] step 786300/1000000 | loss=2.1531 vocab=1.1820 token=1.2233 | lr=8.91e-05 | 6.16 steps/s | VRAM=31.78 Go [2026-07-18 11:57:33] step 786350/1000000 | loss=2.1689 vocab=0.8953 token=1.1969 | lr=8.90e-05 | 6.12 steps/s | VRAM=28.07 Go [2026-07-18 11:57:42] step 786400/1000000 | loss=2.1240 vocab=0.9047 token=1.5403 | lr=8.90e-05 | 6.11 steps/s | VRAM=28.54 Go [2026-07-18 11:57:50] step 786450/1000000 | loss=2.0918 vocab=1.1158 token=1.2699 | lr=8.90e-05 | 6.16 steps/s | VRAM=31.39 Go [2026-07-18 11:57:58] step 786500/1000000 | loss=1.9985 vocab=0.8618 token=1.2876 | lr=8.90e-05 | 6.15 steps/s | VRAM=24.58 Go [2026-07-18 11:58:06] step 786550/1000000 | loss=2.1971 vocab=0.5029 token=1.1414 | lr=8.90e-05 | 6.12 steps/s | VRAM=27.93 Go [2026-07-18 11:58:14] step 786600/1000000 | loss=2.1666 vocab=1.1182 token=1.5158 | lr=8.90e-05 | 6.13 steps/s | VRAM=32.64 Go [2026-07-18 11:58:22] step 786650/1000000 | loss=2.2001 vocab=0.5643 token=1.0976 | lr=8.89e-05 | 6.16 steps/s | VRAM=31.51 Go [2026-07-18 11:58:31] step 786700/1000000 | loss=2.0933 vocab=1.0697 token=1.4553 | lr=8.89e-05 | 6.12 steps/s | VRAM=29.50 Go [2026-07-18 11:58:39] step 786750/1000000 | loss=2.1184 vocab=0.8889 token=1.3600 | lr=8.89e-05 | 6.13 steps/s | VRAM=25.78 Go [2026-07-18 11:58:47] step 786800/1000000 | loss=2.1616 vocab=0.4772 token=1.0195 | lr=8.89e-05 | 6.15 steps/s | VRAM=26.25 Go [2026-07-18 11:58:55] step 786850/1000000 | loss=2.0871 vocab=1.0524 token=1.1881 | lr=8.89e-05 | 6.14 steps/s | VRAM=29.00 Go [2026-07-18 11:59:03] step 786900/1000000 | loss=2.1513 vocab=0.6802 token=1.1952 | lr=8.88e-05 | 6.12 steps/s | VRAM=29.19 Go [2026-07-18 11:59:11] step 786950/1000000 | loss=2.2200 vocab=0.9855 token=1.5055 | lr=8.88e-05 | 6.15 steps/s | VRAM=32.78 Go [2026-07-18 11:59:19] step 787000/1000000 | loss=2.1517 vocab=0.6245 token=1.3276 | lr=8.88e-05 | 6.13 steps/s | VRAM=29.17 Go [2026-07-18 11:59:28] step 787050/1000000 | loss=2.0092 vocab=0.8487 token=1.3160 | lr=8.88e-05 | 6.13 steps/s | VRAM=32.43 Go [2026-07-18 11:59:36] step 787100/1000000 | loss=2.1808 vocab=1.1572 token=1.4610 | lr=8.88e-05 | 6.13 steps/s | VRAM=31.44 Go [2026-07-18 11:59:44] step 787150/1000000 | loss=2.2304 vocab=0.5943 token=1.2858 | lr=8.88e-05 | 6.13 steps/s | VRAM=25.96 Go [2026-07-18 11:59:52] step 787200/1000000 | loss=2.1094 vocab=1.3565 token=1.1406 | lr=8.87e-05 | 6.13 steps/s | VRAM=28.45 Go [2026-07-18 12:00:00] step 787250/1000000 | loss=2.1056 vocab=0.6664 token=1.2630 | lr=8.87e-05 | 6.16 steps/s | VRAM=24.30 Go [2026-07-18 12:00:08] step 787300/1000000 | loss=2.1532 vocab=1.9911 token=1.6182 | lr=8.87e-05 | 6.14 steps/s | VRAM=26.38 Go [2026-07-18 12:00:16] step 787350/1000000 | loss=2.1303 vocab=0.9314 token=1.2933 | lr=8.87e-05 | 6.13 steps/s | VRAM=28.53 Go [2026-07-18 12:00:25] step 787400/1000000 | loss=2.1685 vocab=0.7994 token=1.2935 | lr=8.87e-05 | 6.15 steps/s | VRAM=25.06 Go [2026-07-18 12:00:33] step 787450/1000000 | loss=2.1415 vocab=0.9011 token=1.2671 | lr=8.86e-05 | 6.14 steps/s | VRAM=32.76 Go [2026-07-18 12:00:41] step 787500/1000000 | loss=2.1171 vocab=0.6584 token=1.2861 | lr=8.86e-05 | 6.14 steps/s | VRAM=32.48 Go [2026-07-18 12:00:49] step 787550/1000000 | loss=2.1896 vocab=0.7465 token=1.4921 | lr=8.86e-05 | 6.14 steps/s | VRAM=28.67 Go [2026-07-18 12:00:57] step 787600/1000000 | loss=1.9806 vocab=0.8623 token=1.2230 | lr=8.86e-05 | 6.14 steps/s | VRAM=29.38 Go [2026-07-18 12:01:05] step 787650/1000000 | loss=2.1112 vocab=1.3561 token=1.6244 | lr=8.86e-05 | 6.12 steps/s | VRAM=28.82 Go [2026-07-18 12:01:14] step 787700/1000000 | loss=2.2030 vocab=0.8318 token=1.3218 | lr=8.86e-05 | 6.13 steps/s | VRAM=29.10 Go [2026-07-18 12:01:22] step 787750/1000000 | loss=2.0930 vocab=1.1205 token=1.5888 | lr=8.85e-05 | 6.15 steps/s | VRAM=26.57 Go [2026-07-18 12:01:30] step 787800/1000000 | loss=2.1696 vocab=0.7366 token=1.3684 | lr=8.85e-05 | 6.15 steps/s | VRAM=31.07 Go [2026-07-18 12:01:38] step 787850/1000000 | loss=2.2102 vocab=0.6769 token=1.4832 | lr=8.85e-05 | 6.13 steps/s | VRAM=28.57 Go [2026-07-18 12:01:46] step 787900/1000000 | loss=2.2545 vocab=1.0359 token=1.2413 | lr=8.85e-05 | 6.12 steps/s | VRAM=28.89 Go [2026-07-18 12:01:54] step 787950/1000000 | loss=2.0933 vocab=1.0467 token=1.3658 | lr=8.85e-05 | 6.15 steps/s | VRAM=27.75 Go [2026-07-18 12:02:02] step 788000/1000000 | loss=2.1326 vocab=0.7735 token=1.3010 | lr=8.84e-05 | 6.16 steps/s | VRAM=25.30 Go [2026-07-18 12:02:10] step 788050/1000000 | loss=2.2212 vocab=1.1792 token=1.1769 | lr=8.84e-05 | 6.14 steps/s | VRAM=32.72 Go [2026-07-18 12:02:19] step 788100/1000000 | loss=2.1184 vocab=1.2343 token=1.7063 | lr=8.84e-05 | 6.13 steps/s | VRAM=28.74 Go [2026-07-18 12:02:27] step 788150/1000000 | loss=2.1551 vocab=1.0805 token=1.3971 | lr=8.84e-05 | 6.11 steps/s | VRAM=28.85 Go [2026-07-18 12:02:35] step 788200/1000000 | loss=2.1856 vocab=1.0298 token=1.1609 | lr=8.84e-05 | 6.15 steps/s | VRAM=31.38 Go [2026-07-18 12:02:43] step 788250/1000000 | loss=2.2098 vocab=0.8419 token=1.2255 | lr=8.84e-05 | 6.16 steps/s | VRAM=31.31 Go [2026-07-18 12:02:51] step 788300/1000000 | loss=2.1909 vocab=1.7103 token=2.0032 | lr=8.83e-05 | 6.15 steps/s | VRAM=32.31 Go [2026-07-18 12:02:59] step 788350/1000000 | loss=2.1804 vocab=0.9079 token=1.2059 | lr=8.83e-05 | 6.17 steps/s | VRAM=32.29 Go [2026-07-18 12:03:07] step 788400/1000000 | loss=2.1961 vocab=0.6405 token=1.4398 | lr=8.83e-05 | 6.13 steps/s | VRAM=28.76 Go [2026-07-18 12:03:16] step 788450/1000000 | loss=2.2339 vocab=1.3371 token=1.7284 | lr=8.83e-05 | 6.16 steps/s | VRAM=28.12 Go [2026-07-18 12:03:24] step 788500/1000000 | loss=2.2483 vocab=0.5984 token=1.5241 | lr=8.83e-05 | 6.14 steps/s | VRAM=32.36 Go [2026-07-18 12:03:32] step 788550/1000000 | loss=2.1043 vocab=1.1398 token=1.6577 | lr=8.82e-05 | 6.14 steps/s | VRAM=29.79 Go [2026-07-18 12:03:40] step 788600/1000000 | loss=2.1210 vocab=0.8660 token=1.1608 | lr=8.82e-05 | 6.14 steps/s | VRAM=28.58 Go [2026-07-18 12:03:48] step 788650/1000000 | loss=2.2279 vocab=0.8730 token=1.1633 | lr=8.82e-05 | 6.11 steps/s | VRAM=26.51 Go [2026-07-18 12:03:56] step 788700/1000000 | loss=2.1251 vocab=0.8682 token=1.3826 | lr=8.82e-05 | 6.15 steps/s | VRAM=32.62 Go [2026-07-18 12:04:04] step 788750/1000000 | loss=2.1412 vocab=1.7841 token=1.4537 | lr=8.82e-05 | 6.16 steps/s | VRAM=28.89 Go [2026-07-18 12:04:13] step 788800/1000000 | loss=2.2087 vocab=0.8606 token=1.1289 | lr=8.81e-05 | 6.11 steps/s | VRAM=32.26 Go [2026-07-18 12:04:21] step 788850/1000000 | loss=2.0871 vocab=0.8143 token=1.0596 | lr=8.81e-05 | 6.13 steps/s | VRAM=30.35 Go [2026-07-18 12:04:29] step 788900/1000000 | loss=2.1556 vocab=0.9570 token=1.7063 | lr=8.81e-05 | 6.12 steps/s | VRAM=32.92 Go [2026-07-18 12:04:37] step 788950/1000000 | loss=2.2270 vocab=0.8855 token=1.5284 | lr=8.81e-05 | 6.17 steps/s | VRAM=32.31 Go [2026-07-18 12:04:45] step 789000/1000000 | loss=2.2005 vocab=1.1626 token=1.6174 | lr=8.81e-05 | 6.14 steps/s | VRAM=31.59 Go [2026-07-18 12:04:53] step 789050/1000000 | loss=2.1790 vocab=1.2128 token=1.3954 | lr=8.81e-05 | 6.16 steps/s | VRAM=32.58 Go [2026-07-18 12:05:02] step 789100/1000000 | loss=2.2484 vocab=0.8557 token=1.5056 | lr=8.80e-05 | 6.11 steps/s | VRAM=28.55 Go [2026-07-18 12:05:10] step 789150/1000000 | loss=2.1055 vocab=0.7140 token=1.4369 | lr=8.80e-05 | 6.14 steps/s | VRAM=28.81 Go [2026-07-18 12:05:18] step 789200/1000000 | loss=2.1721 vocab=1.3378 token=1.8908 | lr=8.80e-05 | 6.12 steps/s | VRAM=32.00 Go [2026-07-18 12:05:26] step 789250/1000000 | loss=2.1161 vocab=0.5485 token=1.1493 | lr=8.80e-05 | 6.14 steps/s | VRAM=32.18 Go [2026-07-18 12:05:34] step 789300/1000000 | loss=2.1508 vocab=0.5098 token=0.7949 | lr=8.80e-05 | 6.10 steps/s | VRAM=29.46 Go [2026-07-18 12:05:42] step 789350/1000000 | loss=2.1689 vocab=0.7898 token=1.2688 | lr=8.79e-05 | 6.10 steps/s | VRAM=25.69 Go [2026-07-18 12:05:51] step 789400/1000000 | loss=2.1633 vocab=0.6648 token=1.3867 | lr=8.79e-05 | 6.10 steps/s | VRAM=26.37 Go [2026-07-18 12:05:59] step 789450/1000000 | loss=2.1274 vocab=0.4001 token=0.8759 | lr=8.79e-05 | 6.15 steps/s | VRAM=31.82 Go [2026-07-18 12:06:07] step 789500/1000000 | loss=2.1130 vocab=0.7404 token=1.5420 | lr=8.79e-05 | 6.15 steps/s | VRAM=32.54 Go [2026-07-18 12:06:15] step 789550/1000000 | loss=2.2951 vocab=1.1579 token=1.0136 | lr=8.79e-05 | 6.13 steps/s | VRAM=26.22 Go [2026-07-18 12:06:23] step 789600/1000000 | loss=2.1934 vocab=0.8245 token=1.5222 | lr=8.79e-05 | 6.15 steps/s | VRAM=25.48 Go [2026-07-18 12:06:31] step 789650/1000000 | loss=2.0873 vocab=0.7706 token=0.9677 | lr=8.78e-05 | 6.14 steps/s | VRAM=29.18 Go [2026-07-18 12:06:39] step 789700/1000000 | loss=2.1364 vocab=1.2321 token=1.1298 | lr=8.78e-05 | 6.16 steps/s | VRAM=28.25 Go [2026-07-18 12:06:48] step 789750/1000000 | loss=2.1347 vocab=0.7217 token=1.8449 | lr=8.78e-05 | 6.13 steps/s | VRAM=32.32 Go [2026-07-18 12:06:56] step 789800/1000000 | loss=2.1937 vocab=0.4214 token=1.1196 | lr=8.78e-05 | 6.15 steps/s | VRAM=31.60 Go [2026-07-18 12:07:04] step 789850/1000000 | loss=2.1850 vocab=2.2289 token=1.5731 | lr=8.78e-05 | 6.14 steps/s | VRAM=32.69 Go [2026-07-18 12:07:12] step 789900/1000000 | loss=2.2356 vocab=1.3527 token=1.4949 | lr=8.77e-05 | 6.10 steps/s | VRAM=28.65 Go [2026-07-18 12:07:20] step 789950/1000000 | loss=2.2010 vocab=0.5195 token=0.7972 | lr=8.77e-05 | 6.10 steps/s | VRAM=29.73 Go [2026-07-18 12:07:28] step 790000/1000000 | loss=2.1561 vocab=0.7019 token=1.0167 | lr=8.77e-05 | 6.14 steps/s | VRAM=31.61 Go [2026-07-18 12:07:36] step 790050/1000000 | loss=2.1839 vocab=1.5493 token=1.3962 | lr=8.77e-05 | 6.15 steps/s | VRAM=25.82 Go [2026-07-18 12:07:45] step 790100/1000000 | loss=2.1279 vocab=0.7994 token=1.4425 | lr=8.77e-05 | 6.14 steps/s | VRAM=26.19 Go [2026-07-18 12:07:53] step 790150/1000000 | loss=2.2143 vocab=0.5147 token=1.0210 | lr=8.76e-05 | 6.15 steps/s | VRAM=32.00 Go [2026-07-18 12:08:01] step 790200/1000000 | loss=2.1771 vocab=0.9921 token=1.3227 | lr=8.76e-05 | 6.11 steps/s | VRAM=29.24 Go [2026-07-18 12:08:09] step 790250/1000000 | loss=2.2009 vocab=0.6375 token=1.2349 | lr=8.76e-05 | 6.12 steps/s | VRAM=32.70 Go [2026-07-18 12:08:17] step 790300/1000000 | loss=2.1572 vocab=0.5813 token=1.2072 | lr=8.76e-05 | 6.14 steps/s | VRAM=28.14 Go [2026-07-18 12:08:25] step 790350/1000000 | loss=2.1178 vocab=0.8049 token=0.9021 | lr=8.76e-05 | 6.12 steps/s | VRAM=28.50 Go [2026-07-18 12:08:34] step 790400/1000000 | loss=2.2097 vocab=1.1379 token=1.6056 | lr=8.75e-05 | 6.12 steps/s | VRAM=25.20 Go [2026-07-18 12:08:42] step 790450/1000000 | loss=2.1906 vocab=1.0940 token=1.7150 | lr=8.75e-05 | 6.14 steps/s | VRAM=31.63 Go [2026-07-18 12:08:50] step 790500/1000000 | loss=2.1893 vocab=1.2130 token=1.3427 | lr=8.75e-05 | 6.12 steps/s | VRAM=26.00 Go [2026-07-18 12:08:58] step 790550/1000000 | loss=2.1265 vocab=0.5444 token=1.0319 | lr=8.75e-05 | 6.16 steps/s | VRAM=31.72 Go [2026-07-18 12:09:06] step 790600/1000000 | loss=2.1087 vocab=0.7766 token=1.1670 | lr=8.75e-05 | 6.13 steps/s | VRAM=29.66 Go [2026-07-18 12:09:14] step 790650/1000000 | loss=2.1949 vocab=0.6527 token=1.4010 | lr=8.75e-05 | 6.14 steps/s | VRAM=25.86 Go [2026-07-18 12:09:22] step 790700/1000000 | loss=2.1420 vocab=0.7763 token=1.3800 | lr=8.74e-05 | 6.17 steps/s | VRAM=28.85 Go [2026-07-18 12:09:31] step 790750/1000000 | loss=2.1040 vocab=0.8807 token=1.4038 | lr=8.74e-05 | 6.13 steps/s | VRAM=28.88 Go [2026-07-18 12:09:39] step 790800/1000000 | loss=2.1815 vocab=0.6557 token=1.2213 | lr=8.74e-05 | 6.14 steps/s | VRAM=32.12 Go [2026-07-18 12:09:47] step 790850/1000000 | loss=2.1643 vocab=0.8665 token=1.3011 | lr=8.74e-05 | 6.13 steps/s | VRAM=26.15 Go [2026-07-18 12:09:55] step 790900/1000000 | loss=2.2160 vocab=1.1820 token=1.6521 | lr=8.74e-05 | 6.13 steps/s | VRAM=28.85 Go [2026-07-18 12:10:03] step 790950/1000000 | loss=2.2868 vocab=0.6990 token=1.2596 | lr=8.73e-05 | 6.15 steps/s | VRAM=32.81 Go [2026-07-18 12:10:11] step 791000/1000000 | loss=2.1602 vocab=0.7654 token=1.1168 | lr=8.73e-05 | 6.14 steps/s | VRAM=31.52 Go [2026-07-18 12:10:19] step 791050/1000000 | loss=2.1370 vocab=0.6771 token=1.3348 | lr=8.73e-05 | 6.15 steps/s | VRAM=25.42 Go [2026-07-18 12:10:28] step 791100/1000000 | loss=2.2058 vocab=1.1936 token=1.2133 | lr=8.73e-05 | 6.10 steps/s | VRAM=30.22 Go [2026-07-18 12:10:36] step 791150/1000000 | loss=2.2273 vocab=1.1670 token=1.6540 | lr=8.73e-05 | 6.14 steps/s | VRAM=32.13 Go [2026-07-18 12:10:44] step 791200/1000000 | loss=2.0721 vocab=1.2782 token=1.0152 | lr=8.72e-05 | 6.13 steps/s | VRAM=32.74 Go [2026-07-18 12:10:52] step 791250/1000000 | loss=2.1949 vocab=0.6624 token=1.3927 | lr=8.72e-05 | 6.16 steps/s | VRAM=32.08 Go [2026-07-18 12:11:00] step 791300/1000000 | loss=2.1100 vocab=0.7884 token=1.1583 | lr=8.72e-05 | 6.14 steps/s | VRAM=24.78 Go [2026-07-18 12:11:08] step 791350/1000000 | loss=2.1362 vocab=0.7922 token=1.2528 | lr=8.72e-05 | 6.13 steps/s | VRAM=28.29 Go [2026-07-18 12:11:16] step 791400/1000000 | loss=2.1961 vocab=1.1044 token=1.3046 | lr=8.72e-05 | 6.13 steps/s | VRAM=28.12 Go [2026-07-18 12:11:25] step 791450/1000000 | loss=2.2055 vocab=0.4873 token=1.1345 | lr=8.71e-05 | 6.11 steps/s | VRAM=27.67 Go [2026-07-18 12:11:33] step 791500/1000000 | loss=2.2044 vocab=0.8809 token=1.4487 | lr=8.71e-05 | 6.12 steps/s | VRAM=28.80 Go [2026-07-18 12:11:41] step 791550/1000000 | loss=2.2045 vocab=2.0957 token=1.0019 | lr=8.71e-05 | 6.15 steps/s | VRAM=24.39 Go [2026-07-18 12:11:49] step 791600/1000000 | loss=2.1651 vocab=1.2349 token=1.5773 | lr=8.71e-05 | 6.15 steps/s | VRAM=25.83 Go [2026-07-18 12:11:57] step 791650/1000000 | loss=2.2189 vocab=0.7300 token=1.2760 | lr=8.71e-05 | 6.15 steps/s | VRAM=28.29 Go [2026-07-18 12:12:05] step 791700/1000000 | loss=2.1179 vocab=0.7015 token=1.1626 | lr=8.71e-05 | 6.13 steps/s | VRAM=28.94 Go [2026-07-18 12:12:14] step 791750/1000000 | loss=2.1338 vocab=0.9147 token=1.2587 | lr=8.70e-05 | 6.11 steps/s | VRAM=29.13 Go [2026-07-18 12:12:22] step 791800/1000000 | loss=2.1714 vocab=0.7785 token=1.0265 | lr=8.70e-05 | 6.08 steps/s | VRAM=25.83 Go [2026-07-18 12:12:30] step 791850/1000000 | loss=2.2049 vocab=0.9477 token=1.1399 | lr=8.70e-05 | 6.12 steps/s | VRAM=26.06 Go [2026-07-18 12:12:38] step 791900/1000000 | loss=2.1967 vocab=0.5406 token=1.0216 | lr=8.70e-05 | 6.11 steps/s | VRAM=26.24 Go [2026-07-18 12:12:46] step 791950/1000000 | loss=2.1735 vocab=1.1077 token=1.4978 | lr=8.70e-05 | 6.17 steps/s | VRAM=28.41 Go [2026-07-18 12:12:54] step 792000/1000000 | loss=2.1857 vocab=0.7725 token=1.1499 | lr=8.69e-05 | 6.11 steps/s | VRAM=33.01 Go [2026-07-18 12:13:03] step 792050/1000000 | loss=2.2096 vocab=1.0725 token=1.3161 | lr=8.69e-05 | 6.12 steps/s | VRAM=26.31 Go [2026-07-18 12:13:11] step 792100/1000000 | loss=2.1739 vocab=0.9281 token=1.1433 | lr=8.69e-05 | 6.10 steps/s | VRAM=28.68 Go [2026-07-18 12:13:19] step 792150/1000000 | loss=2.2208 vocab=0.9596 token=1.1311 | lr=8.69e-05 | 6.12 steps/s | VRAM=31.06 Go [2026-07-18 12:13:27] step 792200/1000000 | loss=2.1543 vocab=0.9630 token=1.0396 | lr=8.69e-05 | 6.13 steps/s | VRAM=30.93 Go [2026-07-18 12:13:35] step 792250/1000000 | loss=2.1037 vocab=0.8616 token=1.2956 | lr=8.68e-05 | 6.12 steps/s | VRAM=28.29 Go [2026-07-18 12:13:43] step 792300/1000000 | loss=2.1480 vocab=0.9501 token=1.2941 | lr=8.68e-05 | 6.13 steps/s | VRAM=25.01 Go [2026-07-18 12:13:52] step 792350/1000000 | loss=2.0824 vocab=1.2138 token=1.2406 | lr=8.68e-05 | 6.11 steps/s | VRAM=28.20 Go [2026-07-18 12:14:00] step 792400/1000000 | loss=2.0843 vocab=1.0441 token=1.1866 | lr=8.68e-05 | 6.15 steps/s | VRAM=32.42 Go [2026-07-18 12:14:08] step 792450/1000000 | loss=2.0734 vocab=0.9426 token=1.4143 | lr=8.68e-05 | 6.13 steps/s | VRAM=32.18 Go [2026-07-18 12:14:16] step 792500/1000000 | loss=2.1367 vocab=0.8164 token=1.2561 | lr=8.67e-05 | 6.13 steps/s | VRAM=25.01 Go [2026-07-18 12:14:24] step 792550/1000000 | loss=2.1753 vocab=0.5586 token=1.1912 | lr=8.67e-05 | 6.11 steps/s | VRAM=26.07 Go [2026-07-18 12:14:32] step 792600/1000000 | loss=2.1488 vocab=0.8934 token=1.4307 | lr=8.67e-05 | 6.13 steps/s | VRAM=29.94 Go [2026-07-18 12:14:41] step 792650/1000000 | loss=2.0898 vocab=0.7800 token=0.8985 | lr=8.67e-05 | 6.15 steps/s | VRAM=31.75 Go [2026-07-18 12:14:49] step 792700/1000000 | loss=2.1570 vocab=0.9579 token=1.1731 | lr=8.67e-05 | 6.13 steps/s | VRAM=29.33 Go [2026-07-18 12:14:57] step 792750/1000000 | loss=2.2008 vocab=0.9672 token=1.2365 | lr=8.66e-05 | 6.14 steps/s | VRAM=28.31 Go [2026-07-18 12:15:05] step 792800/1000000 | loss=2.0614 vocab=0.6431 token=1.3325 | lr=8.66e-05 | 6.10 steps/s | VRAM=28.44 Go [2026-07-18 12:15:13] step 792850/1000000 | loss=2.1972 vocab=0.9409 token=1.6100 | lr=8.66e-05 | 6.14 steps/s | VRAM=28.58 Go [2026-07-18 12:15:21] step 792900/1000000 | loss=2.1985 vocab=0.5671 token=1.0596 | lr=8.66e-05 | 6.14 steps/s | VRAM=24.31 Go [2026-07-18 12:15:30] step 792950/1000000 | loss=2.0912 vocab=1.3058 token=1.2201 | lr=8.66e-05 | 6.09 steps/s | VRAM=28.82 Go [2026-07-18 12:15:38] step 793000/1000000 | loss=2.2412 vocab=0.7006 token=1.2364 | lr=8.65e-05 | 6.15 steps/s | VRAM=32.58 Go [2026-07-18 12:15:46] step 793050/1000000 | loss=2.2071 vocab=0.8118 token=1.5259 | lr=8.65e-05 | 6.12 steps/s | VRAM=32.14 Go [2026-07-18 12:15:54] step 793100/1000000 | loss=2.1307 vocab=0.6482 token=1.1688 | lr=8.65e-05 | 6.08 steps/s | VRAM=26.74 Go [2026-07-18 12:16:02] step 793150/1000000 | loss=2.1077 vocab=1.0893 token=1.2826 | lr=8.65e-05 | 6.14 steps/s | VRAM=28.64 Go [2026-07-18 12:16:10] step 793200/1000000 | loss=2.2405 vocab=1.0719 token=1.5319 | lr=8.65e-05 | 6.14 steps/s | VRAM=32.55 Go [2026-07-18 12:16:19] step 793250/1000000 | loss=2.1246 vocab=1.2939 token=1.0683 | lr=8.65e-05 | 6.10 steps/s | VRAM=28.15 Go [2026-07-18 12:16:27] step 793300/1000000 | loss=2.1768 vocab=0.9391 token=1.7218 | lr=8.64e-05 | 6.11 steps/s | VRAM=26.19 Go [2026-07-18 12:16:35] step 793350/1000000 | loss=2.1942 vocab=1.0626 token=1.5417 | lr=8.64e-05 | 6.11 steps/s | VRAM=28.92 Go [2026-07-18 12:16:43] step 793400/1000000 | loss=2.1565 vocab=1.5025 token=1.5669 | lr=8.64e-05 | 6.12 steps/s | VRAM=28.00 Go [2026-07-18 12:16:51] step 793450/1000000 | loss=2.1610 vocab=0.8753 token=0.8450 | lr=8.64e-05 | 6.12 steps/s | VRAM=32.97 Go [2026-07-18 12:16:59] step 793500/1000000 | loss=2.1374 vocab=0.9802 token=1.5164 | lr=8.64e-05 | 6.12 steps/s | VRAM=26.57 Go [2026-07-18 12:17:08] step 793550/1000000 | loss=2.1412 vocab=1.0604 token=1.3178 | lr=8.63e-05 | 6.13 steps/s | VRAM=32.17 Go [2026-07-18 12:17:16] step 793600/1000000 | loss=2.2053 vocab=1.0774 token=1.4615 | lr=8.63e-05 | 6.14 steps/s | VRAM=28.44 Go [2026-07-18 12:17:24] step 793650/1000000 | loss=2.0493 vocab=0.8437 token=1.2451 | lr=8.63e-05 | 6.13 steps/s | VRAM=26.99 Go [2026-07-18 12:17:32] step 793700/1000000 | loss=2.1737 vocab=0.8522 token=1.4198 | lr=8.63e-05 | 6.16 steps/s | VRAM=31.24 Go [2026-07-18 12:17:40] step 793750/1000000 | loss=2.1656 vocab=0.5093 token=1.2158 | lr=8.63e-05 | 6.13 steps/s | VRAM=31.85 Go [2026-07-18 12:17:48] step 793800/1000000 | loss=2.1988 vocab=0.9149 token=1.1167 | lr=8.62e-05 | 6.13 steps/s | VRAM=28.25 Go [2026-07-18 12:17:56] step 793850/1000000 | loss=2.1677 vocab=0.4640 token=1.4451 | lr=8.62e-05 | 6.17 steps/s | VRAM=31.47 Go [2026-07-18 12:18:05] step 793900/1000000 | loss=2.1685 vocab=0.5834 token=0.6852 | lr=8.62e-05 | 6.14 steps/s | VRAM=29.11 Go [2026-07-18 12:18:13] step 793950/1000000 | loss=2.1547 vocab=1.4214 token=1.5802 | lr=8.62e-05 | 6.19 steps/s | VRAM=25.20 Go [2026-07-18 12:18:21] step 794000/1000000 | loss=2.0653 vocab=1.0935 token=1.1648 | lr=8.62e-05 | 6.18 steps/s | VRAM=32.22 Go [2026-07-18 12:18:29] step 794050/1000000 | loss=2.2125 vocab=1.1989 token=1.6088 | lr=8.61e-05 | 6.14 steps/s | VRAM=32.45 Go [2026-07-18 12:18:37] step 794100/1000000 | loss=2.2319 vocab=1.1266 token=1.5627 | lr=8.61e-05 | 6.11 steps/s | VRAM=28.24 Go [2026-07-18 12:18:45] step 794150/1000000 | loss=2.0529 vocab=0.6679 token=1.0745 | lr=8.61e-05 | 6.15 steps/s | VRAM=25.71 Go [2026-07-18 12:18:53] step 794200/1000000 | loss=2.1167 vocab=0.5279 token=0.8249 | lr=8.61e-05 | 6.12 steps/s | VRAM=25.26 Go [2026-07-18 12:19:02] step 794250/1000000 | loss=2.1316 vocab=0.9482 token=1.4160 | lr=8.61e-05 | 6.08 steps/s | VRAM=26.51 Go [2026-07-18 12:19:10] step 794300/1000000 | loss=2.0912 vocab=0.7336 token=1.3013 | lr=8.60e-05 | 6.14 steps/s | VRAM=28.98 Go [2026-07-18 12:19:18] step 794350/1000000 | loss=2.1535 vocab=0.7211 token=1.4084 | lr=8.60e-05 | 6.10 steps/s | VRAM=28.65 Go [2026-07-18 12:19:26] step 794400/1000000 | loss=2.2097 vocab=0.7007 token=1.4492 | lr=8.60e-05 | 6.11 steps/s | VRAM=26.28 Go [2026-07-18 12:19:34] step 794450/1000000 | loss=2.1688 vocab=0.6656 token=1.2664 | lr=8.60e-05 | 6.15 steps/s | VRAM=27.96 Go [2026-07-18 12:19:42] step 794500/1000000 | loss=2.1418 vocab=1.2014 token=0.9513 | lr=8.60e-05 | 6.12 steps/s | VRAM=26.62 Go [2026-07-18 12:19:51] step 794550/1000000 | loss=2.1832 vocab=0.5429 token=1.3502 | lr=8.59e-05 | 6.12 steps/s | VRAM=26.42 Go [2026-07-18 12:19:59] step 794600/1000000 | loss=2.1749 vocab=0.6654 token=1.1755 | lr=8.59e-05 | 6.13 steps/s | VRAM=26.32 Go [2026-07-18 12:20:07] step 794650/1000000 | loss=2.1234 vocab=0.7422 token=0.8488 | lr=8.59e-05 | 6.11 steps/s | VRAM=29.37 Go [2026-07-18 12:20:15] step 794700/1000000 | loss=2.1281 vocab=0.8622 token=1.3313 | lr=8.59e-05 | 6.13 steps/s | VRAM=31.84 Go [2026-07-18 12:20:23] step 794750/1000000 | loss=2.0906 vocab=0.6655 token=1.1420 | lr=8.59e-05 | 6.09 steps/s | VRAM=29.06 Go [2026-07-18 12:20:32] step 794800/1000000 | loss=2.1970 vocab=0.3504 token=0.8871 | lr=8.58e-05 | 6.10 steps/s | VRAM=25.67 Go [2026-07-18 12:20:40] step 794850/1000000 | loss=2.1708 vocab=1.2483 token=1.6123 | lr=8.58e-05 | 6.10 steps/s | VRAM=28.71 Go [2026-07-18 12:20:48] step 794900/1000000 | loss=2.1973 vocab=0.8139 token=1.3328 | lr=8.58e-05 | 6.14 steps/s | VRAM=32.03 Go [2026-07-18 12:20:56] step 794950/1000000 | loss=2.1846 vocab=0.7373 token=1.0994 | lr=8.58e-05 | 6.13 steps/s | VRAM=28.65 Go [2026-07-18 12:21:04] step 795000/1000000 | loss=2.1052 vocab=0.7459 token=1.2919 | lr=8.58e-05 | 6.11 steps/s | VRAM=26.33 Go [2026-07-18 12:21:12] step 795050/1000000 | loss=2.2035 vocab=1.0327 token=1.5063 | lr=8.57e-05 | 6.14 steps/s | VRAM=29.17 Go [2026-07-18 12:21:21] step 795100/1000000 | loss=2.0762 vocab=1.4266 token=1.4562 | lr=8.57e-05 | 6.10 steps/s | VRAM=28.05 Go [2026-07-18 12:21:29] step 795150/1000000 | loss=2.0748 vocab=0.6505 token=1.2273 | lr=8.57e-05 | 6.13 steps/s | VRAM=31.68 Go [2026-07-18 12:21:37] step 795200/1000000 | loss=2.1756 vocab=1.1147 token=1.3779 | lr=8.57e-05 | 6.11 steps/s | VRAM=32.09 Go [2026-07-18 12:21:45] step 795250/1000000 | loss=2.2701 vocab=1.2205 token=1.2802 | lr=8.57e-05 | 6.10 steps/s | VRAM=26.05 Go [2026-07-18 12:21:53] step 795300/1000000 | loss=2.1475 vocab=0.5456 token=1.2261 | lr=8.56e-05 | 6.11 steps/s | VRAM=29.01 Go [2026-07-18 12:22:01] step 795350/1000000 | loss=2.1243 vocab=0.7115 token=1.2636 | lr=8.56e-05 | 6.16 steps/s | VRAM=31.24 Go [2026-07-18 12:22:10] step 795400/1000000 | loss=2.1443 vocab=0.8363 token=1.4901 | lr=8.56e-05 | 6.10 steps/s | VRAM=26.66 Go [2026-07-18 12:22:18] step 795450/1000000 | loss=2.1708 vocab=0.4691 token=1.0898 | lr=8.56e-05 | 6.15 steps/s | VRAM=29.67 Go [2026-07-18 12:22:26] step 795500/1000000 | loss=2.1531 vocab=0.7345 token=1.1644 | lr=8.56e-05 | 6.12 steps/s | VRAM=28.57 Go [2026-07-18 12:22:34] step 795550/1000000 | loss=2.1721 vocab=1.3544 token=1.8474 | lr=8.55e-05 | 6.13 steps/s | VRAM=32.50 Go [2026-07-18 12:22:42] step 795600/1000000 | loss=2.1995 vocab=1.0701 token=0.8340 | lr=8.55e-05 | 6.14 steps/s | VRAM=31.57 Go [2026-07-18 12:22:50] step 795650/1000000 | loss=2.1057 vocab=1.0695 token=1.2528 | lr=8.55e-05 | 6.13 steps/s | VRAM=28.75 Go [2026-07-18 12:22:59] step 795700/1000000 | loss=2.1302 vocab=0.7861 token=1.0795 | lr=8.55e-05 | 6.08 steps/s | VRAM=27.89 Go [2026-07-18 12:23:07] step 795750/1000000 | loss=2.1983 vocab=0.5793 token=0.9488 | lr=8.55e-05 | 6.16 steps/s | VRAM=29.02 Go [2026-07-18 12:23:15] step 795800/1000000 | loss=2.1610 vocab=0.9477 token=1.1865 | lr=8.54e-05 | 6.13 steps/s | VRAM=27.66 Go [2026-07-18 12:23:23] step 795850/1000000 | loss=2.2372 vocab=0.9915 token=1.3385 | lr=8.54e-05 | 6.15 steps/s | VRAM=28.10 Go [2026-07-18 12:23:31] step 795900/1000000 | loss=2.2166 vocab=0.9624 token=1.2702 | lr=8.54e-05 | 6.09 steps/s | VRAM=25.68 Go [2026-07-18 12:23:39] step 795950/1000000 | loss=2.1060 vocab=0.9075 token=1.3880 | lr=8.54e-05 | 6.12 steps/s | VRAM=26.53 Go [2026-07-18 12:23:48] step 796000/1000000 | loss=2.1713 vocab=0.8087 token=1.5148 | lr=8.54e-05 | 6.09 steps/s | VRAM=28.50 Go [2026-07-18 12:23:56] step 796050/1000000 | loss=2.1298 vocab=0.6717 token=1.3165 | lr=8.53e-05 | 6.14 steps/s | VRAM=30.91 Go [2026-07-18 12:24:04] step 796100/1000000 | loss=2.1870 vocab=0.7555 token=1.3387 | lr=8.53e-05 | 6.14 steps/s | VRAM=28.74 Go [2026-07-18 12:24:12] step 796150/1000000 | loss=2.1201 vocab=0.7075 token=1.4470 | lr=8.53e-05 | 6.13 steps/s | VRAM=26.31 Go [2026-07-18 12:24:20] step 796200/1000000 | loss=2.1029 vocab=0.6711 token=0.9108 | lr=8.53e-05 | 6.13 steps/s | VRAM=28.96 Go [2026-07-18 12:24:28] step 796250/1000000 | loss=2.1979 vocab=0.5553 token=0.5007 | lr=8.53e-05 | 6.12 steps/s | VRAM=26.05 Go [2026-07-18 12:24:36] step 796300/1000000 | loss=2.1336 vocab=0.8557 token=1.1954 | lr=8.52e-05 | 6.12 steps/s | VRAM=26.20 Go [2026-07-18 12:24:45] step 796350/1000000 | loss=2.0548 vocab=1.0674 token=1.5132 | lr=8.52e-05 | 6.14 steps/s | VRAM=32.09 Go [2026-07-18 12:24:53] step 796400/1000000 | loss=2.2469 vocab=1.5031 token=1.6323 | lr=8.52e-05 | 6.13 steps/s | VRAM=28.42 Go [2026-07-18 12:25:01] step 796450/1000000 | loss=2.1288 vocab=0.8386 token=1.0112 | lr=8.52e-05 | 6.13 steps/s | VRAM=31.77 Go [2026-07-18 12:25:09] step 796500/1000000 | loss=2.1750 vocab=0.4744 token=0.9355 | lr=8.52e-05 | 6.11 steps/s | VRAM=28.75 Go [2026-07-18 12:25:17] step 796550/1000000 | loss=2.1464 vocab=1.1633 token=1.3726 | lr=8.51e-05 | 6.15 steps/s | VRAM=28.03 Go [2026-07-18 12:25:25] step 796600/1000000 | loss=2.1540 vocab=1.0913 token=1.5961 | lr=8.51e-05 | 6.14 steps/s | VRAM=28.20 Go [2026-07-18 12:25:34] step 796650/1000000 | loss=2.1339 vocab=0.4790 token=0.8400 | lr=8.51e-05 | 6.11 steps/s | VRAM=26.28 Go [2026-07-18 12:25:42] step 796700/1000000 | loss=2.1989 vocab=0.7620 token=0.9217 | lr=8.51e-05 | 6.15 steps/s | VRAM=31.84 Go [2026-07-18 12:25:50] step 796750/1000000 | loss=2.2118 vocab=0.7623 token=1.5039 | lr=8.50e-05 | 6.15 steps/s | VRAM=30.40 Go [2026-07-18 12:25:58] step 796800/1000000 | loss=2.0700 vocab=0.7490 token=1.4862 | lr=8.50e-05 | 6.11 steps/s | VRAM=29.71 Go [2026-07-18 12:26:06] step 796850/1000000 | loss=2.1665 vocab=1.4901 token=1.3696 | lr=8.50e-05 | 6.12 steps/s | VRAM=29.28 Go [2026-07-18 12:26:14] step 796900/1000000 | loss=2.1420 vocab=1.0346 token=0.7880 | lr=8.50e-05 | 6.12 steps/s | VRAM=25.75 Go [2026-07-18 12:26:23] step 796950/1000000 | loss=2.1454 vocab=0.8859 token=1.3544 | lr=8.50e-05 | 6.12 steps/s | VRAM=29.01 Go [2026-07-18 12:26:31] step 797000/1000000 | loss=2.1606 vocab=1.1082 token=1.2594 | lr=8.49e-05 | 6.12 steps/s | VRAM=28.63 Go [2026-07-18 12:26:39] step 797050/1000000 | loss=2.1910 vocab=1.0514 token=1.4511 | lr=8.49e-05 | 6.11 steps/s | VRAM=28.97 Go [2026-07-18 12:26:47] step 797100/1000000 | loss=2.1365 vocab=0.8294 token=0.8163 | lr=8.49e-05 | 6.14 steps/s | VRAM=32.37 Go [2026-07-18 12:26:55] step 797150/1000000 | loss=2.1026 vocab=0.9439 token=1.3415 | lr=8.49e-05 | 6.10 steps/s | VRAM=25.77 Go [2026-07-18 12:27:03] step 797200/1000000 | loss=2.2159 vocab=1.0342 token=1.5464 | lr=8.49e-05 | 6.13 steps/s | VRAM=28.52 Go [2026-07-18 12:27:12] step 797250/1000000 | loss=2.1718 vocab=0.7416 token=1.4270 | lr=8.48e-05 | 6.13 steps/s | VRAM=28.46 Go [2026-07-18 12:27:20] step 797300/1000000 | loss=2.2112 vocab=0.8027 token=1.2421 | lr=8.48e-05 | 6.13 steps/s | VRAM=28.88 Go [2026-07-18 12:27:28] step 797350/1000000 | loss=2.1538 vocab=0.5609 token=0.8891 | lr=8.48e-05 | 6.15 steps/s | VRAM=30.93 Go [2026-07-18 12:27:36] step 797400/1000000 | loss=2.1200 vocab=0.8602 token=1.0833 | lr=8.48e-05 | 6.13 steps/s | VRAM=26.13 Go [2026-07-18 12:27:44] step 797450/1000000 | loss=2.1901 vocab=0.9835 token=1.2656 | lr=8.48e-05 | 6.14 steps/s | VRAM=32.88 Go [2026-07-18 12:27:52] step 797500/1000000 | loss=2.2380 vocab=0.9145 token=1.4227 | lr=8.47e-05 | 6.13 steps/s | VRAM=26.01 Go [2026-07-18 12:28:00] step 797550/1000000 | loss=2.1258 vocab=1.0937 token=1.3360 | lr=8.47e-05 | 6.12 steps/s | VRAM=28.16 Go [2026-07-18 12:28:09] step 797600/1000000 | loss=2.1001 vocab=0.8444 token=1.2984 | lr=8.47e-05 | 6.12 steps/s | VRAM=29.88 Go [2026-07-18 12:28:17] step 797650/1000000 | loss=2.2127 vocab=0.5514 token=1.1076 | lr=8.47e-05 | 6.11 steps/s | VRAM=28.25 Go [2026-07-18 12:28:25] step 797700/1000000 | loss=2.0941 vocab=0.8012 token=1.1757 | lr=8.47e-05 | 6.14 steps/s | VRAM=31.37 Go [2026-07-18 12:28:33] step 797750/1000000 | loss=2.0613 vocab=0.8437 token=1.1668 | lr=8.46e-05 | 6.09 steps/s | VRAM=28.36 Go [2026-07-18 12:28:41] step 797800/1000000 | loss=2.2756 vocab=0.4270 token=1.0179 | lr=8.46e-05 | 6.10 steps/s | VRAM=25.10 Go [2026-07-18 12:28:49] step 797850/1000000 | loss=2.1801 vocab=0.9539 token=1.8481 | lr=8.46e-05 | 6.14 steps/s | VRAM=27.36 Go [2026-07-18 12:28:58] step 797900/1000000 | loss=2.1013 vocab=0.5463 token=1.1103 | lr=8.46e-05 | 6.10 steps/s | VRAM=29.08 Go [2026-07-18 12:29:06] step 797950/1000000 | loss=2.1218 vocab=0.7764 token=1.2347 | lr=8.46e-05 | 6.09 steps/s | VRAM=29.36 Go [2026-07-18 12:29:14] step 798000/1000000 | loss=2.1927 vocab=0.5918 token=1.0949 | lr=8.45e-05 | 6.12 steps/s | VRAM=29.03 Go [2026-07-18 12:29:22] step 798050/1000000 | loss=2.2077 vocab=0.6236 token=1.4232 | lr=8.45e-05 | 6.12 steps/s | VRAM=26.79 Go [2026-07-18 12:29:30] step 798100/1000000 | loss=2.2246 vocab=0.5677 token=1.0451 | lr=8.45e-05 | 6.13 steps/s | VRAM=25.10 Go [2026-07-18 12:29:39] step 798150/1000000 | loss=2.2338 vocab=1.1173 token=1.1429 | lr=8.45e-05 | 6.12 steps/s | VRAM=28.48 Go [2026-07-18 12:29:47] step 798200/1000000 | loss=2.0990 vocab=0.7963 token=1.4425 | lr=8.45e-05 | 6.16 steps/s | VRAM=26.90 Go [2026-07-18 12:29:55] step 798250/1000000 | loss=2.1979 vocab=0.5514 token=1.3893 | lr=8.44e-05 | 6.17 steps/s | VRAM=29.01 Go [2026-07-18 12:30:03] step 798300/1000000 | loss=2.1041 vocab=0.9502 token=1.1647 | lr=8.44e-05 | 6.12 steps/s | VRAM=32.75 Go [2026-07-18 12:30:11] step 798350/1000000 | loss=2.1622 vocab=0.8153 token=1.4287 | lr=8.44e-05 | 6.12 steps/s | VRAM=28.67 Go [2026-07-18 12:30:19] step 798400/1000000 | loss=2.0945 vocab=0.7478 token=1.2401 | lr=8.44e-05 | 6.14 steps/s | VRAM=28.53 Go [2026-07-18 12:30:27] step 798450/1000000 | loss=2.0859 vocab=1.1107 token=1.2712 | lr=8.43e-05 | 6.09 steps/s | VRAM=26.94 Go [2026-07-18 12:30:36] step 798500/1000000 | loss=2.1269 vocab=1.0730 token=1.5729 | lr=8.43e-05 | 6.17 steps/s | VRAM=30.79 Go [2026-07-18 12:30:44] step 798550/1000000 | loss=2.1289 vocab=1.6276 token=1.0088 | lr=8.43e-05 | 6.13 steps/s | VRAM=25.57 Go [2026-07-18 12:30:52] step 798600/1000000 | loss=2.0842 vocab=0.9700 token=0.9085 | lr=8.43e-05 | 6.11 steps/s | VRAM=28.68 Go [2026-07-18 12:31:00] step 798650/1000000 | loss=2.1795 vocab=0.5225 token=1.5520 | lr=8.43e-05 | 6.13 steps/s | VRAM=29.19 Go [2026-07-18 12:31:08] step 798700/1000000 | loss=2.2169 vocab=1.0286 token=1.3696 | lr=8.42e-05 | 6.10 steps/s | VRAM=29.55 Go [2026-07-18 12:31:16] step 798750/1000000 | loss=2.2853 vocab=1.1405 token=1.4485 | lr=8.42e-05 | 6.12 steps/s | VRAM=29.24 Go [2026-07-18 12:31:25] step 798800/1000000 | loss=2.0967 vocab=1.3054 token=0.9796 | lr=8.42e-05 | 6.13 steps/s | VRAM=26.43 Go [2026-07-18 12:31:33] step 798850/1000000 | loss=2.2167 vocab=0.6795 token=1.0886 | lr=8.42e-05 | 6.15 steps/s | VRAM=31.31 Go [2026-07-18 12:31:41] step 798900/1000000 | loss=2.1762 vocab=1.5911 token=1.6983 | lr=8.42e-05 | 6.08 steps/s | VRAM=26.39 Go [2026-07-18 12:31:49] step 798950/1000000 | loss=2.1801 vocab=0.7173 token=0.9594 | lr=8.41e-05 | 6.11 steps/s | VRAM=32.04 Go [2026-07-18 12:31:57] step 799000/1000000 | loss=2.1099 vocab=0.8122 token=1.1176 | lr=8.41e-05 | 6.11 steps/s | VRAM=26.46 Go [2026-07-18 12:32:05] step 799050/1000000 | loss=2.1463 vocab=1.6318 token=1.0630 | lr=8.41e-05 | 6.11 steps/s | VRAM=29.06 Go [2026-07-18 12:32:14] step 799100/1000000 | loss=2.2727 vocab=0.8084 token=1.4201 | lr=8.41e-05 | 6.12 steps/s | VRAM=29.19 Go [2026-07-18 12:32:22] step 799150/1000000 | loss=2.0585 vocab=1.5473 token=1.3678 | lr=8.41e-05 | 6.11 steps/s | VRAM=26.19 Go [2026-07-18 12:32:30] step 799200/1000000 | loss=2.1514 vocab=0.9130 token=1.2860 | lr=8.40e-05 | 6.13 steps/s | VRAM=29.19 Go [2026-07-18 12:32:38] step 799250/1000000 | loss=2.0779 vocab=1.4997 token=1.2945 | lr=8.40e-05 | 6.11 steps/s | VRAM=26.50 Go [2026-07-18 12:32:46] step 799300/1000000 | loss=2.0850 vocab=1.0656 token=1.2698 | lr=8.40e-05 | 6.14 steps/s | VRAM=31.27 Go [2026-07-18 12:32:54] step 799350/1000000 | loss=2.2552 vocab=0.5971 token=1.3409 | lr=8.40e-05 | 6.13 steps/s | VRAM=30.67 Go [2026-07-18 12:33:03] step 799400/1000000 | loss=2.2556 vocab=1.3303 token=1.4932 | lr=8.40e-05 | 6.09 steps/s | VRAM=27.13 Go [2026-07-18 12:33:11] step 799450/1000000 | loss=2.1376 vocab=0.5618 token=1.0953 | lr=8.39e-05 | 6.13 steps/s | VRAM=25.60 Go [2026-07-18 12:33:19] step 799500/1000000 | loss=2.0498 vocab=0.7051 token=1.1448 | lr=8.39e-05 | 6.16 steps/s | VRAM=29.00 Go [2026-07-18 12:33:27] step 799550/1000000 | loss=2.1288 vocab=1.3710 token=1.1744 | lr=8.39e-05 | 6.14 steps/s | VRAM=28.48 Go [2026-07-18 12:33:35] step 799600/1000000 | loss=2.1879 vocab=0.8685 token=0.8776 | lr=8.39e-05 | 6.11 steps/s | VRAM=28.43 Go [2026-07-18 12:33:43] step 799650/1000000 | loss=2.1104 vocab=1.3003 token=1.3568 | lr=8.38e-05 | 6.16 steps/s | VRAM=30.88 Go [2026-07-18 12:33:52] step 799700/1000000 | loss=2.1333 vocab=0.6951 token=1.4279 | lr=8.38e-05 | 6.14 steps/s | VRAM=28.15 Go [2026-07-18 12:34:00] step 799750/1000000 | loss=2.2078 vocab=1.1807 token=1.7795 | lr=8.38e-05 | 6.07 steps/s | VRAM=26.78 Go [2026-07-18 12:34:08] step 799800/1000000 | loss=2.1036 vocab=0.5882 token=1.0596 | lr=8.38e-05 | 6.14 steps/s | VRAM=26.63 Go [2026-07-18 12:34:16] step 799850/1000000 | loss=2.2306 vocab=0.6427 token=1.3829 | lr=8.38e-05 | 6.14 steps/s | VRAM=27.32 Go [2026-07-18 12:34:24] step 799900/1000000 | loss=2.2023 vocab=0.3861 token=1.2183 | lr=8.37e-05 | 6.15 steps/s | VRAM=25.59 Go [2026-07-18 12:34:32] step 799950/1000000 | loss=2.1443 vocab=1.1926 token=1.4161 | lr=8.37e-05 | 6.17 steps/s | VRAM=32.23 Go [2026-07-18 12:34:40] step 800000/1000000 | loss=2.1801 vocab=0.8872 token=1.4603 | lr=8.37e-05 | 6.14 steps/s | VRAM=27.55 Go [accuracy] step 800000 : 62.54% (log: /workspace/plbert_train/precision_log.csv) [push] step 800000 OK (loss_token=1.4603, accuracy=62.54%) [2026-07-18 12:36:16] step 800050/1000000 | loss=2.1832 vocab=0.6847 token=1.3973 | lr=8.37e-05 | 0.52 steps/s | VRAM=30.78 Go [2026-07-18 12:36:24] step 800100/1000000 | loss=2.0919 vocab=0.5951 token=0.9988 | lr=8.37e-05 | 6.19 steps/s | VRAM=29.79 Go [2026-07-18 12:36:32] step 800150/1000000 | loss=2.1502 vocab=0.6998 token=1.0114 | lr=8.36e-05 | 6.21 steps/s | VRAM=32.45 Go [2026-07-18 12:36:40] step 800200/1000000 | loss=2.2838 vocab=1.0691 token=1.2890 | lr=8.36e-05 | 6.15 steps/s | VRAM=25.35 Go [2026-07-18 12:36:48] step 800250/1000000 | loss=2.1857 vocab=1.1819 token=1.2674 | lr=8.36e-05 | 6.18 steps/s | VRAM=32.60 Go [2026-07-18 12:36:56] step 800300/1000000 | loss=2.1551 vocab=0.8016 token=1.4273 | lr=8.36e-05 | 6.17 steps/s | VRAM=29.52 Go [2026-07-18 12:37:04] step 800350/1000000 | loss=2.0982 vocab=0.9839 token=1.1927 | lr=8.35e-05 | 6.18 steps/s | VRAM=31.71 Go [2026-07-18 12:37:13] step 800400/1000000 | loss=2.0601 vocab=0.7145 token=1.2072 | lr=8.35e-05 | 6.16 steps/s | VRAM=26.14 Go [2026-07-18 12:37:21] step 800450/1000000 | loss=2.1707 vocab=1.1584 token=1.2502 | lr=8.35e-05 | 6.15 steps/s | VRAM=28.82 Go [2026-07-18 12:37:29] step 800500/1000000 | loss=2.2192 vocab=0.9407 token=1.3384 | lr=8.35e-05 | 6.13 steps/s | VRAM=25.65 Go [2026-07-18 12:37:37] step 800550/1000000 | loss=2.1424 vocab=1.1327 token=1.4234 | lr=8.35e-05 | 6.15 steps/s | VRAM=31.95 Go [2026-07-18 12:37:45] step 800600/1000000 | loss=2.2297 vocab=1.2076 token=1.4313 | lr=8.34e-05 | 6.13 steps/s | VRAM=29.92 Go [2026-07-18 12:37:53] step 800650/1000000 | loss=2.0864 vocab=1.1165 token=1.8008 | lr=8.34e-05 | 6.13 steps/s | VRAM=28.97 Go [2026-07-18 12:38:01] step 800700/1000000 | loss=2.0637 vocab=0.6680 token=0.7818 | lr=8.34e-05 | 6.11 steps/s | VRAM=28.22 Go [2026-07-18 12:38:10] step 800750/1000000 | loss=2.1901 vocab=0.6035 token=1.4681 | lr=8.34e-05 | 6.14 steps/s | VRAM=25.17 Go [2026-07-18 12:38:18] step 800800/1000000 | loss=2.1235 vocab=0.4915 token=1.3161 | lr=8.34e-05 | 6.13 steps/s | VRAM=31.77 Go [2026-07-18 12:38:26] step 800850/1000000 | loss=2.2241 vocab=0.6605 token=1.2038 | lr=8.33e-05 | 6.12 steps/s | VRAM=31.94 Go [2026-07-18 12:38:34] step 800900/1000000 | loss=2.1264 vocab=1.1886 token=1.4675 | lr=8.33e-05 | 6.13 steps/s | VRAM=32.48 Go [2026-07-18 12:38:42] step 800950/1000000 | loss=2.1556 vocab=0.8516 token=1.4584 | lr=8.33e-05 | 6.10 steps/s | VRAM=28.68 Go [2026-07-18 12:38:50] step 801000/1000000 | loss=2.1130 vocab=1.1077 token=1.1323 | lr=8.33e-05 | 6.15 steps/s | VRAM=31.08 Go [2026-07-18 12:38:59] step 801050/1000000 | loss=2.1388 vocab=0.5914 token=1.5262 | lr=8.33e-05 | 6.12 steps/s | VRAM=26.16 Go [2026-07-18 12:39:07] step 801100/1000000 | loss=2.1762 vocab=1.1367 token=1.5948 | lr=8.32e-05 | 6.11 steps/s | VRAM=28.61 Go [2026-07-18 12:39:15] step 801150/1000000 | loss=2.1448 vocab=0.8354 token=1.3030 | lr=8.32e-05 | 6.14 steps/s | VRAM=28.62 Go [2026-07-18 12:39:23] step 801200/1000000 | loss=2.2269 vocab=1.3903 token=1.4258 | lr=8.32e-05 | 6.09 steps/s | VRAM=28.65 Go [2026-07-18 12:39:31] step 801250/1000000 | loss=2.0738 vocab=0.8286 token=1.0670 | lr=8.32e-05 | 6.14 steps/s | VRAM=25.50 Go [2026-07-18 12:39:39] step 801300/1000000 | loss=2.1894 vocab=0.6763 token=1.2724 | lr=8.31e-05 | 6.13 steps/s | VRAM=32.69 Go [2026-07-18 12:39:48] step 801350/1000000 | loss=2.1742 vocab=1.1793 token=1.8196 | lr=8.31e-05 | 6.08 steps/s | VRAM=32.80 Go [2026-07-18 12:39:56] step 801400/1000000 | loss=2.1080 vocab=0.6914 token=0.8398 | lr=8.31e-05 | 6.13 steps/s | VRAM=26.22 Go [2026-07-18 12:40:04] step 801450/1000000 | loss=2.0410 vocab=0.7459 token=0.9771 | lr=8.31e-05 | 6.16 steps/s | VRAM=30.76 Go [2026-07-18 12:40:12] step 801500/1000000 | loss=2.1045 vocab=1.0240 token=1.4925 | lr=8.31e-05 | 6.13 steps/s | VRAM=29.64 Go [2026-07-18 12:40:20] step 801550/1000000 | loss=2.0345 vocab=0.9925 token=1.4812 | lr=8.30e-05 | 6.12 steps/s | VRAM=31.41 Go [2026-07-18 12:40:28] step 801600/1000000 | loss=2.0430 vocab=0.7367 token=1.2772 | lr=8.30e-05 | 6.13 steps/s | VRAM=25.79 Go [2026-07-18 12:40:37] step 801650/1000000 | loss=2.1465 vocab=0.9490 token=1.3031 | lr=8.30e-05 | 6.13 steps/s | VRAM=26.61 Go [2026-07-18 12:40:45] step 801700/1000000 | loss=2.1345 vocab=0.8986 token=1.4350 | lr=8.30e-05 | 6.11 steps/s | VRAM=29.50 Go [2026-07-18 12:40:53] step 801750/1000000 | loss=2.1414 vocab=1.1803 token=1.1163 | lr=8.29e-05 | 6.12 steps/s | VRAM=29.20 Go [2026-07-18 12:41:01] step 801800/1000000 | loss=2.1345 vocab=0.8400 token=1.2069 | lr=8.29e-05 | 6.15 steps/s | VRAM=31.86 Go [2026-07-18 12:41:09] step 801850/1000000 | loss=2.1643 vocab=1.4884 token=1.6159 | lr=8.29e-05 | 6.14 steps/s | VRAM=29.22 Go [2026-07-18 12:41:17] step 801900/1000000 | loss=2.0907 vocab=0.8893 token=1.3636 | lr=8.29e-05 | 6.11 steps/s | VRAM=26.08 Go [2026-07-18 12:41:26] step 801950/1000000 | loss=2.1984 vocab=0.8703 token=1.0596 | lr=8.29e-05 | 6.11 steps/s | VRAM=26.43 Go [2026-07-18 12:41:34] step 802000/1000000 | loss=2.1804 vocab=1.2502 token=1.4742 | lr=8.28e-05 | 6.09 steps/s | VRAM=28.55 Go [2026-07-18 12:41:42] step 802050/1000000 | loss=2.2391 vocab=0.6513 token=1.0267 | lr=8.28e-05 | 6.12 steps/s | VRAM=29.89 Go [2026-07-18 12:41:50] step 802100/1000000 | loss=2.1639 vocab=0.5386 token=1.4998 | lr=8.28e-05 | 6.14 steps/s | VRAM=28.81 Go [2026-07-18 12:41:58] step 802150/1000000 | loss=2.1796 vocab=1.0442 token=2.0791 | lr=8.28e-05 | 6.18 steps/s | VRAM=32.00 Go [2026-07-18 12:42:06] step 802200/1000000 | loss=2.2033 vocab=1.0049 token=1.3953 | lr=8.28e-05 | 6.12 steps/s | VRAM=24.74 Go [2026-07-18 12:42:14] step 802250/1000000 | loss=2.1518 vocab=0.9551 token=1.4856 | lr=8.27e-05 | 6.13 steps/s | VRAM=25.93 Go [2026-07-18 12:42:23] step 802300/1000000 | loss=2.1674 vocab=0.4566 token=1.0968 | lr=8.27e-05 | 6.11 steps/s | VRAM=26.55 Go [2026-07-18 12:42:31] step 802350/1000000 | loss=2.2965 vocab=1.0566 token=1.5155 | lr=8.27e-05 | 6.16 steps/s | VRAM=31.37 Go [2026-07-18 12:42:39] step 802400/1000000 | loss=2.1671 vocab=1.0380 token=0.9619 | lr=8.27e-05 | 6.16 steps/s | VRAM=31.30 Go [2026-07-18 12:42:47] step 802450/1000000 | loss=2.1877 vocab=0.5524 token=0.9152 | lr=8.26e-05 | 6.13 steps/s | VRAM=29.12 Go [2026-07-18 12:42:55] step 802500/1000000 | loss=2.2232 vocab=1.0930 token=1.3556 | lr=8.26e-05 | 6.13 steps/s | VRAM=26.61 Go [2026-07-18 12:43:03] step 802550/1000000 | loss=2.2059 vocab=0.6087 token=0.9321 | lr=8.26e-05 | 6.11 steps/s | VRAM=28.66 Go [2026-07-18 12:43:12] step 802600/1000000 | loss=2.1608 vocab=0.9513 token=1.4951 | lr=8.26e-05 | 6.13 steps/s | VRAM=32.99 Go [2026-07-18 12:43:20] step 802650/1000000 | loss=2.1152 vocab=1.3947 token=1.2962 | lr=8.26e-05 | 6.12 steps/s | VRAM=26.95 Go [2026-07-18 12:43:28] step 802700/1000000 | loss=2.1551 vocab=0.7600 token=0.8129 | lr=8.25e-05 | 6.13 steps/s | VRAM=28.49 Go [2026-07-18 12:43:36] step 802750/1000000 | loss=2.1670 vocab=0.9774 token=1.4025 | lr=8.25e-05 | 6.15 steps/s | VRAM=32.40 Go [2026-07-18 12:43:44] step 802800/1000000 | loss=2.1965 vocab=0.6102 token=0.9975 | lr=8.25e-05 | 6.11 steps/s | VRAM=32.66 Go [2026-07-18 12:43:52] step 802850/1000000 | loss=2.1204 vocab=1.1911 token=1.3202 | lr=8.25e-05 | 6.13 steps/s | VRAM=27.99 Go [2026-07-18 12:44:01] step 802900/1000000 | loss=2.1262 vocab=0.7097 token=1.1730 | lr=8.25e-05 | 6.13 steps/s | VRAM=29.62 Go [2026-07-18 12:44:09] step 802950/1000000 | loss=2.1784 vocab=0.4738 token=1.1941 | lr=8.24e-05 | 6.15 steps/s | VRAM=30.76 Go [2026-07-18 12:44:17] step 803000/1000000 | loss=2.1863 vocab=0.6228 token=1.6144 | lr=8.24e-05 | 6.17 steps/s | VRAM=24.40 Go [2026-07-18 12:44:25] step 803050/1000000 | loss=2.0792 vocab=0.5307 token=0.8193 | lr=8.24e-05 | 6.14 steps/s | VRAM=28.15 Go [2026-07-18 12:44:33] step 803100/1000000 | loss=2.1792 vocab=0.5271 token=0.9098 | lr=8.24e-05 | 6.14 steps/s | VRAM=31.43 Go [2026-07-18 12:44:41] step 803150/1000000 | loss=2.1253 vocab=0.4969 token=1.2985 | lr=8.23e-05 | 6.12 steps/s | VRAM=26.11 Go [2026-07-18 12:44:49] step 803200/1000000 | loss=2.1958 vocab=0.6140 token=1.1145 | lr=8.23e-05 | 6.14 steps/s | VRAM=28.26 Go [2026-07-18 12:44:58] step 803250/1000000 | loss=2.1304 vocab=0.7346 token=0.9798 | lr=8.23e-05 | 6.13 steps/s | VRAM=26.14 Go [2026-07-18 12:45:06] step 803300/1000000 | loss=2.1053 vocab=0.6258 token=0.9531 | lr=8.23e-05 | 6.16 steps/s | VRAM=27.87 Go [2026-07-18 12:45:14] step 803350/1000000 | loss=2.1101 vocab=1.0362 token=1.3253 | lr=8.23e-05 | 6.18 steps/s | VRAM=31.07 Go [2026-07-18 12:45:22] step 803400/1000000 | loss=2.1255 vocab=0.7361 token=1.3455 | lr=8.22e-05 | 6.14 steps/s | VRAM=32.61 Go [2026-07-18 12:45:30] step 803450/1000000 | loss=2.1320 vocab=0.8991 token=1.2060 | lr=8.22e-05 | 6.15 steps/s | VRAM=31.38 Go [2026-07-18 12:45:38] step 803500/1000000 | loss=2.1619 vocab=0.7036 token=1.1232 | lr=8.22e-05 | 6.14 steps/s | VRAM=26.04 Go [2026-07-18 12:45:46] step 803550/1000000 | loss=2.2342 vocab=1.3274 token=1.4245 | lr=8.22e-05 | 6.13 steps/s | VRAM=28.57 Go [2026-07-18 12:45:54] step 803600/1000000 | loss=2.1248 vocab=0.5377 token=1.0573 | lr=8.21e-05 | 6.15 steps/s | VRAM=28.50 Go [2026-07-18 12:46:03] step 803650/1000000 | loss=2.2458 vocab=1.2973 token=1.5100 | lr=8.21e-05 | 6.15 steps/s | VRAM=31.70 Go [2026-07-18 12:46:11] step 803700/1000000 | loss=2.0589 vocab=0.7850 token=1.3640 | lr=8.21e-05 | 6.11 steps/s | VRAM=26.04 Go [2026-07-18 12:46:19] step 803750/1000000 | loss=2.0663 vocab=1.1759 token=1.1144 | lr=8.21e-05 | 6.14 steps/s | VRAM=31.58 Go [2026-07-18 12:46:27] step 803800/1000000 | loss=2.1248 vocab=0.6338 token=1.1463 | lr=8.21e-05 | 6.11 steps/s | VRAM=28.28 Go [2026-07-18 12:46:35] step 803850/1000000 | loss=2.1022 vocab=1.1751 token=1.5153 | lr=8.20e-05 | 6.13 steps/s | VRAM=28.20 Go [2026-07-18 12:46:43] step 803900/1000000 | loss=2.2045 vocab=0.8033 token=0.9907 | lr=8.20e-05 | 6.14 steps/s | VRAM=29.56 Go [2026-07-18 12:46:51] step 803950/1000000 | loss=2.1295 vocab=0.8672 token=1.1761 | lr=8.20e-05 | 6.17 steps/s | VRAM=24.41 Go [2026-07-18 12:47:00] step 804000/1000000 | loss=2.2444 vocab=1.0306 token=1.3449 | lr=8.20e-05 | 6.16 steps/s | VRAM=32.31 Go [2026-07-18 12:47:08] step 804050/1000000 | loss=2.2403 vocab=1.1467 token=1.3785 | lr=8.19e-05 | 6.10 steps/s | VRAM=25.47 Go [2026-07-18 12:47:16] step 804100/1000000 | loss=2.2147 vocab=0.7287 token=1.2992 | lr=8.19e-05 | 6.13 steps/s | VRAM=31.75 Go [2026-07-18 12:47:24] step 804150/1000000 | loss=2.0706 vocab=0.4905 token=1.0444 | lr=8.19e-05 | 6.11 steps/s | VRAM=29.90 Go [2026-07-18 12:47:32] step 804200/1000000 | loss=2.1137 vocab=0.8896 token=1.2968 | lr=8.19e-05 | 6.16 steps/s | VRAM=31.39 Go [2026-07-18 12:47:40] step 804250/1000000 | loss=2.1052 vocab=1.3670 token=1.6119 | lr=8.19e-05 | 6.11 steps/s | VRAM=28.18 Go [2026-07-18 12:47:49] step 804300/1000000 | loss=2.1006 vocab=0.4493 token=1.0267 | lr=8.18e-05 | 6.10 steps/s | VRAM=28.60 Go [2026-07-18 12:47:57] step 804350/1000000 | loss=2.0527 vocab=0.7287 token=0.9469 | lr=8.18e-05 | 6.11 steps/s | VRAM=29.89 Go [2026-07-18 12:48:05] step 804400/1000000 | loss=2.1169 vocab=1.1013 token=1.4014 | lr=8.18e-05 | 6.14 steps/s | VRAM=27.92 Go [2026-07-18 12:48:13] step 804450/1000000 | loss=2.0874 vocab=0.7071 token=0.9511 | lr=8.18e-05 | 6.13 steps/s | VRAM=28.30 Go [2026-07-18 12:48:21] step 804500/1000000 | loss=2.1680 vocab=0.5359 token=1.3092 | lr=8.17e-05 | 6.11 steps/s | VRAM=26.07 Go [2026-07-18 12:48:29] step 804550/1000000 | loss=2.1673 vocab=1.1450 token=1.6243 | lr=8.17e-05 | 6.12 steps/s | VRAM=29.15 Go [2026-07-18 12:48:38] step 804600/1000000 | loss=2.1451 vocab=0.9429 token=1.2520 | lr=8.17e-05 | 6.15 steps/s | VRAM=28.72 Go [2026-07-18 12:48:46] step 804650/1000000 | loss=2.2126 vocab=0.9108 token=1.2684 | lr=8.17e-05 | 6.14 steps/s | VRAM=31.86 Go [2026-07-18 12:48:54] step 804700/1000000 | loss=2.1074 vocab=0.7871 token=1.3053 | lr=8.17e-05 | 6.10 steps/s | VRAM=32.60 Go [2026-07-18 12:49:02] step 804750/1000000 | loss=2.1624 vocab=0.9238 token=1.3815 | lr=8.16e-05 | 6.15 steps/s | VRAM=32.53 Go [2026-07-18 12:49:10] step 804800/1000000 | loss=2.2023 vocab=0.7020 token=1.3035 | lr=8.16e-05 | 6.13 steps/s | VRAM=32.26 Go [2026-07-18 12:49:18] step 804850/1000000 | loss=2.1018 vocab=0.4234 token=1.0345 | lr=8.16e-05 | 6.15 steps/s | VRAM=32.03 Go [2026-07-18 12:49:26] step 804900/1000000 | loss=2.2145 vocab=0.5649 token=1.2790 | lr=8.16e-05 | 6.13 steps/s | VRAM=28.20 Go [2026-07-18 12:49:35] step 804950/1000000 | loss=2.0981 vocab=0.5900 token=1.1347 | lr=8.15e-05 | 6.14 steps/s | VRAM=28.68 Go [2026-07-18 12:49:43] step 805000/1000000 | loss=2.1756 vocab=0.6848 token=1.1060 | lr=8.15e-05 | 6.14 steps/s | VRAM=32.08 Go [2026-07-18 12:49:51] step 805050/1000000 | loss=2.1106 vocab=1.1699 token=1.5736 | lr=8.15e-05 | 6.13 steps/s | VRAM=28.97 Go [2026-07-18 12:49:59] step 805100/1000000 | loss=2.1660 vocab=0.7735 token=1.2737 | lr=8.15e-05 | 6.15 steps/s | VRAM=25.41 Go [2026-07-18 12:50:07] step 805150/1000000 | loss=2.2626 vocab=0.8952 token=1.2828 | lr=8.15e-05 | 6.12 steps/s | VRAM=29.65 Go [2026-07-18 12:50:15] step 805200/1000000 | loss=2.2186 vocab=0.9308 token=1.0363 | lr=8.14e-05 | 6.10 steps/s | VRAM=26.60 Go [2026-07-18 12:50:24] step 805250/1000000 | loss=2.2311 vocab=0.7467 token=1.4644 | lr=8.14e-05 | 6.14 steps/s | VRAM=32.31 Go [2026-07-18 12:50:32] step 805300/1000000 | loss=2.1895 vocab=0.8110 token=1.1286 | lr=8.14e-05 | 6.10 steps/s | VRAM=28.39 Go [2026-07-18 12:50:40] step 805350/1000000 | loss=2.1889 vocab=1.3118 token=1.5253 | lr=8.14e-05 | 6.12 steps/s | VRAM=29.34 Go [2026-07-18 12:50:48] step 805400/1000000 | loss=2.2380 vocab=1.3482 token=1.7605 | lr=8.13e-05 | 6.12 steps/s | VRAM=29.29 Go [2026-07-18 12:50:56] step 805450/1000000 | loss=2.2062 vocab=0.8819 token=1.2689 | lr=8.13e-05 | 6.09 steps/s | VRAM=28.74 Go [2026-07-18 12:51:05] step 805500/1000000 | loss=2.1117 vocab=0.8658 token=1.2085 | lr=8.13e-05 | 6.10 steps/s | VRAM=29.72 Go [2026-07-18 12:51:13] step 805550/1000000 | loss=2.1608 vocab=0.8408 token=1.1246 | lr=8.13e-05 | 6.15 steps/s | VRAM=32.05 Go [2026-07-18 12:51:21] step 805600/1000000 | loss=2.0729 vocab=1.3592 token=1.5945 | lr=8.13e-05 | 6.16 steps/s | VRAM=28.78 Go [2026-07-18 12:51:29] step 805650/1000000 | loss=2.1306 vocab=0.9059 token=1.3439 | lr=8.12e-05 | 6.14 steps/s | VRAM=32.73 Go [2026-07-18 12:51:37] step 805700/1000000 | loss=2.1808 vocab=0.5374 token=1.5324 | lr=8.12e-05 | 6.13 steps/s | VRAM=25.13 Go [2026-07-18 12:51:45] step 805750/1000000 | loss=2.2362 vocab=1.0574 token=1.5625 | lr=8.12e-05 | 6.13 steps/s | VRAM=23.90 Go [2026-07-18 12:51:53] step 805800/1000000 | loss=2.1277 vocab=1.8076 token=1.6341 | lr=8.12e-05 | 6.14 steps/s | VRAM=28.82 Go [2026-07-18 12:52:01] step 805850/1000000 | loss=2.2210 vocab=1.0467 token=1.2976 | lr=8.11e-05 | 6.16 steps/s | VRAM=27.85 Go [2026-07-18 12:52:10] step 805900/1000000 | loss=2.1643 vocab=0.4952 token=0.8260 | lr=8.11e-05 | 6.14 steps/s | VRAM=27.87 Go [2026-07-18 12:52:18] step 805950/1000000 | loss=2.1273 vocab=1.0131 token=1.1136 | lr=8.11e-05 | 6.11 steps/s | VRAM=28.53 Go [2026-07-18 12:52:26] step 806000/1000000 | loss=2.1372 vocab=0.6822 token=1.0413 | lr=8.11e-05 | 6.14 steps/s | VRAM=29.58 Go [2026-07-18 12:52:34] step 806050/1000000 | loss=2.1851 vocab=0.8940 token=1.3339 | lr=8.11e-05 | 6.12 steps/s | VRAM=30.17 Go [2026-07-18 12:52:42] step 806100/1000000 | loss=2.2002 vocab=0.8171 token=0.8793 | lr=8.10e-05 | 6.16 steps/s | VRAM=32.96 Go [2026-07-18 12:52:50] step 806150/1000000 | loss=2.2166 vocab=0.8377 token=1.2526 | lr=8.10e-05 | 6.13 steps/s | VRAM=27.87 Go [2026-07-18 12:52:59] step 806200/1000000 | loss=2.1308 vocab=1.0840 token=1.0180 | lr=8.10e-05 | 6.13 steps/s | VRAM=25.83 Go [2026-07-18 12:53:07] step 806250/1000000 | loss=2.1138 vocab=0.3751 token=1.0202 | lr=8.10e-05 | 6.15 steps/s | VRAM=28.33 Go [2026-07-18 12:53:15] step 806300/1000000 | loss=2.1679 vocab=0.8981 token=1.3574 | lr=8.09e-05 | 6.15 steps/s | VRAM=31.77 Go [2026-07-18 12:53:23] step 806350/1000000 | loss=2.1606 vocab=0.7607 token=0.9386 | lr=8.09e-05 | 6.10 steps/s | VRAM=26.28 Go [2026-07-18 12:53:31] step 806400/1000000 | loss=2.2084 vocab=1.1196 token=1.3010 | lr=8.09e-05 | 6.16 steps/s | VRAM=28.33 Go [2026-07-18 12:53:39] step 806450/1000000 | loss=2.1858 vocab=0.6141 token=1.0371 | lr=8.09e-05 | 6.14 steps/s | VRAM=28.97 Go [2026-07-18 12:53:47] step 806500/1000000 | loss=2.0976 vocab=0.7372 token=1.0905 | lr=8.09e-05 | 6.12 steps/s | VRAM=28.31 Go [2026-07-18 12:53:56] step 806550/1000000 | loss=2.2231 vocab=1.5100 token=1.3160 | lr=8.08e-05 | 6.12 steps/s | VRAM=29.66 Go [2026-07-18 12:54:04] step 806600/1000000 | loss=2.2256 vocab=0.6540 token=1.4668 | lr=8.08e-05 | 6.12 steps/s | VRAM=28.76 Go [2026-07-18 12:54:12] step 806650/1000000 | loss=2.1741 vocab=1.0999 token=1.1596 | lr=8.08e-05 | 6.15 steps/s | VRAM=28.94 Go [2026-07-18 12:54:20] step 806700/1000000 | loss=2.2249 vocab=1.4379 token=1.1563 | lr=8.08e-05 | 6.13 steps/s | VRAM=29.24 Go [2026-07-18 12:54:28] step 806750/1000000 | loss=2.1059 vocab=1.1472 token=1.3139 | lr=8.07e-05 | 6.13 steps/s | VRAM=29.69 Go [2026-07-18 12:54:36] step 806800/1000000 | loss=2.2859 vocab=1.2774 token=1.2474 | lr=8.07e-05 | 6.14 steps/s | VRAM=29.04 Go [2026-07-18 12:54:45] step 806850/1000000 | loss=2.1199 vocab=0.8016 token=0.7654 | lr=8.07e-05 | 6.13 steps/s | VRAM=32.22 Go [2026-07-18 12:54:53] step 806900/1000000 | loss=2.1310 vocab=0.7912 token=1.4583 | lr=8.07e-05 | 6.12 steps/s | VRAM=26.56 Go [2026-07-18 12:55:01] step 806950/1000000 | loss=2.0974 vocab=0.8231 token=0.9887 | lr=8.07e-05 | 6.14 steps/s | VRAM=28.47 Go [2026-07-18 12:55:09] step 807000/1000000 | loss=2.2080 vocab=0.9066 token=1.0095 | lr=8.06e-05 | 6.13 steps/s | VRAM=25.78 Go [2026-07-18 12:55:17] step 807050/1000000 | loss=2.0882 vocab=0.7996 token=1.2826 | lr=8.06e-05 | 6.15 steps/s | VRAM=29.66 Go [2026-07-18 12:55:25] step 807100/1000000 | loss=2.1803 vocab=0.8500 token=1.1459 | lr=8.06e-05 | 6.15 steps/s | VRAM=29.53 Go [2026-07-18 12:55:33] step 807150/1000000 | loss=2.1508 vocab=0.5924 token=1.3778 | lr=8.06e-05 | 6.12 steps/s | VRAM=28.93 Go [2026-07-18 12:55:42] step 807200/1000000 | loss=2.1697 vocab=0.5673 token=1.3427 | lr=8.05e-05 | 6.17 steps/s | VRAM=31.65 Go [2026-07-18 12:55:50] step 807250/1000000 | loss=2.2303 vocab=1.2350 token=1.1984 | lr=8.05e-05 | 6.14 steps/s | VRAM=29.15 Go [2026-07-18 12:55:58] step 807300/1000000 | loss=2.1175 vocab=0.6707 token=0.7820 | lr=8.05e-05 | 6.15 steps/s | VRAM=28.31 Go [2026-07-18 12:56:06] step 807350/1000000 | loss=2.1907 vocab=0.6653 token=1.1016 | lr=8.05e-05 | 6.15 steps/s | VRAM=27.96 Go [2026-07-18 12:56:14] step 807400/1000000 | loss=2.0751 vocab=1.2612 token=1.1509 | lr=8.04e-05 | 6.14 steps/s | VRAM=25.01 Go [2026-07-18 12:56:22] step 807450/1000000 | loss=2.0751 vocab=0.5424 token=1.0803 | lr=8.04e-05 | 6.15 steps/s | VRAM=25.16 Go [2026-07-18 12:56:30] step 807500/1000000 | loss=2.1127 vocab=0.6096 token=1.0738 | lr=8.04e-05 | 6.13 steps/s | VRAM=29.26 Go [2026-07-18 12:56:38] step 807550/1000000 | loss=2.1432 vocab=0.8875 token=1.3737 | lr=8.04e-05 | 6.15 steps/s | VRAM=27.98 Go [2026-07-18 12:56:47] step 807600/1000000 | loss=2.1634 vocab=1.3873 token=1.6260 | lr=8.04e-05 | 6.14 steps/s | VRAM=29.22 Go [2026-07-18 12:56:55] step 807650/1000000 | loss=2.1491 vocab=0.7647 token=1.2532 | lr=8.03e-05 | 6.14 steps/s | VRAM=29.63 Go [2026-07-18 12:57:03] step 807700/1000000 | loss=2.1985 vocab=1.1371 token=0.9798 | lr=8.03e-05 | 6.11 steps/s | VRAM=26.70 Go [2026-07-18 12:57:11] step 807750/1000000 | loss=2.1086 vocab=0.7846 token=1.4545 | lr=8.03e-05 | 6.16 steps/s | VRAM=31.08 Go [2026-07-18 12:57:19] step 807800/1000000 | loss=2.1754 vocab=0.8651 token=1.0647 | lr=8.03e-05 | 6.12 steps/s | VRAM=29.56 Go [2026-07-18 12:57:27] step 807850/1000000 | loss=2.1729 vocab=0.2720 token=0.9442 | lr=8.02e-05 | 6.12 steps/s | VRAM=32.75 Go [2026-07-18 12:57:36] step 807900/1000000 | loss=2.1969 vocab=0.8955 token=0.9659 | lr=8.02e-05 | 6.14 steps/s | VRAM=29.70 Go [2026-07-18 12:57:44] step 807950/1000000 | loss=2.1592 vocab=0.4932 token=1.1650 | lr=8.02e-05 | 6.13 steps/s | VRAM=28.94 Go [2026-07-18 12:57:52] step 808000/1000000 | loss=2.0427 vocab=0.6303 token=1.3134 | lr=8.02e-05 | 6.16 steps/s | VRAM=32.31 Go [2026-07-18 12:58:00] step 808050/1000000 | loss=2.2073 vocab=0.6538 token=1.2779 | lr=8.02e-05 | 6.13 steps/s | VRAM=27.96 Go [2026-07-18 12:58:08] step 808100/1000000 | loss=2.1116 vocab=0.3688 token=1.4070 | lr=8.01e-05 | 6.14 steps/s | VRAM=27.81 Go [2026-07-18 12:58:16] step 808150/1000000 | loss=2.1362 vocab=0.9225 token=1.4318 | lr=8.01e-05 | 6.16 steps/s | VRAM=31.30 Go [2026-07-18 12:58:24] step 808200/1000000 | loss=2.2668 vocab=0.9046 token=0.9738 | lr=8.01e-05 | 6.11 steps/s | VRAM=29.83 Go [2026-07-18 12:58:33] step 808250/1000000 | loss=2.2233 vocab=0.5466 token=1.4148 | lr=8.01e-05 | 6.12 steps/s | VRAM=28.09 Go [2026-07-18 12:58:41] step 808300/1000000 | loss=2.2454 vocab=0.9530 token=1.2319 | lr=8.00e-05 | 6.15 steps/s | VRAM=28.81 Go [2026-07-18 12:58:49] step 808350/1000000 | loss=2.2288 vocab=0.7920 token=1.7964 | lr=8.00e-05 | 6.14 steps/s | VRAM=29.96 Go [2026-07-18 12:58:57] step 808400/1000000 | loss=2.2101 vocab=1.2329 token=1.1580 | lr=8.00e-05 | 6.11 steps/s | VRAM=27.89 Go [2026-07-18 12:59:05] step 808450/1000000 | loss=2.1222 vocab=0.5359 token=1.1618 | lr=8.00e-05 | 6.15 steps/s | VRAM=28.22 Go [2026-07-18 12:59:13] step 808500/1000000 | loss=2.0776 vocab=0.7803 token=1.4141 | lr=7.99e-05 | 6.15 steps/s | VRAM=32.18 Go [2026-07-18 12:59:21] step 808550/1000000 | loss=2.3193 vocab=1.6737 token=1.4492 | lr=7.99e-05 | 6.16 steps/s | VRAM=31.31 Go [2026-07-18 12:59:30] step 808600/1000000 | loss=2.1755 vocab=0.4685 token=0.8847 | lr=7.99e-05 | 6.15 steps/s | VRAM=25.99 Go [2026-07-18 12:59:38] step 808650/1000000 | loss=2.1380 vocab=1.0647 token=1.5733 | lr=7.99e-05 | 6.14 steps/s | VRAM=29.43 Go [2026-07-18 12:59:46] step 808700/1000000 | loss=2.2480 vocab=0.7202 token=1.1530 | lr=7.99e-05 | 6.11 steps/s | VRAM=24.76 Go [2026-07-18 12:59:54] step 808750/1000000 | loss=2.1562 vocab=1.1771 token=1.3853 | lr=7.98e-05 | 6.14 steps/s | VRAM=29.03 Go [2026-07-18 13:00:02] step 808800/1000000 | loss=2.2362 vocab=0.5858 token=1.0886 | lr=7.98e-05 | 6.11 steps/s | VRAM=25.37 Go [2026-07-18 13:00:10] step 808850/1000000 | loss=2.1243 vocab=0.8086 token=1.1680 | lr=7.98e-05 | 6.16 steps/s | VRAM=32.68 Go [2026-07-18 13:00:18] step 808900/1000000 | loss=2.1570 vocab=0.9515 token=1.2911 | lr=7.98e-05 | 6.15 steps/s | VRAM=28.32 Go [2026-07-18 13:00:27] step 808950/1000000 | loss=2.1851 vocab=1.3496 token=1.6763 | lr=7.97e-05 | 6.15 steps/s | VRAM=28.47 Go [2026-07-18 13:00:35] step 809000/1000000 | loss=2.2244 vocab=0.9703 token=1.6272 | lr=7.97e-05 | 6.12 steps/s | VRAM=26.71 Go [2026-07-18 13:00:43] step 809050/1000000 | loss=2.1132 vocab=0.7474 token=1.4809 | lr=7.97e-05 | 6.15 steps/s | VRAM=28.97 Go [2026-07-18 13:00:51] step 809100/1000000 | loss=2.1567 vocab=0.3618 token=0.8838 | lr=7.97e-05 | 6.14 steps/s | VRAM=32.62 Go [2026-07-18 13:00:59] step 809150/1000000 | loss=2.1115 vocab=1.3416 token=1.4036 | lr=7.96e-05 | 6.15 steps/s | VRAM=28.67 Go [2026-07-18 13:01:07] step 809200/1000000 | loss=2.1257 vocab=0.8095 token=1.5484 | lr=7.96e-05 | 6.13 steps/s | VRAM=32.61 Go [2026-07-18 13:01:15] step 809250/1000000 | loss=2.2520 vocab=0.5275 token=1.3437 | lr=7.96e-05 | 6.15 steps/s | VRAM=28.20 Go [2026-07-18 13:01:24] step 809300/1000000 | loss=2.1892 vocab=0.7310 token=1.3774 | lr=7.96e-05 | 6.14 steps/s | VRAM=25.63 Go [2026-07-18 13:01:32] step 809350/1000000 | loss=2.1511 vocab=1.1636 token=1.3757 | lr=7.96e-05 | 6.12 steps/s | VRAM=28.57 Go [2026-07-18 13:01:40] step 809400/1000000 | loss=2.0914 vocab=0.6822 token=1.3854 | lr=7.95e-05 | 6.13 steps/s | VRAM=26.04 Go [2026-07-18 13:01:48] step 809450/1000000 | loss=2.1654 vocab=0.8678 token=1.2277 | lr=7.95e-05 | 6.15 steps/s | VRAM=32.20 Go [2026-07-18 13:01:56] step 809500/1000000 | loss=2.1415 vocab=0.5766 token=0.7970 | lr=7.95e-05 | 6.17 steps/s | VRAM=27.76 Go [2026-07-18 13:02:04] step 809550/1000000 | loss=2.1933 vocab=1.1818 token=1.4095 | lr=7.95e-05 | 6.14 steps/s | VRAM=28.30 Go [2026-07-18 13:02:12] step 809600/1000000 | loss=2.2326 vocab=1.4537 token=1.3879 | lr=7.94e-05 | 6.13 steps/s | VRAM=29.02 Go [2026-07-18 13:02:21] step 809650/1000000 | loss=2.2087 vocab=0.8784 token=1.3340 | lr=7.94e-05 | 6.16 steps/s | VRAM=27.61 Go [2026-07-18 13:02:29] step 809700/1000000 | loss=2.1658 vocab=1.7301 token=1.4456 | lr=7.94e-05 | 6.16 steps/s | VRAM=31.90 Go [2026-07-18 13:02:37] step 809750/1000000 | loss=2.1087 vocab=1.4847 token=1.5968 | lr=7.94e-05 | 6.17 steps/s | VRAM=28.08 Go [2026-07-18 13:02:45] step 809800/1000000 | loss=2.2687 vocab=0.8824 token=1.6988 | lr=7.93e-05 | 6.16 steps/s | VRAM=31.71 Go [2026-07-18 13:02:53] step 809850/1000000 | loss=2.0864 vocab=1.2669 token=1.4054 | lr=7.93e-05 | 6.17 steps/s | VRAM=32.07 Go [2026-07-18 13:03:01] step 809900/1000000 | loss=2.0725 vocab=1.1549 token=1.0764 | lr=7.93e-05 | 6.16 steps/s | VRAM=25.22 Go [2026-07-18 13:03:09] step 809950/1000000 | loss=2.1246 vocab=1.1230 token=1.5268 | lr=7.93e-05 | 6.12 steps/s | VRAM=28.77 Go [2026-07-18 13:03:17] step 810000/1000000 | loss=2.1372 vocab=0.8032 token=1.2400 | lr=7.93e-05 | 6.14 steps/s | VRAM=31.14 Go [2026-07-18 13:03:26] step 810050/1000000 | loss=2.0897 vocab=0.4983 token=1.1855 | lr=7.92e-05 | 6.17 steps/s | VRAM=32.00 Go [2026-07-18 13:03:34] step 810100/1000000 | loss=2.2096 vocab=0.7192 token=0.9932 | lr=7.92e-05 | 6.15 steps/s | VRAM=27.84 Go [2026-07-18 13:03:42] step 810150/1000000 | loss=2.0773 vocab=0.5262 token=1.1425 | lr=7.92e-05 | 6.14 steps/s | VRAM=32.30 Go [2026-07-18 13:03:50] step 810200/1000000 | loss=2.0582 vocab=0.7104 token=1.0722 | lr=7.92e-05 | 6.14 steps/s | VRAM=32.02 Go [2026-07-18 13:03:58] step 810250/1000000 | loss=2.2491 vocab=1.2956 token=1.1957 | lr=7.91e-05 | 6.12 steps/s | VRAM=28.29 Go [2026-07-18 13:04:06] step 810300/1000000 | loss=2.1280 vocab=0.4729 token=1.4513 | lr=7.91e-05 | 6.15 steps/s | VRAM=26.77 Go [2026-07-18 13:04:14] step 810350/1000000 | loss=2.2500 vocab=1.2722 token=1.4267 | lr=7.91e-05 | 6.16 steps/s | VRAM=30.63 Go [2026-07-18 13:04:23] step 810400/1000000 | loss=2.1185 vocab=0.7211 token=1.0884 | lr=7.91e-05 | 6.11 steps/s | VRAM=28.63 Go [2026-07-18 13:04:31] step 810450/1000000 | loss=2.1376 vocab=0.9510 token=1.4168 | lr=7.90e-05 | 6.15 steps/s | VRAM=32.76 Go [2026-07-18 13:04:39] step 810500/1000000 | loss=2.1144 vocab=0.6798 token=1.2048 | lr=7.90e-05 | 6.14 steps/s | VRAM=28.33 Go [2026-07-18 13:04:47] step 810550/1000000 | loss=2.2863 vocab=0.8692 token=1.1720 | lr=7.90e-05 | 6.15 steps/s | VRAM=28.32 Go [2026-07-18 13:04:55] step 810600/1000000 | loss=2.1388 vocab=1.0468 token=1.6731 | lr=7.90e-05 | 6.14 steps/s | VRAM=30.66 Go [2026-07-18 13:05:03] step 810650/1000000 | loss=2.1699 vocab=1.4965 token=1.3857 | lr=7.89e-05 | 6.12 steps/s | VRAM=27.10 Go [2026-07-18 13:05:11] step 810700/1000000 | loss=2.1098 vocab=0.6488 token=1.5367 | lr=7.89e-05 | 6.17 steps/s | VRAM=28.48 Go [2026-07-18 13:05:20] step 810750/1000000 | loss=2.1982 vocab=1.2233 token=1.3898 | lr=7.89e-05 | 6.16 steps/s | VRAM=31.79 Go [2026-07-18 13:05:28] step 810800/1000000 | loss=2.2004 vocab=1.2269 token=1.3196 | lr=7.89e-05 | 6.13 steps/s | VRAM=26.80 Go [2026-07-18 13:05:36] step 810850/1000000 | loss=2.1134 vocab=0.7249 token=1.1745 | lr=7.89e-05 | 6.16 steps/s | VRAM=25.08 Go [2026-07-18 13:05:44] step 810900/1000000 | loss=2.1160 vocab=0.8894 token=1.6062 | lr=7.88e-05 | 6.15 steps/s | VRAM=28.78 Go [2026-07-18 13:05:52] step 810950/1000000 | loss=2.1437 vocab=1.4156 token=0.8479 | lr=7.88e-05 | 6.15 steps/s | VRAM=28.45 Go [2026-07-18 13:06:00] step 811000/1000000 | loss=2.0875 vocab=0.7842 token=1.1243 | lr=7.88e-05 | 6.11 steps/s | VRAM=26.28 Go [2026-07-18 13:06:08] step 811050/1000000 | loss=2.2730 vocab=1.0589 token=1.3199 | lr=7.88e-05 | 6.14 steps/s | VRAM=28.49 Go [2026-07-18 13:06:17] step 811100/1000000 | loss=2.1894 vocab=1.0510 token=1.3558 | lr=7.87e-05 | 6.14 steps/s | VRAM=28.16 Go [2026-07-18 13:06:25] step 811150/1000000 | loss=1.9775 vocab=0.8063 token=1.3798 | lr=7.87e-05 | 6.14 steps/s | VRAM=25.77 Go [2026-07-18 13:06:33] step 811200/1000000 | loss=2.2043 vocab=0.8112 token=1.3594 | lr=7.87e-05 | 6.10 steps/s | VRAM=27.79 Go [2026-07-18 13:06:41] step 811250/1000000 | loss=2.1070 vocab=0.7854 token=1.1572 | lr=7.87e-05 | 6.15 steps/s | VRAM=29.44 Go [2026-07-18 13:06:49] step 811300/1000000 | loss=2.0969 vocab=0.8196 token=1.2131 | lr=7.86e-05 | 6.13 steps/s | VRAM=25.76 Go [2026-07-18 13:06:57] step 811350/1000000 | loss=2.2599 vocab=0.4530 token=1.0446 | lr=7.86e-05 | 6.15 steps/s | VRAM=31.41 Go [2026-07-18 13:07:05] step 811400/1000000 | loss=2.2049 vocab=0.9384 token=1.3637 | lr=7.86e-05 | 6.13 steps/s | VRAM=25.11 Go [2026-07-18 13:07:14] step 811450/1000000 | loss=2.1425 vocab=1.2479 token=1.5346 | lr=7.86e-05 | 6.13 steps/s | VRAM=28.08 Go [2026-07-18 13:07:22] step 811500/1000000 | loss=2.1635 vocab=0.5855 token=1.1212 | lr=7.86e-05 | 6.15 steps/s | VRAM=28.69 Go [2026-07-18 13:07:30] step 811550/1000000 | loss=2.0920 vocab=0.8361 token=1.3136 | lr=7.85e-05 | 6.13 steps/s | VRAM=32.24 Go [2026-07-18 13:07:38] step 811600/1000000 | loss=2.2361 vocab=0.4795 token=1.1285 | lr=7.85e-05 | 6.13 steps/s | VRAM=28.91 Go [2026-07-18 13:07:46] step 811650/1000000 | loss=2.0782 vocab=0.6564 token=1.1484 | lr=7.85e-05 | 6.16 steps/s | VRAM=28.69 Go [2026-07-18 13:07:54] step 811700/1000000 | loss=2.1348 vocab=1.2386 token=1.4786 | lr=7.85e-05 | 6.12 steps/s | VRAM=30.12 Go [2026-07-18 13:08:02] step 811750/1000000 | loss=2.0840 vocab=0.6228 token=1.4703 | lr=7.84e-05 | 6.15 steps/s | VRAM=28.80 Go [2026-07-18 13:08:11] step 811800/1000000 | loss=2.1933 vocab=1.1200 token=1.5669 | lr=7.84e-05 | 6.13 steps/s | VRAM=28.04 Go [2026-07-18 13:08:19] step 811850/1000000 | loss=2.1076 vocab=0.4633 token=1.0212 | lr=7.84e-05 | 6.15 steps/s | VRAM=31.49 Go [2026-07-18 13:08:27] step 811900/1000000 | loss=2.0922 vocab=0.7989 token=1.2815 | lr=7.84e-05 | 6.15 steps/s | VRAM=32.00 Go [2026-07-18 13:08:35] step 811950/1000000 | loss=2.2191 vocab=1.2379 token=1.4944 | lr=7.83e-05 | 6.15 steps/s | VRAM=31.99 Go [2026-07-18 13:08:43] step 812000/1000000 | loss=2.0439 vocab=0.7543 token=1.3960 | lr=7.83e-05 | 6.14 steps/s | VRAM=25.96 Go [2026-07-18 13:08:51] step 812050/1000000 | loss=1.9998 vocab=0.7445 token=1.0275 | lr=7.83e-05 | 6.17 steps/s | VRAM=32.50 Go [2026-07-18 13:08:59] step 812100/1000000 | loss=2.1819 vocab=0.5591 token=1.2003 | lr=7.83e-05 | 6.14 steps/s | VRAM=31.77 Go [2026-07-18 13:09:08] step 812150/1000000 | loss=2.2323 vocab=0.7324 token=1.0009 | lr=7.82e-05 | 6.14 steps/s | VRAM=28.57 Go [2026-07-18 13:09:16] step 812200/1000000 | loss=2.2165 vocab=0.6892 token=1.1713 | lr=7.82e-05 | 6.16 steps/s | VRAM=32.39 Go [2026-07-18 13:09:24] step 812250/1000000 | loss=2.0582 vocab=0.7416 token=1.9254 | lr=7.82e-05 | 6.17 steps/s | VRAM=29.17 Go [2026-07-18 13:09:32] step 812300/1000000 | loss=2.1389 vocab=0.5638 token=0.8249 | lr=7.82e-05 | 6.14 steps/s | VRAM=31.72 Go [2026-07-18 13:09:40] step 812350/1000000 | loss=2.2399 vocab=0.5973 token=1.0385 | lr=7.82e-05 | 6.15 steps/s | VRAM=24.66 Go [2026-07-18 13:09:48] step 812400/1000000 | loss=2.2253 vocab=1.0847 token=1.7272 | lr=7.81e-05 | 6.14 steps/s | VRAM=28.44 Go [2026-07-18 13:09:56] step 812450/1000000 | loss=2.1172 vocab=0.6327 token=1.4387 | lr=7.81e-05 | 6.10 steps/s | VRAM=26.07 Go [2026-07-18 13:10:05] step 812500/1000000 | loss=2.0549 vocab=0.6919 token=0.9415 | lr=7.81e-05 | 6.14 steps/s | VRAM=28.53 Go [2026-07-18 13:10:13] step 812550/1000000 | loss=2.2156 vocab=0.8448 token=1.6113 | lr=7.81e-05 | 6.13 steps/s | VRAM=28.88 Go [2026-07-18 13:10:21] step 812600/1000000 | loss=2.0952 vocab=1.0458 token=1.2102 | lr=7.80e-05 | 6.13 steps/s | VRAM=28.33 Go [2026-07-18 13:10:29] step 812650/1000000 | loss=2.1430 vocab=0.5444 token=1.0149 | lr=7.80e-05 | 6.14 steps/s | VRAM=32.72 Go [2026-07-18 13:10:37] step 812700/1000000 | loss=2.1893 vocab=1.0171 token=1.3709 | lr=7.80e-05 | 6.12 steps/s | VRAM=26.43 Go [2026-07-18 13:10:45] step 812750/1000000 | loss=2.0939 vocab=0.6622 token=0.6442 | lr=7.80e-05 | 6.15 steps/s | VRAM=27.81 Go [2026-07-18 13:10:53] step 812800/1000000 | loss=2.1535 vocab=0.5087 token=1.0562 | lr=7.79e-05 | 6.16 steps/s | VRAM=32.12 Go [2026-07-18 13:11:02] step 812850/1000000 | loss=2.1544 vocab=1.0143 token=1.3592 | lr=7.79e-05 | 6.15 steps/s | VRAM=29.40 Go [2026-07-18 13:11:10] step 812900/1000000 | loss=2.1480 vocab=1.7859 token=1.2322 | lr=7.79e-05 | 6.13 steps/s | VRAM=26.68 Go [2026-07-18 13:11:18] step 812950/1000000 | loss=2.0897 vocab=0.6254 token=1.3460 | lr=7.79e-05 | 6.15 steps/s | VRAM=28.32 Go [2026-07-18 13:11:26] step 813000/1000000 | loss=2.1703 vocab=0.9376 token=1.6973 | lr=7.78e-05 | 6.15 steps/s | VRAM=32.52 Go [2026-07-18 13:11:34] step 813050/1000000 | loss=2.2070 vocab=0.8690 token=1.3969 | lr=7.78e-05 | 6.12 steps/s | VRAM=26.23 Go [2026-07-18 13:11:42] step 813100/1000000 | loss=2.1205 vocab=0.5586 token=0.9390 | lr=7.78e-05 | 6.17 steps/s | VRAM=26.12 Go [2026-07-18 13:11:50] step 813150/1000000 | loss=2.0769 vocab=1.0001 token=1.2612 | lr=7.78e-05 | 6.18 steps/s | VRAM=25.60 Go [2026-07-18 13:11:58] step 813200/1000000 | loss=2.2346 vocab=1.1581 token=1.5780 | lr=7.77e-05 | 6.15 steps/s | VRAM=28.80 Go [2026-07-18 13:12:07] step 813250/1000000 | loss=2.2128 vocab=1.2001 token=1.5988 | lr=7.77e-05 | 6.15 steps/s | VRAM=24.38 Go [2026-07-18 13:12:15] step 813300/1000000 | loss=1.9807 vocab=1.0216 token=1.1483 | lr=7.77e-05 | 6.15 steps/s | VRAM=26.37 Go [2026-07-18 13:12:23] step 813350/1000000 | loss=2.0860 vocab=0.8022 token=1.2374 | lr=7.77e-05 | 6.12 steps/s | VRAM=29.78 Go [2026-07-18 13:12:31] step 813400/1000000 | loss=2.1938 vocab=1.0930 token=1.3515 | lr=7.77e-05 | 6.14 steps/s | VRAM=28.78 Go [2026-07-18 13:12:39] step 813450/1000000 | loss=2.1253 vocab=0.9230 token=1.0482 | lr=7.76e-05 | 6.13 steps/s | VRAM=28.85 Go [2026-07-18 13:12:47] step 813500/1000000 | loss=2.3116 vocab=1.0098 token=1.2879 | lr=7.76e-05 | 6.15 steps/s | VRAM=31.78 Go [2026-07-18 13:12:55] step 813550/1000000 | loss=2.1914 vocab=1.1374 token=1.0977 | lr=7.76e-05 | 6.16 steps/s | VRAM=25.69 Go [2026-07-18 13:13:04] step 813600/1000000 | loss=2.1979 vocab=0.8830 token=1.3384 | lr=7.76e-05 | 6.12 steps/s | VRAM=30.85 Go [2026-07-18 13:13:12] step 813650/1000000 | loss=2.1740 vocab=0.3288 token=0.7686 | lr=7.75e-05 | 6.16 steps/s | VRAM=27.94 Go [2026-07-18 13:13:20] step 813700/1000000 | loss=2.2071 vocab=0.7424 token=1.3824 | lr=7.75e-05 | 6.12 steps/s | VRAM=26.84 Go [2026-07-18 13:13:28] step 813750/1000000 | loss=2.2182 vocab=0.8674 token=0.8600 | lr=7.75e-05 | 6.13 steps/s | VRAM=32.41 Go [2026-07-18 13:13:36] step 813800/1000000 | loss=2.1045 vocab=1.0196 token=1.7024 | lr=7.75e-05 | 6.11 steps/s | VRAM=29.21 Go [2026-07-18 13:13:44] step 813850/1000000 | loss=2.2220 vocab=1.1754 token=1.2988 | lr=7.74e-05 | 6.15 steps/s | VRAM=31.85 Go [2026-07-18 13:13:52] step 813900/1000000 | loss=2.1368 vocab=1.1740 token=1.4684 | lr=7.74e-05 | 6.15 steps/s | VRAM=29.69 Go [2026-07-18 13:14:01] step 813950/1000000 | loss=2.1455 vocab=1.0384 token=1.3252 | lr=7.74e-05 | 6.15 steps/s | VRAM=31.04 Go [2026-07-18 13:14:09] step 814000/1000000 | loss=2.1712 vocab=0.7025 token=1.4854 | lr=7.74e-05 | 6.14 steps/s | VRAM=28.76 Go [2026-07-18 13:14:17] step 814050/1000000 | loss=2.1494 vocab=0.8697 token=1.0928 | lr=7.73e-05 | 6.16 steps/s | VRAM=32.92 Go [2026-07-18 13:14:25] step 814100/1000000 | loss=2.1793 vocab=0.9734 token=1.2784 | lr=7.73e-05 | 6.15 steps/s | VRAM=29.73 Go [2026-07-18 13:14:33] step 814150/1000000 | loss=2.1095 vocab=1.0107 token=1.2251 | lr=7.73e-05 | 6.14 steps/s | VRAM=31.76 Go [2026-07-18 13:14:41] step 814200/1000000 | loss=2.0830 vocab=1.0997 token=1.4104 | lr=7.73e-05 | 6.12 steps/s | VRAM=26.58 Go [2026-07-18 13:14:49] step 814250/1000000 | loss=2.0690 vocab=0.7914 token=1.2571 | lr=7.72e-05 | 6.14 steps/s | VRAM=29.07 Go [2026-07-18 13:14:58] step 814300/1000000 | loss=2.1514 vocab=1.0353 token=1.1510 | lr=7.72e-05 | 6.13 steps/s | VRAM=27.65 Go [2026-07-18 13:15:06] step 814350/1000000 | loss=2.0844 vocab=0.8877 token=0.7943 | lr=7.72e-05 | 6.16 steps/s | VRAM=25.91 Go [2026-07-18 13:15:14] step 814400/1000000 | loss=2.0577 vocab=0.8938 token=1.3755 | lr=7.72e-05 | 6.16 steps/s | VRAM=31.29 Go [2026-07-18 13:15:22] step 814450/1000000 | loss=2.1129 vocab=1.4827 token=1.0401 | lr=7.71e-05 | 6.12 steps/s | VRAM=26.04 Go [2026-07-18 13:15:30] step 814500/1000000 | loss=2.0809 vocab=0.9655 token=1.2993 | lr=7.71e-05 | 6.14 steps/s | VRAM=32.57 Go [2026-07-18 13:15:38] step 814550/1000000 | loss=2.2340 vocab=1.0371 token=1.3557 | lr=7.71e-05 | 6.11 steps/s | VRAM=26.32 Go [2026-07-18 13:15:46] step 814600/1000000 | loss=2.1324 vocab=0.6587 token=1.2908 | lr=7.71e-05 | 6.16 steps/s | VRAM=27.79 Go [2026-07-18 13:15:55] step 814650/1000000 | loss=2.0798 vocab=0.7658 token=1.1319 | lr=7.71e-05 | 6.14 steps/s | VRAM=26.83 Go [2026-07-18 13:16:03] step 814700/1000000 | loss=2.0879 vocab=0.9517 token=1.2009 | lr=7.70e-05 | 6.17 steps/s | VRAM=29.96 Go [2026-07-18 13:16:11] step 814750/1000000 | loss=2.1908 vocab=0.6566 token=1.3443 | lr=7.70e-05 | 6.12 steps/s | VRAM=25.66 Go [2026-07-18 13:16:19] step 814800/1000000 | loss=2.1787 vocab=0.9386 token=1.1102 | lr=7.70e-05 | 6.16 steps/s | VRAM=29.28 Go [2026-07-18 13:16:27] step 814850/1000000 | loss=2.1387 vocab=0.9021 token=1.3550 | lr=7.70e-05 | 6.18 steps/s | VRAM=31.90 Go [2026-07-18 13:16:35] step 814900/1000000 | loss=2.1919 vocab=0.3178 token=0.8232 | lr=7.69e-05 | 6.15 steps/s | VRAM=32.40 Go [2026-07-18 13:16:43] step 814950/1000000 | loss=2.1495 vocab=0.9093 token=1.1039 | lr=7.69e-05 | 6.12 steps/s | VRAM=25.73 Go [2026-07-18 13:16:52] step 815000/1000000 | loss=2.1824 vocab=0.9761 token=1.5411 | lr=7.69e-05 | 6.16 steps/s | VRAM=25.93 Go [2026-07-18 13:17:00] step 815050/1000000 | loss=2.1179 vocab=1.1982 token=1.4323 | lr=7.69e-05 | 6.17 steps/s | VRAM=32.27 Go [2026-07-18 13:17:08] step 815100/1000000 | loss=2.1344 vocab=1.0272 token=1.4074 | lr=7.68e-05 | 6.16 steps/s | VRAM=31.30 Go [2026-07-18 13:17:16] step 815150/1000000 | loss=2.1113 vocab=0.8893 token=1.3347 | lr=7.68e-05 | 6.13 steps/s | VRAM=31.26 Go [2026-07-18 13:17:24] step 815200/1000000 | loss=2.2073 vocab=0.5503 token=1.3270 | lr=7.68e-05 | 6.13 steps/s | VRAM=28.44 Go [2026-07-18 13:17:32] step 815250/1000000 | loss=2.1331 vocab=0.7433 token=1.1363 | lr=7.68e-05 | 6.13 steps/s | VRAM=25.63 Go [2026-07-18 13:17:40] step 815300/1000000 | loss=2.0314 vocab=0.5756 token=1.1734 | lr=7.67e-05 | 6.11 steps/s | VRAM=28.87 Go [2026-07-18 13:17:49] step 815350/1000000 | loss=2.1802 vocab=0.9830 token=1.3394 | lr=7.67e-05 | 6.13 steps/s | VRAM=28.79 Go [2026-07-18 13:17:57] step 815400/1000000 | loss=2.1758 vocab=0.7382 token=1.1659 | lr=7.67e-05 | 6.12 steps/s | VRAM=29.23 Go [2026-07-18 13:18:05] step 815450/1000000 | loss=2.0941 vocab=0.7984 token=1.4255 | lr=7.67e-05 | 6.15 steps/s | VRAM=31.77 Go [2026-07-18 13:18:13] step 815500/1000000 | loss=2.1049 vocab=0.9378 token=1.2207 | lr=7.66e-05 | 6.13 steps/s | VRAM=32.91 Go [2026-07-18 13:18:21] step 815550/1000000 | loss=2.1712 vocab=1.1359 token=1.2699 | lr=7.66e-05 | 6.14 steps/s | VRAM=29.47 Go [2026-07-18 13:18:29] step 815600/1000000 | loss=2.0908 vocab=0.7190 token=1.0753 | lr=7.66e-05 | 6.15 steps/s | VRAM=26.08 Go [2026-07-18 13:18:37] step 815650/1000000 | loss=2.1137 vocab=0.8905 token=1.0914 | lr=7.66e-05 | 6.16 steps/s | VRAM=32.95 Go [2026-07-18 13:18:46] step 815700/1000000 | loss=2.0219 vocab=1.6242 token=1.5389 | lr=7.65e-05 | 6.14 steps/s | VRAM=28.45 Go [2026-07-18 13:18:54] step 815750/1000000 | loss=2.2391 vocab=1.0240 token=1.2774 | lr=7.65e-05 | 6.15 steps/s | VRAM=29.09 Go [2026-07-18 13:19:02] step 815800/1000000 | loss=2.1364 vocab=0.8931 token=1.2655 | lr=7.65e-05 | 6.14 steps/s | VRAM=26.48 Go [2026-07-18 13:19:10] step 815850/1000000 | loss=2.1740 vocab=1.1244 token=1.1385 | lr=7.65e-05 | 6.14 steps/s | VRAM=28.80 Go [2026-07-18 13:19:18] step 815900/1000000 | loss=2.1813 vocab=0.8032 token=1.2369 | lr=7.64e-05 | 6.15 steps/s | VRAM=28.41 Go [2026-07-18 13:19:26] step 815950/1000000 | loss=2.2089 vocab=1.2621 token=1.5016 | lr=7.64e-05 | 6.13 steps/s | VRAM=28.14 Go [2026-07-18 13:19:34] step 816000/1000000 | loss=2.1278 vocab=1.0665 token=1.1750 | lr=7.64e-05 | 6.15 steps/s | VRAM=28.30 Go [2026-07-18 13:19:42] step 816050/1000000 | loss=2.1159 vocab=1.0848 token=1.5123 | lr=7.64e-05 | 6.17 steps/s | VRAM=31.36 Go [2026-07-18 13:19:51] step 816100/1000000 | loss=2.1462 vocab=0.8178 token=1.2054 | lr=7.64e-05 | 6.14 steps/s | VRAM=27.06 Go [2026-07-18 13:19:59] step 816150/1000000 | loss=2.1310 vocab=1.6309 token=1.6495 | lr=7.63e-05 | 6.13 steps/s | VRAM=29.57 Go [2026-07-18 13:20:07] step 816200/1000000 | loss=2.2111 vocab=0.5916 token=1.2563 | lr=7.63e-05 | 6.14 steps/s | VRAM=28.52 Go [2026-07-18 13:20:15] step 816250/1000000 | loss=2.2356 vocab=1.1795 token=1.6825 | lr=7.63e-05 | 6.15 steps/s | VRAM=28.01 Go [2026-07-18 13:20:23] step 816300/1000000 | loss=2.1500 vocab=1.0400 token=1.6793 | lr=7.63e-05 | 6.14 steps/s | VRAM=23.24 Go [2026-07-18 13:20:31] step 816350/1000000 | loss=2.2361 vocab=0.6622 token=1.1131 | lr=7.62e-05 | 6.12 steps/s | VRAM=29.46 Go [2026-07-18 13:20:40] step 816400/1000000 | loss=2.2354 vocab=1.0861 token=1.5197 | lr=7.62e-05 | 6.13 steps/s | VRAM=24.97 Go [2026-07-18 13:20:48] step 816450/1000000 | loss=2.1710 vocab=0.9848 token=1.5712 | lr=7.62e-05 | 6.12 steps/s | VRAM=31.35 Go [2026-07-18 13:20:56] step 816500/1000000 | loss=2.2250 vocab=0.9375 token=1.6991 | lr=7.62e-05 | 6.09 steps/s | VRAM=26.24 Go [2026-07-18 13:21:04] step 816550/1000000 | loss=2.1122 vocab=0.7578 token=0.8382 | lr=7.61e-05 | 6.16 steps/s | VRAM=27.86 Go [2026-07-18 13:21:12] step 816600/1000000 | loss=2.1777 vocab=0.8634 token=1.1681 | lr=7.61e-05 | 6.11 steps/s | VRAM=23.19 Go [2026-07-18 13:21:20] step 816650/1000000 | loss=2.1936 vocab=1.1088 token=1.0778 | lr=7.61e-05 | 6.16 steps/s | VRAM=31.03 Go [2026-07-18 13:21:28] step 816700/1000000 | loss=2.0811 vocab=1.1332 token=1.5948 | lr=7.61e-05 | 6.13 steps/s | VRAM=28.14 Go [2026-07-18 13:21:37] step 816750/1000000 | loss=2.1316 vocab=1.2309 token=1.3168 | lr=7.60e-05 | 6.16 steps/s | VRAM=27.51 Go [2026-07-18 13:21:45] step 816800/1000000 | loss=2.0891 vocab=1.1756 token=1.3253 | lr=7.60e-05 | 6.15 steps/s | VRAM=28.06 Go [2026-07-18 13:21:53] step 816850/1000000 | loss=2.1728 vocab=1.2359 token=1.5658 | lr=7.60e-05 | 6.14 steps/s | VRAM=28.77 Go [2026-07-18 13:22:01] step 816900/1000000 | loss=2.1817 vocab=1.2814 token=1.3021 | lr=7.60e-05 | 6.14 steps/s | VRAM=28.63 Go [2026-07-18 13:22:09] step 816950/1000000 | loss=2.1450 vocab=0.7560 token=1.1798 | lr=7.59e-05 | 6.16 steps/s | VRAM=32.51 Go [2026-07-18 13:22:17] step 817000/1000000 | loss=2.1955 vocab=0.7494 token=1.1378 | lr=7.59e-05 | 6.13 steps/s | VRAM=29.83 Go [2026-07-18 13:22:25] step 817050/1000000 | loss=2.0654 vocab=0.9055 token=1.3390 | lr=7.59e-05 | 6.14 steps/s | VRAM=29.22 Go [2026-07-18 13:22:34] step 817100/1000000 | loss=2.2079 vocab=0.7392 token=1.1269 | lr=7.59e-05 | 6.16 steps/s | VRAM=26.28 Go [2026-07-18 13:22:42] step 817150/1000000 | loss=2.1161 vocab=0.9862 token=1.6647 | lr=7.58e-05 | 6.18 steps/s | VRAM=24.63 Go [2026-07-18 13:22:50] step 817200/1000000 | loss=2.1722 vocab=1.0587 token=1.2663 | lr=7.58e-05 | 6.14 steps/s | VRAM=29.75 Go [2026-07-18 13:22:58] step 817250/1000000 | loss=2.1546 vocab=0.6956 token=0.8971 | lr=7.58e-05 | 6.16 steps/s | VRAM=31.83 Go [2026-07-18 13:23:06] step 817300/1000000 | loss=2.1923 vocab=0.9445 token=1.0358 | lr=7.58e-05 | 6.11 steps/s | VRAM=30.04 Go [2026-07-18 13:23:14] step 817350/1000000 | loss=2.1555 vocab=0.4762 token=1.3311 | lr=7.57e-05 | 6.14 steps/s | VRAM=25.05 Go [2026-07-18 13:23:22] step 817400/1000000 | loss=2.0228 vocab=0.8704 token=1.0681 | lr=7.57e-05 | 6.15 steps/s | VRAM=28.45 Go [2026-07-18 13:23:30] step 817450/1000000 | loss=2.1547 vocab=0.5952 token=1.3036 | lr=7.57e-05 | 6.15 steps/s | VRAM=29.43 Go [2026-07-18 13:23:39] step 817500/1000000 | loss=2.1608 vocab=0.6174 token=0.9870 | lr=7.57e-05 | 6.15 steps/s | VRAM=31.66 Go [2026-07-18 13:23:47] step 817550/1000000 | loss=2.1480 vocab=1.2977 token=1.4793 | lr=7.56e-05 | 6.15 steps/s | VRAM=26.69 Go [2026-07-18 13:23:55] step 817600/1000000 | loss=2.1152 vocab=0.5495 token=1.0115 | lr=7.56e-05 | 6.15 steps/s | VRAM=27.82 Go [2026-07-18 13:24:03] step 817650/1000000 | loss=2.0240 vocab=1.0277 token=1.2764 | lr=7.56e-05 | 6.14 steps/s | VRAM=30.11 Go [2026-07-18 13:24:11] step 817700/1000000 | loss=2.1413 vocab=0.6471 token=1.0179 | lr=7.56e-05 | 6.14 steps/s | VRAM=26.67 Go [2026-07-18 13:24:19] step 817750/1000000 | loss=2.0695 vocab=0.5913 token=1.1217 | lr=7.55e-05 | 6.14 steps/s | VRAM=28.74 Go [2026-07-18 13:24:27] step 817800/1000000 | loss=2.1526 vocab=0.7162 token=1.3182 | lr=7.55e-05 | 6.18 steps/s | VRAM=31.30 Go [2026-07-18 13:24:36] step 817850/1000000 | loss=2.1127 vocab=0.6810 token=1.4102 | lr=7.55e-05 | 6.15 steps/s | VRAM=28.99 Go [2026-07-18 13:24:44] step 817900/1000000 | loss=2.1718 vocab=0.9680 token=1.2577 | lr=7.55e-05 | 6.16 steps/s | VRAM=27.50 Go [2026-07-18 13:24:52] step 817950/1000000 | loss=2.1780 vocab=0.5355 token=0.9911 | lr=7.54e-05 | 6.14 steps/s | VRAM=26.31 Go [2026-07-18 13:25:00] step 818000/1000000 | loss=2.2399 vocab=0.9598 token=1.5549 | lr=7.54e-05 | 6.14 steps/s | VRAM=28.25 Go [2026-07-18 13:25:08] step 818050/1000000 | loss=2.1521 vocab=0.8815 token=1.2633 | lr=7.54e-05 | 6.15 steps/s | VRAM=29.42 Go [2026-07-18 13:25:16] step 818100/1000000 | loss=2.1929 vocab=0.5380 token=1.2538 | lr=7.54e-05 | 6.14 steps/s | VRAM=24.96 Go [2026-07-18 13:25:24] step 818150/1000000 | loss=2.0869 vocab=1.1405 token=1.3934 | lr=7.53e-05 | 6.13 steps/s | VRAM=29.94 Go [2026-07-18 13:25:33] step 818200/1000000 | loss=2.2335 vocab=0.8131 token=1.3873 | lr=7.53e-05 | 6.15 steps/s | VRAM=30.79 Go [2026-07-18 13:25:41] step 818250/1000000 | loss=2.0942 vocab=1.2707 token=1.5143 | lr=7.53e-05 | 6.15 steps/s | VRAM=26.11 Go [2026-07-18 13:25:49] step 818300/1000000 | loss=2.0826 vocab=1.4663 token=1.4081 | lr=7.53e-05 | 6.14 steps/s | VRAM=29.02 Go [2026-07-18 13:25:57] step 818350/1000000 | loss=2.1401 vocab=0.8242 token=1.6604 | lr=7.52e-05 | 6.16 steps/s | VRAM=31.25 Go [2026-07-18 13:26:05] step 818400/1000000 | loss=2.0781 vocab=0.7902 token=1.1189 | lr=7.52e-05 | 6.17 steps/s | VRAM=24.19 Go [2026-07-18 13:26:13] step 818450/1000000 | loss=2.2440 vocab=0.9546 token=1.5129 | lr=7.52e-05 | 6.17 steps/s | VRAM=31.02 Go [2026-07-18 13:26:21] step 818500/1000000 | loss=2.0271 vocab=0.9742 token=1.2355 | lr=7.52e-05 | 6.16 steps/s | VRAM=28.09 Go [2026-07-18 13:26:29] step 818550/1000000 | loss=2.0728 vocab=0.8523 token=1.0770 | lr=7.51e-05 | 6.16 steps/s | VRAM=28.65 Go [2026-07-18 13:26:37] step 818600/1000000 | loss=2.2435 vocab=1.1944 token=1.5653 | lr=7.51e-05 | 6.16 steps/s | VRAM=32.18 Go [2026-07-18 13:26:46] step 818650/1000000 | loss=2.2100 vocab=0.9312 token=1.1925 | lr=7.51e-05 | 6.16 steps/s | VRAM=29.95 Go [2026-07-18 13:26:54] step 818700/1000000 | loss=2.1174 vocab=0.8524 token=1.1183 | lr=7.51e-05 | 6.16 steps/s | VRAM=32.48 Go [2026-07-18 13:27:02] step 818750/1000000 | loss=2.0852 vocab=1.0667 token=1.2545 | lr=7.50e-05 | 6.14 steps/s | VRAM=31.91 Go [2026-07-18 13:27:10] step 818800/1000000 | loss=2.1401 vocab=0.7211 token=1.2996 | lr=7.50e-05 | 6.14 steps/s | VRAM=24.95 Go [2026-07-18 13:27:18] step 818850/1000000 | loss=2.1223 vocab=0.6279 token=1.3148 | lr=7.50e-05 | 6.18 steps/s | VRAM=28.64 Go [2026-07-18 13:27:26] step 818900/1000000 | loss=2.2281 vocab=1.0223 token=1.3933 | lr=7.50e-05 | 6.14 steps/s | VRAM=26.16 Go [2026-07-18 13:27:34] step 818950/1000000 | loss=2.1464 vocab=0.8213 token=1.1549 | lr=7.49e-05 | 6.17 steps/s | VRAM=28.27 Go [2026-07-18 13:27:42] step 819000/1000000 | loss=2.1647 vocab=0.5895 token=1.0451 | lr=7.49e-05 | 6.12 steps/s | VRAM=25.72 Go [2026-07-18 13:27:51] step 819050/1000000 | loss=2.1159 vocab=0.7751 token=1.1847 | lr=7.49e-05 | 6.14 steps/s | VRAM=28.59 Go [2026-07-18 13:27:59] step 819100/1000000 | loss=2.0217 vocab=0.4949 token=1.0568 | lr=7.49e-05 | 6.17 steps/s | VRAM=32.71 Go [2026-07-18 13:28:07] step 819150/1000000 | loss=2.0888 vocab=0.6994 token=1.3462 | lr=7.48e-05 | 6.14 steps/s | VRAM=28.84 Go [2026-07-18 13:28:15] step 819200/1000000 | loss=2.1243 vocab=0.6846 token=1.6638 | lr=7.48e-05 | 6.15 steps/s | VRAM=24.87 Go [2026-07-18 13:28:23] step 819250/1000000 | loss=2.1615 vocab=0.4498 token=1.4427 | lr=7.48e-05 | 6.16 steps/s | VRAM=29.84 Go [2026-07-18 13:28:31] step 819300/1000000 | loss=2.1530 vocab=0.7246 token=1.2672 | lr=7.48e-05 | 6.12 steps/s | VRAM=32.11 Go [2026-07-18 13:28:39] step 819350/1000000 | loss=2.1371 vocab=0.5955 token=1.2180 | lr=7.48e-05 | 6.12 steps/s | VRAM=26.46 Go [2026-07-18 13:28:48] step 819400/1000000 | loss=2.1834 vocab=0.4867 token=1.0138 | lr=7.47e-05 | 6.14 steps/s | VRAM=28.18 Go [2026-07-18 13:28:56] step 819450/1000000 | loss=2.1569 vocab=0.8996 token=1.3570 | lr=7.47e-05 | 6.11 steps/s | VRAM=28.93 Go [2026-07-18 13:29:04] step 819500/1000000 | loss=2.0792 vocab=1.1587 token=1.4121 | lr=7.47e-05 | 6.11 steps/s | VRAM=25.62 Go [2026-07-18 13:29:12] step 819550/1000000 | loss=2.1202 vocab=0.8537 token=1.6157 | lr=7.47e-05 | 6.12 steps/s | VRAM=28.44 Go [2026-07-18 13:29:20] step 819600/1000000 | loss=2.2814 vocab=0.6127 token=1.1610 | lr=7.46e-05 | 6.15 steps/s | VRAM=32.12 Go [2026-07-18 13:29:28] step 819650/1000000 | loss=2.1999 vocab=1.1963 token=1.3735 | lr=7.46e-05 | 6.15 steps/s | VRAM=32.65 Go [2026-07-18 13:29:37] step 819700/1000000 | loss=2.0009 vocab=0.7234 token=1.0896 | lr=7.46e-05 | 6.12 steps/s | VRAM=26.45 Go [2026-07-18 13:29:45] step 819750/1000000 | loss=2.2181 vocab=1.1918 token=1.4379 | lr=7.46e-05 | 6.13 steps/s | VRAM=31.82 Go [2026-07-18 13:29:53] step 819800/1000000 | loss=2.1586 vocab=0.8236 token=1.0006 | lr=7.45e-05 | 6.11 steps/s | VRAM=25.90 Go [2026-07-18 13:30:01] step 819850/1000000 | loss=2.2146 vocab=1.0507 token=1.2548 | lr=7.45e-05 | 6.13 steps/s | VRAM=26.10 Go [2026-07-18 13:30:09] step 819900/1000000 | loss=2.1025 vocab=0.8009 token=1.1607 | lr=7.45e-05 | 6.12 steps/s | VRAM=32.61 Go [2026-07-18 13:30:17] step 819950/1000000 | loss=2.0909 vocab=0.9547 token=1.3300 | lr=7.45e-05 | 6.14 steps/s | VRAM=28.79 Go [2026-07-18 13:30:26] step 820000/1000000 | loss=2.1130 vocab=1.2041 token=1.5704 | lr=7.44e-05 | 6.15 steps/s | VRAM=29.01 Go [2026-07-18 13:30:34] step 820050/1000000 | loss=2.1486 vocab=0.5893 token=1.1035 | lr=7.44e-05 | 6.11 steps/s | VRAM=28.44 Go [2026-07-18 13:30:42] step 820100/1000000 | loss=2.1281 vocab=0.8065 token=0.8830 | lr=7.44e-05 | 6.11 steps/s | VRAM=27.62 Go [2026-07-18 13:30:50] step 820150/1000000 | loss=2.2106 vocab=0.5478 token=0.9506 | lr=7.44e-05 | 6.11 steps/s | VRAM=29.18 Go [2026-07-18 13:30:58] step 820200/1000000 | loss=2.1327 vocab=0.3936 token=0.8479 | lr=7.43e-05 | 6.11 steps/s | VRAM=27.83 Go [2026-07-18 13:31:06] step 820250/1000000 | loss=2.1213 vocab=1.0115 token=1.3410 | lr=7.43e-05 | 6.14 steps/s | VRAM=26.36 Go [2026-07-18 13:31:15] step 820300/1000000 | loss=2.2905 vocab=1.1640 token=1.5043 | lr=7.43e-05 | 6.13 steps/s | VRAM=28.38 Go [2026-07-18 13:31:23] step 820350/1000000 | loss=2.1285 vocab=0.8291 token=1.3030 | lr=7.43e-05 | 6.14 steps/s | VRAM=28.86 Go [2026-07-18 13:31:31] step 820400/1000000 | loss=2.1299 vocab=1.2950 token=1.3853 | lr=7.42e-05 | 6.14 steps/s | VRAM=25.03 Go [2026-07-18 13:31:39] step 820450/1000000 | loss=2.1519 vocab=0.8576 token=1.3696 | lr=7.42e-05 | 6.16 steps/s | VRAM=28.22 Go [2026-07-18 13:31:47] step 820500/1000000 | loss=2.1252 vocab=1.4111 token=1.9652 | lr=7.42e-05 | 6.15 steps/s | VRAM=25.69 Go [2026-07-18 13:31:55] step 820550/1000000 | loss=2.2353 vocab=0.6850 token=1.0864 | lr=7.42e-05 | 6.15 steps/s | VRAM=32.02 Go [2026-07-18 13:32:03] step 820600/1000000 | loss=2.1900 vocab=0.7324 token=1.3206 | lr=7.41e-05 | 6.12 steps/s | VRAM=29.78 Go [2026-07-18 13:32:11] step 820650/1000000 | loss=2.1313 vocab=0.9807 token=1.5195 | lr=7.41e-05 | 6.16 steps/s | VRAM=28.29 Go [2026-07-18 13:32:20] step 820700/1000000 | loss=2.0714 vocab=0.5726 token=0.6286 | lr=7.41e-05 | 6.17 steps/s | VRAM=28.44 Go [2026-07-18 13:32:28] step 820750/1000000 | loss=2.0941 vocab=0.7593 token=0.8217 | lr=7.41e-05 | 6.17 steps/s | VRAM=28.95 Go [2026-07-18 13:32:36] step 820800/1000000 | loss=2.1990 vocab=1.0414 token=1.4955 | lr=7.40e-05 | 6.16 steps/s | VRAM=32.33 Go [2026-07-18 13:32:44] step 820850/1000000 | loss=2.0786 vocab=0.5802 token=0.9653 | lr=7.40e-05 | 6.13 steps/s | VRAM=28.13 Go [2026-07-18 13:32:52] step 820900/1000000 | loss=2.2198 vocab=0.9543 token=1.2455 | lr=7.40e-05 | 6.13 steps/s | VRAM=33.02 Go [2026-07-18 13:33:00] step 820950/1000000 | loss=1.9504 vocab=0.5321 token=1.2254 | lr=7.40e-05 | 6.16 steps/s | VRAM=28.99 Go [2026-07-18 13:33:08] step 821000/1000000 | loss=2.1818 vocab=0.7375 token=0.8097 | lr=7.39e-05 | 6.11 steps/s | VRAM=26.37 Go [2026-07-18 13:33:17] step 821050/1000000 | loss=2.2114 vocab=1.2509 token=1.0622 | lr=7.39e-05 | 6.16 steps/s | VRAM=31.49 Go [2026-07-18 13:33:25] step 821100/1000000 | loss=2.1240 vocab=0.8697 token=1.3318 | lr=7.39e-05 | 6.11 steps/s | VRAM=26.37 Go [2026-07-18 13:33:33] step 821150/1000000 | loss=2.2075 vocab=1.7779 token=1.5885 | lr=7.38e-05 | 6.15 steps/s | VRAM=24.89 Go [2026-07-18 13:33:41] step 821200/1000000 | loss=2.0681 vocab=1.1553 token=1.3219 | lr=7.38e-05 | 6.15 steps/s | VRAM=26.42 Go [2026-07-18 13:33:49] step 821250/1000000 | loss=2.2805 vocab=0.8459 token=1.4567 | lr=7.38e-05 | 6.09 steps/s | VRAM=26.47 Go [2026-07-18 13:33:57] step 821300/1000000 | loss=2.1393 vocab=0.5696 token=0.9692 | lr=7.38e-05 | 6.13 steps/s | VRAM=31.56 Go [2026-07-18 13:34:06] step 821350/1000000 | loss=2.1469 vocab=1.2619 token=1.6608 | lr=7.37e-05 | 6.10 steps/s | VRAM=29.16 Go [2026-07-18 13:34:14] step 821400/1000000 | loss=2.0698 vocab=0.9643 token=1.3021 | lr=7.37e-05 | 6.11 steps/s | VRAM=29.07 Go [2026-07-18 13:34:22] step 821450/1000000 | loss=2.0725 vocab=0.6377 token=0.8463 | lr=7.37e-05 | 6.15 steps/s | VRAM=28.70 Go [2026-07-18 13:34:30] step 821500/1000000 | loss=2.2775 vocab=0.6437 token=1.4309 | lr=7.37e-05 | 6.16 steps/s | VRAM=31.43 Go [2026-07-18 13:34:38] step 821550/1000000 | loss=2.1612 vocab=0.5635 token=1.4530 | lr=7.36e-05 | 6.12 steps/s | VRAM=25.84 Go [2026-07-18 13:34:46] step 821600/1000000 | loss=2.2003 vocab=0.9618 token=1.3172 | lr=7.36e-05 | 6.14 steps/s | VRAM=27.66 Go [2026-07-18 13:34:54] step 821650/1000000 | loss=2.2517 vocab=1.0727 token=1.6379 | lr=7.36e-05 | 6.12 steps/s | VRAM=31.34 Go [2026-07-18 13:35:03] step 821700/1000000 | loss=2.0532 vocab=0.9090 token=1.1788 | lr=7.36e-05 | 6.13 steps/s | VRAM=28.14 Go [2026-07-18 13:35:11] step 821750/1000000 | loss=2.1441 vocab=0.7873 token=1.5554 | lr=7.35e-05 | 6.15 steps/s | VRAM=28.84 Go [2026-07-18 13:35:19] step 821800/1000000 | loss=2.2673 vocab=0.8432 token=1.3493 | lr=7.35e-05 | 6.13 steps/s | VRAM=26.33 Go [2026-07-18 13:35:27] step 821850/1000000 | loss=2.1683 vocab=1.2668 token=1.2632 | lr=7.35e-05 | 6.13 steps/s | VRAM=29.13 Go [2026-07-18 13:35:35] step 821900/1000000 | loss=2.1958 vocab=0.7474 token=1.4635 | lr=7.35e-05 | 6.16 steps/s | VRAM=31.20 Go [2026-07-18 13:35:43] step 821950/1000000 | loss=2.2319 vocab=0.9055 token=1.2711 | lr=7.34e-05 | 6.13 steps/s | VRAM=32.69 Go [2026-07-18 13:35:52] step 822000/1000000 | loss=2.1057 vocab=0.7895 token=1.1725 | lr=7.34e-05 | 6.12 steps/s | VRAM=28.84 Go [2026-07-18 13:36:00] step 822050/1000000 | loss=2.1574 vocab=0.9011 token=1.2591 | lr=7.34e-05 | 6.12 steps/s | VRAM=25.82 Go [2026-07-18 13:36:08] step 822100/1000000 | loss=2.1174 vocab=1.0557 token=1.4225 | lr=7.34e-05 | 6.13 steps/s | VRAM=28.90 Go [2026-07-18 13:36:16] step 822150/1000000 | loss=2.2450 vocab=0.8998 token=1.3059 | lr=7.33e-05 | 6.12 steps/s | VRAM=28.83 Go [2026-07-18 13:36:24] step 822200/1000000 | loss=2.1238 vocab=0.8050 token=1.3041 | lr=7.33e-05 | 6.16 steps/s | VRAM=25.18 Go [2026-07-18 13:36:32] step 822250/1000000 | loss=2.1444 vocab=0.9606 token=1.5782 | lr=7.33e-05 | 6.13 steps/s | VRAM=31.63 Go [2026-07-18 13:36:40] step 822300/1000000 | loss=2.1387 vocab=1.0403 token=1.4593 | lr=7.33e-05 | 6.10 steps/s | VRAM=29.23 Go [2026-07-18 13:36:49] step 822350/1000000 | loss=2.1416 vocab=0.9899 token=1.1312 | lr=7.32e-05 | 6.11 steps/s | VRAM=28.80 Go [2026-07-18 13:36:57] step 822400/1000000 | loss=2.2155 vocab=1.0479 token=1.5283 | lr=7.32e-05 | 6.08 steps/s | VRAM=26.19 Go [2026-07-18 13:37:05] step 822450/1000000 | loss=2.1606 vocab=0.9302 token=1.3944 | lr=7.32e-05 | 6.12 steps/s | VRAM=32.52 Go [2026-07-18 13:37:13] step 822500/1000000 | loss=2.2732 vocab=1.6790 token=1.4567 | lr=7.32e-05 | 6.08 steps/s | VRAM=28.67 Go [2026-07-18 13:37:21] step 822550/1000000 | loss=2.1693 vocab=1.1004 token=1.1365 | lr=7.31e-05 | 6.13 steps/s | VRAM=28.56 Go [2026-07-18 13:37:30] step 822600/1000000 | loss=2.1071 vocab=1.0753 token=1.2117 | lr=7.31e-05 | 6.13 steps/s | VRAM=26.12 Go [2026-07-18 13:37:38] step 822650/1000000 | loss=2.1990 vocab=0.7645 token=0.9038 | lr=7.31e-05 | 6.16 steps/s | VRAM=31.65 Go [2026-07-18 13:37:46] step 822700/1000000 | loss=2.2426 vocab=0.3282 token=1.2213 | lr=7.31e-05 | 6.10 steps/s | VRAM=30.73 Go [2026-07-18 13:37:54] step 822750/1000000 | loss=2.1704 vocab=0.7100 token=1.4712 | lr=7.30e-05 | 6.13 steps/s | VRAM=28.99 Go [2026-07-18 13:38:02] step 822800/1000000 | loss=2.1133 vocab=1.0897 token=1.1352 | lr=7.30e-05 | 6.12 steps/s | VRAM=26.10 Go [2026-07-18 13:38:10] step 822850/1000000 | loss=2.0401 vocab=0.6348 token=1.3361 | lr=7.30e-05 | 6.09 steps/s | VRAM=29.27 Go [2026-07-18 13:38:19] step 822900/1000000 | loss=2.1393 vocab=0.8432 token=1.5176 | lr=7.30e-05 | 6.14 steps/s | VRAM=31.29 Go [2026-07-18 13:38:27] step 822950/1000000 | loss=2.0693 vocab=1.0763 token=1.3953 | lr=7.29e-05 | 6.14 steps/s | VRAM=27.76 Go [2026-07-18 13:38:35] step 823000/1000000 | loss=2.1831 vocab=0.4185 token=1.0276 | lr=7.29e-05 | 6.11 steps/s | VRAM=29.14 Go [2026-07-18 13:38:43] step 823050/1000000 | loss=2.0970 vocab=0.8491 token=1.1771 | lr=7.29e-05 | 6.07 steps/s | VRAM=25.77 Go [2026-07-18 13:38:51] step 823100/1000000 | loss=2.1533 vocab=0.4735 token=1.2914 | lr=7.29e-05 | 6.15 steps/s | VRAM=27.99 Go [2026-07-18 13:38:59] step 823150/1000000 | loss=2.1426 vocab=1.1421 token=1.2918 | lr=7.28e-05 | 6.13 steps/s | VRAM=26.21 Go [2026-07-18 13:39:08] step 823200/1000000 | loss=2.1739 vocab=0.4268 token=0.8791 | lr=7.28e-05 | 6.15 steps/s | VRAM=30.75 Go [2026-07-18 13:39:16] step 823250/1000000 | loss=2.1126 vocab=0.8336 token=1.1896 | lr=7.28e-05 | 6.16 steps/s | VRAM=28.49 Go [2026-07-18 13:39:24] step 823300/1000000 | loss=2.1826 vocab=1.2647 token=1.7146 | lr=7.28e-05 | 6.12 steps/s | VRAM=25.99 Go [2026-07-18 13:39:32] step 823350/1000000 | loss=2.1579 vocab=0.7884 token=1.3683 | lr=7.27e-05 | 6.14 steps/s | VRAM=24.79 Go [2026-07-18 13:39:40] step 823400/1000000 | loss=2.1119 vocab=1.2400 token=1.4979 | lr=7.27e-05 | 6.13 steps/s | VRAM=32.51 Go [2026-07-18 13:39:48] step 823450/1000000 | loss=2.1143 vocab=0.5874 token=1.2981 | lr=7.27e-05 | 6.14 steps/s | VRAM=29.22 Go [2026-07-18 13:39:56] step 823500/1000000 | loss=2.1581 vocab=1.1167 token=1.6094 | lr=7.27e-05 | 6.16 steps/s | VRAM=32.48 Go [2026-07-18 13:40:05] step 823550/1000000 | loss=2.1446 vocab=0.8610 token=1.4051 | lr=7.26e-05 | 6.15 steps/s | VRAM=32.12 Go [2026-07-18 13:40:13] step 823600/1000000 | loss=2.0079 vocab=0.5253 token=1.0936 | lr=7.26e-05 | 6.14 steps/s | VRAM=32.24 Go [2026-07-18 13:40:21] step 823650/1000000 | loss=2.1669 vocab=0.8807 token=1.2299 | lr=7.26e-05 | 6.13 steps/s | VRAM=32.25 Go [2026-07-18 13:40:29] step 823700/1000000 | loss=2.0672 vocab=1.2550 token=1.3697 | lr=7.26e-05 | 6.13 steps/s | VRAM=29.16 Go [2026-07-18 13:40:37] step 823750/1000000 | loss=2.1300 vocab=0.5751 token=1.3167 | lr=7.25e-05 | 6.09 steps/s | VRAM=26.56 Go [2026-07-18 13:40:45] step 823800/1000000 | loss=2.0766 vocab=0.8842 token=1.0145 | lr=7.25e-05 | 6.09 steps/s | VRAM=28.09 Go [2026-07-18 13:40:54] step 823850/1000000 | loss=2.1686 vocab=1.4113 token=1.3074 | lr=7.25e-05 | 6.16 steps/s | VRAM=27.76 Go [2026-07-18 13:41:02] step 823900/1000000 | loss=2.1743 vocab=1.3011 token=1.0899 | lr=7.25e-05 | 6.13 steps/s | VRAM=28.63 Go [2026-07-18 13:41:10] step 823950/1000000 | loss=2.1606 vocab=0.8308 token=1.3224 | lr=7.24e-05 | 6.13 steps/s | VRAM=25.41 Go [2026-07-18 13:41:18] step 824000/1000000 | loss=2.1060 vocab=1.1319 token=1.6357 | lr=7.24e-05 | 6.12 steps/s | VRAM=29.61 Go [2026-07-18 13:41:26] step 824050/1000000 | loss=2.1951 vocab=0.9673 token=1.1099 | lr=7.24e-05 | 6.15 steps/s | VRAM=24.30 Go [2026-07-18 13:41:34] step 824100/1000000 | loss=2.2136 vocab=0.8263 token=1.5297 | lr=7.24e-05 | 6.14 steps/s | VRAM=28.70 Go [2026-07-18 13:41:43] step 824150/1000000 | loss=2.2096 vocab=1.2985 token=1.2772 | lr=7.23e-05 | 6.11 steps/s | VRAM=29.82 Go [2026-07-18 13:41:51] step 824200/1000000 | loss=2.1303 vocab=0.8858 token=1.4522 | lr=7.23e-05 | 6.12 steps/s | VRAM=28.84 Go [2026-07-18 13:41:59] step 824250/1000000 | loss=2.0857 vocab=0.8321 token=1.5189 | lr=7.23e-05 | 6.13 steps/s | VRAM=28.93 Go [2026-07-18 13:42:07] step 824300/1000000 | loss=2.0930 vocab=0.4283 token=0.9172 | lr=7.22e-05 | 6.16 steps/s | VRAM=32.44 Go [2026-07-18 13:42:15] step 824350/1000000 | loss=2.1267 vocab=0.9919 token=1.5250 | lr=7.22e-05 | 6.12 steps/s | VRAM=24.96 Go [2026-07-18 13:42:23] step 824400/1000000 | loss=2.2003 vocab=1.1158 token=1.6430 | lr=7.22e-05 | 6.09 steps/s | VRAM=30.23 Go [2026-07-18 13:42:31] step 824450/1000000 | loss=2.1601 vocab=0.5634 token=1.4628 | lr=7.22e-05 | 6.15 steps/s | VRAM=31.99 Go [2026-07-18 13:42:40] step 824500/1000000 | loss=2.1517 vocab=0.7460 token=1.2406 | lr=7.21e-05 | 6.11 steps/s | VRAM=32.47 Go [2026-07-18 13:42:48] step 824550/1000000 | loss=2.0204 vocab=0.5810 token=1.1351 | lr=7.21e-05 | 6.14 steps/s | VRAM=28.06 Go [2026-07-18 13:42:56] step 824600/1000000 | loss=2.1354 vocab=1.3008 token=1.2672 | lr=7.21e-05 | 6.12 steps/s | VRAM=28.53 Go [2026-07-18 13:43:04] step 824650/1000000 | loss=2.1392 vocab=1.0179 token=1.1851 | lr=7.21e-05 | 6.14 steps/s | VRAM=24.96 Go [2026-07-18 13:43:12] step 824700/1000000 | loss=2.1929 vocab=0.4916 token=1.2402 | lr=7.20e-05 | 6.12 steps/s | VRAM=27.60 Go [2026-07-18 13:43:20] step 824750/1000000 | loss=2.0805 vocab=0.9951 token=1.2755 | lr=7.20e-05 | 6.17 steps/s | VRAM=31.51 Go [2026-07-18 13:43:29] step 824800/1000000 | loss=2.1591 vocab=1.0826 token=1.7982 | lr=7.20e-05 | 6.09 steps/s | VRAM=30.58 Go [2026-07-18 13:43:37] step 824850/1000000 | loss=2.1395 vocab=0.8755 token=1.4348 | lr=7.20e-05 | 6.14 steps/s | VRAM=31.01 Go [2026-07-18 13:43:45] step 824900/1000000 | loss=2.1686 vocab=0.7868 token=1.2071 | lr=7.19e-05 | 6.10 steps/s | VRAM=29.09 Go [2026-07-18 13:43:53] step 824950/1000000 | loss=2.1939 vocab=0.8492 token=1.3176 | lr=7.19e-05 | 6.15 steps/s | VRAM=28.38 Go [2026-07-18 13:44:01] step 825000/1000000 | loss=2.1583 vocab=1.4109 token=1.3442 | lr=7.19e-05 | 6.10 steps/s | VRAM=30.31 Go [2026-07-18 13:44:09] step 825050/1000000 | loss=2.1917 vocab=1.1123 token=1.5042 | lr=7.19e-05 | 6.15 steps/s | VRAM=32.17 Go [2026-07-18 13:44:18] step 825100/1000000 | loss=2.0880 vocab=0.7281 token=1.2075 | lr=7.18e-05 | 6.09 steps/s | VRAM=29.10 Go [2026-07-18 13:44:26] step 825150/1000000 | loss=2.1730 vocab=0.9729 token=1.1182 | lr=7.18e-05 | 6.15 steps/s | VRAM=28.94 Go [2026-07-18 13:44:34] step 825200/1000000 | loss=2.1964 vocab=0.4906 token=0.8554 | lr=7.18e-05 | 6.14 steps/s | VRAM=28.57 Go [2026-07-18 13:44:42] step 825250/1000000 | loss=2.1716 vocab=0.8073 token=0.9944 | lr=7.18e-05 | 6.14 steps/s | VRAM=29.27 Go [2026-07-18 13:44:50] step 825300/1000000 | loss=2.1270 vocab=0.6886 token=0.9778 | lr=7.17e-05 | 6.14 steps/s | VRAM=32.47 Go [2026-07-18 13:44:58] step 825350/1000000 | loss=2.0829 vocab=1.0059 token=1.3391 | lr=7.17e-05 | 6.08 steps/s | VRAM=26.61 Go [2026-07-18 13:45:07] step 825400/1000000 | loss=2.1332 vocab=0.7463 token=1.1000 | lr=7.17e-05 | 6.12 steps/s | VRAM=32.01 Go [2026-07-18 13:45:15] step 825450/1000000 | loss=2.2483 vocab=0.7061 token=1.2138 | lr=7.17e-05 | 6.13 steps/s | VRAM=32.29 Go [2026-07-18 13:45:23] step 825500/1000000 | loss=2.1899 vocab=0.8206 token=1.4789 | lr=7.16e-05 | 6.12 steps/s | VRAM=32.66 Go [2026-07-18 13:45:31] step 825550/1000000 | loss=2.0233 vocab=1.2215 token=1.4823 | lr=7.16e-05 | 6.11 steps/s | VRAM=26.27 Go [2026-07-18 13:45:39] step 825600/1000000 | loss=2.1737 vocab=0.6463 token=1.0533 | lr=7.16e-05 | 6.12 steps/s | VRAM=28.85 Go [2026-07-18 13:45:47] step 825650/1000000 | loss=2.2775 vocab=1.5736 token=1.2764 | lr=7.16e-05 | 6.11 steps/s | VRAM=28.17 Go [2026-07-18 13:45:56] step 825700/1000000 | loss=2.1045 vocab=0.8981 token=1.6099 | lr=7.15e-05 | 6.13 steps/s | VRAM=26.99 Go [2026-07-18 13:46:04] step 825750/1000000 | loss=2.0201 vocab=0.8227 token=1.3820 | lr=7.15e-05 | 6.13 steps/s | VRAM=32.23 Go [2026-07-18 13:46:12] step 825800/1000000 | loss=2.1587 vocab=0.5438 token=1.2096 | lr=7.15e-05 | 6.14 steps/s | VRAM=28.15 Go [2026-07-18 13:46:20] step 825850/1000000 | loss=2.1549 vocab=1.1547 token=1.3720 | lr=7.15e-05 | 6.14 steps/s | VRAM=27.73 Go [2026-07-18 13:46:28] step 825900/1000000 | loss=2.2140 vocab=1.2275 token=1.1345 | lr=7.14e-05 | 6.14 steps/s | VRAM=25.94 Go [2026-07-18 13:46:36] step 825950/1000000 | loss=2.1315 vocab=1.2086 token=1.5884 | lr=7.14e-05 | 6.11 steps/s | VRAM=32.40 Go [2026-07-18 13:46:45] step 826000/1000000 | loss=2.0701 vocab=1.0599 token=1.4014 | lr=7.14e-05 | 6.10 steps/s | VRAM=28.58 Go [2026-07-18 13:46:53] step 826050/1000000 | loss=2.1409 vocab=0.6672 token=1.2471 | lr=7.13e-05 | 6.12 steps/s | VRAM=27.94 Go [2026-07-18 13:47:01] step 826100/1000000 | loss=2.1363 vocab=0.9112 token=1.3464 | lr=7.13e-05 | 6.15 steps/s | VRAM=28.54 Go [2026-07-18 13:47:09] step 826150/1000000 | loss=2.1019 vocab=0.4951 token=0.8696 | lr=7.13e-05 | 6.10 steps/s | VRAM=28.17 Go [2026-07-18 13:47:17] step 826200/1000000 | loss=2.1844 vocab=1.5543 token=1.1441 | lr=7.13e-05 | 6.10 steps/s | VRAM=28.42 Go [2026-07-18 13:47:25] step 826250/1000000 | loss=2.0329 vocab=0.8854 token=1.1981 | lr=7.12e-05 | 6.12 steps/s | VRAM=25.27 Go [2026-07-18 13:47:34] step 826300/1000000 | loss=2.1559 vocab=0.9154 token=1.1612 | lr=7.12e-05 | 6.13 steps/s | VRAM=29.37 Go [2026-07-18 13:47:42] step 826350/1000000 | loss=2.0895 vocab=0.9415 token=1.3331 | lr=7.12e-05 | 6.15 steps/s | VRAM=31.86 Go [2026-07-18 13:47:50] step 826400/1000000 | loss=2.1141 vocab=0.8143 token=0.9379 | lr=7.12e-05 | 6.14 steps/s | VRAM=32.80 Go [2026-07-18 13:47:58] step 826450/1000000 | loss=2.1958 vocab=0.9176 token=1.6847 | lr=7.11e-05 | 6.08 steps/s | VRAM=26.57 Go [2026-07-18 13:48:06] step 826500/1000000 | loss=2.1895 vocab=0.8636 token=1.3042 | lr=7.11e-05 | 6.13 steps/s | VRAM=32.81 Go [2026-07-18 13:48:14] step 826550/1000000 | loss=2.1341 vocab=0.9970 token=1.0823 | lr=7.11e-05 | 6.16 steps/s | VRAM=31.30 Go [2026-07-18 13:48:23] step 826600/1000000 | loss=2.0934 vocab=0.9908 token=1.4027 | lr=7.11e-05 | 6.11 steps/s | VRAM=28.34 Go [2026-07-18 13:48:31] step 826650/1000000 | loss=2.0368 vocab=0.6711 token=0.5962 | lr=7.10e-05 | 6.14 steps/s | VRAM=28.74 Go [2026-07-18 13:48:39] step 826700/1000000 | loss=2.3101 vocab=0.5758 token=0.9543 | lr=7.10e-05 | 6.11 steps/s | VRAM=32.90 Go [2026-07-18 13:48:47] step 826750/1000000 | loss=2.0859 vocab=0.8528 token=1.7074 | lr=7.10e-05 | 6.11 steps/s | VRAM=29.39 Go [2026-07-18 13:48:55] step 826800/1000000 | loss=2.2060 vocab=1.4231 token=1.4284 | lr=7.10e-05 | 6.10 steps/s | VRAM=28.71 Go [2026-07-18 13:49:03] step 826850/1000000 | loss=2.1827 vocab=0.8873 token=1.0191 | lr=7.09e-05 | 6.15 steps/s | VRAM=30.74 Go [2026-07-18 13:49:12] step 826900/1000000 | loss=2.0893 vocab=0.7263 token=1.2395 | lr=7.09e-05 | 6.11 steps/s | VRAM=28.07 Go [2026-07-18 13:49:20] step 826950/1000000 | loss=2.1890 vocab=1.4285 token=1.2374 | lr=7.09e-05 | 6.12 steps/s | VRAM=32.40 Go [2026-07-18 13:49:28] step 827000/1000000 | loss=2.1118 vocab=0.2920 token=0.5678 | lr=7.09e-05 | 6.16 steps/s | VRAM=28.57 Go [2026-07-18 13:49:36] step 827050/1000000 | loss=2.1912 vocab=0.9879 token=1.4635 | lr=7.08e-05 | 6.15 steps/s | VRAM=31.33 Go [2026-07-18 13:49:44] step 827100/1000000 | loss=2.1142 vocab=0.5272 token=1.2182 | lr=7.08e-05 | 6.11 steps/s | VRAM=31.83 Go [2026-07-18 13:49:52] step 827150/1000000 | loss=2.1226 vocab=0.5878 token=1.2411 | lr=7.08e-05 | 6.11 steps/s | VRAM=26.43 Go [2026-07-18 13:50:01] step 827200/1000000 | loss=2.2113 vocab=0.6982 token=1.2402 | lr=7.08e-05 | 6.10 steps/s | VRAM=28.56 Go [2026-07-18 13:50:09] step 827250/1000000 | loss=2.1548 vocab=0.7878 token=0.7969 | lr=7.07e-05 | 6.16 steps/s | VRAM=31.12 Go [2026-07-18 13:50:17] step 827300/1000000 | loss=2.1690 vocab=1.1399 token=1.2459 | lr=7.07e-05 | 6.10 steps/s | VRAM=32.42 Go [2026-07-18 13:50:25] step 827350/1000000 | loss=2.1374 vocab=0.7237 token=1.2030 | lr=7.07e-05 | 6.15 steps/s | VRAM=26.58 Go [2026-07-18 13:50:33] step 827400/1000000 | loss=2.1683 vocab=0.8666 token=1.4787 | lr=7.06e-05 | 6.08 steps/s | VRAM=29.34 Go [2026-07-18 13:50:41] step 827450/1000000 | loss=2.2185 vocab=0.4800 token=1.2866 | lr=7.06e-05 | 6.14 steps/s | VRAM=30.05 Go [2026-07-18 13:50:50] step 827500/1000000 | loss=2.2259 vocab=1.1907 token=1.6814 | lr=7.06e-05 | 6.14 steps/s | VRAM=31.69 Go [2026-07-18 13:50:58] step 827550/1000000 | loss=2.0635 vocab=0.6146 token=1.2904 | lr=7.06e-05 | 6.12 steps/s | VRAM=28.97 Go [2026-07-18 13:51:06] step 827600/1000000 | loss=2.0470 vocab=0.5600 token=1.1378 | lr=7.05e-05 | 6.13 steps/s | VRAM=29.23 Go [2026-07-18 13:51:14] step 827650/1000000 | loss=2.1174 vocab=0.9282 token=1.1529 | lr=7.05e-05 | 6.11 steps/s | VRAM=28.53 Go [2026-07-18 13:51:22] step 827700/1000000 | loss=2.0927 vocab=0.7189 token=0.9599 | lr=7.05e-05 | 6.10 steps/s | VRAM=29.19 Go [2026-07-18 13:51:30] step 827750/1000000 | loss=2.0441 vocab=0.7950 token=1.5065 | lr=7.05e-05 | 6.13 steps/s | VRAM=28.99 Go [2026-07-18 13:51:39] step 827800/1000000 | loss=2.1839 vocab=1.0660 token=1.3260 | lr=7.04e-05 | 6.14 steps/s | VRAM=32.47 Go [2026-07-18 13:51:47] step 827850/1000000 | loss=2.0546 vocab=0.7485 token=1.2991 | lr=7.04e-05 | 6.14 steps/s | VRAM=25.78 Go [2026-07-18 13:51:55] step 827900/1000000 | loss=2.1337 vocab=1.3005 token=1.3392 | lr=7.04e-05 | 6.14 steps/s | VRAM=31.70 Go [2026-07-18 13:52:03] step 827950/1000000 | loss=2.1023 vocab=0.8473 token=0.8672 | lr=7.04e-05 | 6.11 steps/s | VRAM=28.20 Go [2026-07-18 13:52:11] step 828000/1000000 | loss=2.1556 vocab=0.7207 token=0.9389 | lr=7.03e-05 | 6.10 steps/s | VRAM=27.13 Go [2026-07-18 13:52:19] step 828050/1000000 | loss=2.1697 vocab=1.0731 token=1.2137 | lr=7.03e-05 | 6.14 steps/s | VRAM=28.47 Go [2026-07-18 13:52:27] step 828100/1000000 | loss=2.1208 vocab=1.2681 token=1.1439 | lr=7.03e-05 | 6.13 steps/s | VRAM=28.43 Go [2026-07-18 13:52:36] step 828150/1000000 | loss=2.1612 vocab=0.8759 token=1.2741 | lr=7.03e-05 | 6.15 steps/s | VRAM=24.43 Go [2026-07-18 13:52:44] step 828200/1000000 | loss=2.1918 vocab=0.8826 token=1.0594 | lr=7.02e-05 | 6.15 steps/s | VRAM=30.92 Go [2026-07-18 13:52:52] step 828250/1000000 | loss=2.2016 vocab=1.0970 token=2.1322 | lr=7.02e-05 | 6.14 steps/s | VRAM=31.52 Go [2026-07-18 13:53:00] step 828300/1000000 | loss=2.2428 vocab=0.6694 token=1.1976 | lr=7.02e-05 | 6.14 steps/s | VRAM=30.93 Go [2026-07-18 13:53:08] step 828350/1000000 | loss=2.1534 vocab=0.7536 token=0.9885 | lr=7.02e-05 | 6.13 steps/s | VRAM=31.88 Go [2026-07-18 13:53:16] step 828400/1000000 | loss=2.1372 vocab=0.7277 token=1.5585 | lr=7.01e-05 | 6.08 steps/s | VRAM=29.98 Go [2026-07-18 13:53:25] step 828450/1000000 | loss=2.1150 vocab=0.5802 token=0.9526 | lr=7.01e-05 | 6.15 steps/s | VRAM=31.64 Go [2026-07-18 13:53:33] step 828500/1000000 | loss=2.1387 vocab=0.8701 token=1.3125 | lr=7.01e-05 | 6.14 steps/s | VRAM=32.07 Go [2026-07-18 13:53:41] step 828550/1000000 | loss=2.1565 vocab=1.3592 token=1.2503 | lr=7.00e-05 | 6.09 steps/s | VRAM=29.77 Go [2026-07-18 13:53:49] step 828600/1000000 | loss=2.1139 vocab=1.1714 token=1.4380 | lr=7.00e-05 | 6.11 steps/s | VRAM=28.29 Go [2026-07-18 13:53:57] step 828650/1000000 | loss=2.1391 vocab=1.2238 token=1.2836 | lr=7.00e-05 | 6.12 steps/s | VRAM=28.69 Go [2026-07-18 13:54:05] step 828700/1000000 | loss=2.2085 vocab=0.6069 token=1.0410 | lr=7.00e-05 | 6.09 steps/s | VRAM=26.19 Go [2026-07-18 13:54:14] step 828750/1000000 | loss=2.1793 vocab=0.7365 token=0.6811 | lr=6.99e-05 | 6.13 steps/s | VRAM=28.27 Go [2026-07-18 13:54:22] step 828800/1000000 | loss=2.1325 vocab=0.8485 token=1.7818 | lr=6.99e-05 | 6.11 steps/s | VRAM=29.33 Go [2026-07-18 13:54:30] step 828850/1000000 | loss=2.1377 vocab=0.3282 token=0.8039 | lr=6.99e-05 | 6.15 steps/s | VRAM=31.42 Go [2026-07-18 13:54:38] step 828900/1000000 | loss=2.1803 vocab=1.0368 token=1.1941 | lr=6.99e-05 | 6.10 steps/s | VRAM=28.96 Go [2026-07-18 13:54:46] step 828950/1000000 | loss=2.2470 vocab=0.9598 token=1.5785 | lr=6.98e-05 | 6.16 steps/s | VRAM=31.76 Go [2026-07-18 13:54:54] step 829000/1000000 | loss=2.1221 vocab=0.6828 token=1.3037 | lr=6.98e-05 | 6.12 steps/s | VRAM=32.73 Go [2026-07-18 13:55:03] step 829050/1000000 | loss=2.1060 vocab=0.9604 token=1.3010 | lr=6.98e-05 | 6.12 steps/s | VRAM=25.77 Go [2026-07-18 13:55:11] step 829100/1000000 | loss=2.0325 vocab=1.0462 token=1.3642 | lr=6.98e-05 | 6.14 steps/s | VRAM=32.02 Go [2026-07-18 13:55:19] step 829150/1000000 | loss=2.0626 vocab=0.6930 token=1.1818 | lr=6.97e-05 | 6.13 steps/s | VRAM=29.10 Go [2026-07-18 13:55:27] step 829200/1000000 | loss=2.0867 vocab=0.9247 token=1.3392 | lr=6.97e-05 | 6.11 steps/s | VRAM=26.61 Go [2026-07-18 13:55:35] step 829250/1000000 | loss=2.0594 vocab=0.5635 token=1.1137 | lr=6.97e-05 | 6.11 steps/s | VRAM=28.71 Go [2026-07-18 13:55:43] step 829300/1000000 | loss=1.9942 vocab=0.6443 token=1.1955 | lr=6.97e-05 | 6.14 steps/s | VRAM=29.49 Go [2026-07-18 13:55:52] step 829350/1000000 | loss=2.1730 vocab=0.6755 token=1.4590 | lr=6.96e-05 | 6.13 steps/s | VRAM=27.74 Go [2026-07-18 13:56:00] step 829400/1000000 | loss=2.0505 vocab=1.1186 token=1.6819 | lr=6.96e-05 | 6.15 steps/s | VRAM=28.42 Go [2026-07-18 13:56:08] step 829450/1000000 | loss=2.2127 vocab=1.1696 token=1.5110 | lr=6.96e-05 | 6.12 steps/s | VRAM=28.80 Go [2026-07-18 13:56:16] step 829500/1000000 | loss=2.1110 vocab=0.5288 token=1.2734 | lr=6.95e-05 | 6.12 steps/s | VRAM=26.13 Go [2026-07-18 13:56:24] step 829550/1000000 | loss=2.0387 vocab=0.7026 token=1.1223 | lr=6.95e-05 | 6.14 steps/s | VRAM=24.59 Go [2026-07-18 13:56:32] step 829600/1000000 | loss=2.1300 vocab=0.7929 token=1.2466 | lr=6.95e-05 | 6.12 steps/s | VRAM=28.38 Go [2026-07-18 13:56:40] step 829650/1000000 | loss=2.2547 vocab=0.5277 token=1.0488 | lr=6.95e-05 | 6.11 steps/s | VRAM=27.98 Go [2026-07-18 13:56:49] step 829700/1000000 | loss=2.1384 vocab=1.2174 token=1.3146 | lr=6.94e-05 | 6.14 steps/s | VRAM=24.86 Go [2026-07-18 13:56:57] step 829750/1000000 | loss=2.1881 vocab=0.4254 token=0.9935 | lr=6.94e-05 | 6.14 steps/s | VRAM=25.76 Go [2026-07-18 13:57:05] step 829800/1000000 | loss=2.1146 vocab=0.6714 token=1.3278 | lr=6.94e-05 | 6.10 steps/s | VRAM=29.14 Go [2026-07-18 13:57:13] step 829850/1000000 | loss=2.0453 vocab=0.8798 token=1.5286 | lr=6.94e-05 | 6.10 steps/s | VRAM=28.36 Go [2026-07-18 13:57:21] step 829900/1000000 | loss=2.1139 vocab=0.6714 token=1.1344 | lr=6.93e-05 | 6.11 steps/s | VRAM=32.28 Go [2026-07-18 13:57:29] step 829950/1000000 | loss=2.0752 vocab=0.8001 token=1.2759 | lr=6.93e-05 | 6.13 steps/s | VRAM=31.76 Go [2026-07-18 13:57:38] step 830000/1000000 | loss=2.1217 vocab=0.7937 token=0.8079 | lr=6.93e-05 | 6.13 steps/s | VRAM=32.22 Go [2026-07-18 13:57:46] step 830050/1000000 | loss=2.0444 vocab=1.1407 token=1.3859 | lr=6.93e-05 | 6.08 steps/s | VRAM=26.05 Go [2026-07-18 13:57:54] step 830100/1000000 | loss=2.1967 vocab=0.8720 token=1.2829 | lr=6.92e-05 | 6.13 steps/s | VRAM=28.02 Go [2026-07-18 13:58:02] step 830150/1000000 | loss=2.1294 vocab=0.6782 token=1.1880 | lr=6.92e-05 | 6.12 steps/s | VRAM=32.24 Go [2026-07-18 13:58:10] step 830200/1000000 | loss=2.1052 vocab=0.7895 token=1.1459 | lr=6.92e-05 | 6.10 steps/s | VRAM=28.55 Go [2026-07-18 13:58:19] step 830250/1000000 | loss=2.1681 vocab=1.1343 token=1.2698 | lr=6.92e-05 | 6.13 steps/s | VRAM=29.16 Go [2026-07-18 13:58:27] step 830300/1000000 | loss=2.2336 vocab=0.9390 token=1.2917 | lr=6.91e-05 | 6.13 steps/s | VRAM=29.76 Go [2026-07-18 13:58:35] step 830350/1000000 | loss=2.0653 vocab=0.5275 token=0.9762 | lr=6.91e-05 | 6.11 steps/s | VRAM=28.94 Go [2026-07-18 13:58:43] step 830400/1000000 | loss=2.1169 vocab=1.1376 token=1.4874 | lr=6.91e-05 | 6.12 steps/s | VRAM=27.83 Go [2026-07-18 13:58:51] step 830450/1000000 | loss=2.1035 vocab=1.0768 token=1.5068 | lr=6.90e-05 | 6.13 steps/s | VRAM=31.70 Go [2026-07-18 13:58:59] step 830500/1000000 | loss=2.1253 vocab=0.4275 token=0.7133 | lr=6.90e-05 | 6.12 steps/s | VRAM=26.03 Go [2026-07-18 13:59:08] step 830550/1000000 | loss=2.2491 vocab=1.7465 token=1.4886 | lr=6.90e-05 | 6.13 steps/s | VRAM=29.08 Go [2026-07-18 13:59:16] step 830600/1000000 | loss=2.2210 vocab=0.8013 token=1.2799 | lr=6.90e-05 | 6.12 steps/s | VRAM=27.10 Go [2026-07-18 13:59:24] step 830650/1000000 | loss=2.0640 vocab=1.1106 token=1.1701 | lr=6.89e-05 | 6.14 steps/s | VRAM=31.68 Go [2026-07-18 13:59:32] step 830700/1000000 | loss=2.1698 vocab=0.9816 token=1.2396 | lr=6.89e-05 | 6.12 steps/s | VRAM=32.79 Go [2026-07-18 13:59:40] step 830750/1000000 | loss=2.0651 vocab=0.8841 token=0.7662 | lr=6.89e-05 | 6.11 steps/s | VRAM=32.85 Go [2026-07-18 13:59:48] step 830800/1000000 | loss=2.1438 vocab=0.9417 token=1.3158 | lr=6.89e-05 | 6.13 steps/s | VRAM=31.98 Go [2026-07-18 13:59:57] step 830850/1000000 | loss=2.0747 vocab=0.9238 token=1.4741 | lr=6.88e-05 | 6.11 steps/s | VRAM=29.21 Go [2026-07-18 14:00:05] step 830900/1000000 | loss=2.1548 vocab=0.8736 token=1.4307 | lr=6.88e-05 | 6.13 steps/s | VRAM=28.81 Go [2026-07-18 14:00:13] step 830950/1000000 | loss=2.1448 vocab=0.5263 token=1.0918 | lr=6.88e-05 | 6.13 steps/s | VRAM=29.31 Go [2026-07-18 14:00:21] step 831000/1000000 | loss=2.1682 vocab=1.2287 token=1.6108 | lr=6.88e-05 | 6.10 steps/s | VRAM=28.38 Go [2026-07-18 14:00:29] step 831050/1000000 | loss=2.0487 vocab=1.1205 token=1.1680 | lr=6.87e-05 | 6.14 steps/s | VRAM=28.49 Go [2026-07-18 14:00:37] step 831100/1000000 | loss=2.0614 vocab=0.5537 token=1.3195 | lr=6.87e-05 | 6.08 steps/s | VRAM=26.85 Go [2026-07-18 14:00:46] step 831150/1000000 | loss=2.0882 vocab=0.9354 token=1.1772 | lr=6.87e-05 | 6.14 steps/s | VRAM=25.69 Go [2026-07-18 14:00:54] step 831200/1000000 | loss=2.0804 vocab=1.0812 token=1.6928 | lr=6.86e-05 | 6.14 steps/s | VRAM=25.80 Go [2026-07-18 14:01:02] step 831250/1000000 | loss=2.2658 vocab=0.9683 token=1.2970 | lr=6.86e-05 | 6.16 steps/s | VRAM=32.08 Go [2026-07-18 14:01:10] step 831300/1000000 | loss=2.2100 vocab=0.7489 token=1.2681 | lr=6.86e-05 | 6.13 steps/s | VRAM=32.86 Go [2026-07-18 14:01:18] step 831350/1000000 | loss=2.0721 vocab=0.7227 token=1.1297 | lr=6.86e-05 | 6.08 steps/s | VRAM=28.84 Go [2026-07-18 14:01:26] step 831400/1000000 | loss=2.1346 vocab=1.2254 token=1.0163 | lr=6.85e-05 | 6.11 steps/s | VRAM=26.34 Go [2026-07-18 14:01:35] step 831450/1000000 | loss=2.0460 vocab=0.7104 token=0.8296 | lr=6.85e-05 | 6.12 steps/s | VRAM=28.61 Go [2026-07-18 14:01:43] step 831500/1000000 | loss=2.0782 vocab=1.2216 token=1.5826 | lr=6.85e-05 | 6.11 steps/s | VRAM=26.47 Go [2026-07-18 14:01:51] step 831550/1000000 | loss=2.0967 vocab=0.7803 token=1.6475 | lr=6.85e-05 | 6.13 steps/s | VRAM=29.03 Go [2026-07-18 14:01:59] step 831600/1000000 | loss=2.1284 vocab=1.2847 token=1.5226 | lr=6.84e-05 | 6.12 steps/s | VRAM=28.90 Go [2026-07-18 14:02:07] step 831650/1000000 | loss=2.1532 vocab=1.2800 token=1.3204 | lr=6.84e-05 | 6.14 steps/s | VRAM=24.58 Go [2026-07-18 14:02:15] step 831700/1000000 | loss=2.1141 vocab=0.5219 token=0.9588 | lr=6.84e-05 | 6.15 steps/s | VRAM=28.22 Go [2026-07-18 14:02:23] step 831750/1000000 | loss=2.1110 vocab=0.8062 token=1.5573 | lr=6.84e-05 | 6.12 steps/s | VRAM=26.52 Go [2026-07-18 14:02:32] step 831800/1000000 | loss=2.0662 vocab=0.5953 token=1.2189 | lr=6.83e-05 | 6.15 steps/s | VRAM=31.25 Go [2026-07-18 14:02:40] step 831850/1000000 | loss=2.0723 vocab=1.3681 token=1.4971 | lr=6.83e-05 | 6.15 steps/s | VRAM=25.73 Go [2026-07-18 14:02:48] step 831900/1000000 | loss=1.9936 vocab=0.8058 token=1.3141 | lr=6.83e-05 | 6.12 steps/s | VRAM=28.36 Go [2026-07-18 14:02:56] step 831950/1000000 | loss=2.1673 vocab=1.3306 token=1.2201 | lr=6.82e-05 | 6.13 steps/s | VRAM=26.01 Go [2026-07-18 14:03:04] step 832000/1000000 | loss=2.0384 vocab=0.8526 token=1.2145 | lr=6.82e-05 | 6.15 steps/s | VRAM=32.74 Go [2026-07-18 14:03:12] step 832050/1000000 | loss=2.1190 vocab=0.6901 token=0.7982 | lr=6.82e-05 | 6.16 steps/s | VRAM=31.39 Go [2026-07-18 14:03:21] step 832100/1000000 | loss=2.1157 vocab=0.6001 token=1.5648 | lr=6.82e-05 | 6.11 steps/s | VRAM=26.37 Go [2026-07-18 14:03:29] step 832150/1000000 | loss=2.1564 vocab=0.5162 token=0.6201 | lr=6.81e-05 | 6.13 steps/s | VRAM=28.37 Go [2026-07-18 14:03:37] step 832200/1000000 | loss=2.1352 vocab=0.4446 token=0.9174 | lr=6.81e-05 | 6.13 steps/s | VRAM=28.00 Go [2026-07-18 14:03:45] step 832250/1000000 | loss=2.0244 vocab=0.6109 token=1.1580 | lr=6.81e-05 | 6.12 steps/s | VRAM=28.98 Go [2026-07-18 14:03:53] step 832300/1000000 | loss=2.0851 vocab=1.0578 token=1.8887 | lr=6.81e-05 | 6.14 steps/s | VRAM=31.45 Go [2026-07-18 14:04:01] step 832350/1000000 | loss=2.1900 vocab=0.9064 token=1.1530 | lr=6.80e-05 | 6.16 steps/s | VRAM=32.89 Go [2026-07-18 14:04:09] step 832400/1000000 | loss=2.0715 vocab=1.1118 token=1.2223 | lr=6.80e-05 | 6.15 steps/s | VRAM=32.90 Go [2026-07-18 14:04:17] step 832450/1000000 | loss=2.1412 vocab=1.1330 token=1.1658 | lr=6.80e-05 | 6.16 steps/s | VRAM=28.95 Go [2026-07-18 14:04:26] step 832500/1000000 | loss=2.3049 vocab=1.1756 token=1.8182 | lr=6.80e-05 | 6.12 steps/s | VRAM=26.53 Go [2026-07-18 14:04:34] step 832550/1000000 | loss=2.1139 vocab=0.5909 token=0.9580 | lr=6.79e-05 | 6.12 steps/s | VRAM=28.97 Go [2026-07-18 14:04:42] step 832600/1000000 | loss=2.0571 vocab=0.7223 token=1.2879 | lr=6.79e-05 | 6.14 steps/s | VRAM=28.96 Go [2026-07-18 14:04:50] step 832650/1000000 | loss=2.1098 vocab=0.5544 token=1.3767 | lr=6.79e-05 | 6.17 steps/s | VRAM=31.78 Go [2026-07-18 14:04:58] step 832700/1000000 | loss=2.1238 vocab=0.5658 token=0.9713 | lr=6.78e-05 | 6.13 steps/s | VRAM=25.77 Go [2026-07-18 14:05:06] step 832750/1000000 | loss=2.1156 vocab=0.4866 token=1.1472 | lr=6.78e-05 | 6.11 steps/s | VRAM=28.98 Go [2026-07-18 14:05:15] step 832800/1000000 | loss=2.2178 vocab=0.8480 token=1.3833 | lr=6.78e-05 | 6.15 steps/s | VRAM=26.27 Go [2026-07-18 14:05:23] step 832850/1000000 | loss=2.1376 vocab=0.9035 token=1.2169 | lr=6.78e-05 | 6.15 steps/s | VRAM=32.97 Go [2026-07-18 14:05:31] step 832900/1000000 | loss=2.0566 vocab=0.7545 token=1.4015 | lr=6.77e-05 | 6.15 steps/s | VRAM=28.22 Go [2026-07-18 14:05:39] step 832950/1000000 | loss=2.2285 vocab=0.4287 token=1.1060 | lr=6.77e-05 | 6.14 steps/s | VRAM=25.67 Go [2026-07-18 14:05:47] step 833000/1000000 | loss=2.1411 vocab=0.8642 token=1.1496 | lr=6.77e-05 | 6.12 steps/s | VRAM=32.24 Go [2026-07-18 14:05:55] step 833050/1000000 | loss=2.0655 vocab=0.7035 token=1.2483 | lr=6.77e-05 | 6.11 steps/s | VRAM=29.34 Go [2026-07-18 14:06:03] step 833100/1000000 | loss=2.1404 vocab=0.6751 token=0.7630 | lr=6.76e-05 | 6.12 steps/s | VRAM=28.12 Go [2026-07-18 14:06:12] step 833150/1000000 | loss=2.1423 vocab=0.8763 token=1.3788 | lr=6.76e-05 | 6.14 steps/s | VRAM=28.37 Go [2026-07-18 14:06:20] step 833200/1000000 | loss=2.1234 vocab=0.6606 token=1.4650 | lr=6.76e-05 | 6.11 steps/s | VRAM=29.45 Go [2026-07-18 14:06:28] step 833250/1000000 | loss=2.1783 vocab=1.3828 token=1.4878 | lr=6.76e-05 | 6.13 steps/s | VRAM=29.03 Go [2026-07-18 14:06:36] step 833300/1000000 | loss=2.2165 vocab=0.7086 token=1.1867 | lr=6.75e-05 | 6.13 steps/s | VRAM=29.13 Go [2026-07-18 14:06:44] step 833350/1000000 | loss=2.0794 vocab=0.9715 token=1.6968 | lr=6.75e-05 | 6.14 steps/s | VRAM=28.58 Go [2026-07-18 14:06:52] step 833400/1000000 | loss=2.1006 vocab=1.2602 token=1.1849 | lr=6.75e-05 | 6.14 steps/s | VRAM=32.77 Go [2026-07-18 14:07:01] step 833450/1000000 | loss=2.0094 vocab=0.8298 token=1.1390 | lr=6.74e-05 | 6.14 steps/s | VRAM=31.63 Go [2026-07-18 14:07:09] step 833500/1000000 | loss=2.1535 vocab=0.7877 token=0.9653 | lr=6.74e-05 | 6.09 steps/s | VRAM=29.06 Go [2026-07-18 14:07:17] step 833550/1000000 | loss=2.3694 vocab=1.1882 token=1.7562 | lr=6.74e-05 | 6.12 steps/s | VRAM=32.83 Go [2026-07-18 14:07:25] step 833600/1000000 | loss=2.1062 vocab=1.1025 token=1.3052 | lr=6.74e-05 | 6.11 steps/s | VRAM=29.11 Go [2026-07-18 14:07:33] step 833650/1000000 | loss=2.1435 vocab=1.3694 token=1.4154 | lr=6.73e-05 | 6.14 steps/s | VRAM=24.34 Go [2026-07-18 14:07:41] step 833700/1000000 | loss=2.1323 vocab=0.7673 token=1.3741 | lr=6.73e-05 | 6.14 steps/s | VRAM=32.73 Go [2026-07-18 14:07:50] step 833750/1000000 | loss=2.1687 vocab=1.0380 token=1.5830 | lr=6.73e-05 | 6.14 steps/s | VRAM=25.95 Go [2026-07-18 14:07:58] step 833800/1000000 | loss=2.1082 vocab=1.5212 token=1.4673 | lr=6.73e-05 | 6.16 steps/s | VRAM=28.31 Go [2026-07-18 14:08:06] step 833850/1000000 | loss=2.2304 vocab=1.5936 token=1.4498 | lr=6.72e-05 | 6.14 steps/s | VRAM=29.88 Go [2026-07-18 14:08:14] step 833900/1000000 | loss=2.1611 vocab=0.7795 token=1.3814 | lr=6.72e-05 | 6.14 steps/s | VRAM=26.23 Go [2026-07-18 14:08:22] step 833950/1000000 | loss=2.1667 vocab=0.6625 token=1.2010 | lr=6.72e-05 | 6.13 steps/s | VRAM=28.48 Go [2026-07-18 14:08:30] step 834000/1000000 | loss=2.0580 vocab=0.5803 token=1.0789 | lr=6.72e-05 | 6.15 steps/s | VRAM=28.95 Go [2026-07-18 14:08:38] step 834050/1000000 | loss=2.0561 vocab=0.5534 token=1.2611 | lr=6.71e-05 | 6.14 steps/s | VRAM=32.30 Go [2026-07-18 14:08:47] step 834100/1000000 | loss=2.0920 vocab=1.0764 token=1.2736 | lr=6.71e-05 | 6.12 steps/s | VRAM=31.71 Go [2026-07-18 14:08:55] step 834150/1000000 | loss=2.1771 vocab=0.7758 token=1.2461 | lr=6.71e-05 | 6.11 steps/s | VRAM=29.64 Go [2026-07-18 14:09:03] step 834200/1000000 | loss=2.1362 vocab=0.7822 token=1.7067 | lr=6.70e-05 | 6.15 steps/s | VRAM=24.77 Go [2026-07-18 14:09:11] step 834250/1000000 | loss=2.1202 vocab=0.8078 token=1.3478 | lr=6.70e-05 | 6.12 steps/s | VRAM=26.47 Go [2026-07-18 14:09:19] step 834300/1000000 | loss=2.1516 vocab=0.4076 token=1.0659 | lr=6.70e-05 | 6.18 steps/s | VRAM=31.33 Go [2026-07-18 14:09:27] step 834350/1000000 | loss=2.0794 vocab=0.4629 token=0.9173 | lr=6.70e-05 | 6.16 steps/s | VRAM=26.16 Go [2026-07-18 14:09:35] step 834400/1000000 | loss=2.1793 vocab=0.4833 token=1.2396 | lr=6.69e-05 | 6.14 steps/s | VRAM=25.11 Go [2026-07-18 14:09:44] step 834450/1000000 | loss=2.0606 vocab=0.7054 token=1.1021 | lr=6.69e-05 | 6.15 steps/s | VRAM=30.53 Go [2026-07-18 14:09:52] step 834500/1000000 | loss=2.1597 vocab=0.5858 token=1.4465 | lr=6.69e-05 | 6.16 steps/s | VRAM=25.47 Go [2026-07-18 14:10:00] step 834550/1000000 | loss=2.2205 vocab=0.9238 token=1.3246 | lr=6.69e-05 | 6.15 steps/s | VRAM=31.60 Go [2026-07-18 14:10:08] step 834600/1000000 | loss=2.1836 vocab=0.9076 token=1.4863 | lr=6.68e-05 | 6.14 steps/s | VRAM=25.87 Go [2026-07-18 14:10:16] step 834650/1000000 | loss=2.0022 vocab=0.6939 token=0.7603 | lr=6.68e-05 | 6.16 steps/s | VRAM=26.69 Go [2026-07-18 14:10:24] step 834700/1000000 | loss=2.0752 vocab=0.2643 token=0.7497 | lr=6.68e-05 | 6.14 steps/s | VRAM=29.30 Go [2026-07-18 14:10:32] step 834750/1000000 | loss=2.1704 vocab=1.1230 token=1.2743 | lr=6.67e-05 | 6.17 steps/s | VRAM=28.97 Go [2026-07-18 14:10:40] step 834800/1000000 | loss=2.1394 vocab=0.9511 token=1.6009 | lr=6.67e-05 | 6.17 steps/s | VRAM=24.47 Go [2026-07-18 14:10:49] step 834850/1000000 | loss=2.1704 vocab=1.3762 token=1.3428 | lr=6.67e-05 | 6.14 steps/s | VRAM=24.96 Go [2026-07-18 14:10:57] step 834900/1000000 | loss=2.1521 vocab=0.8512 token=1.0966 | lr=6.67e-05 | 6.15 steps/s | VRAM=32.23 Go [2026-07-18 14:11:05] step 834950/1000000 | loss=2.0419 vocab=0.6969 token=1.1878 | lr=6.66e-05 | 6.13 steps/s | VRAM=29.25 Go [2026-07-18 14:11:13] step 835000/1000000 | loss=2.0607 vocab=0.6413 token=1.4251 | lr=6.66e-05 | 6.14 steps/s | VRAM=29.89 Go [2026-07-18 14:11:21] step 835050/1000000 | loss=2.1334 vocab=1.3526 token=1.0858 | lr=6.66e-05 | 6.14 steps/s | VRAM=28.89 Go [2026-07-18 14:11:29] step 835100/1000000 | loss=1.9919 vocab=0.7044 token=1.3437 | lr=6.66e-05 | 6.13 steps/s | VRAM=29.66 Go [2026-07-18 14:11:37] step 835150/1000000 | loss=2.1450 vocab=1.1076 token=1.2459 | lr=6.65e-05 | 6.14 steps/s | VRAM=32.60 Go [2026-07-18 14:11:46] step 835200/1000000 | loss=2.0697 vocab=0.8319 token=1.4592 | lr=6.65e-05 | 6.11 steps/s | VRAM=28.30 Go [2026-07-18 14:11:54] step 835250/1000000 | loss=2.0659 vocab=1.2623 token=1.3801 | lr=6.65e-05 | 6.14 steps/s | VRAM=32.31 Go [2026-07-18 14:12:02] step 835300/1000000 | loss=2.1172 vocab=1.0079 token=1.1483 | lr=6.65e-05 | 6.13 steps/s | VRAM=32.43 Go [2026-07-18 14:12:10] step 835350/1000000 | loss=2.0980 vocab=1.2364 token=1.5851 | lr=6.64e-05 | 6.14 steps/s | VRAM=31.87 Go [2026-07-18 14:12:18] step 835400/1000000 | loss=2.2274 vocab=1.4218 token=1.3698 | lr=6.64e-05 | 6.14 steps/s | VRAM=28.85 Go [2026-07-18 14:12:26] step 835450/1000000 | loss=2.1573 vocab=1.2730 token=1.6177 | lr=6.64e-05 | 6.15 steps/s | VRAM=28.21 Go [2026-07-18 14:12:34] step 835500/1000000 | loss=2.1737 vocab=0.7660 token=1.0136 | lr=6.63e-05 | 6.15 steps/s | VRAM=32.08 Go [2026-07-18 14:12:43] step 835550/1000000 | loss=2.0719 vocab=0.8347 token=1.4766 | lr=6.63e-05 | 6.08 steps/s | VRAM=26.38 Go [2026-07-18 14:12:51] step 835600/1000000 | loss=2.0941 vocab=1.3357 token=0.9181 | lr=6.63e-05 | 6.14 steps/s | VRAM=28.82 Go [2026-07-18 14:12:59] step 835650/1000000 | loss=2.1725 vocab=1.0566 token=0.9834 | lr=6.63e-05 | 6.14 steps/s | VRAM=32.70 Go [2026-07-18 14:13:07] step 835700/1000000 | loss=1.9685 vocab=1.0018 token=1.5105 | lr=6.62e-05 | 6.10 steps/s | VRAM=26.44 Go [2026-07-18 14:13:15] step 835750/1000000 | loss=2.1150 vocab=0.5840 token=1.1287 | lr=6.62e-05 | 6.14 steps/s | VRAM=28.14 Go [2026-07-18 14:13:23] step 835800/1000000 | loss=2.1929 vocab=0.7004 token=1.4950 | lr=6.62e-05 | 6.10 steps/s | VRAM=25.25 Go [2026-07-18 14:13:32] step 835850/1000000 | loss=2.2101 vocab=1.4554 token=1.3421 | lr=6.62e-05 | 6.12 steps/s | VRAM=25.84 Go [2026-07-18 14:13:40] step 835900/1000000 | loss=2.1078 vocab=0.7190 token=1.2513 | lr=6.61e-05 | 6.13 steps/s | VRAM=31.40 Go [2026-07-18 14:13:48] step 835950/1000000 | loss=2.1075 vocab=0.7423 token=1.1543 | lr=6.61e-05 | 6.11 steps/s | VRAM=24.64 Go [2026-07-18 14:13:56] step 836000/1000000 | loss=2.0959 vocab=0.6631 token=1.2139 | lr=6.61e-05 | 6.12 steps/s | VRAM=28.53 Go [2026-07-18 14:14:04] step 836050/1000000 | loss=2.0893 vocab=0.4366 token=1.2705 | lr=6.60e-05 | 6.12 steps/s | VRAM=32.38 Go [2026-07-18 14:14:13] step 836100/1000000 | loss=2.0830 vocab=1.3064 token=1.5587 | lr=6.60e-05 | 6.11 steps/s | VRAM=26.35 Go [2026-07-18 14:14:21] step 836150/1000000 | loss=2.0377 vocab=0.5257 token=0.9952 | lr=6.60e-05 | 6.12 steps/s | VRAM=28.53 Go [2026-07-18 14:14:29] step 836200/1000000 | loss=2.1600 vocab=1.5375 token=1.5861 | lr=6.60e-05 | 6.11 steps/s | VRAM=29.75 Go [2026-07-18 14:14:37] step 836250/1000000 | loss=2.1155 vocab=0.5849 token=0.9299 | lr=6.59e-05 | 6.14 steps/s | VRAM=33.00 Go [2026-07-18 14:14:45] step 836300/1000000 | loss=2.2275 vocab=1.1630 token=1.1582 | lr=6.59e-05 | 6.11 steps/s | VRAM=26.52 Go [2026-07-18 14:14:53] step 836350/1000000 | loss=2.1131 vocab=0.9818 token=1.2548 | lr=6.59e-05 | 6.13 steps/s | VRAM=25.70 Go [2026-07-18 14:15:02] step 836400/1000000 | loss=2.1295 vocab=0.7422 token=0.9421 | lr=6.59e-05 | 6.10 steps/s | VRAM=28.50 Go [2026-07-18 14:15:10] step 836450/1000000 | loss=2.1454 vocab=0.7079 token=1.1168 | lr=6.58e-05 | 6.13 steps/s | VRAM=24.13 Go [2026-07-18 14:15:18] step 836500/1000000 | loss=2.0383 vocab=0.9646 token=1.3420 | lr=6.58e-05 | 6.09 steps/s | VRAM=27.90 Go [2026-07-18 14:15:26] step 836550/1000000 | loss=1.9964 vocab=0.5438 token=1.1332 | lr=6.58e-05 | 6.12 steps/s | VRAM=32.25 Go [2026-07-18 14:15:34] step 836600/1000000 | loss=2.0139 vocab=0.5281 token=0.8936 | lr=6.57e-05 | 6.08 steps/s | VRAM=31.69 Go [2026-07-18 14:15:43] step 836650/1000000 | loss=2.1206 vocab=1.1861 token=1.2578 | lr=6.57e-05 | 6.10 steps/s | VRAM=26.46 Go [2026-07-18 14:15:51] step 836700/1000000 | loss=2.0855 vocab=0.7877 token=1.4490 | lr=6.57e-05 | 6.12 steps/s | VRAM=25.95 Go [2026-07-18 14:15:59] step 836750/1000000 | loss=2.0572 vocab=1.0888 token=1.2994 | lr=6.57e-05 | 6.16 steps/s | VRAM=31.63 Go [2026-07-18 14:16:07] step 836800/1000000 | loss=2.1247 vocab=1.0532 token=1.2810 | lr=6.56e-05 | 6.13 steps/s | VRAM=26.84 Go [2026-07-18 14:16:15] step 836850/1000000 | loss=2.0086 vocab=1.0553 token=1.2895 | lr=6.56e-05 | 6.12 steps/s | VRAM=31.68 Go [2026-07-18 14:16:23] step 836900/1000000 | loss=2.1103 vocab=0.4469 token=1.0946 | lr=6.56e-05 | 6.13 steps/s | VRAM=27.45 Go [2026-07-18 14:16:31] step 836950/1000000 | loss=1.9747 vocab=1.4071 token=1.2803 | lr=6.56e-05 | 6.14 steps/s | VRAM=31.51 Go [2026-07-18 14:16:40] step 837000/1000000 | loss=2.0916 vocab=0.4773 token=1.0254 | lr=6.55e-05 | 6.13 steps/s | VRAM=32.22 Go [2026-07-18 14:16:48] step 837050/1000000 | loss=2.0208 vocab=0.9351 token=0.8365 | lr=6.55e-05 | 6.14 steps/s | VRAM=28.20 Go [2026-07-18 14:16:56] step 837100/1000000 | loss=2.1037 vocab=1.0138 token=1.2303 | lr=6.55e-05 | 6.10 steps/s | VRAM=25.86 Go [2026-07-18 14:17:04] step 837150/1000000 | loss=2.1370 vocab=1.3098 token=1.6223 | lr=6.54e-05 | 6.12 steps/s | VRAM=24.49 Go [2026-07-18 14:17:12] step 837200/1000000 | loss=2.1196 vocab=0.6228 token=1.1656 | lr=6.54e-05 | 6.13 steps/s | VRAM=26.00 Go [2026-07-18 14:17:20] step 837250/1000000 | loss=2.0915 vocab=0.7252 token=0.9466 | lr=6.54e-05 | 6.15 steps/s | VRAM=28.11 Go [2026-07-18 14:17:29] step 837300/1000000 | loss=2.1879 vocab=0.5579 token=1.0550 | lr=6.54e-05 | 6.10 steps/s | VRAM=29.39 Go [2026-07-18 14:17:37] step 837350/1000000 | loss=1.9891 vocab=0.4379 token=1.1182 | lr=6.53e-05 | 6.10 steps/s | VRAM=25.22 Go [2026-07-18 14:17:45] step 837400/1000000 | loss=2.1510 vocab=0.9681 token=1.4109 | lr=6.53e-05 | 6.15 steps/s | VRAM=31.20 Go [2026-07-18 14:17:53] step 837450/1000000 | loss=2.1471 vocab=0.6804 token=1.0265 | lr=6.53e-05 | 6.11 steps/s | VRAM=26.29 Go [2026-07-18 14:18:01] step 837500/1000000 | loss=2.1709 vocab=0.9172 token=1.7077 | lr=6.53e-05 | 6.12 steps/s | VRAM=29.42 Go [2026-07-18 14:18:09] step 837550/1000000 | loss=2.1348 vocab=0.8468 token=1.2473 | lr=6.52e-05 | 6.13 steps/s | VRAM=32.37 Go [2026-07-18 14:18:18] step 837600/1000000 | loss=2.0937 vocab=0.8840 token=1.6109 | lr=6.52e-05 | 6.11 steps/s | VRAM=28.22 Go [2026-07-18 14:18:26] step 837650/1000000 | loss=2.1364 vocab=0.9175 token=1.2121 | lr=6.52e-05 | 6.14 steps/s | VRAM=32.66 Go [2026-07-18 14:18:34] step 837700/1000000 | loss=2.1436 vocab=1.0734 token=1.1796 | lr=6.51e-05 | 6.15 steps/s | VRAM=28.55 Go [2026-07-18 14:18:42] step 837750/1000000 | loss=2.1500 vocab=1.0010 token=1.2685 | lr=6.51e-05 | 6.09 steps/s | VRAM=28.34 Go [2026-07-18 14:18:50] step 837800/1000000 | loss=2.1529 vocab=1.2814 token=1.5643 | lr=6.51e-05 | 6.13 steps/s | VRAM=28.08 Go [2026-07-18 14:18:58] step 837850/1000000 | loss=2.1883 vocab=0.5964 token=1.3706 | lr=6.51e-05 | 6.12 steps/s | VRAM=26.94 Go [2026-07-18 14:19:07] step 837900/1000000 | loss=2.0741 vocab=1.0514 token=1.4153 | lr=6.50e-05 | 6.13 steps/s | VRAM=29.10 Go [2026-07-18 14:19:15] step 837950/1000000 | loss=2.1574 vocab=0.9093 token=0.8730 | lr=6.50e-05 | 6.13 steps/s | VRAM=28.60 Go [2026-07-18 14:19:23] step 838000/1000000 | loss=2.0057 vocab=0.7190 token=1.3934 | lr=6.50e-05 | 6.13 steps/s | VRAM=32.11 Go [2026-07-18 14:19:31] step 838050/1000000 | loss=2.0338 vocab=0.8994 token=1.2396 | lr=6.50e-05 | 6.14 steps/s | VRAM=32.28 Go [2026-07-18 14:19:39] step 838100/1000000 | loss=2.2369 vocab=0.9900 token=1.4945 | lr=6.49e-05 | 6.12 steps/s | VRAM=32.87 Go [2026-07-18 14:19:47] step 838150/1000000 | loss=2.0677 vocab=0.8001 token=1.7473 | lr=6.49e-05 | 6.14 steps/s | VRAM=25.66 Go [2026-07-18 14:19:56] step 838200/1000000 | loss=2.1119 vocab=0.3326 token=0.8848 | lr=6.49e-05 | 6.14 steps/s | VRAM=28.99 Go [2026-07-18 14:20:04] step 838250/1000000 | loss=2.1622 vocab=1.1189 token=1.2637 | lr=6.48e-05 | 6.16 steps/s | VRAM=31.45 Go [2026-07-18 14:20:12] step 838300/1000000 | loss=2.1756 vocab=0.8129 token=1.1430 | lr=6.48e-05 | 6.10 steps/s | VRAM=32.90 Go [2026-07-18 14:20:20] step 838350/1000000 | loss=2.1288 vocab=1.1021 token=1.1746 | lr=6.48e-05 | 6.10 steps/s | VRAM=29.24 Go [2026-07-18 14:20:28] step 838400/1000000 | loss=2.1233 vocab=1.2682 token=1.0944 | lr=6.48e-05 | 6.15 steps/s | VRAM=25.16 Go [2026-07-18 14:20:36] step 838450/1000000 | loss=2.3082 vocab=1.0672 token=1.5697 | lr=6.47e-05 | 6.14 steps/s | VRAM=28.33 Go [2026-07-18 14:20:44] step 838500/1000000 | loss=2.1577 vocab=0.4676 token=1.1376 | lr=6.47e-05 | 6.12 steps/s | VRAM=29.53 Go [2026-07-18 14:20:53] step 838550/1000000 | loss=2.1915 vocab=1.0113 token=1.7908 | lr=6.47e-05 | 6.13 steps/s | VRAM=28.27 Go [2026-07-18 14:21:01] step 838600/1000000 | loss=2.0228 vocab=0.8833 token=1.5071 | lr=6.47e-05 | 6.12 steps/s | VRAM=28.66 Go [2026-07-18 14:21:09] step 838650/1000000 | loss=2.1062 vocab=0.6774 token=1.2298 | lr=6.46e-05 | 6.10 steps/s | VRAM=28.59 Go [2026-07-18 14:21:17] step 838700/1000000 | loss=2.1527 vocab=0.3940 token=0.7927 | lr=6.46e-05 | 6.09 steps/s | VRAM=29.13 Go [2026-07-18 14:21:25] step 838750/1000000 | loss=2.1968 vocab=1.2887 token=1.4678 | lr=6.46e-05 | 6.12 steps/s | VRAM=32.43 Go [2026-07-18 14:21:33] step 838800/1000000 | loss=2.0819 vocab=0.9647 token=1.5243 | lr=6.45e-05 | 6.17 steps/s | VRAM=31.39 Go [2026-07-18 14:21:42] step 838850/1000000 | loss=2.1353 vocab=0.6245 token=1.0323 | lr=6.45e-05 | 6.13 steps/s | VRAM=25.98 Go [2026-07-18 14:21:50] step 838900/1000000 | loss=2.1737 vocab=0.9312 token=1.0318 | lr=6.45e-05 | 6.14 steps/s | VRAM=31.89 Go [2026-07-18 14:21:58] step 838950/1000000 | loss=2.1679 vocab=0.8154 token=1.2541 | lr=6.45e-05 | 6.09 steps/s | VRAM=28.45 Go [2026-07-18 14:22:06] step 839000/1000000 | loss=2.1965 vocab=1.4698 token=1.6402 | lr=6.44e-05 | 6.11 steps/s | VRAM=32.45 Go [2026-07-18 14:22:14] step 839050/1000000 | loss=2.1305 vocab=0.8887 token=1.0603 | lr=6.44e-05 | 6.13 steps/s | VRAM=31.41 Go [2026-07-18 14:22:22] step 839100/1000000 | loss=2.0399 vocab=0.5012 token=0.9760 | lr=6.44e-05 | 6.13 steps/s | VRAM=28.85 Go [2026-07-18 14:22:31] step 839150/1000000 | loss=2.0167 vocab=0.6261 token=1.2649 | lr=6.44e-05 | 6.14 steps/s | VRAM=31.32 Go [2026-07-18 14:22:39] step 839200/1000000 | loss=2.1479 vocab=1.1251 token=1.5807 | lr=6.43e-05 | 6.15 steps/s | VRAM=32.30 Go [2026-07-18 14:22:47] step 839250/1000000 | loss=2.1368 vocab=0.6815 token=1.2172 | lr=6.43e-05 | 6.15 steps/s | VRAM=31.22 Go [2026-07-18 14:22:55] step 839300/1000000 | loss=2.0990 vocab=1.0338 token=1.0418 | lr=6.43e-05 | 6.14 steps/s | VRAM=29.11 Go [2026-07-18 14:23:03] step 839350/1000000 | loss=2.1443 vocab=0.6507 token=1.2726 | lr=6.42e-05 | 6.12 steps/s | VRAM=29.27 Go [2026-07-18 14:23:11] step 839400/1000000 | loss=2.1780 vocab=1.0747 token=1.1992 | lr=6.42e-05 | 6.11 steps/s | VRAM=29.98 Go [2026-07-18 14:23:20] step 839450/1000000 | loss=1.9883 vocab=0.8157 token=1.2277 | lr=6.42e-05 | 6.12 steps/s | VRAM=28.78 Go [2026-07-18 14:23:28] step 839500/1000000 | loss=2.1970 vocab=1.0450 token=1.4562 | lr=6.42e-05 | 6.13 steps/s | VRAM=27.59 Go [2026-07-18 14:23:36] step 839550/1000000 | loss=2.1961 vocab=0.6634 token=1.1660 | lr=6.41e-05 | 6.11 steps/s | VRAM=26.89 Go [2026-07-18 14:23:44] step 839600/1000000 | loss=2.1381 vocab=0.5362 token=1.5207 | lr=6.41e-05 | 6.11 steps/s | VRAM=29.82 Go [2026-07-18 14:23:52] step 839650/1000000 | loss=2.1542 vocab=0.7506 token=1.2363 | lr=6.41e-05 | 6.14 steps/s | VRAM=31.87 Go [2026-07-18 14:24:00] step 839700/1000000 | loss=2.0821 vocab=0.9100 token=1.0214 | lr=6.41e-05 | 6.11 steps/s | VRAM=32.10 Go [2026-07-18 14:24:09] step 839750/1000000 | loss=2.2510 vocab=1.1303 token=1.6179 | lr=6.40e-05 | 6.15 steps/s | VRAM=28.03 Go [2026-07-18 14:24:17] step 839800/1000000 | loss=2.0329 vocab=0.8605 token=1.1153 | lr=6.40e-05 | 6.14 steps/s | VRAM=32.22 Go [2026-07-18 14:24:25] step 839850/1000000 | loss=2.1578 vocab=0.7329 token=0.8332 | lr=6.40e-05 | 6.14 steps/s | VRAM=26.41 Go [2026-07-18 14:24:33] step 839900/1000000 | loss=2.0612 vocab=0.7726 token=1.1848 | lr=6.39e-05 | 6.16 steps/s | VRAM=28.34 Go [2026-07-18 14:24:41] step 839950/1000000 | loss=2.1145 vocab=0.7281 token=0.9908 | lr=6.39e-05 | 6.14 steps/s | VRAM=26.76 Go [2026-07-18 14:24:49] step 840000/1000000 | loss=2.1645 vocab=1.0324 token=1.3501 | lr=6.39e-05 | 6.11 steps/s | VRAM=29.00 Go [2026-07-18 14:24:57] step 840050/1000000 | loss=2.2630 vocab=1.4578 token=1.3752 | lr=6.39e-05 | 6.13 steps/s | VRAM=29.84 Go [2026-07-18 14:25:06] step 840100/1000000 | loss=2.0982 vocab=0.5920 token=0.9837 | lr=6.38e-05 | 6.14 steps/s | VRAM=25.96 Go [2026-07-18 14:25:14] step 840150/1000000 | loss=2.0746 vocab=0.5418 token=0.9443 | lr=6.38e-05 | 6.17 steps/s | VRAM=28.60 Go [2026-07-18 14:25:22] step 840200/1000000 | loss=2.1130 vocab=0.8126 token=1.0929 | lr=6.38e-05 | 6.14 steps/s | VRAM=32.38 Go [2026-07-18 14:25:30] step 840250/1000000 | loss=2.0911 vocab=0.6657 token=1.0046 | lr=6.38e-05 | 6.14 steps/s | VRAM=32.30 Go [2026-07-18 14:25:38] step 840300/1000000 | loss=2.2018 vocab=0.7402 token=1.0282 | lr=6.37e-05 | 6.12 steps/s | VRAM=31.74 Go [2026-07-18 14:25:46] step 840350/1000000 | loss=2.1144 vocab=0.8338 token=1.4692 | lr=6.37e-05 | 6.12 steps/s | VRAM=25.65 Go [2026-07-18 14:25:54] step 840400/1000000 | loss=2.0859 vocab=1.1788 token=1.2546 | lr=6.37e-05 | 6.13 steps/s | VRAM=25.14 Go [2026-07-18 14:26:03] step 840450/1000000 | loss=2.0920 vocab=0.6160 token=0.9876 | lr=6.36e-05 | 6.10 steps/s | VRAM=28.48 Go [2026-07-18 14:26:11] step 840500/1000000 | loss=2.0493 vocab=1.0861 token=0.9348 | lr=6.36e-05 | 6.17 steps/s | VRAM=28.07 Go [2026-07-18 14:26:19] step 840550/1000000 | loss=2.1681 vocab=0.6944 token=1.2853 | lr=6.36e-05 | 6.14 steps/s | VRAM=29.45 Go [2026-07-18 14:26:27] step 840600/1000000 | loss=2.2196 vocab=1.0855 token=1.4678 | lr=6.36e-05 | 6.11 steps/s | VRAM=28.49 Go [2026-07-18 14:26:35] step 840650/1000000 | loss=2.0658 vocab=0.8059 token=1.1159 | lr=6.35e-05 | 6.16 steps/s | VRAM=31.89 Go [2026-07-18 14:26:43] step 840700/1000000 | loss=2.2076 vocab=0.9518 token=1.4633 | lr=6.35e-05 | 6.14 steps/s | VRAM=32.34 Go [2026-07-18 14:26:51] step 840750/1000000 | loss=2.2151 vocab=2.2370 token=1.6288 | lr=6.35e-05 | 6.14 steps/s | VRAM=32.06 Go [2026-07-18 14:27:00] step 840800/1000000 | loss=2.1818 vocab=1.0406 token=1.4412 | lr=6.34e-05 | 6.12 steps/s | VRAM=28.69 Go [2026-07-18 14:27:08] step 840850/1000000 | loss=2.0562 vocab=0.5641 token=0.9713 | lr=6.34e-05 | 6.15 steps/s | VRAM=29.60 Go [2026-07-18 14:27:16] step 840900/1000000 | loss=2.0174 vocab=0.9663 token=1.3788 | lr=6.34e-05 | 6.15 steps/s | VRAM=28.97 Go [2026-07-18 14:27:24] step 840950/1000000 | loss=2.1028 vocab=1.1238 token=1.4039 | lr=6.34e-05 | 6.16 steps/s | VRAM=31.15 Go [2026-07-18 14:27:32] step 841000/1000000 | loss=2.1763 vocab=1.0799 token=1.4753 | lr=6.33e-05 | 6.13 steps/s | VRAM=27.23 Go [2026-07-18 14:27:40] step 841050/1000000 | loss=2.2144 vocab=0.9439 token=1.2094 | lr=6.33e-05 | 6.11 steps/s | VRAM=27.94 Go [2026-07-18 14:27:49] step 841100/1000000 | loss=2.1467 vocab=0.3923 token=0.5682 | lr=6.33e-05 | 6.13 steps/s | VRAM=32.04 Go [2026-07-18 14:27:57] step 841150/1000000 | loss=2.2019 vocab=0.9780 token=1.3334 | lr=6.33e-05 | 6.12 steps/s | VRAM=26.87 Go [2026-07-18 14:28:05] step 841200/1000000 | loss=2.0735 vocab=0.6280 token=0.9346 | lr=6.32e-05 | 6.15 steps/s | VRAM=27.98 Go [2026-07-18 14:28:13] step 841250/1000000 | loss=1.9991 vocab=1.3641 token=1.3574 | lr=6.32e-05 | 6.17 steps/s | VRAM=26.64 Go [2026-07-18 14:28:21] step 841300/1000000 | loss=2.1163 vocab=0.8443 token=1.4419 | lr=6.32e-05 | 6.15 steps/s | VRAM=29.24 Go [2026-07-18 14:28:29] step 841350/1000000 | loss=2.1875 vocab=1.2528 token=1.4202 | lr=6.31e-05 | 6.17 steps/s | VRAM=30.66 Go [2026-07-18 14:28:37] step 841400/1000000 | loss=2.1844 vocab=1.2437 token=1.2009 | lr=6.31e-05 | 6.12 steps/s | VRAM=28.58 Go [2026-07-18 14:28:46] step 841450/1000000 | loss=2.0726 vocab=0.7787 token=0.8755 | lr=6.31e-05 | 6.13 steps/s | VRAM=27.47 Go [2026-07-18 14:28:54] step 841500/1000000 | loss=2.2210 vocab=0.6324 token=1.1966 | lr=6.31e-05 | 6.13 steps/s | VRAM=26.28 Go [2026-07-18 14:29:02] step 841550/1000000 | loss=2.1143 vocab=0.7925 token=1.0703 | lr=6.30e-05 | 6.15 steps/s | VRAM=31.41 Go [2026-07-18 14:29:10] step 841600/1000000 | loss=2.0920 vocab=0.6454 token=0.8731 | lr=6.30e-05 | 6.16 steps/s | VRAM=25.22 Go [2026-07-18 14:29:18] step 841650/1000000 | loss=2.1777 vocab=0.6566 token=1.2394 | lr=6.30e-05 | 6.11 steps/s | VRAM=26.66 Go [2026-07-18 14:29:26] step 841700/1000000 | loss=2.0772 vocab=1.4307 token=1.0777 | lr=6.30e-05 | 6.13 steps/s | VRAM=32.28 Go [2026-07-18 14:29:34] step 841750/1000000 | loss=2.1765 vocab=0.9956 token=1.4625 | lr=6.29e-05 | 6.11 steps/s | VRAM=29.46 Go [2026-07-18 14:29:43] step 841800/1000000 | loss=2.2255 vocab=1.0085 token=1.1371 | lr=6.29e-05 | 6.13 steps/s | VRAM=29.62 Go [2026-07-18 14:29:51] step 841850/1000000 | loss=2.2065 vocab=0.8415 token=1.2774 | lr=6.29e-05 | 6.13 steps/s | VRAM=28.34 Go [2026-07-18 14:29:59] step 841900/1000000 | loss=2.2137 vocab=1.0028 token=1.3166 | lr=6.28e-05 | 6.10 steps/s | VRAM=23.43 Go [2026-07-18 14:30:07] step 841950/1000000 | loss=2.1154 vocab=0.4749 token=0.9840 | lr=6.28e-05 | 6.15 steps/s | VRAM=31.87 Go [2026-07-18 14:30:15] step 842000/1000000 | loss=2.0610 vocab=0.6282 token=1.1740 | lr=6.28e-05 | 6.13 steps/s | VRAM=28.32 Go [2026-07-18 14:30:23] step 842050/1000000 | loss=2.0794 vocab=1.2052 token=1.0719 | lr=6.28e-05 | 6.11 steps/s | VRAM=29.49 Go [2026-07-18 14:30:32] step 842100/1000000 | loss=2.2109 vocab=1.2111 token=1.1231 | lr=6.27e-05 | 6.13 steps/s | VRAM=26.79 Go [2026-07-18 14:30:40] step 842150/1000000 | loss=2.2159 vocab=1.1407 token=1.0152 | lr=6.27e-05 | 6.13 steps/s | VRAM=28.11 Go [2026-07-18 14:30:48] step 842200/1000000 | loss=2.1787 vocab=0.9130 token=1.0969 | lr=6.27e-05 | 6.15 steps/s | VRAM=25.11 Go [2026-07-18 14:30:56] step 842250/1000000 | loss=2.1945 vocab=0.5158 token=1.1997 | lr=6.26e-05 | 6.17 steps/s | VRAM=31.95 Go [2026-07-18 14:31:04] step 842300/1000000 | loss=2.0896 vocab=0.8290 token=0.9175 | lr=6.26e-05 | 6.11 steps/s | VRAM=29.13 Go [2026-07-18 14:31:12] step 842350/1000000 | loss=2.1023 vocab=1.2854 token=1.2698 | lr=6.26e-05 | 6.12 steps/s | VRAM=28.44 Go [2026-07-18 14:31:20] step 842400/1000000 | loss=2.0713 vocab=0.7910 token=0.8803 | lr=6.26e-05 | 6.16 steps/s | VRAM=29.16 Go [2026-07-18 14:31:29] step 842450/1000000 | loss=2.0988 vocab=1.2211 token=1.4329 | lr=6.25e-05 | 6.14 steps/s | VRAM=28.22 Go [2026-07-18 14:31:37] step 842500/1000000 | loss=2.1088 vocab=0.5204 token=0.8823 | lr=6.25e-05 | 6.14 steps/s | VRAM=29.45 Go [2026-07-18 14:31:45] step 842550/1000000 | loss=2.1496 vocab=0.6013 token=1.1715 | lr=6.25e-05 | 6.13 steps/s | VRAM=31.35 Go [2026-07-18 14:31:53] step 842600/1000000 | loss=2.1277 vocab=1.8925 token=2.1206 | lr=6.25e-05 | 6.12 steps/s | VRAM=29.02 Go [2026-07-18 14:32:01] step 842650/1000000 | loss=2.0218 vocab=0.6630 token=1.3384 | lr=6.24e-05 | 6.15 steps/s | VRAM=31.98 Go [2026-07-18 14:32:09] step 842700/1000000 | loss=2.1209 vocab=0.9071 token=1.5227 | lr=6.24e-05 | 6.14 steps/s | VRAM=29.56 Go [2026-07-18 14:32:17] step 842750/1000000 | loss=2.2464 vocab=1.2672 token=1.2944 | lr=6.24e-05 | 6.13 steps/s | VRAM=28.54 Go [2026-07-18 14:32:26] step 842800/1000000 | loss=2.1565 vocab=1.0447 token=1.7431 | lr=6.23e-05 | 6.11 steps/s | VRAM=25.90 Go [2026-07-18 14:32:34] step 842850/1000000 | loss=2.0868 vocab=0.6692 token=1.3312 | lr=6.23e-05 | 6.15 steps/s | VRAM=32.62 Go [2026-07-18 14:32:42] step 842900/1000000 | loss=2.1755 vocab=1.0275 token=0.7910 | lr=6.23e-05 | 6.13 steps/s | VRAM=25.74 Go [2026-07-18 14:32:50] step 842950/1000000 | loss=2.0152 vocab=1.1830 token=1.5835 | lr=6.23e-05 | 6.12 steps/s | VRAM=26.68 Go [2026-07-18 14:32:58] step 843000/1000000 | loss=2.1097 vocab=0.9221 token=0.9071 | lr=6.22e-05 | 6.15 steps/s | VRAM=28.08 Go [2026-07-18 14:33:06] step 843050/1000000 | loss=2.0605 vocab=1.1284 token=1.1088 | lr=6.22e-05 | 6.14 steps/s | VRAM=28.87 Go [2026-07-18 14:33:15] step 843100/1000000 | loss=2.0182 vocab=0.5024 token=0.7898 | lr=6.22e-05 | 6.16 steps/s | VRAM=31.62 Go [2026-07-18 14:33:23] step 843150/1000000 | loss=2.1976 vocab=1.7126 token=1.2779 | lr=6.22e-05 | 6.12 steps/s | VRAM=25.94 Go [2026-07-18 14:33:31] step 843200/1000000 | loss=2.1520 vocab=1.5189 token=1.1523 | lr=6.21e-05 | 6.15 steps/s | VRAM=29.62 Go [2026-07-18 14:33:39] step 843250/1000000 | loss=2.0969 vocab=0.8469 token=1.2550 | lr=6.21e-05 | 6.13 steps/s | VRAM=26.69 Go [2026-07-18 14:33:47] step 843300/1000000 | loss=2.1370 vocab=0.5941 token=1.2787 | lr=6.21e-05 | 6.16 steps/s | VRAM=28.22 Go [2026-07-18 14:33:55] step 843350/1000000 | loss=2.1938 vocab=0.9399 token=0.9452 | lr=6.20e-05 | 6.15 steps/s | VRAM=29.42 Go [2026-07-18 14:34:03] step 843400/1000000 | loss=2.1682 vocab=1.2033 token=1.5875 | lr=6.20e-05 | 6.16 steps/s | VRAM=28.61 Go [2026-07-18 14:34:11] step 843450/1000000 | loss=2.1318 vocab=1.4467 token=1.2774 | lr=6.20e-05 | 6.17 steps/s | VRAM=27.79 Go [2026-07-18 14:34:20] step 843500/1000000 | loss=2.0750 vocab=0.9810 token=1.0005 | lr=6.20e-05 | 6.14 steps/s | VRAM=25.24 Go [2026-07-18 14:34:28] step 843550/1000000 | loss=2.2715 vocab=0.9236 token=1.5331 | lr=6.19e-05 | 6.14 steps/s | VRAM=32.45 Go [2026-07-18 14:34:36] step 843600/1000000 | loss=2.1765 vocab=0.9201 token=0.9289 | lr=6.19e-05 | 6.12 steps/s | VRAM=32.58 Go [2026-07-18 14:34:44] step 843650/1000000 | loss=2.1234 vocab=0.8143 token=1.1624 | lr=6.19e-05 | 6.16 steps/s | VRAM=31.52 Go [2026-07-18 14:34:52] step 843700/1000000 | loss=2.2342 vocab=0.9462 token=0.9248 | lr=6.18e-05 | 6.16 steps/s | VRAM=32.17 Go [2026-07-18 14:35:00] step 843750/1000000 | loss=2.1167 vocab=0.9089 token=1.1759 | lr=6.18e-05 | 6.15 steps/s | VRAM=28.20 Go [2026-07-18 14:35:08] step 843800/1000000 | loss=2.1416 vocab=1.1028 token=1.4626 | lr=6.18e-05 | 6.14 steps/s | VRAM=29.56 Go [2026-07-18 14:35:17] step 843850/1000000 | loss=2.2255 vocab=1.0754 token=1.4556 | lr=6.18e-05 | 6.16 steps/s | VRAM=31.98 Go [2026-07-18 14:35:25] step 843900/1000000 | loss=2.1148 vocab=1.3943 token=1.6873 | lr=6.17e-05 | 6.14 steps/s | VRAM=28.48 Go [2026-07-18 14:35:33] step 843950/1000000 | loss=2.1692 vocab=0.7565 token=1.2995 | lr=6.17e-05 | 6.13 steps/s | VRAM=32.84 Go [2026-07-18 14:35:41] step 844000/1000000 | loss=2.0565 vocab=0.6345 token=1.1704 | lr=6.17e-05 | 6.14 steps/s | VRAM=32.66 Go [2026-07-18 14:35:49] step 844050/1000000 | loss=2.1144 vocab=1.3050 token=0.8981 | lr=6.17e-05 | 6.13 steps/s | VRAM=32.21 Go [2026-07-18 14:35:57] step 844100/1000000 | loss=2.0977 vocab=1.0049 token=1.3853 | lr=6.16e-05 | 6.15 steps/s | VRAM=24.97 Go [2026-07-18 14:36:05] step 844150/1000000 | loss=2.0615 vocab=0.9104 token=1.0953 | lr=6.16e-05 | 6.12 steps/s | VRAM=28.53 Go [2026-07-18 14:36:14] step 844200/1000000 | loss=2.0504 vocab=0.9758 token=1.2919 | lr=6.16e-05 | 6.10 steps/s | VRAM=26.34 Go [2026-07-18 14:36:22] step 844250/1000000 | loss=2.0901 vocab=0.9126 token=1.3436 | lr=6.15e-05 | 6.15 steps/s | VRAM=32.64 Go [2026-07-18 14:36:30] step 844300/1000000 | loss=2.2582 vocab=1.2475 token=1.5802 | lr=6.15e-05 | 6.15 steps/s | VRAM=28.02 Go [2026-07-18 14:36:38] step 844350/1000000 | loss=1.9614 vocab=0.5119 token=1.3358 | lr=6.15e-05 | 6.15 steps/s | VRAM=31.97 Go [2026-07-18 14:36:46] step 844400/1000000 | loss=2.1075 vocab=1.0225 token=1.3736 | lr=6.15e-05 | 6.13 steps/s | VRAM=29.20 Go [2026-07-18 14:36:54] step 844450/1000000 | loss=2.2205 vocab=0.5544 token=1.1456 | lr=6.14e-05 | 6.14 steps/s | VRAM=26.59 Go [2026-07-18 14:37:03] step 844500/1000000 | loss=2.0660 vocab=0.7743 token=0.9806 | lr=6.14e-05 | 6.13 steps/s | VRAM=32.64 Go [2026-07-18 14:37:11] step 844550/1000000 | loss=2.0569 vocab=0.6396 token=1.0485 | lr=6.14e-05 | 6.14 steps/s | VRAM=32.96 Go [2026-07-18 14:37:19] step 844600/1000000 | loss=2.0519 vocab=0.8947 token=0.9902 | lr=6.13e-05 | 6.12 steps/s | VRAM=28.24 Go [2026-07-18 14:37:27] step 844650/1000000 | loss=2.1359 vocab=1.3039 token=1.1849 | lr=6.13e-05 | 6.12 steps/s | VRAM=31.66 Go [2026-07-18 14:37:35] step 844700/1000000 | loss=2.1626 vocab=1.4122 token=1.4839 | lr=6.13e-05 | 6.13 steps/s | VRAM=25.34 Go [2026-07-18 14:37:43] step 844750/1000000 | loss=2.0966 vocab=1.0278 token=1.1800 | lr=6.13e-05 | 6.12 steps/s | VRAM=28.46 Go [2026-07-18 14:37:51] step 844800/1000000 | loss=1.9927 vocab=1.0640 token=1.3777 | lr=6.12e-05 | 6.12 steps/s | VRAM=28.69 Go [2026-07-18 14:38:00] step 844850/1000000 | loss=2.1515 vocab=0.7717 token=1.0918 | lr=6.12e-05 | 6.12 steps/s | VRAM=31.75 Go [2026-07-18 14:38:08] step 844900/1000000 | loss=2.1383 vocab=1.4549 token=1.6259 | lr=6.12e-05 | 6.11 steps/s | VRAM=24.91 Go [2026-07-18 14:38:16] step 844950/1000000 | loss=2.1056 vocab=1.3049 token=1.1452 | lr=6.11e-05 | 6.08 steps/s | VRAM=28.02 Go [2026-07-18 14:38:24] step 845000/1000000 | loss=2.0949 vocab=1.1400 token=1.4904 | lr=6.11e-05 | 6.11 steps/s | VRAM=32.11 Go [2026-07-18 14:38:32] step 845050/1000000 | loss=2.0273 vocab=0.6237 token=1.3369 | lr=6.11e-05 | 6.12 steps/s | VRAM=32.72 Go [2026-07-18 14:38:41] step 845100/1000000 | loss=2.1729 vocab=1.2043 token=1.7114 | lr=6.11e-05 | 6.11 steps/s | VRAM=31.23 Go [2026-07-18 14:38:49] step 845150/1000000 | loss=2.1013 vocab=0.6363 token=1.3785 | lr=6.10e-05 | 6.09 steps/s | VRAM=28.37 Go [2026-07-18 14:38:57] step 845200/1000000 | loss=2.1155 vocab=0.7069 token=1.2687 | lr=6.10e-05 | 6.09 steps/s | VRAM=26.33 Go [2026-07-18 14:39:05] step 845250/1000000 | loss=2.1243 vocab=1.4159 token=1.5272 | lr=6.10e-05 | 6.12 steps/s | VRAM=31.93 Go [2026-07-18 14:39:13] step 845300/1000000 | loss=2.1730 vocab=0.8935 token=1.2668 | lr=6.10e-05 | 6.06 steps/s | VRAM=26.50 Go [2026-07-18 14:39:22] step 845350/1000000 | loss=2.2156 vocab=0.6981 token=1.0679 | lr=6.09e-05 | 6.09 steps/s | VRAM=29.14 Go [2026-07-18 14:39:30] step 845400/1000000 | loss=2.0998 vocab=0.7132 token=1.4344 | lr=6.09e-05 | 6.11 steps/s | VRAM=29.86 Go [2026-07-18 14:39:38] step 845450/1000000 | loss=2.1286 vocab=0.4631 token=1.0469 | lr=6.09e-05 | 6.08 steps/s | VRAM=28.62 Go [2026-07-18 14:39:46] step 845500/1000000 | loss=2.1401 vocab=0.5311 token=1.1809 | lr=6.08e-05 | 6.11 steps/s | VRAM=28.61 Go [2026-07-18 14:39:54] step 845550/1000000 | loss=2.1476 vocab=1.3532 token=1.3912 | lr=6.08e-05 | 6.11 steps/s | VRAM=29.50 Go [2026-07-18 14:40:03] step 845600/1000000 | loss=2.2150 vocab=0.8882 token=0.9440 | lr=6.08e-05 | 6.12 steps/s | VRAM=28.18 Go [2026-07-18 14:40:11] step 845650/1000000 | loss=2.0180 vocab=0.6030 token=1.0499 | lr=6.08e-05 | 6.09 steps/s | VRAM=26.76 Go [2026-07-18 14:40:19] step 845700/1000000 | loss=2.1668 vocab=0.8480 token=1.1730 | lr=6.07e-05 | 6.14 steps/s | VRAM=25.18 Go [2026-07-18 14:40:27] step 845750/1000000 | loss=2.1966 vocab=0.9298 token=1.3724 | lr=6.07e-05 | 6.12 steps/s | VRAM=30.06 Go [2026-07-18 14:40:35] step 845800/1000000 | loss=2.0408 vocab=0.9132 token=1.0399 | lr=6.07e-05 | 6.11 steps/s | VRAM=28.86 Go [2026-07-18 14:40:43] step 845850/1000000 | loss=2.0683 vocab=0.6887 token=1.2738 | lr=6.06e-05 | 6.12 steps/s | VRAM=27.88 Go [2026-07-18 14:40:52] step 845900/1000000 | loss=2.0222 vocab=0.4797 token=1.2098 | lr=6.06e-05 | 6.14 steps/s | VRAM=31.89 Go [2026-07-18 14:41:00] step 845950/1000000 | loss=2.1320 vocab=1.0755 token=1.0331 | lr=6.06e-05 | 6.13 steps/s | VRAM=23.91 Go [2026-07-18 14:41:08] step 846000/1000000 | loss=2.1039 vocab=1.8492 token=1.2637 | lr=6.06e-05 | 6.17 steps/s | VRAM=32.09 Go [2026-07-18 14:41:16] step 846050/1000000 | loss=2.1448 vocab=0.6695 token=0.8219 | lr=6.05e-05 | 6.15 steps/s | VRAM=32.68 Go [2026-07-18 14:41:24] step 846100/1000000 | loss=2.0763 vocab=1.3504 token=1.4620 | lr=6.05e-05 | 6.13 steps/s | VRAM=28.31 Go [2026-07-18 14:41:32] step 846150/1000000 | loss=2.1177 vocab=0.9630 token=0.8492 | lr=6.05e-05 | 6.13 steps/s | VRAM=24.35 Go [2026-07-18 14:41:40] step 846200/1000000 | loss=2.0899 vocab=0.9301 token=1.2739 | lr=6.05e-05 | 6.14 steps/s | VRAM=28.98 Go [2026-07-18 14:41:49] step 846250/1000000 | loss=2.0808 vocab=0.6157 token=1.0171 | lr=6.04e-05 | 6.10 steps/s | VRAM=26.06 Go [2026-07-18 14:41:57] step 846300/1000000 | loss=2.1414 vocab=0.8853 token=1.2029 | lr=6.04e-05 | 6.16 steps/s | VRAM=27.58 Go [2026-07-18 14:42:05] step 846350/1000000 | loss=2.1523 vocab=0.9985 token=1.3958 | lr=6.04e-05 | 6.12 steps/s | VRAM=28.31 Go [2026-07-18 14:42:13] step 846400/1000000 | loss=2.1929 vocab=1.0735 token=1.2683 | lr=6.03e-05 | 6.11 steps/s | VRAM=32.43 Go [2026-07-18 14:42:21] step 846450/1000000 | loss=2.0767 vocab=0.6584 token=1.2600 | lr=6.03e-05 | 6.12 steps/s | VRAM=29.43 Go [2026-07-18 14:42:29] step 846500/1000000 | loss=2.0276 vocab=0.4599 token=1.0285 | lr=6.03e-05 | 6.14 steps/s | VRAM=25.81 Go [2026-07-18 14:42:38] step 846550/1000000 | loss=2.1535 vocab=0.6439 token=1.0033 | lr=6.03e-05 | 6.14 steps/s | VRAM=32.91 Go [2026-07-18 14:42:46] step 846600/1000000 | loss=2.0711 vocab=0.8684 token=1.2983 | lr=6.02e-05 | 6.10 steps/s | VRAM=27.06 Go [2026-07-18 14:42:54] step 846650/1000000 | loss=2.0452 vocab=0.8597 token=1.4457 | lr=6.02e-05 | 6.17 steps/s | VRAM=28.06 Go [2026-07-18 14:43:02] step 846700/1000000 | loss=2.1455 vocab=0.7173 token=1.4284 | lr=6.02e-05 | 6.16 steps/s | VRAM=26.06 Go [2026-07-18 14:43:10] step 846750/1000000 | loss=2.1134 vocab=0.6068 token=0.9982 | lr=6.01e-05 | 6.16 steps/s | VRAM=31.56 Go [2026-07-18 14:43:18] step 846800/1000000 | loss=2.0608 vocab=1.1871 token=1.3661 | lr=6.01e-05 | 6.14 steps/s | VRAM=29.10 Go [2026-07-18 14:43:26] step 846850/1000000 | loss=2.0424 vocab=0.6557 token=0.9377 | lr=6.01e-05 | 6.13 steps/s | VRAM=32.63 Go [2026-07-18 14:43:35] step 846900/1000000 | loss=2.0984 vocab=0.6481 token=1.2062 | lr=6.01e-05 | 6.16 steps/s | VRAM=27.37 Go [2026-07-18 14:43:43] step 846950/1000000 | loss=2.0184 vocab=0.4713 token=0.8721 | lr=6.00e-05 | 6.13 steps/s | VRAM=27.86 Go [2026-07-18 14:43:51] step 847000/1000000 | loss=2.1398 vocab=0.8005 token=1.3195 | lr=6.00e-05 | 6.15 steps/s | VRAM=27.70 Go [2026-07-18 14:43:59] step 847050/1000000 | loss=2.1531 vocab=0.9224 token=1.0666 | lr=6.00e-05 | 6.14 steps/s | VRAM=27.29 Go [2026-07-18 14:44:07] step 847100/1000000 | loss=2.0018 vocab=1.0952 token=0.9083 | lr=5.99e-05 | 6.16 steps/s | VRAM=28.94 Go [2026-07-18 14:44:15] step 847150/1000000 | loss=2.1767 vocab=0.7453 token=1.1274 | lr=5.99e-05 | 6.12 steps/s | VRAM=32.55 Go [2026-07-18 14:44:23] step 847200/1000000 | loss=2.1329 vocab=1.2905 token=1.5790 | lr=5.99e-05 | 6.12 steps/s | VRAM=27.96 Go [2026-07-18 14:44:32] step 847250/1000000 | loss=2.0478 vocab=0.8940 token=1.1119 | lr=5.99e-05 | 6.13 steps/s | VRAM=29.05 Go [2026-07-18 14:44:40] step 847300/1000000 | loss=2.1803 vocab=0.8884 token=0.9695 | lr=5.98e-05 | 6.14 steps/s | VRAM=29.30 Go [2026-07-18 14:44:48] step 847350/1000000 | loss=2.0388 vocab=1.1589 token=1.2262 | lr=5.98e-05 | 6.14 steps/s | VRAM=26.46 Go [2026-07-18 14:44:56] step 847400/1000000 | loss=2.1906 vocab=0.7707 token=1.6580 | lr=5.98e-05 | 6.15 steps/s | VRAM=24.74 Go [2026-07-18 14:45:04] step 847450/1000000 | loss=2.0720 vocab=0.5802 token=1.2087 | lr=5.98e-05 | 6.10 steps/s | VRAM=29.03 Go [2026-07-18 14:45:12] step 847500/1000000 | loss=2.2092 vocab=1.5719 token=1.2246 | lr=5.97e-05 | 6.14 steps/s | VRAM=31.25 Go [2026-07-18 14:45:21] step 847550/1000000 | loss=2.1534 vocab=0.7147 token=1.4900 | lr=5.97e-05 | 6.13 steps/s | VRAM=27.84 Go [2026-07-18 14:45:29] step 847600/1000000 | loss=2.2219 vocab=0.8365 token=1.3915 | lr=5.97e-05 | 6.14 steps/s | VRAM=29.38 Go [2026-07-18 14:45:37] step 847650/1000000 | loss=2.0994 vocab=0.7872 token=1.1061 | lr=5.96e-05 | 6.11 steps/s | VRAM=28.65 Go [2026-07-18 14:45:45] step 847700/1000000 | loss=2.0997 vocab=1.1365 token=1.1860 | lr=5.96e-05 | 6.15 steps/s | VRAM=27.63 Go [2026-07-18 14:45:53] step 847750/1000000 | loss=2.0109 vocab=0.6504 token=1.1436 | lr=5.96e-05 | 6.12 steps/s | VRAM=29.57 Go [2026-07-18 14:46:01] step 847800/1000000 | loss=2.1417 vocab=0.8434 token=1.0339 | lr=5.96e-05 | 6.12 steps/s | VRAM=31.67 Go [2026-07-18 14:46:09] step 847850/1000000 | loss=2.0797 vocab=0.5331 token=1.3001 | lr=5.95e-05 | 6.16 steps/s | VRAM=31.28 Go [2026-07-18 14:46:18] step 847900/1000000 | loss=2.2371 vocab=1.0390 token=1.3799 | lr=5.95e-05 | 6.16 steps/s | VRAM=27.89 Go [2026-07-18 14:46:26] step 847950/1000000 | loss=2.1570 vocab=0.6567 token=1.0441 | lr=5.95e-05 | 6.10 steps/s | VRAM=26.40 Go [2026-07-18 14:46:34] step 848000/1000000 | loss=2.0931 vocab=0.7900 token=1.2147 | lr=5.94e-05 | 6.14 steps/s | VRAM=31.34 Go [2026-07-18 14:46:42] step 848050/1000000 | loss=2.0994 vocab=0.7917 token=1.1099 | lr=5.94e-05 | 6.12 steps/s | VRAM=28.51 Go [2026-07-18 14:46:50] step 848100/1000000 | loss=2.1471 vocab=0.4331 token=0.9752 | lr=5.94e-05 | 6.10 steps/s | VRAM=28.44 Go [2026-07-18 14:46:58] step 848150/1000000 | loss=2.0978 vocab=0.7207 token=1.8325 | lr=5.94e-05 | 6.12 steps/s | VRAM=26.30 Go [2026-07-18 14:47:07] step 848200/1000000 | loss=2.1052 vocab=0.8388 token=1.4464 | lr=5.93e-05 | 6.12 steps/s | VRAM=26.77 Go [2026-07-18 14:47:15] step 848250/1000000 | loss=2.0938 vocab=0.6720 token=1.1047 | lr=5.93e-05 | 6.14 steps/s | VRAM=24.58 Go [2026-07-18 14:47:23] step 848300/1000000 | loss=2.1047 vocab=0.7931 token=1.2397 | lr=5.93e-05 | 6.13 steps/s | VRAM=28.55 Go [2026-07-18 14:47:31] step 848350/1000000 | loss=2.1127 vocab=0.7444 token=1.5407 | lr=5.92e-05 | 6.14 steps/s | VRAM=32.11 Go [2026-07-18 14:47:39] step 848400/1000000 | loss=2.0948 vocab=0.7284 token=0.7216 | lr=5.92e-05 | 6.17 steps/s | VRAM=31.45 Go [2026-07-18 14:47:47] step 848450/1000000 | loss=2.1174 vocab=0.5482 token=1.3237 | lr=5.92e-05 | 6.11 steps/s | VRAM=23.19 Go [2026-07-18 14:47:56] step 848500/1000000 | loss=2.1759 vocab=0.4116 token=1.1262 | lr=5.92e-05 | 6.12 steps/s | VRAM=29.44 Go [2026-07-18 14:48:04] step 848550/1000000 | loss=2.2164 vocab=0.7088 token=1.1745 | lr=5.91e-05 | 6.14 steps/s | VRAM=29.24 Go [2026-07-18 14:48:12] step 848600/1000000 | loss=2.0032 vocab=0.8227 token=1.3383 | lr=5.91e-05 | 6.16 steps/s | VRAM=32.67 Go [2026-07-18 14:48:20] step 848650/1000000 | loss=2.2224 vocab=0.8463 token=1.1884 | lr=5.91e-05 | 6.13 steps/s | VRAM=28.22 Go [2026-07-18 14:48:28] step 848700/1000000 | loss=2.0735 vocab=0.5609 token=1.2300 | lr=5.90e-05 | 6.14 steps/s | VRAM=31.36 Go [2026-07-18 14:48:36] step 848750/1000000 | loss=2.1094 vocab=0.7468 token=1.0131 | lr=5.90e-05 | 6.11 steps/s | VRAM=29.99 Go [2026-07-18 14:48:44] step 848800/1000000 | loss=2.0604 vocab=0.6623 token=1.4954 | lr=5.90e-05 | 6.11 steps/s | VRAM=28.54 Go [2026-07-18 14:48:53] step 848850/1000000 | loss=2.1139 vocab=0.7130 token=1.2226 | lr=5.90e-05 | 6.14 steps/s | VRAM=28.16 Go [2026-07-18 14:49:01] step 848900/1000000 | loss=2.0621 vocab=0.7167 token=0.9029 | lr=5.89e-05 | 6.13 steps/s | VRAM=27.84 Go [2026-07-18 14:49:09] step 848950/1000000 | loss=2.0468 vocab=0.9306 token=1.5341 | lr=5.89e-05 | 6.16 steps/s | VRAM=28.66 Go [2026-07-18 14:49:17] step 849000/1000000 | loss=2.0125 vocab=0.7224 token=1.5063 | lr=5.89e-05 | 6.16 steps/s | VRAM=31.87 Go [2026-07-18 14:49:25] step 849050/1000000 | loss=2.0838 vocab=0.5727 token=0.9851 | lr=5.89e-05 | 6.15 steps/s | VRAM=28.92 Go [2026-07-18 14:49:33] step 849100/1000000 | loss=2.0952 vocab=0.5735 token=0.7487 | lr=5.88e-05 | 6.11 steps/s | VRAM=24.65 Go [2026-07-18 14:49:41] step 849150/1000000 | loss=1.9789 vocab=0.6911 token=1.3166 | lr=5.88e-05 | 6.14 steps/s | VRAM=25.04 Go [2026-07-18 14:49:50] step 849200/1000000 | loss=2.1404 vocab=1.0233 token=1.6768 | lr=5.88e-05 | 6.16 steps/s | VRAM=31.79 Go [2026-07-18 14:49:58] step 849250/1000000 | loss=2.0793 vocab=0.7127 token=1.1703 | lr=5.87e-05 | 6.11 steps/s | VRAM=28.11 Go [2026-07-18 14:50:06] step 849300/1000000 | loss=2.1519 vocab=1.1709 token=1.1609 | lr=5.87e-05 | 6.15 steps/s | VRAM=27.87 Go [2026-07-18 14:50:14] step 849350/1000000 | loss=2.0077 vocab=1.1633 token=1.0601 | lr=5.87e-05 | 6.16 steps/s | VRAM=32.08 Go [2026-07-18 14:50:22] step 849400/1000000 | loss=2.1139 vocab=0.6003 token=1.0382 | lr=5.87e-05 | 6.13 steps/s | VRAM=26.48 Go [2026-07-18 14:50:30] step 849450/1000000 | loss=2.0838 vocab=1.2172 token=1.6514 | lr=5.86e-05 | 6.14 steps/s | VRAM=26.45 Go [2026-07-18 14:50:38] step 849500/1000000 | loss=2.1988 vocab=0.5481 token=0.7448 | lr=5.86e-05 | 6.14 steps/s | VRAM=27.78 Go [2026-07-18 14:50:47] step 849550/1000000 | loss=2.1632 vocab=1.1392 token=1.2873 | lr=5.86e-05 | 6.16 steps/s | VRAM=24.66 Go [2026-07-18 14:50:55] step 849600/1000000 | loss=2.1667 vocab=1.1610 token=1.2990 | lr=5.85e-05 | 6.17 steps/s | VRAM=29.28 Go [2026-07-18 14:51:03] step 849650/1000000 | loss=2.1570 vocab=0.5094 token=1.1888 | lr=5.85e-05 | 6.14 steps/s | VRAM=28.43 Go [2026-07-18 14:51:11] step 849700/1000000 | loss=2.1851 vocab=0.5241 token=0.9296 | lr=5.85e-05 | 6.10 steps/s | VRAM=26.28 Go [2026-07-18 14:51:19] step 849750/1000000 | loss=2.1658 vocab=0.6435 token=1.3578 | lr=5.85e-05 | 6.14 steps/s | VRAM=32.33 Go [2026-07-18 14:51:27] step 849800/1000000 | loss=2.2013 vocab=1.0595 token=1.3553 | lr=5.84e-05 | 6.12 steps/s | VRAM=27.90 Go [2026-07-18 14:51:36] step 849850/1000000 | loss=2.1432 vocab=1.0216 token=1.5546 | lr=5.84e-05 | 6.10 steps/s | VRAM=28.87 Go [2026-07-18 14:51:44] step 849900/1000000 | loss=2.1786 vocab=0.4828 token=1.0024 | lr=5.84e-05 | 6.15 steps/s | VRAM=25.77 Go [2026-07-18 14:51:52] step 849950/1000000 | loss=2.0919 vocab=1.2268 token=1.0633 | lr=5.83e-05 | 6.11 steps/s | VRAM=32.03 Go [2026-07-18 14:52:00] step 850000/1000000 | loss=2.1752 vocab=1.2789 token=1.5865 | lr=5.83e-05 | 6.15 steps/s | VRAM=31.00 Go [accuracy] step 850000 : 63.29% (log: /workspace/plbert_train/precision_log.csv) [push] step 850000 OK (loss_token=1.5865, accuracy=63.29%) [2026-07-18 14:53:47] step 850050/1000000 | loss=2.0714 vocab=0.5216 token=1.2422 | lr=5.83e-05 | 0.47 steps/s | VRAM=30.81 Go [2026-07-18 14:53:55] step 850100/1000000 | loss=2.1030 vocab=0.6277 token=1.2797 | lr=5.83e-05 | 6.17 steps/s | VRAM=23.00 Go [2026-07-18 14:54:03] step 850150/1000000 | loss=2.1364 vocab=0.6307 token=1.1153 | lr=5.82e-05 | 6.20 steps/s | VRAM=31.66 Go [2026-07-18 14:54:11] step 850200/1000000 | loss=2.0531 vocab=0.6709 token=1.4546 | lr=5.82e-05 | 6.17 steps/s | VRAM=24.73 Go [2026-07-18 14:54:19] step 850250/1000000 | loss=2.2113 vocab=0.9743 token=1.1829 | lr=5.82e-05 | 6.17 steps/s | VRAM=29.52 Go [2026-07-18 14:54:27] step 850300/1000000 | loss=2.0608 vocab=0.5805 token=0.9666 | lr=5.81e-05 | 6.16 steps/s | VRAM=29.21 Go [2026-07-18 14:54:36] step 850350/1000000 | loss=2.1240 vocab=1.2320 token=1.3380 | lr=5.81e-05 | 6.16 steps/s | VRAM=27.93 Go [2026-07-18 14:54:44] step 850400/1000000 | loss=2.1071 vocab=0.5817 token=1.0930 | lr=5.81e-05 | 6.13 steps/s | VRAM=28.84 Go [2026-07-18 14:54:52] step 850450/1000000 | loss=2.1746 vocab=0.9030 token=1.1918 | lr=5.81e-05 | 6.17 steps/s | VRAM=25.26 Go [2026-07-18 14:55:00] step 850500/1000000 | loss=2.0918 vocab=0.6993 token=1.4395 | lr=5.80e-05 | 6.12 steps/s | VRAM=26.30 Go [2026-07-18 14:55:08] step 850550/1000000 | loss=2.1245 vocab=1.1111 token=1.2911 | lr=5.80e-05 | 6.15 steps/s | VRAM=27.19 Go [2026-07-18 14:55:16] step 850600/1000000 | loss=2.0718 vocab=0.9743 token=1.3009 | lr=5.80e-05 | 6.15 steps/s | VRAM=29.95 Go [2026-07-18 14:55:24] step 850650/1000000 | loss=2.0580 vocab=1.3201 token=1.0238 | lr=5.80e-05 | 6.13 steps/s | VRAM=32.03 Go [2026-07-18 14:55:33] step 850700/1000000 | loss=2.0453 vocab=0.4984 token=1.0006 | lr=5.79e-05 | 6.10 steps/s | VRAM=28.86 Go [2026-07-18 14:55:41] step 850750/1000000 | loss=2.1042 vocab=0.8375 token=1.3012 | lr=5.79e-05 | 6.15 steps/s | VRAM=32.01 Go [2026-07-18 14:55:49] step 850800/1000000 | loss=2.1114 vocab=0.7246 token=0.9436 | lr=5.79e-05 | 6.10 steps/s | VRAM=26.27 Go [2026-07-18 14:55:57] step 850850/1000000 | loss=2.0748 vocab=0.3139 token=0.7423 | lr=5.78e-05 | 6.13 steps/s | VRAM=29.21 Go [2026-07-18 14:56:05] step 850900/1000000 | loss=2.0963 vocab=0.5014 token=1.2702 | lr=5.78e-05 | 6.10 steps/s | VRAM=29.18 Go [2026-07-18 14:56:13] step 850950/1000000 | loss=2.0881 vocab=0.9467 token=0.7328 | lr=5.78e-05 | 6.12 steps/s | VRAM=28.59 Go [2026-07-18 14:56:22] step 851000/1000000 | loss=2.0486 vocab=0.6913 token=1.6480 | lr=5.78e-05 | 6.13 steps/s | VRAM=32.10 Go [2026-07-18 14:56:30] step 851050/1000000 | loss=2.1637 vocab=0.9019 token=1.1184 | lr=5.77e-05 | 6.15 steps/s | VRAM=27.78 Go [2026-07-18 14:56:38] step 851100/1000000 | loss=2.0445 vocab=0.9134 token=1.1709 | lr=5.77e-05 | 6.11 steps/s | VRAM=26.48 Go [2026-07-18 14:56:46] step 851150/1000000 | loss=2.0188 vocab=0.5497 token=0.9975 | lr=5.77e-05 | 6.13 steps/s | VRAM=31.81 Go [2026-07-18 14:56:54] step 851200/1000000 | loss=2.1670 vocab=0.9381 token=1.1269 | lr=5.76e-05 | 6.15 steps/s | VRAM=32.22 Go [2026-07-18 14:57:02] step 851250/1000000 | loss=2.0228 vocab=0.7057 token=1.1373 | lr=5.76e-05 | 6.14 steps/s | VRAM=31.74 Go [2026-07-18 14:57:11] step 851300/1000000 | loss=2.0822 vocab=1.0836 token=1.2806 | lr=5.76e-05 | 6.12 steps/s | VRAM=30.07 Go [2026-07-18 14:57:19] step 851350/1000000 | loss=2.1092 vocab=0.9621 token=1.2706 | lr=5.76e-05 | 6.11 steps/s | VRAM=26.86 Go [2026-07-18 14:57:27] step 851400/1000000 | loss=2.0026 vocab=0.7967 token=1.5120 | lr=5.75e-05 | 6.15 steps/s | VRAM=31.68 Go [2026-07-18 14:57:35] step 851450/1000000 | loss=2.0827 vocab=0.5845 token=0.8291 | lr=5.75e-05 | 6.14 steps/s | VRAM=24.71 Go [2026-07-18 14:57:43] step 851500/1000000 | loss=2.0470 vocab=0.8578 token=1.2523 | lr=5.75e-05 | 6.12 steps/s | VRAM=25.90 Go [2026-07-18 14:57:51] step 851550/1000000 | loss=2.1194 vocab=0.8248 token=0.9427 | lr=5.74e-05 | 6.10 steps/s | VRAM=23.35 Go [2026-07-18 14:57:59] step 851600/1000000 | loss=1.9773 vocab=0.7464 token=0.9574 | lr=5.74e-05 | 6.14 steps/s | VRAM=27.85 Go [2026-07-18 14:58:08] step 851650/1000000 | loss=2.2752 vocab=0.7476 token=0.9566 | lr=5.74e-05 | 6.11 steps/s | VRAM=28.10 Go [2026-07-18 14:58:16] step 851700/1000000 | loss=2.1637 vocab=1.0712 token=1.4157 | lr=5.74e-05 | 6.15 steps/s | VRAM=32.99 Go [2026-07-18 14:58:24] step 851750/1000000 | loss=2.1085 vocab=0.4068 token=0.6769 | lr=5.73e-05 | 6.13 steps/s | VRAM=26.40 Go [2026-07-18 14:58:32] step 851800/1000000 | loss=2.1022 vocab=0.9830 token=1.3495 | lr=5.73e-05 | 6.12 steps/s | VRAM=25.56 Go [2026-07-18 14:58:40] step 851850/1000000 | loss=2.0852 vocab=0.9262 token=1.4473 | lr=5.73e-05 | 6.12 steps/s | VRAM=28.16 Go [2026-07-18 14:58:48] step 851900/1000000 | loss=2.1532 vocab=0.9509 token=1.2753 | lr=5.72e-05 | 6.11 steps/s | VRAM=28.92 Go [2026-07-18 14:58:57] step 851950/1000000 | loss=2.0852 vocab=0.6278 token=1.0605 | lr=5.72e-05 | 6.10 steps/s | VRAM=28.61 Go [2026-07-18 14:59:05] step 852000/1000000 | loss=2.1852 vocab=0.8833 token=1.3194 | lr=5.72e-05 | 6.13 steps/s | VRAM=26.34 Go [2026-07-18 14:59:13] step 852050/1000000 | loss=2.1098 vocab=0.9611 token=1.0634 | lr=5.72e-05 | 6.12 steps/s | VRAM=28.52 Go [2026-07-18 14:59:21] step 852100/1000000 | loss=2.1169 vocab=1.1164 token=1.2656 | lr=5.71e-05 | 6.12 steps/s | VRAM=28.75 Go [2026-07-18 14:59:29] step 852150/1000000 | loss=2.0700 vocab=0.7588 token=1.0718 | lr=5.71e-05 | 6.12 steps/s | VRAM=28.55 Go [2026-07-18 14:59:37] step 852200/1000000 | loss=2.1055 vocab=0.9323 token=1.6194 | lr=5.71e-05 | 6.12 steps/s | VRAM=23.70 Go [2026-07-18 14:59:46] step 852250/1000000 | loss=2.2154 vocab=1.3772 token=0.9322 | lr=5.70e-05 | 6.15 steps/s | VRAM=28.53 Go [2026-07-18 14:59:54] step 852300/1000000 | loss=2.1076 vocab=0.6731 token=0.9665 | lr=5.70e-05 | 6.13 steps/s | VRAM=27.77 Go [2026-07-18 15:00:02] step 852350/1000000 | loss=2.0879 vocab=0.9551 token=1.4412 | lr=5.70e-05 | 6.12 steps/s | VRAM=32.64 Go [2026-07-18 15:00:10] step 852400/1000000 | loss=2.0563 vocab=0.8848 token=1.5515 | lr=5.70e-05 | 6.13 steps/s | VRAM=28.58 Go [2026-07-18 15:00:18] step 852450/1000000 | loss=2.1940 vocab=0.9801 token=1.3930 | lr=5.69e-05 | 6.14 steps/s | VRAM=26.86 Go [2026-07-18 15:00:26] step 852500/1000000 | loss=2.1523 vocab=0.9545 token=1.2149 | lr=5.69e-05 | 6.13 steps/s | VRAM=27.62 Go [2026-07-18 15:00:35] step 852550/1000000 | loss=2.2208 vocab=1.3321 token=1.0213 | lr=5.69e-05 | 6.12 steps/s | VRAM=32.08 Go [2026-07-18 15:00:43] step 852600/1000000 | loss=2.2419 vocab=0.7682 token=1.4463 | lr=5.68e-05 | 6.14 steps/s | VRAM=29.46 Go [2026-07-18 15:00:51] step 852650/1000000 | loss=2.0770 vocab=0.7041 token=1.0652 | lr=5.68e-05 | 6.11 steps/s | VRAM=25.86 Go [2026-07-18 15:00:59] step 852700/1000000 | loss=2.0951 vocab=0.7604 token=1.1089 | lr=5.68e-05 | 6.16 steps/s | VRAM=28.32 Go [2026-07-18 15:01:07] step 852750/1000000 | loss=2.1604 vocab=0.7590 token=1.1706 | lr=5.68e-05 | 6.13 steps/s | VRAM=26.26 Go [2026-07-18 15:01:15] step 852800/1000000 | loss=2.0766 vocab=0.4047 token=0.9075 | lr=5.67e-05 | 6.12 steps/s | VRAM=29.10 Go [2026-07-18 15:01:24] step 852850/1000000 | loss=2.1946 vocab=0.9143 token=1.4205 | lr=5.67e-05 | 6.13 steps/s | VRAM=32.19 Go [2026-07-18 15:01:32] step 852900/1000000 | loss=2.1665 vocab=0.7419 token=1.0999 | lr=5.67e-05 | 6.13 steps/s | VRAM=31.99 Go [2026-07-18 15:01:40] step 852950/1000000 | loss=2.1430 vocab=0.9978 token=1.4664 | lr=5.67e-05 | 6.11 steps/s | VRAM=29.10 Go [2026-07-18 15:01:48] step 853000/1000000 | loss=2.1701 vocab=1.2415 token=1.4259 | lr=5.66e-05 | 6.13 steps/s | VRAM=28.22 Go [2026-07-18 15:01:56] step 853050/1000000 | loss=2.1472 vocab=1.2191 token=1.2282 | lr=5.66e-05 | 6.15 steps/s | VRAM=28.42 Go [2026-07-18 15:02:04] step 853100/1000000 | loss=2.0685 vocab=0.9740 token=1.2023 | lr=5.66e-05 | 6.17 steps/s | VRAM=29.04 Go [2026-07-18 15:02:12] step 853150/1000000 | loss=2.1583 vocab=0.9825 token=0.9741 | lr=5.65e-05 | 6.14 steps/s | VRAM=28.98 Go [2026-07-18 15:02:21] step 853200/1000000 | loss=2.0402 vocab=0.7218 token=1.1140 | lr=5.65e-05 | 6.13 steps/s | VRAM=25.95 Go [2026-07-18 15:02:29] step 853250/1000000 | loss=2.0674 vocab=1.1041 token=1.1662 | lr=5.65e-05 | 6.15 steps/s | VRAM=29.20 Go [2026-07-18 15:02:37] step 853300/1000000 | loss=2.1048 vocab=0.5164 token=0.8094 | lr=5.65e-05 | 6.13 steps/s | VRAM=32.57 Go [2026-07-18 15:02:45] step 853350/1000000 | loss=2.0563 vocab=1.0570 token=1.1125 | lr=5.64e-05 | 6.15 steps/s | VRAM=32.78 Go [2026-07-18 15:02:53] step 853400/1000000 | loss=2.0817 vocab=0.5667 token=1.2136 | lr=5.64e-05 | 6.12 steps/s | VRAM=27.83 Go [2026-07-18 15:03:01] step 853450/1000000 | loss=2.0983 vocab=1.1200 token=1.5478 | lr=5.64e-05 | 6.14 steps/s | VRAM=32.80 Go [2026-07-18 15:03:09] step 853500/1000000 | loss=2.1454 vocab=0.7207 token=1.0556 | lr=5.63e-05 | 6.12 steps/s | VRAM=28.19 Go [2026-07-18 15:03:18] step 853550/1000000 | loss=2.0040 vocab=0.4960 token=1.0945 | lr=5.63e-05 | 6.15 steps/s | VRAM=28.77 Go [2026-07-18 15:03:26] step 853600/1000000 | loss=2.1934 vocab=1.2379 token=1.2604 | lr=5.63e-05 | 6.14 steps/s | VRAM=28.10 Go [2026-07-18 15:03:34] step 853650/1000000 | loss=2.0964 vocab=0.5962 token=1.5117 | lr=5.63e-05 | 6.13 steps/s | VRAM=28.91 Go [2026-07-18 15:03:42] step 853700/1000000 | loss=2.0891 vocab=0.4797 token=0.9332 | lr=5.62e-05 | 6.12 steps/s | VRAM=32.01 Go [2026-07-18 15:03:50] step 853750/1000000 | loss=2.0657 vocab=0.7475 token=1.3153 | lr=5.62e-05 | 6.11 steps/s | VRAM=31.65 Go [2026-07-18 15:03:58] step 853800/1000000 | loss=2.1798 vocab=1.0887 token=1.3701 | lr=5.62e-05 | 6.14 steps/s | VRAM=31.71 Go [2026-07-18 15:04:07] step 853850/1000000 | loss=2.1560 vocab=0.8473 token=1.3968 | lr=5.61e-05 | 6.13 steps/s | VRAM=28.98 Go [2026-07-18 15:04:15] step 853900/1000000 | loss=2.1026 vocab=0.7838 token=1.6424 | lr=5.61e-05 | 6.12 steps/s | VRAM=28.75 Go [2026-07-18 15:04:23] step 853950/1000000 | loss=2.1169 vocab=0.5311 token=0.8564 | lr=5.61e-05 | 6.14 steps/s | VRAM=26.06 Go [2026-07-18 15:04:31] step 854000/1000000 | loss=2.0814 vocab=0.6866 token=1.0689 | lr=5.61e-05 | 6.15 steps/s | VRAM=29.33 Go [2026-07-18 15:04:39] step 854050/1000000 | loss=2.1785 vocab=0.5762 token=1.0357 | lr=5.60e-05 | 6.11 steps/s | VRAM=29.47 Go [2026-07-18 15:04:47] step 854100/1000000 | loss=2.1301 vocab=0.9484 token=1.1224 | lr=5.60e-05 | 6.15 steps/s | VRAM=32.15 Go [2026-07-18 15:04:55] step 854150/1000000 | loss=2.1198 vocab=1.1202 token=0.9776 | lr=5.60e-05 | 6.11 steps/s | VRAM=29.12 Go [2026-07-18 15:05:04] step 854200/1000000 | loss=2.1778 vocab=0.4198 token=0.6309 | lr=5.59e-05 | 6.09 steps/s | VRAM=28.41 Go [2026-07-18 15:05:12] step 854250/1000000 | loss=2.1976 vocab=0.7076 token=1.1702 | lr=5.59e-05 | 6.16 steps/s | VRAM=29.22 Go [2026-07-18 15:05:20] step 854300/1000000 | loss=2.0901 vocab=0.8315 token=1.1077 | lr=5.59e-05 | 6.11 steps/s | VRAM=28.28 Go [2026-07-18 15:05:28] step 854350/1000000 | loss=2.1361 vocab=0.6936 token=0.8275 | lr=5.59e-05 | 6.11 steps/s | VRAM=26.80 Go [2026-07-18 15:05:36] step 854400/1000000 | loss=2.0374 vocab=0.4925 token=0.9524 | lr=5.58e-05 | 6.14 steps/s | VRAM=28.94 Go [2026-07-18 15:05:44] step 854450/1000000 | loss=1.9996 vocab=0.9712 token=1.1023 | lr=5.58e-05 | 6.16 steps/s | VRAM=24.69 Go [2026-07-18 15:05:53] step 854500/1000000 | loss=2.1284 vocab=0.4313 token=1.2287 | lr=5.58e-05 | 6.13 steps/s | VRAM=32.08 Go [2026-07-18 15:06:01] step 854550/1000000 | loss=2.1705 vocab=0.9196 token=1.4073 | lr=5.57e-05 | 6.12 steps/s | VRAM=29.15 Go [2026-07-18 15:06:09] step 854600/1000000 | loss=2.0772 vocab=0.3908 token=1.0023 | lr=5.57e-05 | 6.12 steps/s | VRAM=28.90 Go [2026-07-18 15:06:17] step 854650/1000000 | loss=2.1663 vocab=0.9459 token=1.4138 | lr=5.57e-05 | 6.10 steps/s | VRAM=26.66 Go [2026-07-18 15:06:25] step 854700/1000000 | loss=2.0870 vocab=0.6976 token=1.2004 | lr=5.57e-05 | 6.15 steps/s | VRAM=32.20 Go [2026-07-18 15:06:33] step 854750/1000000 | loss=2.1343 vocab=1.3600 token=1.3615 | lr=5.56e-05 | 6.09 steps/s | VRAM=28.89 Go [2026-07-18 15:06:42] step 854800/1000000 | loss=2.1734 vocab=1.0888 token=1.4542 | lr=5.56e-05 | 6.13 steps/s | VRAM=32.56 Go [2026-07-18 15:06:50] step 854850/1000000 | loss=2.0908 vocab=1.1161 token=1.3216 | lr=5.56e-05 | 6.12 steps/s | VRAM=28.58 Go [2026-07-18 15:06:58] step 854900/1000000 | loss=2.1036 vocab=0.8949 token=1.2259 | lr=5.55e-05 | 6.12 steps/s | VRAM=29.19 Go [2026-07-18 15:07:06] step 854950/1000000 | loss=2.1239 vocab=0.6547 token=1.0483 | lr=5.55e-05 | 6.11 steps/s | VRAM=28.29 Go [2026-07-18 15:07:14] step 855000/1000000 | loss=2.2008 vocab=0.9097 token=1.5550 | lr=5.55e-05 | 6.12 steps/s | VRAM=32.94 Go [2026-07-18 15:07:22] step 855050/1000000 | loss=1.9944 vocab=1.0371 token=1.3505 | lr=5.55e-05 | 6.12 steps/s | VRAM=32.00 Go [2026-07-18 15:07:31] step 855100/1000000 | loss=2.1750 vocab=0.8323 token=1.1081 | lr=5.54e-05 | 6.13 steps/s | VRAM=28.06 Go [2026-07-18 15:07:39] step 855150/1000000 | loss=2.1029 vocab=0.8125 token=1.5359 | lr=5.54e-05 | 6.12 steps/s | VRAM=28.28 Go [2026-07-18 15:07:47] step 855200/1000000 | loss=2.1298 vocab=0.9257 token=1.1315 | lr=5.54e-05 | 6.12 steps/s | VRAM=32.95 Go [2026-07-18 15:07:55] step 855250/1000000 | loss=2.0949 vocab=0.5312 token=1.2751 | lr=5.53e-05 | 6.16 steps/s | VRAM=28.12 Go [2026-07-18 15:08:03] step 855300/1000000 | loss=2.1478 vocab=0.6790 token=1.0950 | lr=5.53e-05 | 6.13 steps/s | VRAM=28.74 Go [2026-07-18 15:08:11] step 855350/1000000 | loss=2.1304 vocab=0.4419 token=0.9899 | lr=5.53e-05 | 6.11 steps/s | VRAM=33.01 Go [2026-07-18 15:08:20] step 855400/1000000 | loss=2.1779 vocab=1.0427 token=1.2197 | lr=5.53e-05 | 6.12 steps/s | VRAM=26.47 Go [2026-07-18 15:08:28] step 855450/1000000 | loss=2.1320 vocab=0.7990 token=1.0892 | lr=5.52e-05 | 6.12 steps/s | VRAM=32.06 Go [2026-07-18 15:08:36] step 855500/1000000 | loss=2.0670 vocab=0.7862 token=1.1108 | lr=5.52e-05 | 6.07 steps/s | VRAM=23.65 Go [2026-07-18 15:08:44] step 855550/1000000 | loss=2.1357 vocab=1.0907 token=1.3225 | lr=5.52e-05 | 6.14 steps/s | VRAM=31.72 Go [2026-07-18 15:08:52] step 855600/1000000 | loss=2.1882 vocab=1.5398 token=1.3568 | lr=5.51e-05 | 6.13 steps/s | VRAM=28.62 Go [2026-07-18 15:09:01] step 855650/1000000 | loss=2.0835 vocab=0.7994 token=1.1895 | lr=5.51e-05 | 6.09 steps/s | VRAM=31.72 Go [2026-07-18 15:09:09] step 855700/1000000 | loss=2.1690 vocab=0.5321 token=1.0918 | lr=5.51e-05 | 6.10 steps/s | VRAM=25.89 Go [2026-07-18 15:09:17] step 855750/1000000 | loss=2.0473 vocab=0.7755 token=1.1681 | lr=5.51e-05 | 6.14 steps/s | VRAM=24.55 Go [2026-07-18 15:09:25] step 855800/1000000 | loss=2.1593 vocab=1.0906 token=1.2348 | lr=5.50e-05 | 6.11 steps/s | VRAM=30.02 Go [2026-07-18 15:09:33] step 855850/1000000 | loss=2.1450 vocab=0.7643 token=1.0985 | lr=5.50e-05 | 6.14 steps/s | VRAM=29.00 Go [2026-07-18 15:09:41] step 855900/1000000 | loss=2.0103 vocab=1.0124 token=1.2975 | lr=5.50e-05 | 6.13 steps/s | VRAM=28.05 Go [2026-07-18 15:09:49] step 855950/1000000 | loss=1.9953 vocab=0.8838 token=1.1346 | lr=5.49e-05 | 6.15 steps/s | VRAM=29.05 Go [2026-07-18 15:09:58] step 856000/1000000 | loss=2.1690 vocab=1.1039 token=1.2570 | lr=5.49e-05 | 6.11 steps/s | VRAM=25.48 Go [2026-07-18 15:10:06] step 856050/1000000 | loss=2.0152 vocab=0.7087 token=1.2467 | lr=5.49e-05 | 6.11 steps/s | VRAM=32.54 Go [2026-07-18 15:10:14] step 856100/1000000 | loss=2.0959 vocab=0.8615 token=1.3325 | lr=5.49e-05 | 6.09 steps/s | VRAM=28.88 Go [2026-07-18 15:10:22] step 856150/1000000 | loss=2.1252 vocab=1.3988 token=1.2265 | lr=5.48e-05 | 6.12 steps/s | VRAM=32.21 Go [2026-07-18 15:10:30] step 856200/1000000 | loss=2.2211 vocab=0.8217 token=1.3763 | lr=5.48e-05 | 6.10 steps/s | VRAM=29.63 Go [2026-07-18 15:10:39] step 856250/1000000 | loss=2.1569 vocab=1.1650 token=1.5023 | lr=5.48e-05 | 6.15 steps/s | VRAM=31.68 Go [2026-07-18 15:10:47] step 856300/1000000 | loss=2.0073 vocab=0.7503 token=1.3375 | lr=5.48e-05 | 6.14 steps/s | VRAM=32.41 Go [2026-07-18 15:10:55] step 856350/1000000 | loss=2.1585 vocab=1.0000 token=1.4528 | lr=5.47e-05 | 6.09 steps/s | VRAM=28.70 Go [2026-07-18 15:11:03] step 856400/1000000 | loss=2.0484 vocab=0.6288 token=1.2931 | lr=5.47e-05 | 6.14 steps/s | VRAM=28.73 Go [2026-07-18 15:11:11] step 856450/1000000 | loss=2.0695 vocab=0.4239 token=1.1654 | lr=5.47e-05 | 6.11 steps/s | VRAM=26.06 Go [2026-07-18 15:11:19] step 856500/1000000 | loss=2.1025 vocab=0.9973 token=1.2897 | lr=5.46e-05 | 6.13 steps/s | VRAM=28.91 Go [2026-07-18 15:11:28] step 856550/1000000 | loss=2.0477 vocab=1.1786 token=1.2803 | lr=5.46e-05 | 6.13 steps/s | VRAM=28.29 Go [2026-07-18 15:11:36] step 856600/1000000 | loss=2.0994 vocab=0.8405 token=0.9207 | lr=5.46e-05 | 6.16 steps/s | VRAM=31.05 Go [2026-07-18 15:11:44] step 856650/1000000 | loss=2.0581 vocab=1.7940 token=1.4586 | lr=5.46e-05 | 6.09 steps/s | VRAM=31.20 Go [2026-07-18 15:11:52] step 856700/1000000 | loss=2.1573 vocab=0.7989 token=1.1438 | lr=5.45e-05 | 6.13 steps/s | VRAM=31.90 Go [2026-07-18 15:12:00] step 856750/1000000 | loss=2.1000 vocab=0.9011 token=1.1300 | lr=5.45e-05 | 6.15 steps/s | VRAM=25.84 Go [2026-07-18 15:12:08] step 856800/1000000 | loss=2.0486 vocab=1.0523 token=1.3799 | lr=5.45e-05 | 6.12 steps/s | VRAM=26.37 Go [2026-07-18 15:12:17] step 856850/1000000 | loss=2.1223 vocab=0.8674 token=1.2989 | lr=5.44e-05 | 6.12 steps/s | VRAM=32.00 Go [2026-07-18 15:12:25] step 856900/1000000 | loss=1.9481 vocab=0.6703 token=1.0065 | lr=5.44e-05 | 6.14 steps/s | VRAM=26.33 Go [2026-07-18 15:12:33] step 856950/1000000 | loss=2.0850 vocab=0.8860 token=0.8877 | lr=5.44e-05 | 6.12 steps/s | VRAM=25.69 Go [2026-07-18 15:12:41] step 857000/1000000 | loss=2.1012 vocab=0.6452 token=1.1098 | lr=5.44e-05 | 6.16 steps/s | VRAM=32.07 Go [2026-07-18 15:12:49] step 857050/1000000 | loss=2.1115 vocab=0.6105 token=1.3520 | lr=5.43e-05 | 6.10 steps/s | VRAM=29.33 Go [2026-07-18 15:12:57] step 857100/1000000 | loss=2.1527 vocab=0.8686 token=1.2664 | lr=5.43e-05 | 6.11 steps/s | VRAM=28.18 Go [2026-07-18 15:13:05] step 857150/1000000 | loss=2.1379 vocab=0.2994 token=0.9451 | lr=5.43e-05 | 6.15 steps/s | VRAM=31.65 Go [2026-07-18 15:13:14] step 857200/1000000 | loss=2.0797 vocab=0.7831 token=1.7758 | lr=5.42e-05 | 6.11 steps/s | VRAM=27.71 Go [2026-07-18 15:13:22] step 857250/1000000 | loss=2.0003 vocab=0.8092 token=1.1059 | lr=5.42e-05 | 6.11 steps/s | VRAM=26.15 Go [2026-07-18 15:13:30] step 857300/1000000 | loss=2.1490 vocab=1.1987 token=1.3337 | lr=5.42e-05 | 6.11 steps/s | VRAM=28.98 Go [2026-07-18 15:13:38] step 857350/1000000 | loss=2.1824 vocab=1.3434 token=1.6086 | lr=5.42e-05 | 6.10 steps/s | VRAM=26.95 Go [2026-07-18 15:13:46] step 857400/1000000 | loss=2.1296 vocab=1.3397 token=1.5305 | lr=5.41e-05 | 6.15 steps/s | VRAM=28.21 Go [2026-07-18 15:13:54] step 857450/1000000 | loss=2.1869 vocab=0.8115 token=1.2009 | lr=5.41e-05 | 6.13 steps/s | VRAM=27.40 Go [2026-07-18 15:14:03] step 857500/1000000 | loss=2.1195 vocab=1.2897 token=1.8332 | lr=5.41e-05 | 6.13 steps/s | VRAM=32.00 Go [2026-07-18 15:14:11] step 857550/1000000 | loss=2.1045 vocab=1.0328 token=1.3904 | lr=5.40e-05 | 6.12 steps/s | VRAM=24.30 Go [2026-07-18 15:14:19] step 857600/1000000 | loss=2.0639 vocab=0.4995 token=1.2380 | lr=5.40e-05 | 6.12 steps/s | VRAM=32.41 Go [2026-07-18 15:14:27] step 857650/1000000 | loss=2.0027 vocab=0.8059 token=1.2288 | lr=5.40e-05 | 6.15 steps/s | VRAM=25.63 Go [2026-07-18 15:14:35] step 857700/1000000 | loss=1.9729 vocab=0.9152 token=1.1976 | lr=5.40e-05 | 6.15 steps/s | VRAM=28.46 Go [2026-07-18 15:14:43] step 857750/1000000 | loss=2.0245 vocab=1.9639 token=1.4754 | lr=5.39e-05 | 6.10 steps/s | VRAM=32.64 Go [2026-07-18 15:14:52] step 857800/1000000 | loss=2.0806 vocab=0.4758 token=1.0991 | lr=5.39e-05 | 6.13 steps/s | VRAM=32.28 Go [2026-07-18 15:15:00] step 857850/1000000 | loss=2.0870 vocab=1.1217 token=1.2339 | lr=5.39e-05 | 6.14 steps/s | VRAM=31.42 Go [2026-07-18 15:15:08] step 857900/1000000 | loss=2.0595 vocab=0.6117 token=0.9980 | lr=5.38e-05 | 6.16 steps/s | VRAM=32.23 Go [2026-07-18 15:15:16] step 857950/1000000 | loss=2.2608 vocab=0.5339 token=0.8924 | lr=5.38e-05 | 6.14 steps/s | VRAM=27.45 Go [2026-07-18 15:15:24] step 858000/1000000 | loss=2.0676 vocab=1.1692 token=1.2327 | lr=5.38e-05 | 6.11 steps/s | VRAM=23.77 Go [2026-07-18 15:15:32] step 858050/1000000 | loss=2.1067 vocab=1.4298 token=1.3973 | lr=5.38e-05 | 6.16 steps/s | VRAM=28.66 Go [2026-07-18 15:15:40] step 858100/1000000 | loss=2.1182 vocab=0.5631 token=1.4301 | lr=5.37e-05 | 6.12 steps/s | VRAM=29.01 Go [2026-07-18 15:15:49] step 858150/1000000 | loss=2.0908 vocab=1.1652 token=1.5717 | lr=5.37e-05 | 6.13 steps/s | VRAM=24.48 Go [2026-07-18 15:15:57] step 858200/1000000 | loss=2.0413 vocab=1.6113 token=1.4841 | lr=5.37e-05 | 6.16 steps/s | VRAM=31.94 Go [2026-07-18 15:16:05] step 858250/1000000 | loss=2.1386 vocab=1.2135 token=0.8534 | lr=5.36e-05 | 6.12 steps/s | VRAM=28.41 Go [2026-07-18 15:16:13] step 858300/1000000 | loss=2.0481 vocab=0.5758 token=0.6538 | lr=5.36e-05 | 6.12 steps/s | VRAM=28.54 Go [2026-07-18 15:16:21] step 858350/1000000 | loss=2.2039 vocab=0.6429 token=1.1640 | lr=5.36e-05 | 6.08 steps/s | VRAM=26.42 Go [2026-07-18 15:16:29] step 858400/1000000 | loss=2.1001 vocab=0.6839 token=0.9560 | lr=5.36e-05 | 6.12 steps/s | VRAM=31.23 Go [2026-07-18 15:16:38] step 858450/1000000 | loss=2.1416 vocab=0.7519 token=0.7674 | lr=5.35e-05 | 6.12 steps/s | VRAM=33.02 Go [2026-07-18 15:16:46] step 858500/1000000 | loss=2.0715 vocab=1.2775 token=1.2431 | lr=5.35e-05 | 6.12 steps/s | VRAM=28.72 Go [2026-07-18 15:16:54] step 858550/1000000 | loss=2.0699 vocab=0.7620 token=0.9545 | lr=5.35e-05 | 6.14 steps/s | VRAM=32.78 Go [2026-07-18 15:17:02] step 858600/1000000 | loss=2.1423 vocab=0.7835 token=1.2890 | lr=5.34e-05 | 6.12 steps/s | VRAM=28.26 Go [2026-07-18 15:17:10] step 858650/1000000 | loss=2.1224 vocab=1.4216 token=1.1808 | lr=5.34e-05 | 6.11 steps/s | VRAM=31.18 Go [2026-07-18 15:17:18] step 858700/1000000 | loss=2.1117 vocab=0.7998 token=1.2200 | lr=5.34e-05 | 6.14 steps/s | VRAM=27.50 Go [2026-07-18 15:17:27] step 858750/1000000 | loss=2.0056 vocab=1.2215 token=1.2320 | lr=5.34e-05 | 6.14 steps/s | VRAM=28.96 Go [2026-07-18 15:17:35] step 858800/1000000 | loss=2.0264 vocab=1.0230 token=1.1912 | lr=5.33e-05 | 6.11 steps/s | VRAM=30.14 Go [2026-07-18 15:17:43] step 858850/1000000 | loss=2.0745 vocab=1.0218 token=1.1103 | lr=5.33e-05 | 6.12 steps/s | VRAM=28.53 Go [2026-07-18 15:17:51] step 858900/1000000 | loss=2.1139 vocab=0.9848 token=1.6161 | lr=5.33e-05 | 6.15 steps/s | VRAM=25.59 Go [2026-07-18 15:17:59] step 858950/1000000 | loss=2.1626 vocab=0.5322 token=1.3141 | lr=5.32e-05 | 6.11 steps/s | VRAM=26.22 Go [2026-07-18 15:18:07] step 859000/1000000 | loss=2.1441 vocab=0.6655 token=1.2976 | lr=5.32e-05 | 6.11 steps/s | VRAM=26.48 Go [2026-07-18 15:18:16] step 859050/1000000 | loss=2.0890 vocab=0.9224 token=1.5357 | lr=5.32e-05 | 6.12 steps/s | VRAM=25.15 Go [2026-07-18 15:18:24] step 859100/1000000 | loss=2.0482 vocab=0.9305 token=0.8206 | lr=5.32e-05 | 6.13 steps/s | VRAM=29.68 Go [2026-07-18 15:18:32] step 859150/1000000 | loss=2.0370 vocab=0.7187 token=1.2571 | lr=5.31e-05 | 6.12 steps/s | VRAM=25.93 Go [2026-07-18 15:18:40] step 859200/1000000 | loss=2.1289 vocab=0.8825 token=1.3254 | lr=5.31e-05 | 6.10 steps/s | VRAM=28.50 Go [2026-07-18 15:18:48] step 859250/1000000 | loss=2.0605 vocab=0.6899 token=0.9804 | lr=5.31e-05 | 6.12 steps/s | VRAM=26.12 Go [2026-07-18 15:18:57] step 859300/1000000 | loss=2.1214 vocab=1.1359 token=1.3613 | lr=5.30e-05 | 6.09 steps/s | VRAM=29.12 Go [2026-07-18 15:19:05] step 859350/1000000 | loss=2.1052 vocab=0.9203 token=1.4349 | lr=5.30e-05 | 6.14 steps/s | VRAM=32.56 Go [2026-07-18 15:19:13] step 859400/1000000 | loss=2.1672 vocab=1.6125 token=1.4088 | lr=5.30e-05 | 6.15 steps/s | VRAM=31.50 Go [2026-07-18 15:19:21] step 859450/1000000 | loss=2.1969 vocab=1.0719 token=1.4692 | lr=5.30e-05 | 6.13 steps/s | VRAM=32.66 Go [2026-07-18 15:19:29] step 859500/1000000 | loss=2.1559 vocab=0.6522 token=0.9535 | lr=5.29e-05 | 6.10 steps/s | VRAM=29.36 Go [2026-07-18 15:19:37] step 859550/1000000 | loss=2.0521 vocab=1.0237 token=1.1374 | lr=5.29e-05 | 6.11 steps/s | VRAM=28.95 Go [2026-07-18 15:19:45] step 859600/1000000 | loss=2.0721 vocab=0.9423 token=1.3585 | lr=5.29e-05 | 6.12 steps/s | VRAM=28.99 Go [2026-07-18 15:19:54] step 859650/1000000 | loss=2.0998 vocab=1.0558 token=1.1272 | lr=5.28e-05 | 6.15 steps/s | VRAM=31.16 Go [2026-07-18 15:20:02] step 859700/1000000 | loss=2.1632 vocab=1.6920 token=1.4630 | lr=5.28e-05 | 6.13 steps/s | VRAM=32.37 Go [2026-07-18 15:20:10] step 859750/1000000 | loss=2.0756 vocab=0.7069 token=1.2063 | lr=5.28e-05 | 6.12 steps/s | VRAM=25.95 Go [2026-07-18 15:20:18] step 859800/1000000 | loss=2.2230 vocab=1.2695 token=1.2365 | lr=5.28e-05 | 6.12 steps/s | VRAM=28.77 Go [2026-07-18 15:20:26] step 859850/1000000 | loss=2.1830 vocab=0.7866 token=1.1867 | lr=5.27e-05 | 6.14 steps/s | VRAM=31.26 Go [2026-07-18 15:20:34] step 859900/1000000 | loss=2.1433 vocab=0.8677 token=1.6449 | lr=5.27e-05 | 6.12 steps/s | VRAM=29.24 Go [2026-07-18 15:20:43] step 859950/1000000 | loss=2.1368 vocab=1.1262 token=1.2236 | lr=5.27e-05 | 6.16 steps/s | VRAM=31.68 Go [2026-07-18 15:20:51] step 860000/1000000 | loss=2.0628 vocab=0.4911 token=0.9060 | lr=5.26e-05 | 6.10 steps/s | VRAM=28.94 Go [2026-07-18 15:20:59] step 860050/1000000 | loss=2.1500 vocab=0.9605 token=1.2185 | lr=5.26e-05 | 6.14 steps/s | VRAM=28.91 Go [2026-07-18 15:21:07] step 860100/1000000 | loss=2.1376 vocab=0.9221 token=1.2339 | lr=5.26e-05 | 6.10 steps/s | VRAM=29.52 Go [2026-07-18 15:21:15] step 860150/1000000 | loss=2.0631 vocab=1.3876 token=1.2166 | lr=5.26e-05 | 6.11 steps/s | VRAM=32.91 Go [2026-07-18 15:21:23] step 860200/1000000 | loss=2.1868 vocab=1.2818 token=1.1658 | lr=5.25e-05 | 6.13 steps/s | VRAM=30.08 Go [2026-07-18 15:21:32] step 860250/1000000 | loss=2.2009 vocab=1.1954 token=0.7916 | lr=5.25e-05 | 6.15 steps/s | VRAM=28.18 Go [2026-07-18 15:21:40] step 860300/1000000 | loss=2.0374 vocab=0.7727 token=1.4505 | lr=5.25e-05 | 6.13 steps/s | VRAM=31.61 Go [2026-07-18 15:21:48] step 860350/1000000 | loss=2.0760 vocab=1.0656 token=1.3312 | lr=5.24e-05 | 6.10 steps/s | VRAM=32.78 Go [2026-07-18 15:21:56] step 860400/1000000 | loss=2.1927 vocab=0.7459 token=1.4252 | lr=5.24e-05 | 6.14 steps/s | VRAM=24.82 Go [2026-07-18 15:22:04] step 860450/1000000 | loss=2.0000 vocab=0.3109 token=0.5888 | lr=5.24e-05 | 6.15 steps/s | VRAM=28.52 Go [2026-07-18 15:22:12] step 860500/1000000 | loss=2.0365 vocab=0.6690 token=1.0750 | lr=5.24e-05 | 6.11 steps/s | VRAM=29.70 Go [2026-07-18 15:22:21] step 860550/1000000 | loss=2.0008 vocab=0.7628 token=1.3498 | lr=5.23e-05 | 6.11 steps/s | VRAM=23.15 Go [2026-07-18 15:22:29] step 860600/1000000 | loss=2.1424 vocab=1.0062 token=1.6250 | lr=5.23e-05 | 6.11 steps/s | VRAM=24.24 Go [2026-07-18 15:22:37] step 860650/1000000 | loss=2.1410 vocab=1.0762 token=1.5666 | lr=5.23e-05 | 6.11 steps/s | VRAM=29.63 Go [2026-07-18 15:22:45] step 860700/1000000 | loss=2.1746 vocab=1.1242 token=0.9019 | lr=5.22e-05 | 6.11 steps/s | VRAM=29.21 Go [2026-07-18 15:22:53] step 860750/1000000 | loss=2.1129 vocab=0.6936 token=1.1644 | lr=5.22e-05 | 6.12 steps/s | VRAM=28.36 Go [2026-07-18 15:23:01] step 860800/1000000 | loss=2.1864 vocab=1.0897 token=1.6289 | lr=5.22e-05 | 6.13 steps/s | VRAM=32.57 Go [2026-07-18 15:23:10] step 860850/1000000 | loss=2.0218 vocab=1.0794 token=1.2611 | lr=5.22e-05 | 6.12 steps/s | VRAM=28.61 Go [2026-07-18 15:23:18] step 860900/1000000 | loss=2.1390 vocab=0.3270 token=1.0331 | lr=5.21e-05 | 6.12 steps/s | VRAM=28.43 Go [2026-07-18 15:23:26] step 860950/1000000 | loss=2.1033 vocab=0.8460 token=1.5351 | lr=5.21e-05 | 6.12 steps/s | VRAM=33.02 Go [2026-07-18 15:23:34] step 861000/1000000 | loss=2.1243 vocab=1.1589 token=1.4616 | lr=5.21e-05 | 6.11 steps/s | VRAM=29.60 Go [2026-07-18 15:23:42] step 861050/1000000 | loss=2.1321 vocab=0.5664 token=1.2162 | lr=5.20e-05 | 6.17 steps/s | VRAM=28.34 Go [2026-07-18 15:23:50] step 861100/1000000 | loss=2.1511 vocab=1.0947 token=1.4595 | lr=5.20e-05 | 6.10 steps/s | VRAM=29.40 Go [2026-07-18 15:23:59] step 861150/1000000 | loss=2.1076 vocab=0.5816 token=1.3568 | lr=5.20e-05 | 6.12 steps/s | VRAM=26.11 Go [2026-07-18 15:24:07] step 861200/1000000 | loss=2.1088 vocab=0.7444 token=1.1192 | lr=5.20e-05 | 6.13 steps/s | VRAM=28.45 Go [2026-07-18 15:24:15] step 861250/1000000 | loss=2.1447 vocab=1.0129 token=1.2787 | lr=5.19e-05 | 6.11 steps/s | VRAM=32.59 Go [2026-07-18 15:24:23] step 861300/1000000 | loss=2.0628 vocab=1.0299 token=0.8587 | lr=5.19e-05 | 6.14 steps/s | VRAM=28.99 Go [2026-07-18 15:24:31] step 861350/1000000 | loss=2.0349 vocab=0.9813 token=1.4418 | lr=5.19e-05 | 6.15 steps/s | VRAM=25.73 Go [2026-07-18 15:24:39] step 861400/1000000 | loss=2.1162 vocab=0.6673 token=1.0198 | lr=5.18e-05 | 6.13 steps/s | VRAM=32.49 Go [2026-07-18 15:24:47] step 861450/1000000 | loss=2.1181 vocab=0.9444 token=1.2211 | lr=5.18e-05 | 6.14 steps/s | VRAM=27.77 Go [2026-07-18 15:24:56] step 861500/1000000 | loss=2.1275 vocab=0.8466 token=1.3366 | lr=5.18e-05 | 6.13 steps/s | VRAM=32.61 Go [2026-07-18 15:25:04] step 861550/1000000 | loss=2.1133 vocab=0.6549 token=0.5710 | lr=5.18e-05 | 6.13 steps/s | VRAM=28.18 Go [2026-07-18 15:25:12] step 861600/1000000 | loss=2.0244 vocab=0.6475 token=1.0243 | lr=5.17e-05 | 6.15 steps/s | VRAM=32.45 Go [2026-07-18 15:25:20] step 861650/1000000 | loss=2.0668 vocab=0.8542 token=0.9448 | lr=5.17e-05 | 6.15 steps/s | VRAM=31.76 Go [2026-07-18 15:25:28] step 861700/1000000 | loss=2.1445 vocab=0.6507 token=1.1013 | lr=5.17e-05 | 6.12 steps/s | VRAM=29.41 Go [2026-07-18 15:25:36] step 861750/1000000 | loss=2.0555 vocab=0.9542 token=1.3279 | lr=5.16e-05 | 6.13 steps/s | VRAM=29.48 Go [2026-07-18 15:25:44] step 861800/1000000 | loss=2.0481 vocab=1.1285 token=1.4297 | lr=5.16e-05 | 6.16 steps/s | VRAM=31.66 Go [2026-07-18 15:25:53] step 861850/1000000 | loss=2.1063 vocab=0.6169 token=1.1939 | lr=5.16e-05 | 6.12 steps/s | VRAM=25.95 Go [2026-07-18 15:26:01] step 861900/1000000 | loss=2.0843 vocab=0.5676 token=0.7888 | lr=5.16e-05 | 6.14 steps/s | VRAM=31.88 Go [2026-07-18 15:26:09] step 861950/1000000 | loss=2.0952 vocab=0.8383 token=1.3640 | lr=5.15e-05 | 6.10 steps/s | VRAM=32.50 Go [2026-07-18 15:26:17] step 862000/1000000 | loss=2.1516 vocab=1.0334 token=1.4946 | lr=5.15e-05 | 6.11 steps/s | VRAM=25.47 Go [2026-07-18 15:26:25] step 862050/1000000 | loss=2.1819 vocab=0.7740 token=1.5066 | lr=5.15e-05 | 6.13 steps/s | VRAM=28.98 Go [2026-07-18 15:26:34] step 862100/1000000 | loss=2.1537 vocab=1.0748 token=1.3502 | lr=5.14e-05 | 6.10 steps/s | VRAM=32.72 Go [2026-07-18 15:26:42] step 862150/1000000 | loss=2.0597 vocab=0.7881 token=1.3803 | lr=5.14e-05 | 6.11 steps/s | VRAM=29.92 Go [2026-07-18 15:26:50] step 862200/1000000 | loss=2.0026 vocab=0.6802 token=1.1887 | lr=5.14e-05 | 6.11 steps/s | VRAM=28.95 Go [2026-07-18 15:26:58] step 862250/1000000 | loss=2.1719 vocab=0.4172 token=1.2610 | lr=5.14e-05 | 6.12 steps/s | VRAM=32.45 Go [2026-07-18 15:27:06] step 862300/1000000 | loss=2.0619 vocab=0.6136 token=1.3519 | lr=5.13e-05 | 6.12 steps/s | VRAM=28.70 Go [2026-07-18 15:27:14] step 862350/1000000 | loss=2.1743 vocab=0.9999 token=1.1353 | lr=5.13e-05 | 6.11 steps/s | VRAM=28.41 Go [2026-07-18 15:27:23] step 862400/1000000 | loss=2.1441 vocab=0.4892 token=1.1206 | lr=5.13e-05 | 6.09 steps/s | VRAM=27.55 Go [2026-07-18 15:27:31] step 862450/1000000 | loss=2.1029 vocab=0.5581 token=1.0123 | lr=5.12e-05 | 6.11 steps/s | VRAM=32.30 Go [2026-07-18 15:27:39] step 862500/1000000 | loss=2.1460 vocab=1.3428 token=1.5750 | lr=5.12e-05 | 6.13 steps/s | VRAM=31.59 Go [2026-07-18 15:27:47] step 862550/1000000 | loss=2.2576 vocab=0.4089 token=0.6499 | lr=5.12e-05 | 6.12 steps/s | VRAM=30.83 Go [2026-07-18 15:27:55] step 862600/1000000 | loss=2.2241 vocab=0.9714 token=1.0365 | lr=5.12e-05 | 6.14 steps/s | VRAM=29.02 Go [2026-07-18 15:28:03] step 862650/1000000 | loss=2.1410 vocab=0.7988 token=1.0749 | lr=5.11e-05 | 6.13 steps/s | VRAM=26.02 Go [2026-07-18 15:28:12] step 862700/1000000 | loss=2.0245 vocab=1.0357 token=1.4974 | lr=5.11e-05 | 6.13 steps/s | VRAM=28.67 Go [2026-07-18 15:28:20] step 862750/1000000 | loss=2.1647 vocab=1.1330 token=1.2528 | lr=5.11e-05 | 6.11 steps/s | VRAM=32.13 Go [2026-07-18 15:28:28] step 862800/1000000 | loss=2.0447 vocab=0.6633 token=0.7046 | lr=5.10e-05 | 6.12 steps/s | VRAM=28.84 Go [2026-07-18 15:28:36] step 862850/1000000 | loss=2.1947 vocab=0.5209 token=1.2540 | lr=5.10e-05 | 6.10 steps/s | VRAM=26.63 Go [2026-07-18 15:28:44] step 862900/1000000 | loss=2.0069 vocab=1.1647 token=0.8708 | lr=5.10e-05 | 6.15 steps/s | VRAM=31.73 Go [2026-07-18 15:28:52] step 862950/1000000 | loss=1.9994 vocab=0.8670 token=1.1804 | lr=5.10e-05 | 6.14 steps/s | VRAM=28.44 Go [2026-07-18 15:29:01] step 863000/1000000 | loss=2.1484 vocab=0.8385 token=1.6159 | lr=5.09e-05 | 6.11 steps/s | VRAM=26.42 Go [2026-07-18 15:29:09] step 863050/1000000 | loss=2.1169 vocab=1.0796 token=1.2491 | lr=5.09e-05 | 6.13 steps/s | VRAM=29.06 Go [2026-07-18 15:29:17] step 863100/1000000 | loss=2.1263 vocab=0.7901 token=1.5152 | lr=5.09e-05 | 6.13 steps/s | VRAM=31.47 Go [2026-07-18 15:29:25] step 863150/1000000 | loss=2.1193 vocab=0.7851 token=1.4606 | lr=5.08e-05 | 6.15 steps/s | VRAM=24.90 Go [2026-07-18 15:29:33] step 863200/1000000 | loss=2.1527 vocab=0.5419 token=0.6848 | lr=5.08e-05 | 6.13 steps/s | VRAM=33.01 Go [2026-07-18 15:29:41] step 863250/1000000 | loss=2.2239 vocab=1.0152 token=0.9081 | lr=5.08e-05 | 6.13 steps/s | VRAM=29.05 Go [2026-07-18 15:29:50] step 863300/1000000 | loss=2.1029 vocab=0.6269 token=1.1605 | lr=5.08e-05 | 6.09 steps/s | VRAM=28.59 Go [2026-07-18 15:29:58] step 863350/1000000 | loss=2.1356 vocab=1.5783 token=1.5821 | lr=5.07e-05 | 6.14 steps/s | VRAM=28.18 Go [2026-07-18 15:30:06] step 863400/1000000 | loss=2.2666 vocab=1.0321 token=1.4120 | lr=5.07e-05 | 6.11 steps/s | VRAM=28.16 Go [2026-07-18 15:30:14] step 863450/1000000 | loss=2.0704 vocab=0.6617 token=1.0978 | lr=5.07e-05 | 6.14 steps/s | VRAM=27.98 Go [2026-07-18 15:30:22] step 863500/1000000 | loss=2.0826 vocab=0.6517 token=1.3283 | lr=5.06e-05 | 6.13 steps/s | VRAM=28.10 Go [2026-07-18 15:30:30] step 863550/1000000 | loss=1.9869 vocab=0.6121 token=1.0072 | lr=5.06e-05 | 6.12 steps/s | VRAM=29.19 Go [2026-07-18 15:30:39] step 863600/1000000 | loss=2.1568 vocab=1.1322 token=1.5922 | lr=5.06e-05 | 6.11 steps/s | VRAM=32.98 Go [2026-07-18 15:30:47] step 863650/1000000 | loss=2.1459 vocab=0.5388 token=0.9427 | lr=5.06e-05 | 6.14 steps/s | VRAM=32.68 Go [2026-07-18 15:30:55] step 863700/1000000 | loss=2.0199 vocab=0.6415 token=1.2940 | lr=5.05e-05 | 6.10 steps/s | VRAM=28.46 Go [2026-07-18 15:31:03] step 863750/1000000 | loss=2.1458 vocab=0.7433 token=1.1423 | lr=5.05e-05 | 6.10 steps/s | VRAM=26.36 Go [2026-07-18 15:31:11] step 863800/1000000 | loss=2.1056 vocab=1.0390 token=1.3256 | lr=5.05e-05 | 6.14 steps/s | VRAM=28.34 Go [2026-07-18 15:31:19] step 863850/1000000 | loss=2.1062 vocab=0.7761 token=1.4326 | lr=5.04e-05 | 6.14 steps/s | VRAM=28.07 Go [2026-07-18 15:31:28] step 863900/1000000 | loss=2.1356 vocab=0.6573 token=1.2556 | lr=5.04e-05 | 6.12 steps/s | VRAM=24.51 Go [2026-07-18 15:31:36] step 863950/1000000 | loss=2.0712 vocab=0.4632 token=0.8562 | lr=5.04e-05 | 6.10 steps/s | VRAM=32.91 Go [2026-07-18 15:31:44] step 864000/1000000 | loss=2.1107 vocab=1.1292 token=1.4173 | lr=5.04e-05 | 6.13 steps/s | VRAM=31.56 Go [2026-07-18 15:31:52] step 864050/1000000 | loss=2.0510 vocab=0.6644 token=0.8237 | lr=5.03e-05 | 6.13 steps/s | VRAM=29.70 Go [2026-07-18 15:32:00] step 864100/1000000 | loss=2.1988 vocab=1.1202 token=1.0990 | lr=5.03e-05 | 6.12 steps/s | VRAM=29.77 Go [2026-07-18 15:32:08] step 864150/1000000 | loss=2.0920 vocab=0.8196 token=1.2317 | lr=5.03e-05 | 6.14 steps/s | VRAM=28.78 Go [2026-07-18 15:32:17] step 864200/1000000 | loss=2.0683 vocab=0.6065 token=0.6884 | lr=5.02e-05 | 6.13 steps/s | VRAM=32.45 Go [2026-07-18 15:32:25] step 864250/1000000 | loss=2.1453 vocab=0.6792 token=0.9983 | lr=5.02e-05 | 6.11 steps/s | VRAM=26.32 Go [2026-07-18 15:32:33] step 864300/1000000 | loss=2.1777 vocab=1.4405 token=1.3044 | lr=5.02e-05 | 6.11 steps/s | VRAM=32.61 Go [2026-07-18 15:32:41] step 864350/1000000 | loss=2.1412 vocab=0.5245 token=1.0382 | lr=5.02e-05 | 6.13 steps/s | VRAM=31.10 Go [2026-07-18 15:32:49] step 864400/1000000 | loss=2.0356 vocab=0.9053 token=1.1458 | lr=5.01e-05 | 6.12 steps/s | VRAM=23.69 Go [2026-07-18 15:32:57] step 864450/1000000 | loss=2.1426 vocab=0.6809 token=1.3924 | lr=5.01e-05 | 6.13 steps/s | VRAM=29.22 Go [2026-07-18 15:33:06] step 864500/1000000 | loss=2.1207 vocab=0.9256 token=1.3878 | lr=5.01e-05 | 6.11 steps/s | VRAM=31.65 Go [2026-07-18 15:33:14] step 864550/1000000 | loss=2.1949 vocab=1.2045 token=1.5576 | lr=5.00e-05 | 6.12 steps/s | VRAM=29.38 Go [2026-07-18 15:33:22] step 864600/1000000 | loss=2.0684 vocab=0.9257 token=1.4420 | lr=5.00e-05 | 6.10 steps/s | VRAM=29.46 Go [2026-07-18 15:33:30] step 864650/1000000 | loss=2.1467 vocab=0.8010 token=1.2457 | lr=5.00e-05 | 6.08 steps/s | VRAM=28.04 Go [2026-07-18 15:33:38] step 864700/1000000 | loss=2.0989 vocab=0.6486 token=1.7548 | lr=5.00e-05 | 6.14 steps/s | VRAM=30.33 Go [2026-07-18 15:33:46] step 864750/1000000 | loss=2.1404 vocab=1.2171 token=1.2686 | lr=4.99e-05 | 6.12 steps/s | VRAM=26.78 Go [2026-07-18 15:33:55] step 864800/1000000 | loss=2.0780 vocab=0.3088 token=0.9564 | lr=4.99e-05 | 6.13 steps/s | VRAM=28.52 Go [2026-07-18 15:34:03] step 864850/1000000 | loss=2.0784 vocab=0.5129 token=0.8936 | lr=4.99e-05 | 6.12 steps/s | VRAM=31.81 Go [2026-07-18 15:34:11] step 864900/1000000 | loss=2.0759 vocab=1.2915 token=1.4774 | lr=4.98e-05 | 6.13 steps/s | VRAM=31.92 Go [2026-07-18 15:34:19] step 864950/1000000 | loss=2.1434 vocab=1.0759 token=0.7904 | lr=4.98e-05 | 6.10 steps/s | VRAM=28.76 Go [2026-07-18 15:34:27] step 865000/1000000 | loss=2.1715 vocab=0.6443 token=1.1274 | lr=4.98e-05 | 6.13 steps/s | VRAM=32.53 Go [2026-07-18 15:34:35] step 865050/1000000 | loss=2.1393 vocab=0.7867 token=0.8794 | lr=4.98e-05 | 6.10 steps/s | VRAM=31.82 Go [2026-07-18 15:34:44] step 865100/1000000 | loss=2.1013 vocab=0.6796 token=1.0536 | lr=4.97e-05 | 6.12 steps/s | VRAM=32.24 Go [2026-07-18 15:34:52] step 865150/1000000 | loss=2.1017 vocab=0.5985 token=1.0424 | lr=4.97e-05 | 6.12 steps/s | VRAM=32.71 Go [2026-07-18 15:35:00] step 865200/1000000 | loss=2.0571 vocab=0.3477 token=0.9269 | lr=4.97e-05 | 6.14 steps/s | VRAM=29.13 Go [2026-07-18 15:35:08] step 865250/1000000 | loss=2.1334 vocab=0.8365 token=1.0586 | lr=4.96e-05 | 6.14 steps/s | VRAM=28.87 Go [2026-07-18 15:35:16] step 865300/1000000 | loss=2.0672 vocab=1.1835 token=1.5832 | lr=4.96e-05 | 6.15 steps/s | VRAM=27.83 Go [2026-07-18 15:35:24] step 865350/1000000 | loss=2.0904 vocab=0.9402 token=2.0652 | lr=4.96e-05 | 6.13 steps/s | VRAM=32.14 Go [2026-07-18 15:35:33] step 865400/1000000 | loss=2.0290 vocab=0.9290 token=1.1158 | lr=4.96e-05 | 6.15 steps/s | VRAM=31.57 Go [2026-07-18 15:35:41] step 865450/1000000 | loss=2.0476 vocab=0.9611 token=1.4477 | lr=4.95e-05 | 6.12 steps/s | VRAM=28.04 Go [2026-07-18 15:35:49] step 865500/1000000 | loss=2.1594 vocab=0.4889 token=0.8719 | lr=4.95e-05 | 6.10 steps/s | VRAM=26.71 Go [2026-07-18 15:35:57] step 865550/1000000 | loss=2.2077 vocab=1.4678 token=1.2962 | lr=4.95e-05 | 6.11 steps/s | VRAM=29.47 Go [2026-07-18 15:36:05] step 865600/1000000 | loss=2.0961 vocab=0.6734 token=0.8029 | lr=4.94e-05 | 6.14 steps/s | VRAM=28.73 Go [2026-07-18 15:36:13] step 865650/1000000 | loss=2.1950 vocab=0.4853 token=1.1929 | lr=4.94e-05 | 6.11 steps/s | VRAM=29.08 Go [2026-07-18 15:36:22] step 865700/1000000 | loss=2.1788 vocab=1.0339 token=1.5588 | lr=4.94e-05 | 6.15 steps/s | VRAM=31.92 Go [2026-07-18 15:36:30] step 865750/1000000 | loss=2.0603 vocab=0.6226 token=1.3239 | lr=4.94e-05 | 6.10 steps/s | VRAM=32.80 Go [2026-07-18 15:36:38] step 865800/1000000 | loss=2.0802 vocab=0.3654 token=0.7368 | lr=4.93e-05 | 6.13 steps/s | VRAM=31.93 Go [2026-07-18 15:36:46] step 865850/1000000 | loss=2.0587 vocab=0.6141 token=1.3996 | lr=4.93e-05 | 6.16 steps/s | VRAM=31.71 Go [2026-07-18 15:36:54] step 865900/1000000 | loss=2.0745 vocab=0.5029 token=0.8437 | lr=4.93e-05 | 6.10 steps/s | VRAM=27.51 Go [2026-07-18 15:37:02] step 865950/1000000 | loss=2.0803 vocab=0.9327 token=1.6080 | lr=4.92e-05 | 6.17 steps/s | VRAM=28.36 Go [2026-07-18 15:37:10] step 866000/1000000 | loss=2.1288 vocab=1.3490 token=1.3216 | lr=4.92e-05 | 6.13 steps/s | VRAM=29.56 Go [2026-07-18 15:37:19] step 866050/1000000 | loss=2.1622 vocab=0.7511 token=1.1204 | lr=4.92e-05 | 6.15 steps/s | VRAM=27.96 Go [2026-07-18 15:37:27] step 866100/1000000 | loss=2.1659 vocab=0.9970 token=1.0972 | lr=4.92e-05 | 6.16 steps/s | VRAM=29.06 Go [2026-07-18 15:37:35] step 866150/1000000 | loss=2.1015 vocab=1.4483 token=1.7198 | lr=4.91e-05 | 6.13 steps/s | VRAM=31.98 Go [2026-07-18 15:37:43] step 866200/1000000 | loss=2.1174 vocab=0.9347 token=0.9670 | lr=4.91e-05 | 6.10 steps/s | VRAM=30.17 Go [2026-07-18 15:37:51] step 866250/1000000 | loss=2.0544 vocab=0.7723 token=0.8151 | lr=4.91e-05 | 6.09 steps/s | VRAM=26.17 Go [2026-07-18 15:37:59] step 866300/1000000 | loss=2.0373 vocab=0.9417 token=1.2247 | lr=4.90e-05 | 6.16 steps/s | VRAM=32.32 Go [2026-07-18 15:38:08] step 866350/1000000 | loss=2.1194 vocab=0.8254 token=1.1029 | lr=4.90e-05 | 6.14 steps/s | VRAM=30.82 Go [2026-07-18 15:38:16] step 866400/1000000 | loss=2.1909 vocab=0.6440 token=0.8827 | lr=4.90e-05 | 6.14 steps/s | VRAM=26.30 Go [2026-07-18 15:38:24] step 866450/1000000 | loss=2.1205 vocab=0.8758 token=1.2693 | lr=4.90e-05 | 6.13 steps/s | VRAM=28.56 Go [2026-07-18 15:38:32] step 866500/1000000 | loss=2.1071 vocab=1.0572 token=1.2981 | lr=4.89e-05 | 6.12 steps/s | VRAM=28.37 Go [2026-07-18 15:38:40] step 866550/1000000 | loss=2.0668 vocab=1.0892 token=1.4401 | lr=4.89e-05 | 6.15 steps/s | VRAM=29.39 Go [2026-07-18 15:38:48] step 866600/1000000 | loss=2.1890 vocab=0.7494 token=1.2337 | lr=4.89e-05 | 6.15 steps/s | VRAM=24.95 Go [2026-07-18 15:38:56] step 866650/1000000 | loss=2.1932 vocab=0.5892 token=1.0535 | lr=4.88e-05 | 6.12 steps/s | VRAM=28.50 Go [2026-07-18 15:39:05] step 866700/1000000 | loss=2.0789 vocab=0.8731 token=0.9034 | lr=4.88e-05 | 6.12 steps/s | VRAM=27.86 Go [2026-07-18 15:39:13] step 866750/1000000 | loss=2.1467 vocab=1.4028 token=1.3329 | lr=4.88e-05 | 6.11 steps/s | VRAM=30.10 Go [2026-07-18 15:39:21] step 866800/1000000 | loss=2.0422 vocab=0.7703 token=1.1137 | lr=4.88e-05 | 6.11 steps/s | VRAM=28.40 Go [2026-07-18 15:39:29] step 866850/1000000 | loss=2.0958 vocab=0.9093 token=1.6008 | lr=4.87e-05 | 6.12 steps/s | VRAM=29.16 Go [2026-07-18 15:39:37] step 866900/1000000 | loss=2.1921 vocab=1.1555 token=0.9996 | lr=4.87e-05 | 6.14 steps/s | VRAM=28.73 Go [2026-07-18 15:39:45] step 866950/1000000 | loss=2.1898 vocab=1.1014 token=1.5289 | lr=4.87e-05 | 6.10 steps/s | VRAM=28.80 Go [2026-07-18 15:39:54] step 867000/1000000 | loss=2.1729 vocab=0.7758 token=1.0137 | lr=4.86e-05 | 6.13 steps/s | VRAM=28.84 Go [2026-07-18 15:40:02] step 867050/1000000 | loss=2.0374 vocab=1.3371 token=1.2038 | lr=4.86e-05 | 6.14 steps/s | VRAM=32.48 Go [2026-07-18 15:40:10] step 867100/1000000 | loss=2.0867 vocab=0.7645 token=1.2862 | lr=4.86e-05 | 6.15 steps/s | VRAM=29.39 Go [2026-07-18 15:40:18] step 867150/1000000 | loss=2.1379 vocab=0.6522 token=1.1314 | lr=4.86e-05 | 6.15 steps/s | VRAM=25.63 Go [2026-07-18 15:40:26] step 867200/1000000 | loss=2.0611 vocab=0.6346 token=1.3258 | lr=4.85e-05 | 6.15 steps/s | VRAM=31.10 Go [2026-07-18 15:40:34] step 867250/1000000 | loss=2.1101 vocab=0.4875 token=0.7259 | lr=4.85e-05 | 6.13 steps/s | VRAM=26.32 Go [2026-07-18 15:40:43] step 867300/1000000 | loss=2.1105 vocab=0.8741 token=0.9442 | lr=4.85e-05 | 6.15 steps/s | VRAM=28.34 Go [2026-07-18 15:40:51] step 867350/1000000 | loss=2.1073 vocab=1.0461 token=0.9914 | lr=4.85e-05 | 6.12 steps/s | VRAM=32.19 Go [2026-07-18 15:40:59] step 867400/1000000 | loss=2.1174 vocab=1.1150 token=0.9972 | lr=4.84e-05 | 6.14 steps/s | VRAM=30.72 Go [2026-07-18 15:41:07] step 867450/1000000 | loss=1.9826 vocab=0.9289 token=1.0597 | lr=4.84e-05 | 6.14 steps/s | VRAM=27.62 Go [2026-07-18 15:41:15] step 867500/1000000 | loss=2.1459 vocab=0.9872 token=1.1051 | lr=4.84e-05 | 6.14 steps/s | VRAM=28.76 Go [2026-07-18 15:41:23] step 867550/1000000 | loss=2.0328 vocab=0.5705 token=0.9026 | lr=4.83e-05 | 6.16 steps/s | VRAM=26.29 Go [2026-07-18 15:41:31] step 867600/1000000 | loss=2.0231 vocab=0.9950 token=1.6767 | lr=4.83e-05 | 6.16 steps/s | VRAM=32.51 Go [2026-07-18 15:41:39] step 867650/1000000 | loss=1.9598 vocab=0.9052 token=1.6616 | lr=4.83e-05 | 6.16 steps/s | VRAM=29.65 Go [2026-07-18 15:41:48] step 867700/1000000 | loss=2.0692 vocab=0.6108 token=1.0185 | lr=4.83e-05 | 6.16 steps/s | VRAM=28.55 Go [2026-07-18 15:41:56] step 867750/1000000 | loss=2.1210 vocab=0.5264 token=0.7683 | lr=4.82e-05 | 6.16 steps/s | VRAM=29.52 Go [2026-07-18 15:42:04] step 867800/1000000 | loss=2.0742 vocab=1.5451 token=1.8762 | lr=4.82e-05 | 6.15 steps/s | VRAM=28.61 Go [2026-07-18 15:42:12] step 867850/1000000 | loss=2.1971 vocab=1.0060 token=1.3078 | lr=4.82e-05 | 6.15 steps/s | VRAM=29.01 Go [2026-07-18 15:42:20] step 867900/1000000 | loss=1.9787 vocab=0.5671 token=1.1531 | lr=4.81e-05 | 6.11 steps/s | VRAM=28.31 Go [2026-07-18 15:42:28] step 867950/1000000 | loss=2.1742 vocab=0.8435 token=1.3278 | lr=4.81e-05 | 6.14 steps/s | VRAM=29.11 Go [2026-07-18 15:42:36] step 868000/1000000 | loss=2.1268 vocab=0.8947 token=0.9562 | lr=4.81e-05 | 6.14 steps/s | VRAM=29.04 Go [2026-07-18 15:42:45] step 868050/1000000 | loss=2.1033 vocab=1.3068 token=1.1241 | lr=4.81e-05 | 6.15 steps/s | VRAM=32.18 Go [2026-07-18 15:42:53] step 868100/1000000 | loss=2.0663 vocab=1.0744 token=1.5696 | lr=4.80e-05 | 6.13 steps/s | VRAM=25.85 Go [2026-07-18 15:43:01] step 868150/1000000 | loss=2.2168 vocab=0.5509 token=0.8513 | lr=4.80e-05 | 6.11 steps/s | VRAM=29.17 Go [2026-07-18 15:43:09] step 868200/1000000 | loss=2.2383 vocab=1.2891 token=1.3222 | lr=4.80e-05 | 6.10 steps/s | VRAM=28.12 Go [2026-07-18 15:43:17] step 868250/1000000 | loss=2.1598 vocab=1.2267 token=1.5721 | lr=4.79e-05 | 6.12 steps/s | VRAM=26.04 Go [2026-07-18 15:43:25] step 868300/1000000 | loss=2.0224 vocab=0.5603 token=1.1644 | lr=4.79e-05 | 6.14 steps/s | VRAM=27.84 Go [2026-07-18 15:43:34] step 868350/1000000 | loss=2.1130 vocab=1.4820 token=1.3472 | lr=4.79e-05 | 6.14 steps/s | VRAM=28.84 Go [2026-07-18 15:43:42] step 868400/1000000 | loss=2.0493 vocab=1.3073 token=1.4881 | lr=4.79e-05 | 6.13 steps/s | VRAM=32.76 Go [2026-07-18 15:43:50] step 868450/1000000 | loss=2.1940 vocab=0.9914 token=1.3909 | lr=4.78e-05 | 6.14 steps/s | VRAM=32.60 Go [2026-07-18 15:43:58] step 868500/1000000 | loss=2.0009 vocab=1.2974 token=1.1549 | lr=4.78e-05 | 6.14 steps/s | VRAM=31.59 Go [2026-07-18 15:44:06] step 868550/1000000 | loss=2.0392 vocab=0.4915 token=1.0308 | lr=4.78e-05 | 6.13 steps/s | VRAM=31.05 Go [2026-07-18 15:44:14] step 868600/1000000 | loss=2.1815 vocab=0.6022 token=1.2262 | lr=4.77e-05 | 6.12 steps/s | VRAM=28.27 Go [2026-07-18 15:44:22] step 868650/1000000 | loss=2.1299 vocab=0.3691 token=0.9337 | lr=4.77e-05 | 6.15 steps/s | VRAM=27.65 Go [2026-07-18 15:44:31] step 868700/1000000 | loss=2.1061 vocab=0.8955 token=1.4020 | lr=4.77e-05 | 6.14 steps/s | VRAM=32.79 Go [2026-07-18 15:44:39] step 868750/1000000 | loss=2.0006 vocab=0.5703 token=0.9928 | lr=4.77e-05 | 6.14 steps/s | VRAM=32.91 Go [2026-07-18 15:44:47] step 868800/1000000 | loss=2.0876 vocab=0.6748 token=0.9542 | lr=4.76e-05 | 6.12 steps/s | VRAM=28.12 Go [2026-07-18 15:44:55] step 868850/1000000 | loss=2.1095 vocab=1.0404 token=1.2786 | lr=4.76e-05 | 6.15 steps/s | VRAM=27.93 Go [2026-07-18 15:45:03] step 868900/1000000 | loss=2.0357 vocab=0.9201 token=1.3379 | lr=4.76e-05 | 6.14 steps/s | VRAM=28.06 Go [2026-07-18 15:45:11] step 868950/1000000 | loss=2.1682 vocab=1.0882 token=1.3875 | lr=4.75e-05 | 6.10 steps/s | VRAM=30.46 Go [2026-07-18 15:45:20] step 869000/1000000 | loss=2.1595 vocab=0.8346 token=1.1308 | lr=4.75e-05 | 6.16 steps/s | VRAM=31.88 Go [2026-07-18 15:45:28] step 869050/1000000 | loss=2.1366 vocab=0.7977 token=1.1862 | lr=4.75e-05 | 6.14 steps/s | VRAM=32.51 Go [2026-07-18 15:45:36] step 869100/1000000 | loss=2.1515 vocab=0.9063 token=1.3506 | lr=4.75e-05 | 6.11 steps/s | VRAM=26.82 Go [2026-07-18 15:45:44] step 869150/1000000 | loss=2.1027 vocab=0.6740 token=1.1535 | lr=4.74e-05 | 6.16 steps/s | VRAM=24.89 Go [2026-07-18 15:45:52] step 869200/1000000 | loss=2.0899 vocab=0.8515 token=1.3562 | lr=4.74e-05 | 6.13 steps/s | VRAM=29.24 Go [2026-07-18 15:46:00] step 869250/1000000 | loss=2.0244 vocab=0.8889 token=0.9203 | lr=4.74e-05 | 6.15 steps/s | VRAM=28.02 Go [2026-07-18 15:46:08] step 869300/1000000 | loss=2.0880 vocab=1.7055 token=1.3647 | lr=4.73e-05 | 6.15 steps/s | VRAM=24.56 Go [2026-07-18 15:46:17] step 869350/1000000 | loss=2.0154 vocab=0.6568 token=0.9068 | lr=4.73e-05 | 6.16 steps/s | VRAM=31.69 Go [2026-07-18 15:46:25] step 869400/1000000 | loss=2.0088 vocab=0.7550 token=1.0587 | lr=4.73e-05 | 6.13 steps/s | VRAM=27.57 Go [2026-07-18 15:46:33] step 869450/1000000 | loss=2.0886 vocab=0.7676 token=1.3217 | lr=4.73e-05 | 6.13 steps/s | VRAM=29.19 Go [2026-07-18 15:46:41] step 869500/1000000 | loss=2.0864 vocab=0.6427 token=1.0207 | lr=4.72e-05 | 6.13 steps/s | VRAM=31.96 Go [2026-07-18 15:46:49] step 869550/1000000 | loss=2.1039 vocab=0.5037 token=0.8410 | lr=4.72e-05 | 6.12 steps/s | VRAM=32.81 Go [2026-07-18 15:46:57] step 869600/1000000 | loss=2.1109 vocab=0.7968 token=1.2611 | lr=4.72e-05 | 6.09 steps/s | VRAM=28.98 Go [2026-07-18 15:47:06] step 869650/1000000 | loss=2.0943 vocab=0.8390 token=1.1501 | lr=4.71e-05 | 6.13 steps/s | VRAM=25.66 Go [2026-07-18 15:47:14] step 869700/1000000 | loss=2.1512 vocab=1.0567 token=1.6204 | lr=4.71e-05 | 6.11 steps/s | VRAM=26.45 Go [2026-07-18 15:47:22] step 869750/1000000 | loss=2.1166 vocab=0.9814 token=1.0143 | lr=4.71e-05 | 6.16 steps/s | VRAM=25.56 Go [2026-07-18 15:47:30] step 869800/1000000 | loss=2.1109 vocab=1.1108 token=1.3272 | lr=4.71e-05 | 6.13 steps/s | VRAM=27.97 Go [2026-07-18 15:47:38] step 869850/1000000 | loss=2.0137 vocab=1.0480 token=1.5062 | lr=4.70e-05 | 6.15 steps/s | VRAM=31.87 Go [2026-07-18 15:47:46] step 869900/1000000 | loss=2.1473 vocab=1.1418 token=1.3128 | lr=4.70e-05 | 6.10 steps/s | VRAM=27.21 Go [2026-07-18 15:47:54] step 869950/1000000 | loss=2.0724 vocab=1.3907 token=1.5517 | lr=4.70e-05 | 6.13 steps/s | VRAM=29.36 Go [2026-07-18 15:48:03] step 870000/1000000 | loss=2.1554 vocab=0.8143 token=1.2269 | lr=4.69e-05 | 6.09 steps/s | VRAM=28.28 Go [2026-07-18 15:48:11] step 870050/1000000 | loss=2.0841 vocab=0.8878 token=1.2977 | lr=4.69e-05 | 6.11 steps/s | VRAM=26.19 Go [2026-07-18 15:48:19] step 870100/1000000 | loss=2.0590 vocab=1.2105 token=1.3223 | lr=4.69e-05 | 6.14 steps/s | VRAM=32.41 Go [2026-07-18 15:48:27] step 870150/1000000 | loss=2.0914 vocab=0.4921 token=0.8660 | lr=4.69e-05 | 6.12 steps/s | VRAM=28.59 Go [2026-07-18 15:48:35] step 870200/1000000 | loss=2.1228 vocab=0.9029 token=0.9642 | lr=4.68e-05 | 6.11 steps/s | VRAM=29.02 Go [2026-07-18 15:48:43] step 870250/1000000 | loss=2.1262 vocab=1.0580 token=1.1510 | lr=4.68e-05 | 6.15 steps/s | VRAM=31.04 Go [2026-07-18 15:48:52] step 870300/1000000 | loss=2.0505 vocab=1.3227 token=1.5321 | lr=4.68e-05 | 6.13 steps/s | VRAM=30.98 Go [2026-07-18 15:49:00] step 870350/1000000 | loss=2.0721 vocab=0.8708 token=1.3054 | lr=4.67e-05 | 6.13 steps/s | VRAM=32.29 Go [2026-07-18 15:49:08] step 870400/1000000 | loss=1.9936 vocab=0.8831 token=1.5991 | lr=4.67e-05 | 6.14 steps/s | VRAM=31.44 Go [2026-07-18 15:49:16] step 870450/1000000 | loss=2.1373 vocab=0.7531 token=1.3197 | lr=4.67e-05 | 6.12 steps/s | VRAM=28.40 Go [2026-07-18 15:49:24] step 870500/1000000 | loss=2.1008 vocab=0.9548 token=1.3923 | lr=4.67e-05 | 6.12 steps/s | VRAM=25.63 Go [2026-07-18 15:49:32] step 870550/1000000 | loss=2.0070 vocab=1.2547 token=1.1541 | lr=4.66e-05 | 6.14 steps/s | VRAM=28.45 Go [2026-07-18 15:49:41] step 870600/1000000 | loss=2.1421 vocab=1.3222 token=1.2417 | lr=4.66e-05 | 6.14 steps/s | VRAM=32.76 Go [2026-07-18 15:49:49] step 870650/1000000 | loss=2.0503 vocab=0.6297 token=0.6337 | lr=4.66e-05 | 6.11 steps/s | VRAM=28.58 Go [2026-07-18 15:49:57] step 870700/1000000 | loss=2.0429 vocab=0.7652 token=1.0876 | lr=4.65e-05 | 6.11 steps/s | VRAM=28.18 Go [2026-07-18 15:50:05] step 870750/1000000 | loss=2.1018 vocab=0.6733 token=1.0106 | lr=4.65e-05 | 6.12 steps/s | VRAM=26.04 Go [2026-07-18 15:50:13] step 870800/1000000 | loss=2.1228 vocab=0.7558 token=0.8837 | lr=4.65e-05 | 6.06 steps/s | VRAM=32.73 Go [2026-07-18 15:50:22] step 870850/1000000 | loss=2.0842 vocab=0.4588 token=0.6185 | lr=4.65e-05 | 6.15 steps/s | VRAM=28.75 Go [2026-07-18 15:50:30] step 870900/1000000 | loss=2.0529 vocab=0.4741 token=0.8931 | lr=4.64e-05 | 6.13 steps/s | VRAM=29.24 Go [2026-07-18 15:50:38] step 870950/1000000 | loss=2.0779 vocab=0.8426 token=1.1115 | lr=4.64e-05 | 6.11 steps/s | VRAM=25.89 Go [2026-07-18 15:50:46] step 871000/1000000 | loss=2.2098 vocab=1.5730 token=1.3093 | lr=4.64e-05 | 6.07 steps/s | VRAM=23.37 Go [2026-07-18 15:50:54] step 871050/1000000 | loss=2.1092 vocab=0.8904 token=1.3007 | lr=4.63e-05 | 6.13 steps/s | VRAM=27.91 Go [2026-07-18 15:51:02] step 871100/1000000 | loss=2.0590 vocab=0.5512 token=1.2613 | lr=4.63e-05 | 6.15 steps/s | VRAM=28.28 Go [2026-07-18 15:51:11] step 871150/1000000 | loss=2.1570 vocab=0.9374 token=1.5967 | lr=4.63e-05 | 6.12 steps/s | VRAM=30.07 Go [2026-07-18 15:51:19] step 871200/1000000 | loss=2.0906 vocab=0.5861 token=0.6056 | lr=4.63e-05 | 6.14 steps/s | VRAM=26.84 Go [2026-07-18 15:51:27] step 871250/1000000 | loss=2.1384 vocab=1.2868 token=1.4744 | lr=4.62e-05 | 6.14 steps/s | VRAM=26.41 Go [2026-07-18 15:51:35] step 871300/1000000 | loss=2.1529 vocab=1.1524 token=1.0429 | lr=4.62e-05 | 6.11 steps/s | VRAM=28.28 Go [2026-07-18 15:51:43] step 871350/1000000 | loss=2.0725 vocab=0.6572 token=0.8714 | lr=4.62e-05 | 6.15 steps/s | VRAM=28.57 Go [2026-07-18 15:51:51] step 871400/1000000 | loss=2.0983 vocab=0.8620 token=1.1509 | lr=4.61e-05 | 6.10 steps/s | VRAM=28.31 Go [2026-07-18 15:52:00] step 871450/1000000 | loss=2.0043 vocab=0.6083 token=1.3600 | lr=4.61e-05 | 6.12 steps/s | VRAM=30.37 Go [2026-07-18 15:52:08] step 871500/1000000 | loss=2.1185 vocab=0.8536 token=1.0857 | lr=4.61e-05 | 6.11 steps/s | VRAM=32.41 Go [2026-07-18 15:52:16] step 871550/1000000 | loss=1.9889 vocab=0.6449 token=1.0379 | lr=4.61e-05 | 6.16 steps/s | VRAM=29.76 Go [2026-07-18 15:52:24] step 871600/1000000 | loss=2.1800 vocab=0.4608 token=0.9060 | lr=4.60e-05 | 6.11 steps/s | VRAM=28.64 Go [2026-07-18 15:52:32] step 871650/1000000 | loss=2.0781 vocab=0.8182 token=1.3363 | lr=4.60e-05 | 6.12 steps/s | VRAM=26.88 Go [2026-07-18 15:52:40] step 871700/1000000 | loss=2.1336 vocab=0.8305 token=1.3743 | lr=4.60e-05 | 6.15 steps/s | VRAM=26.30 Go [2026-07-18 15:52:48] step 871750/1000000 | loss=2.1410 vocab=1.1483 token=1.2589 | lr=4.59e-05 | 6.12 steps/s | VRAM=29.14 Go [2026-07-18 15:52:57] step 871800/1000000 | loss=2.0177 vocab=0.7993 token=1.2199 | lr=4.59e-05 | 6.09 steps/s | VRAM=28.65 Go [2026-07-18 15:53:05] step 871850/1000000 | loss=2.0456 vocab=0.9046 token=1.4242 | lr=4.59e-05 | 6.14 steps/s | VRAM=32.48 Go [2026-07-18 15:53:13] step 871900/1000000 | loss=2.1069 vocab=1.1158 token=0.8944 | lr=4.59e-05 | 6.15 steps/s | VRAM=32.39 Go [2026-07-18 15:53:21] step 871950/1000000 | loss=2.0503 vocab=0.5116 token=1.1889 | lr=4.58e-05 | 6.12 steps/s | VRAM=28.87 Go [2026-07-18 15:53:29] step 872000/1000000 | loss=2.0987 vocab=1.1843 token=1.5064 | lr=4.58e-05 | 6.16 steps/s | VRAM=29.22 Go [2026-07-18 15:53:37] step 872050/1000000 | loss=2.1269 vocab=0.6810 token=1.0872 | lr=4.58e-05 | 6.13 steps/s | VRAM=32.01 Go [2026-07-18 15:53:46] step 872100/1000000 | loss=2.1126 vocab=1.8801 token=1.5908 | lr=4.57e-05 | 6.11 steps/s | VRAM=31.66 Go [2026-07-18 15:53:54] step 872150/1000000 | loss=2.1554 vocab=0.7700 token=1.2172 | lr=4.57e-05 | 6.14 steps/s | VRAM=32.31 Go [2026-07-18 15:54:02] step 872200/1000000 | loss=2.1323 vocab=0.4487 token=0.9283 | lr=4.57e-05 | 6.12 steps/s | VRAM=28.87 Go [2026-07-18 15:54:10] step 872250/1000000 | loss=2.0774 vocab=1.2378 token=1.5689 | lr=4.57e-05 | 6.15 steps/s | VRAM=29.19 Go [2026-07-18 15:54:18] step 872300/1000000 | loss=2.0670 vocab=0.9345 token=1.3188 | lr=4.56e-05 | 6.11 steps/s | VRAM=28.42 Go [2026-07-18 15:54:26] step 872350/1000000 | loss=2.0681 vocab=0.9197 token=1.1897 | lr=4.56e-05 | 6.06 steps/s | VRAM=28.50 Go [2026-07-18 15:54:35] step 872400/1000000 | loss=2.1100 vocab=0.7469 token=1.3205 | lr=4.56e-05 | 6.12 steps/s | VRAM=31.95 Go [2026-07-18 15:54:43] step 872450/1000000 | loss=2.0541 vocab=0.8915 token=1.4960 | lr=4.55e-05 | 6.12 steps/s | VRAM=32.37 Go [2026-07-18 15:54:51] step 872500/1000000 | loss=2.1726 vocab=1.0630 token=1.2929 | lr=4.55e-05 | 6.10 steps/s | VRAM=28.45 Go [2026-07-18 15:54:59] step 872550/1000000 | loss=1.9839 vocab=0.3938 token=1.1855 | lr=4.55e-05 | 6.12 steps/s | VRAM=29.18 Go [2026-07-18 15:55:07] step 872600/1000000 | loss=2.0644 vocab=0.9434 token=1.0699 | lr=4.55e-05 | 6.08 steps/s | VRAM=26.15 Go [2026-07-18 15:55:16] step 872650/1000000 | loss=2.0607 vocab=0.8684 token=0.9220 | lr=4.54e-05 | 6.09 steps/s | VRAM=27.68 Go [2026-07-18 15:55:24] step 872700/1000000 | loss=2.1370 vocab=1.4062 token=1.1626 | lr=4.54e-05 | 6.16 steps/s | VRAM=31.65 Go [2026-07-18 15:55:32] step 872750/1000000 | loss=2.1554 vocab=0.7991 token=1.2861 | lr=4.54e-05 | 6.14 steps/s | VRAM=32.00 Go [2026-07-18 15:55:40] step 872800/1000000 | loss=2.0791 vocab=1.1221 token=1.2797 | lr=4.53e-05 | 6.11 steps/s | VRAM=25.41 Go [2026-07-18 15:55:48] step 872850/1000000 | loss=2.0540 vocab=0.7899 token=1.3342 | lr=4.53e-05 | 6.13 steps/s | VRAM=29.64 Go [2026-07-18 15:55:56] step 872900/1000000 | loss=2.2521 vocab=0.9745 token=1.6551 | lr=4.53e-05 | 6.13 steps/s | VRAM=28.51 Go [2026-07-18 15:56:04] step 872950/1000000 | loss=2.0970 vocab=0.9837 token=1.3351 | lr=4.53e-05 | 6.15 steps/s | VRAM=32.75 Go [2026-07-18 15:56:13] step 873000/1000000 | loss=2.1604 vocab=0.6812 token=1.0032 | lr=4.52e-05 | 6.17 steps/s | VRAM=28.75 Go [2026-07-18 15:56:21] step 873050/1000000 | loss=2.1138 vocab=1.0360 token=1.3514 | lr=4.52e-05 | 6.11 steps/s | VRAM=25.58 Go [2026-07-18 15:56:29] step 873100/1000000 | loss=2.1224 vocab=0.8192 token=1.4097 | lr=4.52e-05 | 6.14 steps/s | VRAM=32.89 Go [2026-07-18 15:56:37] step 873150/1000000 | loss=2.0804 vocab=0.9298 token=1.5408 | lr=4.51e-05 | 6.16 steps/s | VRAM=24.55 Go [2026-07-18 15:56:45] step 873200/1000000 | loss=2.2260 vocab=0.5422 token=0.7816 | lr=4.51e-05 | 6.12 steps/s | VRAM=26.69 Go [2026-07-18 15:56:53] step 873250/1000000 | loss=2.0220 vocab=0.8604 token=1.2138 | lr=4.51e-05 | 6.12 steps/s | VRAM=27.48 Go [2026-07-18 15:57:02] step 873300/1000000 | loss=2.1651 vocab=0.7515 token=1.3325 | lr=4.51e-05 | 6.13 steps/s | VRAM=29.15 Go [2026-07-18 15:57:10] step 873350/1000000 | loss=2.1092 vocab=1.0566 token=1.4553 | lr=4.50e-05 | 6.12 steps/s | VRAM=28.80 Go [2026-07-18 15:57:18] step 873400/1000000 | loss=2.0896 vocab=1.0117 token=1.3783 | lr=4.50e-05 | 6.10 steps/s | VRAM=32.49 Go [2026-07-18 15:57:26] step 873450/1000000 | loss=2.0539 vocab=0.8478 token=1.3663 | lr=4.50e-05 | 6.08 steps/s | VRAM=25.45 Go [2026-07-18 15:57:34] step 873500/1000000 | loss=2.0423 vocab=0.9236 token=1.2513 | lr=4.50e-05 | 6.12 steps/s | VRAM=27.97 Go [2026-07-18 15:57:42] step 873550/1000000 | loss=1.9790 vocab=0.7385 token=1.1025 | lr=4.49e-05 | 6.12 steps/s | VRAM=28.90 Go [2026-07-18 15:57:51] step 873600/1000000 | loss=2.0688 vocab=0.5700 token=1.0657 | lr=4.49e-05 | 6.13 steps/s | VRAM=31.87 Go [2026-07-18 15:57:59] step 873650/1000000 | loss=2.0542 vocab=0.6480 token=1.3470 | lr=4.49e-05 | 6.09 steps/s | VRAM=27.91 Go [2026-07-18 15:58:07] step 873700/1000000 | loss=2.0765 vocab=0.9096 token=1.1569 | lr=4.48e-05 | 6.11 steps/s | VRAM=27.30 Go [2026-07-18 15:58:15] step 873750/1000000 | loss=2.1948 vocab=0.8773 token=1.1905 | lr=4.48e-05 | 6.14 steps/s | VRAM=28.45 Go [2026-07-18 15:58:23] step 873800/1000000 | loss=2.0687 vocab=1.3456 token=1.4243 | lr=4.48e-05 | 6.14 steps/s | VRAM=29.08 Go [2026-07-18 15:58:31] step 873850/1000000 | loss=2.0498 vocab=1.1169 token=1.1981 | lr=4.48e-05 | 6.12 steps/s | VRAM=30.93 Go [2026-07-18 15:58:40] step 873900/1000000 | loss=2.0819 vocab=0.6430 token=0.8130 | lr=4.47e-05 | 6.13 steps/s | VRAM=28.60 Go [2026-07-18 15:58:48] step 873950/1000000 | loss=2.1371 vocab=0.7657 token=1.4895 | lr=4.47e-05 | 6.10 steps/s | VRAM=32.13 Go [2026-07-18 15:58:56] step 874000/1000000 | loss=2.1505 vocab=0.6953 token=1.2742 | lr=4.47e-05 | 6.14 steps/s | VRAM=32.71 Go [2026-07-18 15:59:04] step 874050/1000000 | loss=2.1089 vocab=0.8891 token=1.1049 | lr=4.46e-05 | 6.10 steps/s | VRAM=28.66 Go [2026-07-18 15:59:12] step 874100/1000000 | loss=2.0849 vocab=0.9798 token=1.3802 | lr=4.46e-05 | 6.12 steps/s | VRAM=29.05 Go [2026-07-18 15:59:20] step 874150/1000000 | loss=2.0483 vocab=0.9692 token=1.6338 | lr=4.46e-05 | 6.13 steps/s | VRAM=30.73 Go [2026-07-18 15:59:29] step 874200/1000000 | loss=2.0483 vocab=0.5187 token=0.9580 | lr=4.46e-05 | 6.07 steps/s | VRAM=29.97 Go [2026-07-18 15:59:37] step 874250/1000000 | loss=2.0859 vocab=1.1634 token=0.9030 | lr=4.45e-05 | 6.12 steps/s | VRAM=28.51 Go [2026-07-18 15:59:45] step 874300/1000000 | loss=1.9852 vocab=0.6822 token=0.8639 | lr=4.45e-05 | 6.10 steps/s | VRAM=26.29 Go [2026-07-18 15:59:53] step 874350/1000000 | loss=2.0088 vocab=0.6633 token=0.9548 | lr=4.45e-05 | 6.12 steps/s | VRAM=28.86 Go [2026-07-18 16:00:01] step 874400/1000000 | loss=2.2077 vocab=0.8067 token=1.1289 | lr=4.44e-05 | 6.12 steps/s | VRAM=25.36 Go [2026-07-18 16:00:10] step 874450/1000000 | loss=2.0713 vocab=0.6570 token=1.2727 | lr=4.44e-05 | 6.12 steps/s | VRAM=28.86 Go [2026-07-18 16:00:18] step 874500/1000000 | loss=2.0072 vocab=0.9263 token=1.2199 | lr=4.44e-05 | 6.11 steps/s | VRAM=25.98 Go [2026-07-18 16:00:26] step 874550/1000000 | loss=2.0623 vocab=0.8064 token=1.0673 | lr=4.44e-05 | 6.11 steps/s | VRAM=32.43 Go [2026-07-18 16:00:34] step 874600/1000000 | loss=2.0154 vocab=0.7549 token=1.5133 | lr=4.43e-05 | 6.10 steps/s | VRAM=26.45 Go [2026-07-18 16:00:42] step 874650/1000000 | loss=2.0138 vocab=1.2925 token=1.4846 | lr=4.43e-05 | 6.12 steps/s | VRAM=31.81 Go [2026-07-18 16:00:50] step 874700/1000000 | loss=2.0576 vocab=0.7384 token=0.5783 | lr=4.43e-05 | 6.12 steps/s | VRAM=25.01 Go [2026-07-18 16:00:59] step 874750/1000000 | loss=1.9889 vocab=0.8191 token=0.9966 | lr=4.42e-05 | 6.11 steps/s | VRAM=26.36 Go [2026-07-18 16:01:07] step 874800/1000000 | loss=2.0805 vocab=1.0166 token=1.3927 | lr=4.42e-05 | 6.13 steps/s | VRAM=31.96 Go [2026-07-18 16:01:15] step 874850/1000000 | loss=2.1345 vocab=0.4732 token=0.8027 | lr=4.42e-05 | 6.16 steps/s | VRAM=24.32 Go [2026-07-18 16:01:23] step 874900/1000000 | loss=2.1168 vocab=0.8280 token=1.1308 | lr=4.42e-05 | 6.10 steps/s | VRAM=32.43 Go [2026-07-18 16:01:31] step 874950/1000000 | loss=1.9244 vocab=0.6469 token=1.1803 | lr=4.41e-05 | 6.13 steps/s | VRAM=28.33 Go [2026-07-18 16:01:39] step 875000/1000000 | loss=2.1108 vocab=0.9098 token=0.7829 | lr=4.41e-05 | 6.12 steps/s | VRAM=25.89 Go [2026-07-18 16:01:48] step 875050/1000000 | loss=2.0061 vocab=0.6513 token=1.0434 | lr=4.41e-05 | 6.14 steps/s | VRAM=28.46 Go [2026-07-18 16:01:56] step 875100/1000000 | loss=2.1007 vocab=0.7808 token=0.9350 | lr=4.40e-05 | 6.09 steps/s | VRAM=29.05 Go [2026-07-18 16:02:04] step 875150/1000000 | loss=2.1020 vocab=0.8695 token=1.5402 | lr=4.40e-05 | 6.13 steps/s | VRAM=32.48 Go [2026-07-18 16:02:12] step 875200/1000000 | loss=2.1303 vocab=1.4074 token=1.5363 | lr=4.40e-05 | 6.10 steps/s | VRAM=28.16 Go [2026-07-18 16:02:20] step 875250/1000000 | loss=2.0519 vocab=0.8473 token=1.0511 | lr=4.40e-05 | 6.12 steps/s | VRAM=29.20 Go [2026-07-18 16:02:29] step 875300/1000000 | loss=2.1629 vocab=0.9954 token=1.4272 | lr=4.39e-05 | 6.11 steps/s | VRAM=28.55 Go [2026-07-18 16:02:37] step 875350/1000000 | loss=2.2021 vocab=1.1050 token=1.4073 | lr=4.39e-05 | 6.11 steps/s | VRAM=25.76 Go [2026-07-18 16:02:45] step 875400/1000000 | loss=2.2853 vocab=0.4865 token=0.9751 | lr=4.39e-05 | 6.12 steps/s | VRAM=31.93 Go [2026-07-18 16:02:53] step 875450/1000000 | loss=2.0783 vocab=1.0945 token=1.3489 | lr=4.38e-05 | 6.12 steps/s | VRAM=31.24 Go [2026-07-18 16:03:01] step 875500/1000000 | loss=2.0829 vocab=0.8306 token=1.2757 | lr=4.38e-05 | 6.08 steps/s | VRAM=27.30 Go [2026-07-18 16:03:09] step 875550/1000000 | loss=1.9910 vocab=1.0598 token=1.3578 | lr=4.38e-05 | 6.13 steps/s | VRAM=27.97 Go [2026-07-18 16:03:18] step 875600/1000000 | loss=2.1391 vocab=0.7306 token=1.2952 | lr=4.38e-05 | 6.12 steps/s | VRAM=29.43 Go [2026-07-18 16:03:26] step 875650/1000000 | loss=2.1003 vocab=1.3087 token=1.3828 | lr=4.37e-05 | 6.12 steps/s | VRAM=27.38 Go [2026-07-18 16:03:34] step 875700/1000000 | loss=2.1872 vocab=1.0364 token=1.4499 | lr=4.37e-05 | 6.13 steps/s | VRAM=23.83 Go [2026-07-18 16:03:42] step 875750/1000000 | loss=2.0383 vocab=1.0033 token=1.2769 | lr=4.37e-05 | 6.08 steps/s | VRAM=26.34 Go [2026-07-18 16:03:50] step 875800/1000000 | loss=2.1978 vocab=0.6484 token=1.3349 | lr=4.36e-05 | 6.14 steps/s | VRAM=26.19 Go [2026-07-18 16:03:58] step 875850/1000000 | loss=2.1933 vocab=1.0679 token=1.1108 | lr=4.36e-05 | 6.13 steps/s | VRAM=25.38 Go [2026-07-18 16:04:07] step 875900/1000000 | loss=2.1902 vocab=0.7393 token=0.7790 | lr=4.36e-05 | 6.12 steps/s | VRAM=28.45 Go [2026-07-18 16:04:15] step 875950/1000000 | loss=2.1425 vocab=0.4945 token=1.0911 | lr=4.36e-05 | 6.12 steps/s | VRAM=25.81 Go [2026-07-18 16:04:23] step 876000/1000000 | loss=2.0773 vocab=0.4907 token=0.9244 | lr=4.35e-05 | 6.12 steps/s | VRAM=32.61 Go [2026-07-18 16:04:31] step 876050/1000000 | loss=2.0069 vocab=1.0597 token=1.1973 | lr=4.35e-05 | 6.17 steps/s | VRAM=32.23 Go [2026-07-18 16:04:39] step 876100/1000000 | loss=2.0095 vocab=1.3920 token=1.4470 | lr=4.35e-05 | 6.12 steps/s | VRAM=31.87 Go [2026-07-18 16:04:47] step 876150/1000000 | loss=2.0630 vocab=0.8847 token=1.2136 | lr=4.35e-05 | 6.14 steps/s | VRAM=31.55 Go [2026-07-18 16:04:56] step 876200/1000000 | loss=2.0879 vocab=1.1448 token=0.9657 | lr=4.34e-05 | 6.14 steps/s | VRAM=32.22 Go [2026-07-18 16:05:04] step 876250/1000000 | loss=2.0105 vocab=0.7856 token=1.1735 | lr=4.34e-05 | 6.13 steps/s | VRAM=33.00 Go [2026-07-18 16:05:12] step 876300/1000000 | loss=1.9784 vocab=0.6359 token=0.9290 | lr=4.34e-05 | 6.14 steps/s | VRAM=32.09 Go [2026-07-18 16:05:20] step 876350/1000000 | loss=2.0591 vocab=1.6004 token=1.3907 | lr=4.33e-05 | 6.14 steps/s | VRAM=25.63 Go [2026-07-18 16:05:28] step 876400/1000000 | loss=2.1111 vocab=0.9825 token=1.5444 | lr=4.33e-05 | 6.15 steps/s | VRAM=26.27 Go [2026-07-18 16:05:36] step 876450/1000000 | loss=2.1538 vocab=0.3526 token=0.7672 | lr=4.33e-05 | 6.13 steps/s | VRAM=28.01 Go [2026-07-18 16:05:44] step 876500/1000000 | loss=2.0884 vocab=0.7851 token=1.0252 | lr=4.33e-05 | 6.14 steps/s | VRAM=29.27 Go [2026-07-18 16:05:53] step 876550/1000000 | loss=2.0434 vocab=0.6349 token=1.1232 | lr=4.32e-05 | 6.14 steps/s | VRAM=28.64 Go [2026-07-18 16:06:01] step 876600/1000000 | loss=2.0821 vocab=0.7569 token=1.2546 | lr=4.32e-05 | 6.13 steps/s | VRAM=28.73 Go [2026-07-18 16:06:09] step 876650/1000000 | loss=2.1079 vocab=1.2160 token=1.0422 | lr=4.32e-05 | 6.15 steps/s | VRAM=30.89 Go [2026-07-18 16:06:17] step 876700/1000000 | loss=2.0360 vocab=0.6459 token=1.2071 | lr=4.31e-05 | 6.12 steps/s | VRAM=29.04 Go [2026-07-18 16:06:25] step 876750/1000000 | loss=2.0849 vocab=0.5477 token=0.9241 | lr=4.31e-05 | 6.15 steps/s | VRAM=31.99 Go [2026-07-18 16:06:33] step 876800/1000000 | loss=2.0912 vocab=1.1566 token=1.3994 | lr=4.31e-05 | 6.10 steps/s | VRAM=28.03 Go [2026-07-18 16:06:41] step 876850/1000000 | loss=1.9914 vocab=0.7765 token=1.1293 | lr=4.31e-05 | 6.15 steps/s | VRAM=29.94 Go [2026-07-18 16:06:50] step 876900/1000000 | loss=2.1020 vocab=1.3382 token=1.8656 | lr=4.30e-05 | 6.12 steps/s | VRAM=26.25 Go [2026-07-18 16:06:58] step 876950/1000000 | loss=1.9913 vocab=0.6212 token=1.1685 | lr=4.30e-05 | 6.17 steps/s | VRAM=32.24 Go [2026-07-18 16:07:06] step 877000/1000000 | loss=2.2010 vocab=1.1099 token=1.0534 | lr=4.30e-05 | 6.14 steps/s | VRAM=28.84 Go [2026-07-18 16:07:14] step 877050/1000000 | loss=2.1274 vocab=0.7552 token=1.1576 | lr=4.29e-05 | 6.11 steps/s | VRAM=28.25 Go [2026-07-18 16:07:22] step 877100/1000000 | loss=2.0532 vocab=0.8940 token=1.2892 | lr=4.29e-05 | 6.13 steps/s | VRAM=28.17 Go [2026-07-18 16:07:30] step 877150/1000000 | loss=2.0353 vocab=0.4388 token=1.2605 | lr=4.29e-05 | 6.13 steps/s | VRAM=30.37 Go [2026-07-18 16:07:38] step 877200/1000000 | loss=2.0968 vocab=0.8145 token=1.1754 | lr=4.29e-05 | 6.15 steps/s | VRAM=29.70 Go [2026-07-18 16:07:47] step 877250/1000000 | loss=2.0645 vocab=0.6744 token=1.0664 | lr=4.28e-05 | 6.12 steps/s | VRAM=28.69 Go [2026-07-18 16:07:55] step 877300/1000000 | loss=2.1073 vocab=1.1086 token=1.2017 | lr=4.28e-05 | 6.15 steps/s | VRAM=32.64 Go [2026-07-18 16:08:03] step 877350/1000000 | loss=2.0883 vocab=1.2647 token=1.4355 | lr=4.28e-05 | 6.14 steps/s | VRAM=25.51 Go [2026-07-18 16:08:11] step 877400/1000000 | loss=2.1420 vocab=0.6229 token=0.7859 | lr=4.27e-05 | 6.13 steps/s | VRAM=29.27 Go [2026-07-18 16:08:19] step 877450/1000000 | loss=2.1209 vocab=0.3219 token=1.2664 | lr=4.27e-05 | 6.17 steps/s | VRAM=31.96 Go [2026-07-18 16:08:27] step 877500/1000000 | loss=2.1364 vocab=0.4553 token=0.7645 | lr=4.27e-05 | 6.11 steps/s | VRAM=28.81 Go [2026-07-18 16:08:36] step 877550/1000000 | loss=2.1545 vocab=0.5611 token=0.9836 | lr=4.27e-05 | 6.11 steps/s | VRAM=23.96 Go [2026-07-18 16:08:44] step 877600/1000000 | loss=2.1381 vocab=0.8650 token=1.0831 | lr=4.26e-05 | 6.14 steps/s | VRAM=28.22 Go [2026-07-18 16:08:52] step 877650/1000000 | loss=1.9636 vocab=0.7750 token=1.0572 | lr=4.26e-05 | 6.11 steps/s | VRAM=26.89 Go [2026-07-18 16:09:00] step 877700/1000000 | loss=2.1822 vocab=1.0548 token=0.8725 | lr=4.26e-05 | 6.15 steps/s | VRAM=28.96 Go [2026-07-18 16:09:08] step 877750/1000000 | loss=2.0084 vocab=1.0354 token=1.3488 | lr=4.25e-05 | 6.13 steps/s | VRAM=32.81 Go [2026-07-18 16:09:16] step 877800/1000000 | loss=2.0481 vocab=1.0310 token=1.4525 | lr=4.25e-05 | 6.14 steps/s | VRAM=29.18 Go [2026-07-18 16:09:24] step 877850/1000000 | loss=2.0115 vocab=1.0190 token=1.0520 | lr=4.25e-05 | 6.13 steps/s | VRAM=29.41 Go [2026-07-18 16:09:33] step 877900/1000000 | loss=2.1272 vocab=1.0438 token=1.1087 | lr=4.25e-05 | 6.16 steps/s | VRAM=25.40 Go [2026-07-18 16:09:41] step 877950/1000000 | loss=2.1289 vocab=0.9368 token=1.3486 | lr=4.24e-05 | 6.13 steps/s | VRAM=32.33 Go [2026-07-18 16:09:49] step 878000/1000000 | loss=2.0892 vocab=0.9411 token=1.1407 | lr=4.24e-05 | 6.13 steps/s | VRAM=28.38 Go [2026-07-18 16:09:57] step 878050/1000000 | loss=2.0313 vocab=0.7923 token=1.1809 | lr=4.24e-05 | 6.14 steps/s | VRAM=25.92 Go [2026-07-18 16:10:05] step 878100/1000000 | loss=2.1365 vocab=0.8034 token=1.1548 | lr=4.24e-05 | 6.14 steps/s | VRAM=24.95 Go [2026-07-18 16:10:13] step 878150/1000000 | loss=2.1555 vocab=0.9598 token=1.5133 | lr=4.23e-05 | 6.17 steps/s | VRAM=31.65 Go [2026-07-18 16:10:21] step 878200/1000000 | loss=2.0722 vocab=1.3185 token=1.3474 | lr=4.23e-05 | 6.16 steps/s | VRAM=25.37 Go [2026-07-18 16:10:30] step 878250/1000000 | loss=2.0973 vocab=1.1480 token=1.1191 | lr=4.23e-05 | 6.11 steps/s | VRAM=25.98 Go [2026-07-18 16:10:38] step 878300/1000000 | loss=2.0637 vocab=0.7852 token=1.2728 | lr=4.22e-05 | 6.15 steps/s | VRAM=33.01 Go [2026-07-18 16:10:46] step 878350/1000000 | loss=2.0370 vocab=0.9230 token=1.2077 | lr=4.22e-05 | 6.12 steps/s | VRAM=29.10 Go [2026-07-18 16:10:54] step 878400/1000000 | loss=2.1179 vocab=0.7738 token=1.4191 | lr=4.22e-05 | 6.15 steps/s | VRAM=27.70 Go [2026-07-18 16:11:02] step 878450/1000000 | loss=2.2157 vocab=0.9126 token=1.3620 | lr=4.22e-05 | 6.15 steps/s | VRAM=31.84 Go [2026-07-18 16:11:10] step 878500/1000000 | loss=2.0866 vocab=0.8562 token=1.2924 | lr=4.21e-05 | 6.14 steps/s | VRAM=29.01 Go [2026-07-18 16:11:18] step 878550/1000000 | loss=2.2718 vocab=0.7695 token=1.4981 | lr=4.21e-05 | 6.11 steps/s | VRAM=28.40 Go [2026-07-18 16:11:27] step 878600/1000000 | loss=2.1348 vocab=0.9974 token=1.6367 | lr=4.21e-05 | 6.11 steps/s | VRAM=26.43 Go [2026-07-18 16:11:35] step 878650/1000000 | loss=2.0833 vocab=1.0886 token=1.4935 | lr=4.20e-05 | 6.11 steps/s | VRAM=28.26 Go [2026-07-18 16:11:43] step 878700/1000000 | loss=2.1456 vocab=0.7392 token=1.3990 | lr=4.20e-05 | 6.13 steps/s | VRAM=28.98 Go [2026-07-18 16:11:51] step 878750/1000000 | loss=2.0659 vocab=0.9790 token=1.1574 | lr=4.20e-05 | 6.10 steps/s | VRAM=25.31 Go [2026-07-18 16:11:59] step 878800/1000000 | loss=2.1091 vocab=0.7980 token=1.0984 | lr=4.20e-05 | 6.12 steps/s | VRAM=29.73 Go [2026-07-18 16:12:07] step 878850/1000000 | loss=2.0240 vocab=0.7463 token=1.1413 | lr=4.19e-05 | 6.15 steps/s | VRAM=28.14 Go [2026-07-18 16:12:16] step 878900/1000000 | loss=2.0146 vocab=1.0674 token=1.2629 | lr=4.19e-05 | 6.13 steps/s | VRAM=28.59 Go [2026-07-18 16:12:24] step 878950/1000000 | loss=2.0428 vocab=0.5803 token=1.5567 | lr=4.19e-05 | 6.14 steps/s | VRAM=29.03 Go [2026-07-18 16:12:32] step 879000/1000000 | loss=2.0948 vocab=0.9771 token=1.3130 | lr=4.18e-05 | 6.16 steps/s | VRAM=27.93 Go [2026-07-18 16:12:40] step 879050/1000000 | loss=2.0355 vocab=0.6397 token=1.1441 | lr=4.18e-05 | 6.15 steps/s | VRAM=28.62 Go [2026-07-18 16:12:48] step 879100/1000000 | loss=2.0811 vocab=0.3967 token=1.7521 | lr=4.18e-05 | 6.12 steps/s | VRAM=30.28 Go [2026-07-18 16:12:56] step 879150/1000000 | loss=2.1287 vocab=0.8580 token=1.1477 | lr=4.18e-05 | 6.11 steps/s | VRAM=32.42 Go [2026-07-18 16:13:05] step 879200/1000000 | loss=2.0995 vocab=0.6002 token=1.0543 | lr=4.17e-05 | 6.14 steps/s | VRAM=29.37 Go [2026-07-18 16:13:13] step 879250/1000000 | loss=2.0942 vocab=0.3246 token=1.1958 | lr=4.17e-05 | 6.14 steps/s | VRAM=25.58 Go [2026-07-18 16:13:21] step 879300/1000000 | loss=2.1339 vocab=0.7582 token=1.0192 | lr=4.17e-05 | 6.14 steps/s | VRAM=25.84 Go [2026-07-18 16:13:29] step 879350/1000000 | loss=2.1699 vocab=1.1366 token=1.4243 | lr=4.16e-05 | 6.12 steps/s | VRAM=28.34 Go [2026-07-18 16:13:37] step 879400/1000000 | loss=2.1762 vocab=1.0120 token=1.6563 | lr=4.16e-05 | 6.12 steps/s | VRAM=33.02 Go [2026-07-18 16:13:45] step 879450/1000000 | loss=2.1433 vocab=0.6098 token=1.2300 | lr=4.16e-05 | 6.12 steps/s | VRAM=31.55 Go [2026-07-18 16:13:53] step 879500/1000000 | loss=2.0595 vocab=1.0263 token=1.3700 | lr=4.16e-05 | 6.15 steps/s | VRAM=31.47 Go [2026-07-18 16:14:02] step 879550/1000000 | loss=2.0810 vocab=0.5918 token=0.9838 | lr=4.15e-05 | 6.13 steps/s | VRAM=25.72 Go [2026-07-18 16:14:10] step 879600/1000000 | loss=2.0597 vocab=0.6419 token=1.4452 | lr=4.15e-05 | 6.10 steps/s | VRAM=28.71 Go [2026-07-18 16:14:18] step 879650/1000000 | loss=2.0344 vocab=0.9990 token=1.3157 | lr=4.15e-05 | 6.11 steps/s | VRAM=28.76 Go [2026-07-18 16:14:26] step 879700/1000000 | loss=2.0475 vocab=0.7956 token=1.2116 | lr=4.15e-05 | 6.11 steps/s | VRAM=24.79 Go [2026-07-18 16:14:34] step 879750/1000000 | loss=1.9772 vocab=0.7862 token=1.1591 | lr=4.14e-05 | 6.14 steps/s | VRAM=31.19 Go [2026-07-18 16:14:42] step 879800/1000000 | loss=2.1542 vocab=1.0364 token=1.3703 | lr=4.14e-05 | 6.11 steps/s | VRAM=32.93 Go [2026-07-18 16:14:51] step 879850/1000000 | loss=1.9999 vocab=0.8128 token=0.9091 | lr=4.14e-05 | 6.12 steps/s | VRAM=29.09 Go [2026-07-18 16:14:59] step 879900/1000000 | loss=2.0396 vocab=0.7733 token=1.0046 | lr=4.13e-05 | 6.12 steps/s | VRAM=26.50 Go [2026-07-18 16:15:07] step 879950/1000000 | loss=2.0640 vocab=0.5373 token=1.4662 | lr=4.13e-05 | 6.15 steps/s | VRAM=28.07 Go [2026-07-18 16:15:15] step 880000/1000000 | loss=2.0283 vocab=0.9708 token=1.4668 | lr=4.13e-05 | 6.12 steps/s | VRAM=26.36 Go [2026-07-18 16:15:23] step 880050/1000000 | loss=2.1259 vocab=0.9554 token=1.6706 | lr=4.13e-05 | 6.13 steps/s | VRAM=28.17 Go [2026-07-18 16:15:31] step 880100/1000000 | loss=2.0156 vocab=0.8563 token=1.0370 | lr=4.12e-05 | 6.14 steps/s | VRAM=27.99 Go [2026-07-18 16:15:40] step 880150/1000000 | loss=2.1723 vocab=0.9850 token=1.5254 | lr=4.12e-05 | 6.15 steps/s | VRAM=31.90 Go [2026-07-18 16:15:48] step 880200/1000000 | loss=2.0953 vocab=0.5423 token=0.8191 | lr=4.12e-05 | 6.14 steps/s | VRAM=28.80 Go [2026-07-18 16:15:56] step 880250/1000000 | loss=2.0788 vocab=0.6372 token=1.0919 | lr=4.11e-05 | 6.13 steps/s | VRAM=24.86 Go [2026-07-18 16:16:04] step 880300/1000000 | loss=2.1258 vocab=0.5626 token=0.7184 | lr=4.11e-05 | 6.12 steps/s | VRAM=25.55 Go [2026-07-18 16:16:12] step 880350/1000000 | loss=2.0567 vocab=0.9687 token=1.1920 | lr=4.11e-05 | 6.08 steps/s | VRAM=29.32 Go [2026-07-18 16:16:20] step 880400/1000000 | loss=1.9566 vocab=1.1838 token=1.2262 | lr=4.11e-05 | 6.13 steps/s | VRAM=32.36 Go [2026-07-18 16:16:29] step 880450/1000000 | loss=1.9819 vocab=0.8450 token=1.3704 | lr=4.10e-05 | 6.10 steps/s | VRAM=28.09 Go [2026-07-18 16:16:37] step 880500/1000000 | loss=2.1660 vocab=0.7622 token=1.3536 | lr=4.10e-05 | 6.13 steps/s | VRAM=28.90 Go [2026-07-18 16:16:45] step 880550/1000000 | loss=2.0616 vocab=0.4818 token=1.1770 | lr=4.10e-05 | 6.13 steps/s | VRAM=27.71 Go [2026-07-18 16:16:53] step 880600/1000000 | loss=2.1364 vocab=1.5748 token=1.2314 | lr=4.09e-05 | 6.12 steps/s | VRAM=26.51 Go [2026-07-18 16:17:01] step 880650/1000000 | loss=2.1232 vocab=1.1451 token=1.4575 | lr=4.09e-05 | 6.13 steps/s | VRAM=29.40 Go [2026-07-18 16:17:09] step 880700/1000000 | loss=2.0914 vocab=0.9739 token=0.7820 | lr=4.09e-05 | 6.11 steps/s | VRAM=25.39 Go [2026-07-18 16:17:18] step 880750/1000000 | loss=2.1432 vocab=0.7832 token=1.0887 | lr=4.09e-05 | 6.11 steps/s | VRAM=32.86 Go [2026-07-18 16:17:26] step 880800/1000000 | loss=2.1252 vocab=0.9091 token=1.4345 | lr=4.08e-05 | 6.13 steps/s | VRAM=30.16 Go [2026-07-18 16:17:34] step 880850/1000000 | loss=1.9958 vocab=0.9419 token=0.9856 | lr=4.08e-05 | 6.13 steps/s | VRAM=28.24 Go [2026-07-18 16:17:42] step 880900/1000000 | loss=2.1507 vocab=1.0072 token=1.3690 | lr=4.08e-05 | 6.09 steps/s | VRAM=26.69 Go [2026-07-18 16:17:50] step 880950/1000000 | loss=2.1172 vocab=0.9106 token=1.2817 | lr=4.08e-05 | 6.15 steps/s | VRAM=31.34 Go [2026-07-18 16:17:58] step 881000/1000000 | loss=2.0912 vocab=0.6881 token=1.5388 | lr=4.07e-05 | 6.14 steps/s | VRAM=28.56 Go [2026-07-18 16:18:07] step 881050/1000000 | loss=2.1147 vocab=0.7965 token=1.0177 | lr=4.07e-05 | 6.13 steps/s | VRAM=32.84 Go [2026-07-18 16:18:15] step 881100/1000000 | loss=1.9986 vocab=0.5615 token=1.2367 | lr=4.07e-05 | 6.10 steps/s | VRAM=26.76 Go [2026-07-18 16:18:23] step 881150/1000000 | loss=2.0003 vocab=0.8445 token=1.3298 | lr=4.06e-05 | 6.15 steps/s | VRAM=25.02 Go [2026-07-18 16:18:31] step 881200/1000000 | loss=2.1451 vocab=1.3202 token=1.5764 | lr=4.06e-05 | 6.15 steps/s | VRAM=28.91 Go [2026-07-18 16:18:39] step 881250/1000000 | loss=2.0410 vocab=0.4126 token=1.3156 | lr=4.06e-05 | 6.13 steps/s | VRAM=32.55 Go [2026-07-18 16:18:47] step 881300/1000000 | loss=2.1484 vocab=0.5576 token=1.4834 | lr=4.06e-05 | 6.13 steps/s | VRAM=29.02 Go [2026-07-18 16:18:56] step 881350/1000000 | loss=2.0282 vocab=0.4742 token=1.2705 | lr=4.05e-05 | 6.14 steps/s | VRAM=25.36 Go [2026-07-18 16:19:04] step 881400/1000000 | loss=2.0747 vocab=0.7971 token=1.5203 | lr=4.05e-05 | 6.14 steps/s | VRAM=32.12 Go [2026-07-18 16:19:12] step 881450/1000000 | loss=2.0926 vocab=1.2090 token=1.3304 | lr=4.05e-05 | 6.12 steps/s | VRAM=29.24 Go [2026-07-18 16:19:20] step 881500/1000000 | loss=2.0127 vocab=1.0157 token=1.1844 | lr=4.04e-05 | 6.16 steps/s | VRAM=31.51 Go [2026-07-18 16:19:28] step 881550/1000000 | loss=2.0838 vocab=0.7479 token=0.9995 | lr=4.04e-05 | 6.11 steps/s | VRAM=26.32 Go [2026-07-18 16:19:36] step 881600/1000000 | loss=2.1768 vocab=0.4152 token=0.7235 | lr=4.04e-05 | 6.14 steps/s | VRAM=28.79 Go [2026-07-18 16:19:44] step 881650/1000000 | loss=2.0848 vocab=0.8200 token=0.8300 | lr=4.04e-05 | 6.13 steps/s | VRAM=29.35 Go [2026-07-18 16:19:53] step 881700/1000000 | loss=2.1841 vocab=1.0458 token=1.3052 | lr=4.03e-05 | 6.14 steps/s | VRAM=29.07 Go [2026-07-18 16:20:01] step 881750/1000000 | loss=2.0233 vocab=0.4081 token=1.0055 | lr=4.03e-05 | 6.15 steps/s | VRAM=32.19 Go [2026-07-18 16:20:09] step 881800/1000000 | loss=2.0738 vocab=0.6302 token=1.3409 | lr=4.03e-05 | 6.13 steps/s | VRAM=28.74 Go [2026-07-18 16:20:17] step 881850/1000000 | loss=2.1679 vocab=0.4666 token=0.9075 | lr=4.02e-05 | 6.13 steps/s | VRAM=29.46 Go [2026-07-18 16:20:25] step 881900/1000000 | loss=2.0892 vocab=0.8991 token=0.9851 | lr=4.02e-05 | 6.14 steps/s | VRAM=32.03 Go [2026-07-18 16:20:33] step 881950/1000000 | loss=2.0763 vocab=0.8737 token=0.9978 | lr=4.02e-05 | 6.12 steps/s | VRAM=26.32 Go [2026-07-18 16:20:42] step 882000/1000000 | loss=2.0170 vocab=1.2945 token=1.3716 | lr=4.02e-05 | 6.12 steps/s | VRAM=31.93 Go [2026-07-18 16:20:50] step 882050/1000000 | loss=2.0020 vocab=0.5050 token=1.3147 | lr=4.01e-05 | 6.14 steps/s | VRAM=32.55 Go [2026-07-18 16:20:58] step 882100/1000000 | loss=2.1359 vocab=0.7342 token=1.0468 | lr=4.01e-05 | 6.14 steps/s | VRAM=32.48 Go [2026-07-18 16:21:06] step 882150/1000000 | loss=2.0605 vocab=1.0080 token=1.1558 | lr=4.01e-05 | 6.16 steps/s | VRAM=31.62 Go [2026-07-18 16:21:14] step 882200/1000000 | loss=2.1215 vocab=0.9696 token=1.1586 | lr=4.01e-05 | 6.13 steps/s | VRAM=25.90 Go [2026-07-18 16:21:22] step 882250/1000000 | loss=2.0255 vocab=0.6758 token=1.2018 | lr=4.00e-05 | 6.15 steps/s | VRAM=25.73 Go [2026-07-18 16:21:30] step 882300/1000000 | loss=2.0669 vocab=1.2548 token=1.2637 | lr=4.00e-05 | 6.14 steps/s | VRAM=29.82 Go [2026-07-18 16:21:39] step 882350/1000000 | loss=2.1293 vocab=0.7788 token=1.2767 | lr=4.00e-05 | 6.14 steps/s | VRAM=28.46 Go [2026-07-18 16:21:47] step 882400/1000000 | loss=2.1347 vocab=0.5221 token=1.2302 | lr=3.99e-05 | 6.14 steps/s | VRAM=28.20 Go [2026-07-18 16:21:55] step 882450/1000000 | loss=2.0938 vocab=1.0891 token=1.6535 | lr=3.99e-05 | 6.15 steps/s | VRAM=28.68 Go [2026-07-18 16:22:03] step 882500/1000000 | loss=2.0967 vocab=1.2890 token=1.5409 | lr=3.99e-05 | 6.13 steps/s | VRAM=26.16 Go [2026-07-18 16:22:11] step 882550/1000000 | loss=1.9964 vocab=1.0649 token=1.6479 | lr=3.99e-05 | 6.09 steps/s | VRAM=28.65 Go [2026-07-18 16:22:19] step 882600/1000000 | loss=2.0824 vocab=0.9545 token=1.4971 | lr=3.98e-05 | 6.15 steps/s | VRAM=26.98 Go [2026-07-18 16:22:27] step 882650/1000000 | loss=2.0281 vocab=0.8159 token=0.7476 | lr=3.98e-05 | 6.13 steps/s | VRAM=31.44 Go [2026-07-18 16:22:36] step 882700/1000000 | loss=2.0684 vocab=0.7719 token=1.5417 | lr=3.98e-05 | 6.10 steps/s | VRAM=29.11 Go [2026-07-18 16:22:44] step 882750/1000000 | loss=2.1282 vocab=0.4624 token=1.1902 | lr=3.97e-05 | 6.09 steps/s | VRAM=28.69 Go [2026-07-18 16:22:52] step 882800/1000000 | loss=2.1107 vocab=0.8537 token=1.2699 | lr=3.97e-05 | 6.13 steps/s | VRAM=30.57 Go [2026-07-18 16:23:00] step 882850/1000000 | loss=2.1113 vocab=1.4548 token=0.9549 | lr=3.97e-05 | 6.12 steps/s | VRAM=26.28 Go [2026-07-18 16:23:08] step 882900/1000000 | loss=1.9506 vocab=1.2457 token=0.9426 | lr=3.97e-05 | 6.16 steps/s | VRAM=28.87 Go [2026-07-18 16:23:16] step 882950/1000000 | loss=2.0779 vocab=0.3476 token=0.8168 | lr=3.96e-05 | 6.11 steps/s | VRAM=28.05 Go [2026-07-18 16:23:25] step 883000/1000000 | loss=2.1055 vocab=1.0501 token=1.2918 | lr=3.96e-05 | 6.11 steps/s | VRAM=26.51 Go [2026-07-18 16:23:33] step 883050/1000000 | loss=2.1239 vocab=0.7363 token=1.0778 | lr=3.96e-05 | 6.15 steps/s | VRAM=31.72 Go [2026-07-18 16:23:41] step 883100/1000000 | loss=2.1668 vocab=0.6322 token=0.9618 | lr=3.96e-05 | 6.13 steps/s | VRAM=31.32 Go [2026-07-18 16:23:49] step 883150/1000000 | loss=2.1836 vocab=1.3009 token=1.0885 | lr=3.95e-05 | 6.13 steps/s | VRAM=24.84 Go [2026-07-18 16:23:57] step 883200/1000000 | loss=2.1355 vocab=0.5664 token=0.9569 | lr=3.95e-05 | 6.11 steps/s | VRAM=29.69 Go [2026-07-18 16:24:05] step 883250/1000000 | loss=2.0384 vocab=0.9015 token=1.1935 | lr=3.95e-05 | 6.14 steps/s | VRAM=29.71 Go [2026-07-18 16:24:14] step 883300/1000000 | loss=2.1003 vocab=0.9707 token=1.3753 | lr=3.94e-05 | 6.14 steps/s | VRAM=24.42 Go [2026-07-18 16:24:22] step 883350/1000000 | loss=2.2157 vocab=0.9877 token=1.1692 | lr=3.94e-05 | 6.13 steps/s | VRAM=29.47 Go [2026-07-18 16:24:30] step 883400/1000000 | loss=2.1876 vocab=0.6344 token=1.4898 | lr=3.94e-05 | 6.12 steps/s | VRAM=26.48 Go [2026-07-18 16:24:38] step 883450/1000000 | loss=2.0746 vocab=0.7465 token=1.2043 | lr=3.94e-05 | 6.13 steps/s | VRAM=29.34 Go [2026-07-18 16:24:46] step 883500/1000000 | loss=2.0457 vocab=1.0911 token=1.4196 | lr=3.93e-05 | 6.12 steps/s | VRAM=29.09 Go [2026-07-18 16:24:54] step 883550/1000000 | loss=2.0685 vocab=1.3599 token=1.0125 | lr=3.93e-05 | 6.13 steps/s | VRAM=33.00 Go [2026-07-18 16:25:03] step 883600/1000000 | loss=2.0176 vocab=0.9469 token=1.3787 | lr=3.93e-05 | 6.13 steps/s | VRAM=26.26 Go [2026-07-18 16:25:11] step 883650/1000000 | loss=2.0711 vocab=0.8292 token=1.1483 | lr=3.92e-05 | 6.15 steps/s | VRAM=26.08 Go [2026-07-18 16:25:19] step 883700/1000000 | loss=2.1414 vocab=0.8100 token=1.0723 | lr=3.92e-05 | 6.12 steps/s | VRAM=26.47 Go [2026-07-18 16:25:27] step 883750/1000000 | loss=2.0278 vocab=1.0199 token=1.1788 | lr=3.92e-05 | 6.13 steps/s | VRAM=30.87 Go [2026-07-18 16:25:35] step 883800/1000000 | loss=2.1331 vocab=1.5354 token=1.6105 | lr=3.92e-05 | 6.13 steps/s | VRAM=28.47 Go [2026-07-18 16:25:43] step 883850/1000000 | loss=2.0467 vocab=0.8575 token=1.2345 | lr=3.91e-05 | 6.13 steps/s | VRAM=27.95 Go [2026-07-18 16:25:52] step 883900/1000000 | loss=2.1383 vocab=0.6303 token=1.2016 | lr=3.91e-05 | 6.13 steps/s | VRAM=25.43 Go [2026-07-18 16:26:00] step 883950/1000000 | loss=2.1593 vocab=0.9028 token=1.1942 | lr=3.91e-05 | 6.15 steps/s | VRAM=27.86 Go [2026-07-18 16:26:08] step 884000/1000000 | loss=2.1452 vocab=0.9580 token=1.6334 | lr=3.91e-05 | 6.15 steps/s | VRAM=32.49 Go [2026-07-18 16:26:16] step 884050/1000000 | loss=2.0328 vocab=1.1802 token=1.0622 | lr=3.90e-05 | 6.13 steps/s | VRAM=25.21 Go [2026-07-18 16:26:24] step 884100/1000000 | loss=2.0734 vocab=1.0211 token=1.2538 | lr=3.90e-05 | 6.15 steps/s | VRAM=26.35 Go [2026-07-18 16:26:32] step 884150/1000000 | loss=2.0549 vocab=0.5277 token=1.0824 | lr=3.90e-05 | 6.16 steps/s | VRAM=31.96 Go [2026-07-18 16:26:40] step 884200/1000000 | loss=2.1112 vocab=0.9754 token=1.1031 | lr=3.89e-05 | 6.11 steps/s | VRAM=28.06 Go [2026-07-18 16:26:49] step 884250/1000000 | loss=2.1792 vocab=0.5982 token=1.1922 | lr=3.89e-05 | 6.13 steps/s | VRAM=32.72 Go [2026-07-18 16:26:57] step 884300/1000000 | loss=2.0196 vocab=0.7283 token=1.0019 | lr=3.89e-05 | 6.14 steps/s | VRAM=32.12 Go [2026-07-18 16:27:05] step 884350/1000000 | loss=2.1299 vocab=0.8896 token=1.3656 | lr=3.89e-05 | 6.11 steps/s | VRAM=29.72 Go [2026-07-18 16:27:13] step 884400/1000000 | loss=2.1371 vocab=1.1159 token=1.3012 | lr=3.88e-05 | 6.14 steps/s | VRAM=32.79 Go [2026-07-18 16:27:21] step 884450/1000000 | loss=2.0503 vocab=0.8163 token=1.1738 | lr=3.88e-05 | 6.13 steps/s | VRAM=31.79 Go [2026-07-18 16:27:29] step 884500/1000000 | loss=2.0557 vocab=1.5417 token=1.0593 | lr=3.88e-05 | 6.13 steps/s | VRAM=28.98 Go [2026-07-18 16:27:37] step 884550/1000000 | loss=2.0200 vocab=1.1411 token=1.1181 | lr=3.87e-05 | 6.12 steps/s | VRAM=28.41 Go [2026-07-18 16:27:46] step 884600/1000000 | loss=2.1210 vocab=1.1909 token=1.4808 | lr=3.87e-05 | 6.14 steps/s | VRAM=28.59 Go [2026-07-18 16:27:54] step 884650/1000000 | loss=2.1478 vocab=1.2144 token=1.1510 | lr=3.87e-05 | 6.15 steps/s | VRAM=28.93 Go [2026-07-18 16:28:02] step 884700/1000000 | loss=2.1194 vocab=0.8404 token=1.3228 | lr=3.87e-05 | 6.11 steps/s | VRAM=28.70 Go [2026-07-18 16:28:10] step 884750/1000000 | loss=2.0947 vocab=1.2858 token=1.3267 | lr=3.86e-05 | 6.11 steps/s | VRAM=25.72 Go [2026-07-18 16:28:18] step 884800/1000000 | loss=2.0858 vocab=0.6739 token=1.0409 | lr=3.86e-05 | 6.15 steps/s | VRAM=31.25 Go [2026-07-18 16:28:26] step 884850/1000000 | loss=2.0854 vocab=0.5859 token=0.9931 | lr=3.86e-05 | 6.13 steps/s | VRAM=29.97 Go [2026-07-18 16:28:35] step 884900/1000000 | loss=1.9679 vocab=0.6307 token=1.0061 | lr=3.86e-05 | 6.09 steps/s | VRAM=29.87 Go [2026-07-18 16:28:43] step 884950/1000000 | loss=2.0742 vocab=0.6317 token=1.1744 | lr=3.85e-05 | 6.14 steps/s | VRAM=24.30 Go [2026-07-18 16:28:51] step 885000/1000000 | loss=2.1683 vocab=0.6310 token=1.2270 | lr=3.85e-05 | 6.11 steps/s | VRAM=32.06 Go [2026-07-18 16:28:59] step 885050/1000000 | loss=2.0191 vocab=0.8257 token=1.1151 | lr=3.85e-05 | 6.10 steps/s | VRAM=28.54 Go [2026-07-18 16:29:07] step 885100/1000000 | loss=2.0646 vocab=0.9950 token=0.9640 | lr=3.84e-05 | 6.12 steps/s | VRAM=26.77 Go [2026-07-18 16:29:15] step 885150/1000000 | loss=2.0851 vocab=0.9172 token=0.9471 | lr=3.84e-05 | 6.12 steps/s | VRAM=26.63 Go [2026-07-18 16:29:24] step 885200/1000000 | loss=2.0633 vocab=1.2141 token=1.2228 | lr=3.84e-05 | 6.13 steps/s | VRAM=31.04 Go [2026-07-18 16:29:32] step 885250/1000000 | loss=2.0342 vocab=0.5799 token=1.0789 | lr=3.84e-05 | 6.11 steps/s | VRAM=26.94 Go [2026-07-18 16:29:40] step 885300/1000000 | loss=2.0343 vocab=1.1720 token=1.4012 | lr=3.83e-05 | 6.13 steps/s | VRAM=25.55 Go [2026-07-18 16:29:48] step 885350/1000000 | loss=2.0771 vocab=0.7951 token=1.2460 | lr=3.83e-05 | 6.15 steps/s | VRAM=31.14 Go [2026-07-18 16:29:56] step 885400/1000000 | loss=2.1217 vocab=1.3583 token=1.3435 | lr=3.83e-05 | 6.12 steps/s | VRAM=28.65 Go [2026-07-18 16:30:04] step 885450/1000000 | loss=2.0048 vocab=0.8618 token=0.9667 | lr=3.82e-05 | 6.12 steps/s | VRAM=29.06 Go [2026-07-18 16:30:13] step 885500/1000000 | loss=2.1508 vocab=0.5602 token=0.8990 | lr=3.82e-05 | 6.13 steps/s | VRAM=29.78 Go [2026-07-18 16:30:21] step 885550/1000000 | loss=2.0330 vocab=0.6598 token=0.7835 | lr=3.82e-05 | 6.10 steps/s | VRAM=26.88 Go [2026-07-18 16:30:29] step 885600/1000000 | loss=2.0618 vocab=1.0316 token=1.7174 | lr=3.82e-05 | 6.13 steps/s | VRAM=32.00 Go [2026-07-18 16:30:37] step 885650/1000000 | loss=2.0599 vocab=0.4192 token=1.1864 | lr=3.81e-05 | 6.15 steps/s | VRAM=28.57 Go [2026-07-18 16:30:45] step 885700/1000000 | loss=2.1887 vocab=0.9235 token=1.4075 | lr=3.81e-05 | 6.13 steps/s | VRAM=26.30 Go [2026-07-18 16:30:53] step 885750/1000000 | loss=2.0889 vocab=0.4532 token=1.4285 | lr=3.81e-05 | 6.12 steps/s | VRAM=29.14 Go [2026-07-18 16:31:02] step 885800/1000000 | loss=2.0309 vocab=0.5147 token=0.9277 | lr=3.81e-05 | 6.14 steps/s | VRAM=25.51 Go [2026-07-18 16:31:10] step 885850/1000000 | loss=2.1081 vocab=1.1996 token=1.3557 | lr=3.80e-05 | 6.09 steps/s | VRAM=26.52 Go [2026-07-18 16:31:18] step 885900/1000000 | loss=2.0653 vocab=0.7988 token=0.9020 | lr=3.80e-05 | 6.15 steps/s | VRAM=28.16 Go [2026-07-18 16:31:26] step 885950/1000000 | loss=2.0807 vocab=1.5345 token=1.3884 | lr=3.80e-05 | 6.15 steps/s | VRAM=25.65 Go [2026-07-18 16:31:34] step 886000/1000000 | loss=2.0776 vocab=0.9457 token=1.2733 | lr=3.79e-05 | 6.11 steps/s | VRAM=29.32 Go [2026-07-18 16:31:42] step 886050/1000000 | loss=2.1021 vocab=0.9157 token=1.3712 | lr=3.79e-05 | 6.11 steps/s | VRAM=26.29 Go [2026-07-18 16:31:51] step 886100/1000000 | loss=2.1271 vocab=0.7046 token=1.0948 | lr=3.79e-05 | 6.12 steps/s | VRAM=26.71 Go [2026-07-18 16:31:59] step 886150/1000000 | loss=2.1645 vocab=0.7580 token=1.1423 | lr=3.79e-05 | 6.13 steps/s | VRAM=32.20 Go [2026-07-18 16:32:07] step 886200/1000000 | loss=2.1240 vocab=0.6312 token=1.1124 | lr=3.78e-05 | 6.12 steps/s | VRAM=28.18 Go [2026-07-18 16:32:15] step 886250/1000000 | loss=2.0248 vocab=0.8613 token=0.6331 | lr=3.78e-05 | 6.12 steps/s | VRAM=27.14 Go [2026-07-18 16:32:23] step 886300/1000000 | loss=2.1730 vocab=0.9914 token=1.5024 | lr=3.78e-05 | 6.13 steps/s | VRAM=28.70 Go [2026-07-18 16:32:31] step 886350/1000000 | loss=2.1345 vocab=1.3808 token=1.4138 | lr=3.78e-05 | 6.13 steps/s | VRAM=26.24 Go [2026-07-18 16:32:40] step 886400/1000000 | loss=2.1830 vocab=1.1472 token=1.4708 | lr=3.77e-05 | 6.10 steps/s | VRAM=24.71 Go [2026-07-18 16:32:48] step 886450/1000000 | loss=2.0413 vocab=0.7925 token=0.9902 | lr=3.77e-05 | 6.11 steps/s | VRAM=30.25 Go [2026-07-18 16:32:56] step 886500/1000000 | loss=1.9862 vocab=0.8339 token=0.9665 | lr=3.77e-05 | 6.12 steps/s | VRAM=25.33 Go [2026-07-18 16:33:04] step 886550/1000000 | loss=2.0802 vocab=1.0878 token=1.2837 | lr=3.76e-05 | 6.13 steps/s | VRAM=27.88 Go [2026-07-18 16:33:12] step 886600/1000000 | loss=2.0892 vocab=0.4047 token=0.9676 | lr=3.76e-05 | 6.10 steps/s | VRAM=26.64 Go [2026-07-18 16:33:20] step 886650/1000000 | loss=2.1447 vocab=0.6334 token=1.1308 | lr=3.76e-05 | 6.14 steps/s | VRAM=32.87 Go [2026-07-18 16:33:29] step 886700/1000000 | loss=2.1682 vocab=0.7899 token=1.4522 | lr=3.76e-05 | 6.14 steps/s | VRAM=24.84 Go [2026-07-18 16:33:37] step 886750/1000000 | loss=2.0626 vocab=1.4820 token=1.5690 | lr=3.75e-05 | 6.15 steps/s | VRAM=27.80 Go [2026-07-18 16:33:45] step 886800/1000000 | loss=2.1205 vocab=0.8443 token=1.1458 | lr=3.75e-05 | 6.12 steps/s | VRAM=32.07 Go [2026-07-18 16:33:53] step 886850/1000000 | loss=2.0834 vocab=1.2884 token=1.4088 | lr=3.75e-05 | 6.15 steps/s | VRAM=28.03 Go [2026-07-18 16:34:01] step 886900/1000000 | loss=2.0466 vocab=0.5893 token=1.4125 | lr=3.74e-05 | 6.11 steps/s | VRAM=27.54 Go [2026-07-18 16:34:09] step 886950/1000000 | loss=2.0880 vocab=0.9889 token=1.5242 | lr=3.74e-05 | 6.13 steps/s | VRAM=31.35 Go [2026-07-18 16:34:18] step 887000/1000000 | loss=2.1012 vocab=0.4999 token=0.7872 | lr=3.74e-05 | 6.14 steps/s | VRAM=27.34 Go [2026-07-18 16:34:26] step 887050/1000000 | loss=2.0254 vocab=0.6259 token=1.0103 | lr=3.74e-05 | 6.14 steps/s | VRAM=29.14 Go [2026-07-18 16:34:34] step 887100/1000000 | loss=2.1049 vocab=0.7183 token=0.9779 | lr=3.73e-05 | 6.12 steps/s | VRAM=32.22 Go [2026-07-18 16:34:42] step 887150/1000000 | loss=2.1308 vocab=1.0957 token=1.3936 | lr=3.73e-05 | 6.13 steps/s | VRAM=31.89 Go [2026-07-18 16:34:50] step 887200/1000000 | loss=2.0182 vocab=0.9571 token=1.1153 | lr=3.73e-05 | 6.13 steps/s | VRAM=25.62 Go [2026-07-18 16:34:58] step 887250/1000000 | loss=2.0894 vocab=0.8200 token=0.9802 | lr=3.73e-05 | 6.12 steps/s | VRAM=24.99 Go [2026-07-18 16:35:06] step 887300/1000000 | loss=2.1610 vocab=0.4863 token=0.9375 | lr=3.72e-05 | 6.14 steps/s | VRAM=32.46 Go [2026-07-18 16:35:15] step 887350/1000000 | loss=1.9860 vocab=1.0899 token=1.3633 | lr=3.72e-05 | 6.14 steps/s | VRAM=31.10 Go [2026-07-18 16:35:23] step 887400/1000000 | loss=2.0151 vocab=0.6961 token=1.5277 | lr=3.72e-05 | 6.12 steps/s | VRAM=32.03 Go [2026-07-18 16:35:31] step 887450/1000000 | loss=2.0457 vocab=0.6123 token=1.1501 | lr=3.71e-05 | 6.10 steps/s | VRAM=28.98 Go [2026-07-18 16:35:39] step 887500/1000000 | loss=2.0563 vocab=0.5658 token=0.7960 | lr=3.71e-05 | 6.11 steps/s | VRAM=26.14 Go [2026-07-18 16:35:47] step 887550/1000000 | loss=2.1039 vocab=1.0101 token=1.3887 | lr=3.71e-05 | 6.13 steps/s | VRAM=29.35 Go [2026-07-18 16:35:55] step 887600/1000000 | loss=2.0211 vocab=0.8459 token=1.3027 | lr=3.71e-05 | 6.11 steps/s | VRAM=31.56 Go [2026-07-18 16:36:04] step 887650/1000000 | loss=2.0471 vocab=0.8105 token=1.4320 | lr=3.70e-05 | 6.07 steps/s | VRAM=27.47 Go [2026-07-18 16:36:12] step 887700/1000000 | loss=2.1098 vocab=0.3299 token=0.9558 | lr=3.70e-05 | 6.11 steps/s | VRAM=28.10 Go [2026-07-18 16:36:20] step 887750/1000000 | loss=2.0837 vocab=0.5879 token=0.8674 | lr=3.70e-05 | 6.10 steps/s | VRAM=25.85 Go [2026-07-18 16:36:28] step 887800/1000000 | loss=2.1507 vocab=0.8725 token=1.2848 | lr=3.70e-05 | 6.14 steps/s | VRAM=30.89 Go [2026-07-18 16:36:36] step 887850/1000000 | loss=2.0790 vocab=0.7257 token=1.2314 | lr=3.69e-05 | 6.11 steps/s | VRAM=28.06 Go [2026-07-18 16:36:45] step 887900/1000000 | loss=2.1579 vocab=0.7365 token=1.3657 | lr=3.69e-05 | 6.09 steps/s | VRAM=26.69 Go [2026-07-18 16:36:53] step 887950/1000000 | loss=2.0807 vocab=0.8667 token=1.0046 | lr=3.69e-05 | 6.11 steps/s | VRAM=32.86 Go [2026-07-18 16:37:01] step 888000/1000000 | loss=2.0788 vocab=0.9440 token=1.2181 | lr=3.68e-05 | 6.11 steps/s | VRAM=28.62 Go [2026-07-18 16:37:09] step 888050/1000000 | loss=2.0517 vocab=1.0117 token=1.5790 | lr=3.68e-05 | 6.13 steps/s | VRAM=27.56 Go [2026-07-18 16:37:17] step 888100/1000000 | loss=2.0287 vocab=0.8055 token=1.1779 | lr=3.68e-05 | 6.12 steps/s | VRAM=27.16 Go [2026-07-18 16:37:26] step 888150/1000000 | loss=2.1485 vocab=0.6739 token=0.8430 | lr=3.68e-05 | 6.13 steps/s | VRAM=30.77 Go [2026-07-18 16:37:34] step 888200/1000000 | loss=2.0332 vocab=0.6729 token=1.1408 | lr=3.67e-05 | 6.11 steps/s | VRAM=28.35 Go [2026-07-18 16:37:42] step 888250/1000000 | loss=2.1294 vocab=0.7863 token=1.1133 | lr=3.67e-05 | 6.12 steps/s | VRAM=31.40 Go [2026-07-18 16:37:50] step 888300/1000000 | loss=2.1287 vocab=1.1619 token=1.5066 | lr=3.67e-05 | 6.11 steps/s | VRAM=29.18 Go [2026-07-18 16:37:58] step 888350/1000000 | loss=2.0236 vocab=0.7603 token=1.4799 | lr=3.67e-05 | 6.14 steps/s | VRAM=31.11 Go [2026-07-18 16:38:06] step 888400/1000000 | loss=2.0273 vocab=1.2680 token=1.0891 | lr=3.66e-05 | 6.11 steps/s | VRAM=26.66 Go [2026-07-18 16:38:15] step 888450/1000000 | loss=1.9959 vocab=0.8415 token=0.9339 | lr=3.66e-05 | 6.12 steps/s | VRAM=29.46 Go [2026-07-18 16:38:23] step 888500/1000000 | loss=2.0867 vocab=0.9466 token=1.6174 | lr=3.66e-05 | 6.14 steps/s | VRAM=32.65 Go [2026-07-18 16:38:31] step 888550/1000000 | loss=2.0208 vocab=0.9099 token=1.4031 | lr=3.65e-05 | 6.13 steps/s | VRAM=28.98 Go [2026-07-18 16:38:39] step 888600/1000000 | loss=2.0738 vocab=0.7700 token=1.3312 | lr=3.65e-05 | 6.11 steps/s | VRAM=26.41 Go [2026-07-18 16:38:47] step 888650/1000000 | loss=2.0638 vocab=1.1630 token=1.5340 | lr=3.65e-05 | 6.11 steps/s | VRAM=29.65 Go [2026-07-18 16:38:55] step 888700/1000000 | loss=2.0735 vocab=0.5547 token=0.8902 | lr=3.65e-05 | 6.10 steps/s | VRAM=28.68 Go [2026-07-18 16:39:04] step 888750/1000000 | loss=2.0916 vocab=1.0221 token=1.2609 | lr=3.64e-05 | 6.11 steps/s | VRAM=28.91 Go [2026-07-18 16:39:12] step 888800/1000000 | loss=2.1227 vocab=0.6972 token=1.0856 | lr=3.64e-05 | 6.14 steps/s | VRAM=24.89 Go [2026-07-18 16:39:20] step 888850/1000000 | loss=2.0333 vocab=0.7308 token=0.9402 | lr=3.64e-05 | 6.12 steps/s | VRAM=29.25 Go [2026-07-18 16:39:28] step 888900/1000000 | loss=2.0906 vocab=0.7936 token=1.1335 | lr=3.64e-05 | 6.13 steps/s | VRAM=29.45 Go [2026-07-18 16:39:36] step 888950/1000000 | loss=2.1080 vocab=0.5602 token=0.9897 | lr=3.63e-05 | 6.08 steps/s | VRAM=26.99 Go [2026-07-18 16:39:44] step 889000/1000000 | loss=2.0817 vocab=1.6257 token=1.3176 | lr=3.63e-05 | 6.12 steps/s | VRAM=28.60 Go [2026-07-18 16:39:53] step 889050/1000000 | loss=2.2055 vocab=0.6998 token=1.3695 | lr=3.63e-05 | 6.11 steps/s | VRAM=26.35 Go [2026-07-18 16:40:01] step 889100/1000000 | loss=2.0216 vocab=0.7890 token=1.2811 | lr=3.62e-05 | 6.12 steps/s | VRAM=29.28 Go [2026-07-18 16:40:09] step 889150/1000000 | loss=2.2317 vocab=1.3801 token=1.8180 | lr=3.62e-05 | 6.10 steps/s | VRAM=28.62 Go [2026-07-18 16:40:17] step 889200/1000000 | loss=2.0875 vocab=0.7773 token=0.8361 | lr=3.62e-05 | 6.16 steps/s | VRAM=31.56 Go [2026-07-18 16:40:25] step 889250/1000000 | loss=2.0498 vocab=0.8863 token=1.3016 | lr=3.62e-05 | 6.12 steps/s | VRAM=29.18 Go [2026-07-18 16:40:34] step 889300/1000000 | loss=1.9427 vocab=0.5032 token=1.2491 | lr=3.61e-05 | 6.11 steps/s | VRAM=28.90 Go [2026-07-18 16:40:42] step 889350/1000000 | loss=2.0699 vocab=0.6300 token=1.0930 | lr=3.61e-05 | 6.13 steps/s | VRAM=28.84 Go [2026-07-18 16:40:50] step 889400/1000000 | loss=2.0469 vocab=0.9790 token=1.3131 | lr=3.61e-05 | 6.11 steps/s | VRAM=28.63 Go [2026-07-18 16:40:58] step 889450/1000000 | loss=1.9704 vocab=0.4617 token=0.8757 | lr=3.61e-05 | 6.10 steps/s | VRAM=26.17 Go [2026-07-18 16:41:06] step 889500/1000000 | loss=2.1053 vocab=0.8873 token=1.4715 | lr=3.60e-05 | 6.14 steps/s | VRAM=32.68 Go [2026-07-18 16:41:14] step 889550/1000000 | loss=2.0544 vocab=0.7828 token=1.2734 | lr=3.60e-05 | 6.10 steps/s | VRAM=29.05 Go [2026-07-18 16:41:22] step 889600/1000000 | loss=2.1529 vocab=0.7261 token=1.1551 | lr=3.60e-05 | 6.16 steps/s | VRAM=32.02 Go [2026-07-18 16:41:31] step 889650/1000000 | loss=2.1152 vocab=1.1329 token=1.3730 | lr=3.59e-05 | 6.11 steps/s | VRAM=25.10 Go [2026-07-18 16:41:39] step 889700/1000000 | loss=2.0694 vocab=0.5590 token=1.1180 | lr=3.59e-05 | 6.11 steps/s | VRAM=26.62 Go [2026-07-18 16:41:47] step 889750/1000000 | loss=2.0412 vocab=0.6806 token=1.1364 | lr=3.59e-05 | 6.15 steps/s | VRAM=24.96 Go [2026-07-18 16:41:55] step 889800/1000000 | loss=2.0641 vocab=1.0976 token=1.3065 | lr=3.59e-05 | 6.12 steps/s | VRAM=26.12 Go [2026-07-18 16:42:03] step 889850/1000000 | loss=2.0658 vocab=1.1644 token=1.1941 | lr=3.58e-05 | 6.11 steps/s | VRAM=29.01 Go [2026-07-18 16:42:12] step 889900/1000000 | loss=2.1301 vocab=1.4572 token=1.1483 | lr=3.58e-05 | 6.13 steps/s | VRAM=27.79 Go [2026-07-18 16:42:20] step 889950/1000000 | loss=2.0492 vocab=0.5058 token=1.0503 | lr=3.58e-05 | 6.10 steps/s | VRAM=29.76 Go [2026-07-18 16:42:28] step 890000/1000000 | loss=2.0895 vocab=0.9756 token=2.0033 | lr=3.58e-05 | 6.15 steps/s | VRAM=27.97 Go [2026-07-18 16:42:36] step 890050/1000000 | loss=2.0236 vocab=1.3649 token=1.5318 | lr=3.57e-05 | 6.16 steps/s | VRAM=31.69 Go [2026-07-18 16:42:44] step 890100/1000000 | loss=2.1112 vocab=0.6466 token=1.0137 | lr=3.57e-05 | 6.14 steps/s | VRAM=25.26 Go [2026-07-18 16:42:52] step 890150/1000000 | loss=2.0643 vocab=1.0476 token=1.6835 | lr=3.57e-05 | 6.12 steps/s | VRAM=29.37 Go [2026-07-18 16:43:00] step 890200/1000000 | loss=1.9325 vocab=0.6081 token=1.2774 | lr=3.56e-05 | 6.14 steps/s | VRAM=31.63 Go [2026-07-18 16:43:09] step 890250/1000000 | loss=1.9449 vocab=0.4764 token=0.7075 | lr=3.56e-05 | 6.13 steps/s | VRAM=28.60 Go [2026-07-18 16:43:17] step 890300/1000000 | loss=2.0716 vocab=1.2156 token=1.2734 | lr=3.56e-05 | 6.09 steps/s | VRAM=28.15 Go [2026-07-18 16:43:25] step 890350/1000000 | loss=2.1863 vocab=0.7215 token=0.8154 | lr=3.56e-05 | 6.12 steps/s | VRAM=32.04 Go [2026-07-18 16:43:33] step 890400/1000000 | loss=2.0493 vocab=0.7740 token=1.0068 | lr=3.55e-05 | 6.14 steps/s | VRAM=28.20 Go [2026-07-18 16:43:41] step 890450/1000000 | loss=2.1091 vocab=0.8494 token=1.1618 | lr=3.55e-05 | 6.14 steps/s | VRAM=28.79 Go [2026-07-18 16:43:49] step 890500/1000000 | loss=2.0763 vocab=0.7466 token=0.7850 | lr=3.55e-05 | 6.13 steps/s | VRAM=29.39 Go [2026-07-18 16:43:58] step 890550/1000000 | loss=2.1189 vocab=0.8313 token=1.0611 | lr=3.55e-05 | 6.12 steps/s | VRAM=26.15 Go [2026-07-18 16:44:06] step 890600/1000000 | loss=2.2361 vocab=1.0068 token=1.3110 | lr=3.54e-05 | 6.11 steps/s | VRAM=32.74 Go [2026-07-18 16:44:14] step 890650/1000000 | loss=2.1023 vocab=1.1559 token=1.1201 | lr=3.54e-05 | 6.12 steps/s | VRAM=32.49 Go [2026-07-18 16:44:22] step 890700/1000000 | loss=2.0575 vocab=1.1831 token=1.5621 | lr=3.54e-05 | 6.10 steps/s | VRAM=26.49 Go [2026-07-18 16:44:30] step 890750/1000000 | loss=2.0541 vocab=0.4595 token=1.1228 | lr=3.53e-05 | 6.12 steps/s | VRAM=28.66 Go [2026-07-18 16:44:38] step 890800/1000000 | loss=2.0981 vocab=0.6770 token=1.0190 | lr=3.53e-05 | 6.11 steps/s | VRAM=28.87 Go [2026-07-18 16:44:47] step 890850/1000000 | loss=2.0824 vocab=0.7718 token=1.1590 | lr=3.53e-05 | 6.15 steps/s | VRAM=30.89 Go [2026-07-18 16:44:55] step 890900/1000000 | loss=2.0414 vocab=1.1330 token=1.2551 | lr=3.53e-05 | 6.13 steps/s | VRAM=24.99 Go [2026-07-18 16:45:03] step 890950/1000000 | loss=2.0783 vocab=0.7000 token=1.3553 | lr=3.52e-05 | 6.14 steps/s | VRAM=31.16 Go [2026-07-18 16:45:11] step 891000/1000000 | loss=2.1787 vocab=1.0347 token=1.1893 | lr=3.52e-05 | 6.13 steps/s | VRAM=28.84 Go [2026-07-18 16:45:19] step 891050/1000000 | loss=2.1047 vocab=1.0737 token=1.5575 | lr=3.52e-05 | 6.13 steps/s | VRAM=28.50 Go [2026-07-18 16:45:27] step 891100/1000000 | loss=1.8745 vocab=0.4849 token=1.3826 | lr=3.52e-05 | 6.12 steps/s | VRAM=32.95 Go [2026-07-18 16:45:36] step 891150/1000000 | loss=1.9982 vocab=0.3611 token=0.7944 | lr=3.51e-05 | 6.09 steps/s | VRAM=28.27 Go [2026-07-18 16:45:44] step 891200/1000000 | loss=2.0408 vocab=0.5531 token=1.0615 | lr=3.51e-05 | 6.09 steps/s | VRAM=26.49 Go [2026-07-18 16:45:52] step 891250/1000000 | loss=1.9676 vocab=0.5707 token=1.0372 | lr=3.51e-05 | 6.12 steps/s | VRAM=29.19 Go [2026-07-18 16:46:00] step 891300/1000000 | loss=2.1027 vocab=1.1465 token=1.5791 | lr=3.50e-05 | 6.11 steps/s | VRAM=29.31 Go [2026-07-18 16:46:08] step 891350/1000000 | loss=2.0472 vocab=1.2510 token=1.1495 | lr=3.50e-05 | 6.13 steps/s | VRAM=28.32 Go [2026-07-18 16:46:17] step 891400/1000000 | loss=1.9696 vocab=0.5653 token=1.4497 | lr=3.50e-05 | 6.11 steps/s | VRAM=27.81 Go [2026-07-18 16:46:25] step 891450/1000000 | loss=1.9933 vocab=0.7579 token=1.3582 | lr=3.50e-05 | 6.13 steps/s | VRAM=28.93 Go [2026-07-18 16:46:33] step 891500/1000000 | loss=2.0073 vocab=0.3706 token=1.1165 | lr=3.49e-05 | 6.08 steps/s | VRAM=26.76 Go [2026-07-18 16:46:41] step 891550/1000000 | loss=2.1860 vocab=1.2856 token=1.3804 | lr=3.49e-05 | 6.10 steps/s | VRAM=28.60 Go [2026-07-18 16:46:49] step 891600/1000000 | loss=2.0696 vocab=1.1812 token=1.1172 | lr=3.49e-05 | 6.09 steps/s | VRAM=28.82 Go [2026-07-18 16:46:58] step 891650/1000000 | loss=2.1640 vocab=0.8959 token=1.2711 | lr=3.49e-05 | 6.10 steps/s | VRAM=27.91 Go [2026-07-18 16:47:06] step 891700/1000000 | loss=1.9737 vocab=0.3360 token=0.8565 | lr=3.48e-05 | 6.10 steps/s | VRAM=31.87 Go [2026-07-18 16:47:14] step 891750/1000000 | loss=2.1026 vocab=0.9398 token=1.3636 | lr=3.48e-05 | 6.12 steps/s | VRAM=32.46 Go [2026-07-18 16:47:22] step 891800/1000000 | loss=1.9968 vocab=0.6588 token=0.9576 | lr=3.48e-05 | 6.07 steps/s | VRAM=23.43 Go [2026-07-18 16:47:30] step 891850/1000000 | loss=2.0830 vocab=0.6256 token=1.0787 | lr=3.47e-05 | 6.12 steps/s | VRAM=29.61 Go [2026-07-18 16:47:38] step 891900/1000000 | loss=2.0074 vocab=0.9813 token=1.1556 | lr=3.47e-05 | 6.13 steps/s | VRAM=29.31 Go [2026-07-18 16:47:47] step 891950/1000000 | loss=2.0850 vocab=0.8301 token=1.1766 | lr=3.47e-05 | 6.12 steps/s | VRAM=26.79 Go [2026-07-18 16:47:55] step 892000/1000000 | loss=2.1009 vocab=1.1989 token=1.5773 | lr=3.47e-05 | 6.11 steps/s | VRAM=28.34 Go [2026-07-18 16:48:03] step 892050/1000000 | loss=2.0464 vocab=0.6603 token=1.1187 | lr=3.46e-05 | 6.12 steps/s | VRAM=29.10 Go [2026-07-18 16:48:11] step 892100/1000000 | loss=1.9760 vocab=0.7354 token=1.3114 | lr=3.46e-05 | 6.11 steps/s | VRAM=25.86 Go [2026-07-18 16:48:19] step 892150/1000000 | loss=2.0255 vocab=0.9016 token=1.4083 | lr=3.46e-05 | 6.12 steps/s | VRAM=26.58 Go [2026-07-18 16:48:27] step 892200/1000000 | loss=2.0449 vocab=1.2401 token=1.6287 | lr=3.46e-05 | 6.12 steps/s | VRAM=29.36 Go [2026-07-18 16:48:36] step 892250/1000000 | loss=2.0874 vocab=1.2431 token=1.1862 | lr=3.45e-05 | 6.11 steps/s | VRAM=26.22 Go [2026-07-18 16:48:44] step 892300/1000000 | loss=2.0397 vocab=0.5689 token=1.0561 | lr=3.45e-05 | 6.14 steps/s | VRAM=28.98 Go [2026-07-18 16:48:52] step 892350/1000000 | loss=2.0579 vocab=0.5552 token=0.7490 | lr=3.45e-05 | 6.11 steps/s | VRAM=28.26 Go [2026-07-18 16:49:00] step 892400/1000000 | loss=2.1374 vocab=0.5234 token=1.2671 | lr=3.44e-05 | 6.11 steps/s | VRAM=29.21 Go [2026-07-18 16:49:08] step 892450/1000000 | loss=1.9782 vocab=1.2287 token=1.5782 | lr=3.44e-05 | 6.13 steps/s | VRAM=28.80 Go [2026-07-18 16:49:16] step 892500/1000000 | loss=2.1063 vocab=0.6711 token=0.7320 | lr=3.44e-05 | 6.15 steps/s | VRAM=25.85 Go [2026-07-18 16:49:25] step 892550/1000000 | loss=1.9965 vocab=0.7795 token=1.6363 | lr=3.44e-05 | 6.14 steps/s | VRAM=28.24 Go [2026-07-18 16:49:33] step 892600/1000000 | loss=2.0219 vocab=0.5307 token=1.1584 | lr=3.43e-05 | 6.11 steps/s | VRAM=31.32 Go [2026-07-18 16:49:41] step 892650/1000000 | loss=2.0452 vocab=1.6074 token=1.3719 | lr=3.43e-05 | 6.08 steps/s | VRAM=26.50 Go [2026-07-18 16:49:49] step 892700/1000000 | loss=2.0574 vocab=0.7734 token=1.0539 | lr=3.43e-05 | 6.14 steps/s | VRAM=27.69 Go [2026-07-18 16:49:57] step 892750/1000000 | loss=2.0545 vocab=1.5155 token=1.1696 | lr=3.43e-05 | 6.15 steps/s | VRAM=32.50 Go [2026-07-18 16:50:05] step 892800/1000000 | loss=1.9574 vocab=0.7959 token=1.3038 | lr=3.42e-05 | 6.15 steps/s | VRAM=25.72 Go [2026-07-18 16:50:14] step 892850/1000000 | loss=2.0106 vocab=0.4465 token=0.8862 | lr=3.42e-05 | 6.15 steps/s | VRAM=25.84 Go [2026-07-18 16:50:22] step 892900/1000000 | loss=2.0805 vocab=0.9404 token=1.5197 | lr=3.42e-05 | 6.15 steps/s | VRAM=28.26 Go [2026-07-18 16:50:30] step 892950/1000000 | loss=2.1179 vocab=0.7953 token=1.0981 | lr=3.42e-05 | 6.15 steps/s | VRAM=28.56 Go [2026-07-18 16:50:38] step 893000/1000000 | loss=2.0775 vocab=0.9579 token=1.1063 | lr=3.41e-05 | 6.15 steps/s | VRAM=24.76 Go [2026-07-18 16:50:46] step 893050/1000000 | loss=2.0875 vocab=0.8485 token=1.4074 | lr=3.41e-05 | 6.13 steps/s | VRAM=25.26 Go [2026-07-18 16:50:54] step 893100/1000000 | loss=2.1072 vocab=0.7601 token=1.0894 | lr=3.41e-05 | 6.14 steps/s | VRAM=26.19 Go [2026-07-18 16:51:02] step 893150/1000000 | loss=2.0315 vocab=0.6195 token=0.8748 | lr=3.40e-05 | 6.16 steps/s | VRAM=32.29 Go [2026-07-18 16:51:10] step 893200/1000000 | loss=2.0462 vocab=0.7858 token=1.2312 | lr=3.40e-05 | 6.14 steps/s | VRAM=32.47 Go [2026-07-18 16:51:19] step 893250/1000000 | loss=1.9754 vocab=1.1838 token=1.3821 | lr=3.40e-05 | 6.14 steps/s | VRAM=28.87 Go [2026-07-18 16:51:27] step 893300/1000000 | loss=2.0647 vocab=1.4358 token=1.2712 | lr=3.40e-05 | 6.14 steps/s | VRAM=32.78 Go [2026-07-18 16:51:35] step 893350/1000000 | loss=2.0532 vocab=0.5091 token=0.7902 | lr=3.39e-05 | 6.13 steps/s | VRAM=28.29 Go [2026-07-18 16:51:43] step 893400/1000000 | loss=2.0659 vocab=0.9460 token=1.2623 | lr=3.39e-05 | 6.13 steps/s | VRAM=29.48 Go [2026-07-18 16:51:51] step 893450/1000000 | loss=2.0859 vocab=1.7226 token=1.4630 | lr=3.39e-05 | 6.14 steps/s | VRAM=30.69 Go [2026-07-18 16:51:59] step 893500/1000000 | loss=2.0607 vocab=0.9350 token=1.1603 | lr=3.39e-05 | 6.13 steps/s | VRAM=32.04 Go [2026-07-18 16:52:08] step 893550/1000000 | loss=2.2273 vocab=0.9581 token=1.4896 | lr=3.38e-05 | 6.11 steps/s | VRAM=26.17 Go [2026-07-18 16:52:16] step 893600/1000000 | loss=2.1716 vocab=0.9501 token=1.1590 | lr=3.38e-05 | 6.14 steps/s | VRAM=28.79 Go [2026-07-18 16:52:24] step 893650/1000000 | loss=2.1359 vocab=0.6154 token=1.1002 | lr=3.38e-05 | 6.14 steps/s | VRAM=31.84 Go [2026-07-18 16:52:32] step 893700/1000000 | loss=2.0756 vocab=0.8840 token=1.4000 | lr=3.37e-05 | 6.10 steps/s | VRAM=26.47 Go [2026-07-18 16:52:40] step 893750/1000000 | loss=1.9971 vocab=1.0137 token=1.1741 | lr=3.37e-05 | 6.14 steps/s | VRAM=32.41 Go [2026-07-18 16:52:48] step 893800/1000000 | loss=2.2265 vocab=0.8768 token=1.2628 | lr=3.37e-05 | 6.13 steps/s | VRAM=29.05 Go [2026-07-18 16:52:56] step 893850/1000000 | loss=2.0460 vocab=0.6545 token=1.1620 | lr=3.37e-05 | 6.14 steps/s | VRAM=25.30 Go [2026-07-18 16:53:05] step 893900/1000000 | loss=2.1314 vocab=0.6600 token=0.8212 | lr=3.36e-05 | 6.14 steps/s | VRAM=30.10 Go [2026-07-18 16:53:13] step 893950/1000000 | loss=2.1281 vocab=0.9047 token=1.6600 | lr=3.36e-05 | 6.13 steps/s | VRAM=29.35 Go [2026-07-18 16:53:21] step 894000/1000000 | loss=2.0172 vocab=0.6180 token=1.0986 | lr=3.36e-05 | 6.13 steps/s | VRAM=24.94 Go [2026-07-18 16:53:29] step 894050/1000000 | loss=2.0223 vocab=0.5643 token=1.0460 | lr=3.36e-05 | 6.11 steps/s | VRAM=29.50 Go [2026-07-18 16:53:37] step 894100/1000000 | loss=2.0897 vocab=0.6565 token=1.1037 | lr=3.35e-05 | 6.13 steps/s | VRAM=31.77 Go [2026-07-18 16:53:45] step 894150/1000000 | loss=2.0213 vocab=0.5041 token=1.1177 | lr=3.35e-05 | 6.09 steps/s | VRAM=29.02 Go [2026-07-18 16:53:54] step 894200/1000000 | loss=2.1008 vocab=0.9666 token=1.2529 | lr=3.35e-05 | 6.14 steps/s | VRAM=28.86 Go [2026-07-18 16:54:02] step 894250/1000000 | loss=2.0310 vocab=1.0350 token=1.2925 | lr=3.35e-05 | 6.12 steps/s | VRAM=28.77 Go [2026-07-18 16:54:10] step 894300/1000000 | loss=2.0919 vocab=0.7936 token=1.6176 | lr=3.34e-05 | 6.12 steps/s | VRAM=28.63 Go [2026-07-18 16:54:18] step 894350/1000000 | loss=2.1893 vocab=0.4918 token=1.4622 | lr=3.34e-05 | 6.12 steps/s | VRAM=29.19 Go [2026-07-18 16:54:26] step 894400/1000000 | loss=1.9613 vocab=0.6974 token=1.1512 | lr=3.34e-05 | 6.16 steps/s | VRAM=29.36 Go [2026-07-18 16:54:34] step 894450/1000000 | loss=2.0956 vocab=0.7948 token=1.3048 | lr=3.33e-05 | 6.10 steps/s | VRAM=28.59 Go [2026-07-18 16:54:43] step 894500/1000000 | loss=2.0526 vocab=0.8430 token=1.2143 | lr=3.33e-05 | 6.13 steps/s | VRAM=29.80 Go [2026-07-18 16:54:51] step 894550/1000000 | loss=1.9580 vocab=0.8371 token=0.9309 | lr=3.33e-05 | 6.17 steps/s | VRAM=28.37 Go [2026-07-18 16:54:59] step 894600/1000000 | loss=2.1040 vocab=0.9518 token=1.1808 | lr=3.33e-05 | 6.13 steps/s | VRAM=29.29 Go [2026-07-18 16:55:07] step 894650/1000000 | loss=2.0260 vocab=1.0916 token=1.4420 | lr=3.32e-05 | 6.13 steps/s | VRAM=26.26 Go [2026-07-18 16:55:15] step 894700/1000000 | loss=2.0676 vocab=0.7413 token=0.9427 | lr=3.32e-05 | 6.13 steps/s | VRAM=28.19 Go [2026-07-18 16:55:23] step 894750/1000000 | loss=2.0312 vocab=0.4948 token=1.1656 | lr=3.32e-05 | 6.13 steps/s | VRAM=28.64 Go [2026-07-18 16:55:32] step 894800/1000000 | loss=2.0794 vocab=0.9441 token=0.8006 | lr=3.32e-05 | 6.11 steps/s | VRAM=25.66 Go [2026-07-18 16:55:40] step 894850/1000000 | loss=1.9825 vocab=0.3081 token=0.7096 | lr=3.31e-05 | 6.16 steps/s | VRAM=31.44 Go [2026-07-18 16:55:48] step 894900/1000000 | loss=2.0709 vocab=0.6408 token=0.9932 | lr=3.31e-05 | 6.16 steps/s | VRAM=29.30 Go [2026-07-18 16:55:56] step 894950/1000000 | loss=1.9986 vocab=1.1274 token=0.9255 | lr=3.31e-05 | 6.12 steps/s | VRAM=29.30 Go [2026-07-18 16:56:04] step 895000/1000000 | loss=2.1584 vocab=0.7199 token=1.1175 | lr=3.31e-05 | 6.12 steps/s | VRAM=32.56 Go [2026-07-18 16:56:12] step 895050/1000000 | loss=2.1645 vocab=0.6200 token=1.5424 | lr=3.30e-05 | 6.17 steps/s | VRAM=31.21 Go [2026-07-18 16:56:20] step 895100/1000000 | loss=2.2444 vocab=0.7483 token=1.2505 | lr=3.30e-05 | 6.15 steps/s | VRAM=28.64 Go [2026-07-18 16:56:29] step 895150/1000000 | loss=2.1417 vocab=0.4772 token=1.5135 | lr=3.30e-05 | 6.15 steps/s | VRAM=31.68 Go [2026-07-18 16:56:37] step 895200/1000000 | loss=1.9996 vocab=0.9412 token=1.1656 | lr=3.29e-05 | 6.15 steps/s | VRAM=32.37 Go [2026-07-18 16:56:45] step 895250/1000000 | loss=2.0461 vocab=1.2696 token=1.3702 | lr=3.29e-05 | 6.14 steps/s | VRAM=32.06 Go [2026-07-18 16:56:53] step 895300/1000000 | loss=2.0916 vocab=1.0735 token=1.6482 | lr=3.29e-05 | 6.13 steps/s | VRAM=32.17 Go [2026-07-18 16:57:01] step 895350/1000000 | loss=2.0623 vocab=0.8663 token=1.3011 | lr=3.29e-05 | 6.13 steps/s | VRAM=23.97 Go [2026-07-18 16:57:09] step 895400/1000000 | loss=2.1738 vocab=0.8036 token=1.0764 | lr=3.28e-05 | 6.13 steps/s | VRAM=28.64 Go [2026-07-18 16:57:17] step 895450/1000000 | loss=2.0787 vocab=1.2912 token=1.3257 | lr=3.28e-05 | 6.17 steps/s | VRAM=31.22 Go [2026-07-18 16:57:26] step 895500/1000000 | loss=2.1979 vocab=0.6776 token=0.9649 | lr=3.28e-05 | 6.09 steps/s | VRAM=27.64 Go [2026-07-18 16:57:34] step 895550/1000000 | loss=2.0530 vocab=0.7343 token=1.0479 | lr=3.28e-05 | 6.15 steps/s | VRAM=31.61 Go [2026-07-18 16:57:42] step 895600/1000000 | loss=2.0171 vocab=0.4021 token=1.2832 | lr=3.27e-05 | 6.14 steps/s | VRAM=32.26 Go [2026-07-18 16:57:50] step 895650/1000000 | loss=1.9866 vocab=0.8039 token=1.1897 | lr=3.27e-05 | 6.12 steps/s | VRAM=25.16 Go [2026-07-18 16:57:58] step 895700/1000000 | loss=2.1433 vocab=0.5227 token=1.0221 | lr=3.27e-05 | 6.18 steps/s | VRAM=27.55 Go [2026-07-18 16:58:06] step 895750/1000000 | loss=2.0325 vocab=0.7709 token=1.2088 | lr=3.27e-05 | 6.14 steps/s | VRAM=28.68 Go [2026-07-18 16:58:14] step 895800/1000000 | loss=2.2651 vocab=0.7965 token=1.3329 | lr=3.26e-05 | 6.11 steps/s | VRAM=29.85 Go [2026-07-18 16:58:23] step 895850/1000000 | loss=1.9823 vocab=0.5068 token=1.1049 | lr=3.26e-05 | 6.14 steps/s | VRAM=28.70 Go [2026-07-18 16:58:31] step 895900/1000000 | loss=2.0220 vocab=0.7884 token=1.1333 | lr=3.26e-05 | 6.14 steps/s | VRAM=29.62 Go [2026-07-18 16:58:39] step 895950/1000000 | loss=2.0769 vocab=0.5037 token=1.0749 | lr=3.25e-05 | 6.17 steps/s | VRAM=31.35 Go [2026-07-18 16:58:47] step 896000/1000000 | loss=2.1323 vocab=0.6844 token=1.4633 | lr=3.25e-05 | 6.11 steps/s | VRAM=26.05 Go [2026-07-18 16:58:55] step 896050/1000000 | loss=2.0726 vocab=1.3352 token=1.2487 | lr=3.25e-05 | 6.11 steps/s | VRAM=32.44 Go [2026-07-18 16:59:03] step 896100/1000000 | loss=2.0305 vocab=1.0147 token=1.2690 | lr=3.25e-05 | 6.15 steps/s | VRAM=26.06 Go [2026-07-18 16:59:11] step 896150/1000000 | loss=2.0769 vocab=1.1862 token=1.5993 | lr=3.24e-05 | 6.14 steps/s | VRAM=28.15 Go [2026-07-18 16:59:20] step 896200/1000000 | loss=2.0468 vocab=0.7765 token=0.9792 | lr=3.24e-05 | 6.11 steps/s | VRAM=29.07 Go [2026-07-18 16:59:28] step 896250/1000000 | loss=2.2455 vocab=1.2243 token=1.7483 | lr=3.24e-05 | 6.14 steps/s | VRAM=29.40 Go [2026-07-18 16:59:36] step 896300/1000000 | loss=2.1454 vocab=0.9318 token=1.2119 | lr=3.24e-05 | 6.13 steps/s | VRAM=32.04 Go [2026-07-18 16:59:44] step 896350/1000000 | loss=2.0227 vocab=0.4725 token=0.9196 | lr=3.23e-05 | 6.11 steps/s | VRAM=28.66 Go [2026-07-18 16:59:52] step 896400/1000000 | loss=2.0981 vocab=1.0038 token=1.0518 | lr=3.23e-05 | 6.10 steps/s | VRAM=32.80 Go [2026-07-18 17:00:00] step 896450/1000000 | loss=2.1549 vocab=0.6908 token=1.1396 | lr=3.23e-05 | 6.11 steps/s | VRAM=28.61 Go [2026-07-18 17:00:09] step 896500/1000000 | loss=2.0365 vocab=0.9497 token=1.1043 | lr=3.23e-05 | 6.14 steps/s | VRAM=32.28 Go [2026-07-18 17:00:17] step 896550/1000000 | loss=2.0989 vocab=0.9615 token=1.3719 | lr=3.22e-05 | 6.15 steps/s | VRAM=29.71 Go [2026-07-18 17:00:25] step 896600/1000000 | loss=2.0519 vocab=0.5425 token=1.2946 | lr=3.22e-05 | 6.17 steps/s | VRAM=28.57 Go [2026-07-18 17:00:33] step 896650/1000000 | loss=2.1243 vocab=0.5731 token=0.7150 | lr=3.22e-05 | 6.17 steps/s | VRAM=25.18 Go [2026-07-18 17:00:41] step 896700/1000000 | loss=2.1023 vocab=1.1022 token=0.8929 | lr=3.21e-05 | 6.16 steps/s | VRAM=27.82 Go [2026-07-18 17:00:49] step 896750/1000000 | loss=2.0170 vocab=0.3596 token=1.0908 | lr=3.21e-05 | 6.17 steps/s | VRAM=27.99 Go [2026-07-18 17:00:57] step 896800/1000000 | loss=2.1108 vocab=1.0154 token=1.2604 | lr=3.21e-05 | 6.14 steps/s | VRAM=25.66 Go [2026-07-18 17:01:05] step 896850/1000000 | loss=2.1233 vocab=1.0332 token=1.4258 | lr=3.21e-05 | 6.17 steps/s | VRAM=32.01 Go [2026-07-18 17:01:14] step 896900/1000000 | loss=2.0368 vocab=0.6953 token=0.8971 | lr=3.20e-05 | 6.16 steps/s | VRAM=32.12 Go [2026-07-18 17:01:22] step 896950/1000000 | loss=2.0767 vocab=0.9839 token=0.9750 | lr=3.20e-05 | 6.14 steps/s | VRAM=28.79 Go [2026-07-18 17:01:30] step 897000/1000000 | loss=2.0882 vocab=0.9328 token=1.3810 | lr=3.20e-05 | 6.15 steps/s | VRAM=28.65 Go [2026-07-18 17:01:38] step 897050/1000000 | loss=2.0010 vocab=1.0504 token=1.2599 | lr=3.20e-05 | 6.13 steps/s | VRAM=26.12 Go [2026-07-18 17:01:46] step 897100/1000000 | loss=2.0430 vocab=0.6858 token=1.2066 | lr=3.19e-05 | 6.12 steps/s | VRAM=33.00 Go [2026-07-18 17:01:54] step 897150/1000000 | loss=2.1188 vocab=0.6011 token=1.0899 | lr=3.19e-05 | 6.10 steps/s | VRAM=28.92 Go [2026-07-18 17:02:02] step 897200/1000000 | loss=2.1059 vocab=1.0424 token=1.3450 | lr=3.19e-05 | 6.16 steps/s | VRAM=28.51 Go [2026-07-18 17:02:11] step 897250/1000000 | loss=2.1102 vocab=0.8871 token=1.3062 | lr=3.19e-05 | 6.16 steps/s | VRAM=31.77 Go [2026-07-18 17:02:19] step 897300/1000000 | loss=2.0708 vocab=0.4356 token=1.2030 | lr=3.18e-05 | 6.17 steps/s | VRAM=31.54 Go [2026-07-18 17:02:27] step 897350/1000000 | loss=1.9089 vocab=0.7260 token=0.8423 | lr=3.18e-05 | 6.14 steps/s | VRAM=28.38 Go [2026-07-18 17:02:35] step 897400/1000000 | loss=2.1473 vocab=1.1364 token=1.4750 | lr=3.18e-05 | 6.15 steps/s | VRAM=24.78 Go [2026-07-18 17:02:43] step 897450/1000000 | loss=2.1408 vocab=1.0453 token=1.2745 | lr=3.18e-05 | 6.14 steps/s | VRAM=30.84 Go [2026-07-18 17:02:51] step 897500/1000000 | loss=2.1233 vocab=0.4049 token=0.7977 | lr=3.17e-05 | 6.10 steps/s | VRAM=26.34 Go [2026-07-18 17:02:59] step 897550/1000000 | loss=1.9744 vocab=0.5812 token=1.2873 | lr=3.17e-05 | 6.12 steps/s | VRAM=26.20 Go [2026-07-18 17:03:08] step 897600/1000000 | loss=2.0875 vocab=1.1621 token=1.1803 | lr=3.17e-05 | 6.15 steps/s | VRAM=32.35 Go [2026-07-18 17:03:16] step 897650/1000000 | loss=2.0186 vocab=0.8737 token=1.1798 | lr=3.16e-05 | 6.12 steps/s | VRAM=26.43 Go [2026-07-18 17:03:24] step 897700/1000000 | loss=2.1424 vocab=0.7954 token=1.5106 | lr=3.16e-05 | 6.14 steps/s | VRAM=28.44 Go [2026-07-18 17:03:32] step 897750/1000000 | loss=1.8908 vocab=0.5414 token=1.1379 | lr=3.16e-05 | 6.16 steps/s | VRAM=28.47 Go [2026-07-18 17:03:40] step 897800/1000000 | loss=2.0370 vocab=0.7208 token=1.1902 | lr=3.16e-05 | 6.18 steps/s | VRAM=32.45 Go [2026-07-18 17:03:48] step 897850/1000000 | loss=2.1257 vocab=0.4968 token=1.2501 | lr=3.15e-05 | 6.15 steps/s | VRAM=32.15 Go [2026-07-18 17:03:56] step 897900/1000000 | loss=2.0734 vocab=0.8656 token=1.4165 | lr=3.15e-05 | 6.13 steps/s | VRAM=32.61 Go [2026-07-18 17:04:05] step 897950/1000000 | loss=2.0376 vocab=0.6605 token=1.1494 | lr=3.15e-05 | 6.12 steps/s | VRAM=29.05 Go [2026-07-18 17:04:13] step 898000/1000000 | loss=2.0629 vocab=0.7790 token=1.5106 | lr=3.15e-05 | 6.12 steps/s | VRAM=29.82 Go [2026-07-18 17:04:21] step 898050/1000000 | loss=1.9191 vocab=0.4896 token=1.1855 | lr=3.14e-05 | 6.15 steps/s | VRAM=31.49 Go [2026-07-18 17:04:29] step 898100/1000000 | loss=2.1145 vocab=0.7894 token=1.0677 | lr=3.14e-05 | 6.16 steps/s | VRAM=32.37 Go [2026-07-18 17:04:37] step 898150/1000000 | loss=2.0342 vocab=0.4324 token=0.8248 | lr=3.14e-05 | 6.16 steps/s | VRAM=28.92 Go [2026-07-18 17:04:45] step 898200/1000000 | loss=2.1166 vocab=0.7120 token=0.8276 | lr=3.14e-05 | 6.15 steps/s | VRAM=31.21 Go [2026-07-18 17:04:53] step 898250/1000000 | loss=2.0489 vocab=0.9824 token=1.1993 | lr=3.13e-05 | 6.12 steps/s | VRAM=31.35 Go [2026-07-18 17:05:02] step 898300/1000000 | loss=2.1843 vocab=0.5280 token=1.0067 | lr=3.13e-05 | 6.13 steps/s | VRAM=32.83 Go [2026-07-18 17:05:10] step 898350/1000000 | loss=1.9919 vocab=1.2088 token=1.4483 | lr=3.13e-05 | 6.10 steps/s | VRAM=26.20 Go [2026-07-18 17:05:18] step 898400/1000000 | loss=2.0373 vocab=0.4309 token=1.1200 | lr=3.13e-05 | 6.12 steps/s | VRAM=25.96 Go [2026-07-18 17:05:26] step 898450/1000000 | loss=2.0932 vocab=1.2875 token=1.4154 | lr=3.12e-05 | 6.12 steps/s | VRAM=27.91 Go [2026-07-18 17:05:34] step 898500/1000000 | loss=2.0526 vocab=0.4828 token=0.8646 | lr=3.12e-05 | 6.15 steps/s | VRAM=27.54 Go [2026-07-18 17:05:42] step 898550/1000000 | loss=2.0870 vocab=0.8716 token=1.0795 | lr=3.12e-05 | 6.13 steps/s | VRAM=28.39 Go [2026-07-18 17:05:51] step 898600/1000000 | loss=2.0382 vocab=0.9129 token=1.1989 | lr=3.11e-05 | 6.12 steps/s | VRAM=28.07 Go [2026-07-18 17:05:59] step 898650/1000000 | loss=2.0022 vocab=1.4586 token=1.2808 | lr=3.11e-05 | 6.11 steps/s | VRAM=28.86 Go [2026-07-18 17:06:07] step 898700/1000000 | loss=2.0760 vocab=1.1244 token=1.5075 | lr=3.11e-05 | 6.12 steps/s | VRAM=26.28 Go [2026-07-18 17:06:15] step 898750/1000000 | loss=2.0588 vocab=1.1571 token=1.0953 | lr=3.11e-05 | 6.14 steps/s | VRAM=24.19 Go [2026-07-18 17:06:23] step 898800/1000000 | loss=2.0274 vocab=0.7415 token=1.1450 | lr=3.10e-05 | 6.11 steps/s | VRAM=28.76 Go [2026-07-18 17:06:31] step 898850/1000000 | loss=2.1025 vocab=0.9768 token=1.0890 | lr=3.10e-05 | 6.17 steps/s | VRAM=31.37 Go [2026-07-18 17:06:40] step 898900/1000000 | loss=1.9990 vocab=0.8705 token=1.5978 | lr=3.10e-05 | 6.13 steps/s | VRAM=29.02 Go [2026-07-18 17:06:48] step 898950/1000000 | loss=2.1881 vocab=0.6927 token=1.1147 | lr=3.10e-05 | 6.14 steps/s | VRAM=31.34 Go [2026-07-18 17:06:56] step 899000/1000000 | loss=1.9610 vocab=0.8772 token=0.9330 | lr=3.09e-05 | 6.14 steps/s | VRAM=31.53 Go [2026-07-18 17:07:04] step 899050/1000000 | loss=2.1124 vocab=1.0413 token=1.1848 | lr=3.09e-05 | 6.15 steps/s | VRAM=31.80 Go [2026-07-18 17:07:12] step 899100/1000000 | loss=2.1128 vocab=1.5103 token=1.1802 | lr=3.09e-05 | 6.13 steps/s | VRAM=32.40 Go [2026-07-18 17:07:20] step 899150/1000000 | loss=2.0665 vocab=1.1044 token=0.7879 | lr=3.09e-05 | 6.13 steps/s | VRAM=29.04 Go [2026-07-18 17:07:28] step 899200/1000000 | loss=2.0921 vocab=0.8003 token=0.9663 | lr=3.08e-05 | 6.11 steps/s | VRAM=31.94 Go [2026-07-18 17:07:37] step 899250/1000000 | loss=2.0684 vocab=0.8310 token=1.1763 | lr=3.08e-05 | 6.09 steps/s | VRAM=26.53 Go [2026-07-18 17:07:45] step 899300/1000000 | loss=2.0442 vocab=0.9729 token=1.1383 | lr=3.08e-05 | 6.09 steps/s | VRAM=29.96 Go [2026-07-18 17:07:53] step 899350/1000000 | loss=2.0789 vocab=1.0221 token=1.3484 | lr=3.08e-05 | 6.17 steps/s | VRAM=31.43 Go [2026-07-18 17:08:01] step 899400/1000000 | loss=2.0208 vocab=0.8650 token=1.3237 | lr=3.07e-05 | 6.12 steps/s | VRAM=28.06 Go [2026-07-18 17:08:09] step 899450/1000000 | loss=2.1226 vocab=0.4488 token=0.7538 | lr=3.07e-05 | 6.14 steps/s | VRAM=28.08 Go [2026-07-18 17:08:17] step 899500/1000000 | loss=2.0051 vocab=0.9452 token=1.3853 | lr=3.07e-05 | 6.12 steps/s | VRAM=28.55 Go [2026-07-18 17:08:26] step 899550/1000000 | loss=2.0148 vocab=0.7971 token=1.2741 | lr=3.06e-05 | 6.12 steps/s | VRAM=26.47 Go [2026-07-18 17:08:34] step 899600/1000000 | loss=2.1041 vocab=0.7935 token=1.0914 | lr=3.06e-05 | 6.14 steps/s | VRAM=28.49 Go [2026-07-18 17:08:42] step 899650/1000000 | loss=2.0810 vocab=1.0772 token=1.6590 | lr=3.06e-05 | 6.16 steps/s | VRAM=28.47 Go [2026-07-18 17:08:50] step 899700/1000000 | loss=2.0987 vocab=1.2523 token=1.4965 | lr=3.06e-05 | 6.13 steps/s | VRAM=26.10 Go [2026-07-18 17:08:58] step 899750/1000000 | loss=1.9439 vocab=0.6831 token=1.2260 | lr=3.05e-05 | 6.14 steps/s | VRAM=31.56 Go [2026-07-18 17:09:06] step 899800/1000000 | loss=2.0339 vocab=1.3193 token=1.2051 | lr=3.05e-05 | 6.16 steps/s | VRAM=28.93 Go [2026-07-18 17:09:14] step 899850/1000000 | loss=2.0168 vocab=1.1165 token=1.3203 | lr=3.05e-05 | 6.15 steps/s | VRAM=25.96 Go [2026-07-18 17:09:23] step 899900/1000000 | loss=2.0150 vocab=1.4523 token=1.8890 | lr=3.05e-05 | 6.15 steps/s | VRAM=28.41 Go [2026-07-18 17:09:31] step 899950/1000000 | loss=2.1830 vocab=0.9729 token=1.3868 | lr=3.04e-05 | 6.11 steps/s | VRAM=28.39 Go [2026-07-18 17:09:39] step 900000/1000000 | loss=2.1201 vocab=0.5536 token=1.2997 | lr=3.04e-05 | 6.12 steps/s | VRAM=29.09 Go [accuracy] step 900000 : 63.92% (log: /workspace/plbert_train/precision_log.csv) [push] step 900000 OK (loss_token=1.2997, accuracy=63.92%) [2026-07-18 17:11:10] step 900050/1000000 | loss=2.1361 vocab=1.0894 token=1.5372 | lr=3.04e-05 | 0.55 steps/s | VRAM=31.20 Go [2026-07-18 17:11:18] step 900100/1000000 | loss=2.1023 vocab=0.8128 token=1.1209 | lr=3.04e-05 | 6.20 steps/s | VRAM=28.12 Go [2026-07-18 17:11:26] step 900150/1000000 | loss=2.1988 vocab=1.0636 token=1.9115 | lr=3.03e-05 | 6.24 steps/s | VRAM=24.74 Go [2026-07-18 17:11:34] step 900200/1000000 | loss=2.1342 vocab=0.8228 token=1.0346 | lr=3.03e-05 | 6.17 steps/s | VRAM=28.00 Go [2026-07-18 17:11:42] step 900250/1000000 | loss=2.1300 vocab=0.8236 token=1.0657 | lr=3.03e-05 | 6.18 steps/s | VRAM=26.38 Go [2026-07-18 17:11:50] step 900300/1000000 | loss=2.0572 vocab=1.0169 token=1.1564 | lr=3.03e-05 | 6.16 steps/s | VRAM=31.34 Go [2026-07-18 17:11:58] step 900350/1000000 | loss=2.0926 vocab=0.5367 token=0.8140 | lr=3.02e-05 | 6.15 steps/s | VRAM=31.16 Go [2026-07-18 17:12:06] step 900400/1000000 | loss=2.1305 vocab=1.2101 token=1.0486 | lr=3.02e-05 | 6.11 steps/s | VRAM=29.19 Go [2026-07-18 17:12:15] step 900450/1000000 | loss=2.0054 vocab=0.9335 token=1.1090 | lr=3.02e-05 | 6.16 steps/s | VRAM=29.14 Go [2026-07-18 17:12:23] step 900500/1000000 | loss=2.0533 vocab=0.9808 token=1.3703 | lr=3.02e-05 | 6.14 steps/s | VRAM=27.38 Go [2026-07-18 17:12:31] step 900550/1000000 | loss=2.0320 vocab=1.0947 token=1.4904 | lr=3.01e-05 | 6.10 steps/s | VRAM=26.54 Go [2026-07-18 17:12:39] step 900600/1000000 | loss=1.9212 vocab=0.8180 token=1.2537 | lr=3.01e-05 | 6.14 steps/s | VRAM=28.77 Go [2026-07-18 17:12:47] step 900650/1000000 | loss=2.0035 vocab=0.8939 token=1.1023 | lr=3.01e-05 | 6.14 steps/s | VRAM=28.88 Go [2026-07-18 17:12:55] step 900700/1000000 | loss=2.0160 vocab=0.9569 token=1.5011 | lr=3.00e-05 | 6.14 steps/s | VRAM=29.15 Go [2026-07-18 17:13:03] step 900750/1000000 | loss=2.1193 vocab=0.7884 token=1.0990 | lr=3.00e-05 | 6.14 steps/s | VRAM=33.02 Go [2026-07-18 17:13:12] step 900800/1000000 | loss=2.1178 vocab=0.7528 token=1.1769 | lr=3.00e-05 | 6.10 steps/s | VRAM=28.65 Go [2026-07-18 17:13:20] step 900850/1000000 | loss=2.0556 vocab=0.7720 token=1.2388 | lr=3.00e-05 | 6.14 steps/s | VRAM=29.52 Go [2026-07-18 17:13:28] step 900900/1000000 | loss=1.9394 vocab=0.4976 token=0.8760 | lr=2.99e-05 | 6.10 steps/s | VRAM=28.42 Go [2026-07-18 17:13:36] step 900950/1000000 | loss=2.0451 vocab=1.0548 token=1.2656 | lr=2.99e-05 | 6.12 steps/s | VRAM=26.08 Go [2026-07-18 17:13:44] step 901000/1000000 | loss=2.0261 vocab=0.8542 token=1.1674 | lr=2.99e-05 | 6.11 steps/s | VRAM=26.55 Go [2026-07-18 17:13:52] step 901050/1000000 | loss=2.1405 vocab=0.9135 token=1.3052 | lr=2.99e-05 | 6.10 steps/s | VRAM=31.33 Go [2026-07-18 17:14:01] step 901100/1000000 | loss=2.0788 vocab=0.7312 token=1.4668 | lr=2.98e-05 | 6.09 steps/s | VRAM=29.08 Go [2026-07-18 17:14:09] step 901150/1000000 | loss=2.1550 vocab=1.0593 token=1.2407 | lr=2.98e-05 | 6.12 steps/s | VRAM=32.48 Go [2026-07-18 17:14:17] step 901200/1000000 | loss=2.0974 vocab=0.7834 token=1.2966 | lr=2.98e-05 | 6.10 steps/s | VRAM=29.44 Go [2026-07-18 17:14:25] step 901250/1000000 | loss=2.1492 vocab=1.7710 token=1.2518 | lr=2.98e-05 | 6.11 steps/s | VRAM=30.60 Go [2026-07-18 17:14:33] step 901300/1000000 | loss=1.9811 vocab=0.6288 token=0.8258 | lr=2.97e-05 | 6.11 steps/s | VRAM=27.10 Go [2026-07-18 17:14:42] step 901350/1000000 | loss=2.0800 vocab=1.2677 token=1.2993 | lr=2.97e-05 | 6.12 steps/s | VRAM=28.74 Go [2026-07-18 17:14:50] step 901400/1000000 | loss=2.0891 vocab=0.6174 token=1.1387 | lr=2.97e-05 | 6.11 steps/s | VRAM=26.81 Go [2026-07-18 17:14:58] step 901450/1000000 | loss=1.8911 vocab=0.5710 token=1.0411 | lr=2.97e-05 | 6.14 steps/s | VRAM=29.38 Go [2026-07-18 17:15:06] step 901500/1000000 | loss=1.9798 vocab=0.4759 token=0.6023 | lr=2.96e-05 | 6.14 steps/s | VRAM=32.73 Go [2026-07-18 17:15:14] step 901550/1000000 | loss=2.1002 vocab=1.2076 token=1.3452 | lr=2.96e-05 | 6.10 steps/s | VRAM=27.74 Go [2026-07-18 17:15:22] step 901600/1000000 | loss=2.1180 vocab=0.6402 token=1.1392 | lr=2.96e-05 | 6.13 steps/s | VRAM=31.99 Go [2026-07-18 17:15:31] step 901650/1000000 | loss=1.9664 vocab=0.8505 token=1.0485 | lr=2.96e-05 | 6.11 steps/s | VRAM=28.81 Go [2026-07-18 17:15:39] step 901700/1000000 | loss=2.0849 vocab=0.8333 token=1.0878 | lr=2.95e-05 | 6.10 steps/s | VRAM=29.78 Go [2026-07-18 17:15:47] step 901750/1000000 | loss=2.1881 vocab=0.6573 token=1.0637 | lr=2.95e-05 | 6.11 steps/s | VRAM=32.75 Go [2026-07-18 17:15:55] step 901800/1000000 | loss=2.1024 vocab=1.0875 token=1.0626 | lr=2.95e-05 | 6.12 steps/s | VRAM=30.88 Go [2026-07-18 17:16:03] step 901850/1000000 | loss=1.9758 vocab=0.9411 token=0.9746 | lr=2.95e-05 | 6.11 steps/s | VRAM=32.67 Go [2026-07-18 17:16:12] step 901900/1000000 | loss=2.0276 vocab=0.7360 token=1.6127 | lr=2.94e-05 | 6.10 steps/s | VRAM=28.80 Go [2026-07-18 17:16:20] step 901950/1000000 | loss=1.8910 vocab=0.6138 token=1.1141 | lr=2.94e-05 | 6.08 steps/s | VRAM=28.87 Go [2026-07-18 17:16:28] step 902000/1000000 | loss=2.1169 vocab=0.8395 token=1.2639 | lr=2.94e-05 | 6.14 steps/s | VRAM=25.78 Go [2026-07-18 17:16:36] step 902050/1000000 | loss=2.0932 vocab=1.3738 token=1.3499 | lr=2.93e-05 | 6.13 steps/s | VRAM=32.43 Go [2026-07-18 17:16:44] step 902100/1000000 | loss=2.1239 vocab=1.2939 token=1.2103 | lr=2.93e-05 | 6.10 steps/s | VRAM=28.30 Go [2026-07-18 17:16:52] step 902150/1000000 | loss=2.0690 vocab=1.4494 token=1.5529 | lr=2.93e-05 | 6.12 steps/s | VRAM=29.26 Go [2026-07-18 17:17:01] step 902200/1000000 | loss=2.0208 vocab=0.8021 token=1.2553 | lr=2.93e-05 | 6.14 steps/s | VRAM=28.00 Go [2026-07-18 17:17:09] step 902250/1000000 | loss=2.0191 vocab=0.9147 token=1.2399 | lr=2.92e-05 | 6.12 steps/s | VRAM=31.65 Go [2026-07-18 17:17:17] step 902300/1000000 | loss=2.0637 vocab=0.3879 token=1.2380 | lr=2.92e-05 | 6.11 steps/s | VRAM=28.63 Go [2026-07-18 17:17:25] step 902350/1000000 | loss=2.0791 vocab=0.7225 token=1.1395 | lr=2.92e-05 | 6.16 steps/s | VRAM=31.52 Go [2026-07-18 17:17:33] step 902400/1000000 | loss=2.0068 vocab=0.8575 token=1.0649 | lr=2.92e-05 | 6.12 steps/s | VRAM=25.50 Go [2026-07-18 17:17:41] step 902450/1000000 | loss=2.0699 vocab=1.5591 token=1.4258 | lr=2.91e-05 | 6.13 steps/s | VRAM=32.13 Go [2026-07-18 17:17:50] step 902500/1000000 | loss=2.0387 vocab=0.5047 token=1.1031 | lr=2.91e-05 | 6.13 steps/s | VRAM=26.44 Go [2026-07-18 17:17:58] step 902550/1000000 | loss=2.0121 vocab=0.8095 token=0.8639 | lr=2.91e-05 | 6.16 steps/s | VRAM=27.71 Go [2026-07-18 17:18:06] step 902600/1000000 | loss=1.8836 vocab=0.8945 token=1.3064 | lr=2.91e-05 | 6.12 steps/s | VRAM=26.42 Go [2026-07-18 17:18:14] step 902650/1000000 | loss=2.0503 vocab=0.7766 token=1.4224 | lr=2.90e-05 | 6.15 steps/s | VRAM=29.26 Go [2026-07-18 17:18:22] step 902700/1000000 | loss=2.0487 vocab=1.1309 token=1.2697 | lr=2.90e-05 | 6.14 steps/s | VRAM=28.85 Go [2026-07-18 17:18:30] step 902750/1000000 | loss=2.0935 vocab=0.7185 token=0.8948 | lr=2.90e-05 | 6.15 steps/s | VRAM=28.07 Go [2026-07-18 17:18:38] step 902800/1000000 | loss=2.0748 vocab=0.9860 token=0.8415 | lr=2.90e-05 | 6.13 steps/s | VRAM=32.19 Go [2026-07-18 17:18:47] step 902850/1000000 | loss=2.0034 vocab=1.1967 token=1.1422 | lr=2.89e-05 | 6.11 steps/s | VRAM=32.10 Go [2026-07-18 17:18:55] step 902900/1000000 | loss=1.9684 vocab=0.8489 token=1.4690 | lr=2.89e-05 | 6.14 steps/s | VRAM=31.04 Go [2026-07-18 17:19:03] step 902950/1000000 | loss=2.0095 vocab=0.9157 token=1.1732 | lr=2.89e-05 | 6.12 steps/s | VRAM=27.01 Go [2026-07-18 17:19:11] step 903000/1000000 | loss=2.0120 vocab=0.7220 token=0.9996 | lr=2.89e-05 | 6.13 steps/s | VRAM=26.26 Go [2026-07-18 17:19:19] step 903050/1000000 | loss=2.1419 vocab=0.6613 token=1.2718 | lr=2.88e-05 | 6.15 steps/s | VRAM=28.73 Go [2026-07-18 17:19:27] step 903100/1000000 | loss=2.0222 vocab=0.5476 token=1.3338 | lr=2.88e-05 | 6.11 steps/s | VRAM=24.31 Go [2026-07-18 17:19:35] step 903150/1000000 | loss=1.9850 vocab=0.7742 token=0.8586 | lr=2.88e-05 | 6.15 steps/s | VRAM=27.78 Go [2026-07-18 17:19:44] step 903200/1000000 | loss=2.0575 vocab=1.0445 token=1.1652 | lr=2.88e-05 | 6.12 steps/s | VRAM=32.48 Go [2026-07-18 17:19:52] step 903250/1000000 | loss=2.1754 vocab=0.8129 token=0.9067 | lr=2.87e-05 | 6.13 steps/s | VRAM=28.26 Go [2026-07-18 17:20:00] step 903300/1000000 | loss=2.0069 vocab=0.9346 token=0.8679 | lr=2.87e-05 | 6.14 steps/s | VRAM=28.11 Go [2026-07-18 17:20:08] step 903350/1000000 | loss=2.0023 vocab=0.4997 token=1.2967 | lr=2.87e-05 | 6.13 steps/s | VRAM=30.17 Go [2026-07-18 17:20:16] step 903400/1000000 | loss=1.9126 vocab=0.6941 token=1.3498 | lr=2.87e-05 | 6.13 steps/s | VRAM=31.63 Go [2026-07-18 17:20:24] step 903450/1000000 | loss=2.0909 vocab=0.8017 token=1.0122 | lr=2.86e-05 | 6.16 steps/s | VRAM=31.88 Go [2026-07-18 17:20:33] step 903500/1000000 | loss=2.0484 vocab=1.2278 token=0.9857 | lr=2.86e-05 | 6.11 steps/s | VRAM=29.31 Go [2026-07-18 17:20:41] step 903550/1000000 | loss=2.0437 vocab=0.8078 token=1.3718 | lr=2.86e-05 | 6.13 steps/s | VRAM=29.09 Go [2026-07-18 17:20:49] step 903600/1000000 | loss=2.0119 vocab=1.0742 token=1.4761 | lr=2.86e-05 | 6.11 steps/s | VRAM=28.84 Go [2026-07-18 17:20:57] step 903650/1000000 | loss=2.0223 vocab=0.9529 token=1.2568 | lr=2.85e-05 | 6.15 steps/s | VRAM=25.82 Go [2026-07-18 17:21:05] step 903700/1000000 | loss=2.0491 vocab=0.9513 token=1.6001 | lr=2.85e-05 | 6.15 steps/s | VRAM=28.46 Go [2026-07-18 17:21:13] step 903750/1000000 | loss=2.1399 vocab=0.9104 token=1.1323 | lr=2.85e-05 | 6.16 steps/s | VRAM=32.36 Go [2026-07-18 17:21:21] step 903800/1000000 | loss=2.1017 vocab=1.3363 token=1.2377 | lr=2.85e-05 | 6.13 steps/s | VRAM=25.04 Go [2026-07-18 17:21:30] step 903850/1000000 | loss=2.0610 vocab=0.8045 token=1.0124 | lr=2.84e-05 | 6.10 steps/s | VRAM=28.93 Go [2026-07-18 17:21:38] step 903900/1000000 | loss=2.0415 vocab=1.1952 token=1.4401 | lr=2.84e-05 | 6.12 steps/s | VRAM=28.82 Go [2026-07-18 17:21:46] step 903950/1000000 | loss=1.9983 vocab=0.7952 token=1.0752 | lr=2.84e-05 | 6.13 steps/s | VRAM=28.93 Go [2026-07-18 17:21:54] step 904000/1000000 | loss=2.0434 vocab=0.7867 token=1.2280 | lr=2.83e-05 | 6.12 steps/s | VRAM=28.35 Go [2026-07-18 17:22:02] step 904050/1000000 | loss=2.0124 vocab=0.8166 token=1.5307 | lr=2.83e-05 | 6.14 steps/s | VRAM=31.86 Go [2026-07-18 17:22:10] step 904100/1000000 | loss=2.1262 vocab=0.8899 token=1.3501 | lr=2.83e-05 | 6.11 steps/s | VRAM=25.60 Go [2026-07-18 17:22:19] step 904150/1000000 | loss=2.1291 vocab=0.7328 token=1.1277 | lr=2.83e-05 | 6.14 steps/s | VRAM=28.16 Go [2026-07-18 17:22:27] step 904200/1000000 | loss=1.9945 vocab=0.7476 token=1.2443 | lr=2.82e-05 | 6.10 steps/s | VRAM=29.88 Go [2026-07-18 17:22:35] step 904250/1000000 | loss=2.0953 vocab=1.0702 token=1.4706 | lr=2.82e-05 | 6.12 steps/s | VRAM=29.07 Go [2026-07-18 17:22:43] step 904300/1000000 | loss=2.1067 vocab=1.0748 token=1.4547 | lr=2.82e-05 | 6.11 steps/s | VRAM=29.62 Go [2026-07-18 17:22:51] step 904350/1000000 | loss=1.9384 vocab=0.4273 token=0.8683 | lr=2.82e-05 | 6.15 steps/s | VRAM=31.95 Go [2026-07-18 17:22:59] step 904400/1000000 | loss=2.1025 vocab=0.9996 token=1.2280 | lr=2.81e-05 | 6.13 steps/s | VRAM=30.91 Go [2026-07-18 17:23:08] step 904450/1000000 | loss=2.1369 vocab=0.6953 token=1.4053 | lr=2.81e-05 | 6.13 steps/s | VRAM=26.02 Go [2026-07-18 17:23:16] step 904500/1000000 | loss=2.0585 vocab=0.9156 token=1.3887 | lr=2.81e-05 | 6.12 steps/s | VRAM=30.19 Go [2026-07-18 17:23:24] step 904550/1000000 | loss=2.1306 vocab=0.7994 token=1.2318 | lr=2.81e-05 | 6.13 steps/s | VRAM=28.60 Go [2026-07-18 17:23:32] step 904600/1000000 | loss=2.0308 vocab=1.2081 token=1.0430 | lr=2.80e-05 | 6.16 steps/s | VRAM=25.88 Go [2026-07-18 17:23:40] step 904650/1000000 | loss=2.0086 vocab=0.9527 token=1.1706 | lr=2.80e-05 | 6.14 steps/s | VRAM=26.40 Go [2026-07-18 17:23:48] step 904700/1000000 | loss=2.0380 vocab=1.0783 token=1.2480 | lr=2.80e-05 | 6.13 steps/s | VRAM=32.58 Go [2026-07-18 17:23:56] step 904750/1000000 | loss=2.2044 vocab=0.9763 token=1.2061 | lr=2.80e-05 | 6.15 steps/s | VRAM=29.36 Go [2026-07-18 17:24:05] step 904800/1000000 | loss=2.0879 vocab=1.2888 token=1.5399 | lr=2.79e-05 | 6.13 steps/s | VRAM=26.10 Go [2026-07-18 17:24:13] step 904850/1000000 | loss=2.1102 vocab=0.5507 token=1.0849 | lr=2.79e-05 | 6.15 steps/s | VRAM=31.45 Go [2026-07-18 17:24:21] step 904900/1000000 | loss=2.0801 vocab=0.5864 token=0.8908 | lr=2.79e-05 | 6.17 steps/s | VRAM=32.43 Go [2026-07-18 17:24:29] step 904950/1000000 | loss=2.0683 vocab=1.2885 token=1.3700 | lr=2.79e-05 | 6.13 steps/s | VRAM=29.58 Go [2026-07-18 17:24:37] step 905000/1000000 | loss=2.0371 vocab=0.6412 token=1.3954 | lr=2.78e-05 | 6.15 steps/s | VRAM=32.86 Go [2026-07-18 17:24:45] step 905050/1000000 | loss=2.0720 vocab=0.7612 token=1.3868 | lr=2.78e-05 | 6.13 steps/s | VRAM=32.36 Go [2026-07-18 17:24:54] step 905100/1000000 | loss=2.1785 vocab=0.7023 token=1.2994 | lr=2.78e-05 | 6.11 steps/s | VRAM=31.59 Go [2026-07-18 17:25:02] step 905150/1000000 | loss=1.9960 vocab=0.9905 token=1.6267 | lr=2.78e-05 | 6.15 steps/s | VRAM=32.50 Go [2026-07-18 17:25:10] step 905200/1000000 | loss=2.1834 vocab=0.9937 token=1.3725 | lr=2.77e-05 | 6.11 steps/s | VRAM=25.46 Go [2026-07-18 17:25:18] step 905250/1000000 | loss=2.0973 vocab=1.2555 token=1.3793 | lr=2.77e-05 | 6.12 steps/s | VRAM=29.04 Go [2026-07-18 17:25:26] step 905300/1000000 | loss=2.0329 vocab=0.6084 token=1.1558 | lr=2.77e-05 | 6.13 steps/s | VRAM=28.18 Go [2026-07-18 17:25:34] step 905350/1000000 | loss=2.0997 vocab=0.6914 token=0.9735 | lr=2.77e-05 | 6.15 steps/s | VRAM=28.51 Go [2026-07-18 17:25:42] step 905400/1000000 | loss=1.9954 vocab=0.4461 token=1.2346 | lr=2.76e-05 | 6.14 steps/s | VRAM=31.49 Go [2026-07-18 17:25:51] step 905450/1000000 | loss=2.0974 vocab=0.8430 token=1.1071 | lr=2.76e-05 | 6.15 steps/s | VRAM=30.99 Go [2026-07-18 17:25:59] step 905500/1000000 | loss=2.0286 vocab=0.6750 token=1.2083 | lr=2.76e-05 | 6.14 steps/s | VRAM=32.35 Go [2026-07-18 17:26:07] step 905550/1000000 | loss=2.1341 vocab=1.3706 token=1.0996 | lr=2.76e-05 | 6.12 steps/s | VRAM=28.51 Go [2026-07-18 17:26:15] step 905600/1000000 | loss=2.0908 vocab=0.9134 token=1.5241 | lr=2.75e-05 | 6.10 steps/s | VRAM=28.37 Go [2026-07-18 17:26:23] step 905650/1000000 | loss=2.0491 vocab=0.3034 token=0.8243 | lr=2.75e-05 | 6.07 steps/s | VRAM=30.11 Go [2026-07-18 17:26:31] step 905700/1000000 | loss=2.1217 vocab=1.1611 token=1.1059 | lr=2.75e-05 | 6.11 steps/s | VRAM=28.85 Go [2026-07-18 17:26:40] step 905750/1000000 | loss=2.0618 vocab=0.7256 token=1.1773 | lr=2.75e-05 | 6.15 steps/s | VRAM=28.37 Go [2026-07-18 17:26:48] step 905800/1000000 | loss=2.1565 vocab=1.3494 token=1.2141 | lr=2.74e-05 | 6.12 steps/s | VRAM=28.44 Go [2026-07-18 17:26:56] step 905850/1000000 | loss=2.1188 vocab=1.0972 token=1.5516 | lr=2.74e-05 | 6.13 steps/s | VRAM=29.40 Go [2026-07-18 17:27:04] step 905900/1000000 | loss=2.0429 vocab=0.5583 token=1.0834 | lr=2.74e-05 | 6.14 steps/s | VRAM=24.96 Go [2026-07-18 17:27:12] step 905950/1000000 | loss=2.1036 vocab=0.5924 token=1.0678 | lr=2.74e-05 | 6.12 steps/s | VRAM=29.71 Go [2026-07-18 17:27:20] step 906000/1000000 | loss=2.0990 vocab=1.2043 token=1.2604 | lr=2.73e-05 | 6.16 steps/s | VRAM=31.33 Go [2026-07-18 17:27:29] step 906050/1000000 | loss=2.0681 vocab=0.7431 token=1.1843 | lr=2.73e-05 | 6.13 steps/s | VRAM=26.74 Go [2026-07-18 17:27:37] step 906100/1000000 | loss=2.0762 vocab=0.3933 token=0.7914 | lr=2.73e-05 | 6.18 steps/s | VRAM=32.17 Go [2026-07-18 17:27:45] step 906150/1000000 | loss=2.0058 vocab=0.7868 token=0.8366 | lr=2.73e-05 | 6.12 steps/s | VRAM=25.51 Go [2026-07-18 17:27:53] step 906200/1000000 | loss=2.0810 vocab=0.7142 token=1.1087 | lr=2.72e-05 | 6.15 steps/s | VRAM=27.91 Go [2026-07-18 17:28:01] step 906250/1000000 | loss=2.0910 vocab=0.7209 token=1.0991 | lr=2.72e-05 | 6.13 steps/s | VRAM=29.28 Go [2026-07-18 17:28:09] step 906300/1000000 | loss=2.1402 vocab=0.6505 token=1.1497 | lr=2.72e-05 | 6.12 steps/s | VRAM=28.52 Go [2026-07-18 17:28:17] step 906350/1000000 | loss=1.9978 vocab=0.4025 token=1.2623 | lr=2.72e-05 | 6.11 steps/s | VRAM=28.78 Go [2026-07-18 17:28:26] step 906400/1000000 | loss=2.0515 vocab=0.3871 token=0.8540 | lr=2.71e-05 | 6.12 steps/s | VRAM=29.31 Go [2026-07-18 17:28:34] step 906450/1000000 | loss=2.1324 vocab=0.9586 token=1.5580 | lr=2.71e-05 | 6.09 steps/s | VRAM=26.40 Go [2026-07-18 17:28:42] step 906500/1000000 | loss=2.0727 vocab=0.3960 token=1.3050 | lr=2.71e-05 | 6.13 steps/s | VRAM=27.93 Go [2026-07-18 17:28:50] step 906550/1000000 | loss=2.0863 vocab=0.7678 token=1.0980 | lr=2.71e-05 | 6.12 steps/s | VRAM=29.13 Go [2026-07-18 17:28:58] step 906600/1000000 | loss=2.1841 vocab=0.8610 token=1.1426 | lr=2.70e-05 | 6.12 steps/s | VRAM=32.65 Go [2026-07-18 17:29:06] step 906650/1000000 | loss=2.0277 vocab=0.4541 token=0.9965 | lr=2.70e-05 | 6.13 steps/s | VRAM=28.58 Go [2026-07-18 17:29:15] step 906700/1000000 | loss=2.0529 vocab=0.8171 token=1.2044 | lr=2.70e-05 | 6.16 steps/s | VRAM=31.80 Go [2026-07-18 17:29:23] step 906750/1000000 | loss=1.9475 vocab=1.2129 token=1.5016 | lr=2.70e-05 | 6.12 steps/s | VRAM=29.10 Go [2026-07-18 17:29:31] step 906800/1000000 | loss=2.0181 vocab=1.0158 token=1.0591 | lr=2.69e-05 | 6.14 steps/s | VRAM=29.51 Go [2026-07-18 17:29:39] step 906850/1000000 | loss=2.0970 vocab=0.9241 token=1.2786 | lr=2.69e-05 | 6.14 steps/s | VRAM=28.81 Go [2026-07-18 17:29:47] step 906900/1000000 | loss=2.0798 vocab=0.4698 token=1.0483 | lr=2.69e-05 | 6.13 steps/s | VRAM=26.31 Go [2026-07-18 17:29:55] step 906950/1000000 | loss=2.1232 vocab=0.5158 token=1.1792 | lr=2.69e-05 | 6.15 steps/s | VRAM=24.59 Go [2026-07-18 17:30:03] step 907000/1000000 | loss=1.9439 vocab=0.7835 token=1.1500 | lr=2.68e-05 | 6.17 steps/s | VRAM=24.84 Go [2026-07-18 17:30:12] step 907050/1000000 | loss=2.0440 vocab=0.4146 token=0.8778 | lr=2.68e-05 | 6.14 steps/s | VRAM=28.93 Go [2026-07-18 17:30:20] step 907100/1000000 | loss=2.0374 vocab=1.0526 token=1.4918 | lr=2.68e-05 | 6.13 steps/s | VRAM=25.76 Go [2026-07-18 17:30:28] step 907150/1000000 | loss=2.0043 vocab=1.7225 token=1.1525 | lr=2.68e-05 | 6.16 steps/s | VRAM=29.71 Go [2026-07-18 17:30:36] step 907200/1000000 | loss=2.1038 vocab=0.8526 token=1.6755 | lr=2.67e-05 | 6.13 steps/s | VRAM=28.50 Go [2026-07-18 17:30:44] step 907250/1000000 | loss=2.0248 vocab=1.1845 token=1.4260 | lr=2.67e-05 | 6.14 steps/s | VRAM=28.31 Go [2026-07-18 17:30:52] step 907300/1000000 | loss=2.2310 vocab=0.9167 token=1.4025 | lr=2.67e-05 | 6.16 steps/s | VRAM=31.46 Go [2026-07-18 17:31:00] step 907350/1000000 | loss=2.0059 vocab=0.9217 token=1.0557 | lr=2.67e-05 | 6.14 steps/s | VRAM=28.88 Go [2026-07-18 17:31:09] step 907400/1000000 | loss=2.0902 vocab=0.9277 token=1.3799 | lr=2.66e-05 | 6.13 steps/s | VRAM=29.01 Go [2026-07-18 17:31:17] step 907450/1000000 | loss=2.1172 vocab=0.7714 token=0.5773 | lr=2.66e-05 | 6.12 steps/s | VRAM=26.49 Go [2026-07-18 17:31:25] step 907500/1000000 | loss=2.1290 vocab=0.7625 token=0.9766 | lr=2.66e-05 | 6.13 steps/s | VRAM=28.98 Go [2026-07-18 17:31:33] step 907550/1000000 | loss=1.9843 vocab=0.9813 token=1.4768 | lr=2.66e-05 | 6.15 steps/s | VRAM=29.23 Go [2026-07-18 17:31:41] step 907600/1000000 | loss=2.1589 vocab=0.8048 token=1.3057 | lr=2.65e-05 | 6.15 steps/s | VRAM=31.44 Go [2026-07-18 17:31:49] step 907650/1000000 | loss=2.0955 vocab=0.8949 token=0.9059 | lr=2.65e-05 | 6.18 steps/s | VRAM=31.36 Go [2026-07-18 17:31:57] step 907700/1000000 | loss=2.0340 vocab=1.1874 token=1.7817 | lr=2.65e-05 | 6.13 steps/s | VRAM=26.13 Go [2026-07-18 17:32:06] step 907750/1000000 | loss=2.0763 vocab=0.7607 token=1.4772 | lr=2.65e-05 | 6.10 steps/s | VRAM=25.24 Go [2026-07-18 17:32:14] step 907800/1000000 | loss=1.9412 vocab=0.8437 token=0.9504 | lr=2.64e-05 | 6.16 steps/s | VRAM=27.87 Go [2026-07-18 17:32:22] step 907850/1000000 | loss=2.0280 vocab=1.2940 token=1.4862 | lr=2.64e-05 | 6.11 steps/s | VRAM=28.49 Go [2026-07-18 17:32:30] step 907900/1000000 | loss=2.1247 vocab=0.4481 token=1.0541 | lr=2.64e-05 | 6.16 steps/s | VRAM=25.14 Go [2026-07-18 17:32:38] step 907950/1000000 | loss=2.1504 vocab=1.1273 token=1.5359 | lr=2.64e-05 | 6.15 steps/s | VRAM=29.33 Go [2026-07-18 17:32:46] step 908000/1000000 | loss=2.2034 vocab=1.5778 token=1.6947 | lr=2.63e-05 | 6.14 steps/s | VRAM=31.70 Go [2026-07-18 17:32:54] step 908050/1000000 | loss=1.9917 vocab=1.0310 token=0.8604 | lr=2.63e-05 | 6.13 steps/s | VRAM=28.45 Go [2026-07-18 17:33:03] step 908100/1000000 | loss=2.0841 vocab=1.0247 token=1.5398 | lr=2.63e-05 | 6.10 steps/s | VRAM=26.05 Go [2026-07-18 17:33:11] step 908150/1000000 | loss=2.0821 vocab=0.4998 token=0.9464 | lr=2.63e-05 | 6.15 steps/s | VRAM=32.24 Go [2026-07-18 17:33:19] step 908200/1000000 | loss=2.0648 vocab=0.5906 token=0.7708 | lr=2.62e-05 | 6.16 steps/s | VRAM=32.69 Go [2026-07-18 17:33:27] step 908250/1000000 | loss=2.0549 vocab=0.8627 token=1.0492 | lr=2.62e-05 | 6.11 steps/s | VRAM=29.07 Go [2026-07-18 17:33:35] step 908300/1000000 | loss=2.0642 vocab=1.3681 token=1.7262 | lr=2.62e-05 | 6.13 steps/s | VRAM=28.17 Go [2026-07-18 17:33:43] step 908350/1000000 | loss=2.0753 vocab=0.7316 token=1.0943 | lr=2.62e-05 | 6.14 steps/s | VRAM=25.69 Go [2026-07-18 17:33:52] step 908400/1000000 | loss=2.1225 vocab=0.9633 token=1.3850 | lr=2.61e-05 | 6.13 steps/s | VRAM=28.78 Go [2026-07-18 17:34:00] step 908450/1000000 | loss=1.9426 vocab=1.1440 token=1.6632 | lr=2.61e-05 | 6.14 steps/s | VRAM=31.46 Go [2026-07-18 17:34:08] step 908500/1000000 | loss=1.9401 vocab=0.9901 token=0.9607 | lr=2.61e-05 | 6.11 steps/s | VRAM=32.09 Go [2026-07-18 17:34:16] step 908550/1000000 | loss=2.1193 vocab=0.7468 token=1.0824 | lr=2.61e-05 | 6.14 steps/s | VRAM=31.03 Go [2026-07-18 17:34:24] step 908600/1000000 | loss=2.0378 vocab=0.7162 token=1.2240 | lr=2.60e-05 | 6.12 steps/s | VRAM=28.75 Go [2026-07-18 17:34:32] step 908650/1000000 | loss=2.0258 vocab=0.7868 token=0.8546 | lr=2.60e-05 | 6.13 steps/s | VRAM=28.93 Go [2026-07-18 17:34:41] step 908700/1000000 | loss=2.1672 vocab=0.3601 token=1.1426 | lr=2.60e-05 | 6.14 steps/s | VRAM=25.79 Go [2026-07-18 17:34:49] step 908750/1000000 | loss=2.0381 vocab=1.1328 token=1.1426 | lr=2.60e-05 | 6.15 steps/s | VRAM=28.82 Go [2026-07-18 17:34:57] step 908800/1000000 | loss=1.9863 vocab=0.7908 token=1.2455 | lr=2.59e-05 | 6.11 steps/s | VRAM=28.44 Go [2026-07-18 17:35:05] step 908850/1000000 | loss=2.0125 vocab=0.9469 token=1.4556 | lr=2.59e-05 | 6.11 steps/s | VRAM=26.47 Go [2026-07-18 17:35:13] step 908900/1000000 | loss=2.1002 vocab=0.5391 token=0.8328 | lr=2.59e-05 | 6.13 steps/s | VRAM=28.12 Go [2026-07-18 17:35:21] step 908950/1000000 | loss=1.9356 vocab=0.6913 token=1.0084 | lr=2.59e-05 | 6.14 steps/s | VRAM=28.82 Go [2026-07-18 17:35:29] step 909000/1000000 | loss=2.0860 vocab=0.9094 token=0.8456 | lr=2.58e-05 | 6.12 steps/s | VRAM=29.00 Go [2026-07-18 17:35:38] step 909050/1000000 | loss=2.0315 vocab=0.5231 token=0.7696 | lr=2.58e-05 | 6.12 steps/s | VRAM=28.35 Go [2026-07-18 17:35:46] step 909100/1000000 | loss=2.0088 vocab=0.6833 token=1.0291 | lr=2.58e-05 | 6.14 steps/s | VRAM=29.39 Go [2026-07-18 17:35:54] step 909150/1000000 | loss=2.0824 vocab=0.9297 token=1.2567 | lr=2.58e-05 | 6.16 steps/s | VRAM=32.77 Go [2026-07-18 17:36:02] step 909200/1000000 | loss=2.1472 vocab=0.6386 token=0.9986 | lr=2.57e-05 | 6.08 steps/s | VRAM=29.70 Go [2026-07-18 17:36:10] step 909250/1000000 | loss=2.0682 vocab=0.5958 token=1.0940 | lr=2.57e-05 | 6.15 steps/s | VRAM=32.35 Go [2026-07-18 17:36:18] step 909300/1000000 | loss=2.1320 vocab=1.1632 token=1.5094 | lr=2.57e-05 | 6.11 steps/s | VRAM=25.78 Go [2026-07-18 17:36:27] step 909350/1000000 | loss=2.0999 vocab=0.6951 token=0.9643 | lr=2.57e-05 | 6.15 steps/s | VRAM=24.76 Go [2026-07-18 17:36:35] step 909400/1000000 | loss=2.0214 vocab=0.9707 token=1.1109 | lr=2.56e-05 | 6.13 steps/s | VRAM=28.16 Go [2026-07-18 17:36:43] step 909450/1000000 | loss=2.0168 vocab=0.8639 token=1.2784 | lr=2.56e-05 | 6.14 steps/s | VRAM=32.22 Go [2026-07-18 17:36:51] step 909500/1000000 | loss=2.2635 vocab=1.0694 token=1.3027 | lr=2.56e-05 | 6.12 steps/s | VRAM=28.44 Go [2026-07-18 17:36:59] step 909550/1000000 | loss=1.9962 vocab=0.8780 token=0.9659 | lr=2.56e-05 | 6.16 steps/s | VRAM=31.71 Go [2026-07-18 17:37:07] step 909600/1000000 | loss=2.0050 vocab=1.0914 token=1.2968 | lr=2.55e-05 | 6.11 steps/s | VRAM=29.30 Go [2026-07-18 17:37:16] step 909650/1000000 | loss=1.9643 vocab=0.7110 token=1.6275 | lr=2.55e-05 | 6.12 steps/s | VRAM=28.63 Go [2026-07-18 17:37:24] step 909700/1000000 | loss=2.0413 vocab=0.7681 token=1.3950 | lr=2.55e-05 | 6.11 steps/s | VRAM=32.71 Go [2026-07-18 17:37:32] step 909750/1000000 | loss=2.0681 vocab=0.7571 token=1.2070 | lr=2.55e-05 | 6.14 steps/s | VRAM=31.36 Go [2026-07-18 17:37:40] step 909800/1000000 | loss=2.1623 vocab=0.8544 token=1.0132 | lr=2.54e-05 | 6.11 steps/s | VRAM=29.30 Go [2026-07-18 17:37:48] step 909850/1000000 | loss=2.0446 vocab=2.1455 token=1.3771 | lr=2.54e-05 | 6.15 steps/s | VRAM=31.39 Go [2026-07-18 17:37:56] step 909900/1000000 | loss=2.0258 vocab=1.1672 token=1.0583 | lr=2.54e-05 | 6.13 steps/s | VRAM=29.36 Go [2026-07-18 17:38:04] step 909950/1000000 | loss=2.0199 vocab=0.9092 token=1.3164 | lr=2.54e-05 | 6.13 steps/s | VRAM=24.99 Go [2026-07-18 17:38:13] step 910000/1000000 | loss=1.9263 vocab=0.7294 token=1.2101 | lr=2.53e-05 | 6.13 steps/s | VRAM=27.75 Go [2026-07-18 17:38:21] step 910050/1000000 | loss=1.9964 vocab=0.3980 token=1.0887 | lr=2.53e-05 | 6.13 steps/s | VRAM=28.85 Go [2026-07-18 17:38:29] step 910100/1000000 | loss=2.0533 vocab=0.9891 token=1.1736 | lr=2.53e-05 | 6.15 steps/s | VRAM=27.52 Go [2026-07-18 17:38:37] step 910150/1000000 | loss=2.0542 vocab=0.6369 token=1.1675 | lr=2.53e-05 | 6.14 steps/s | VRAM=29.02 Go [2026-07-18 17:38:45] step 910200/1000000 | loss=2.0876 vocab=0.8448 token=1.0002 | lr=2.52e-05 | 6.11 steps/s | VRAM=28.23 Go [2026-07-18 17:38:53] step 910250/1000000 | loss=2.0067 vocab=0.7452 token=1.3879 | lr=2.52e-05 | 6.11 steps/s | VRAM=31.78 Go [2026-07-18 17:39:02] step 910300/1000000 | loss=1.9949 vocab=0.6779 token=1.1959 | lr=2.52e-05 | 6.08 steps/s | VRAM=23.74 Go [2026-07-18 17:39:10] step 910350/1000000 | loss=2.0181 vocab=0.7002 token=1.1019 | lr=2.52e-05 | 6.11 steps/s | VRAM=29.14 Go [2026-07-18 17:39:18] step 910400/1000000 | loss=2.0860 vocab=0.6860 token=1.2297 | lr=2.51e-05 | 6.14 steps/s | VRAM=29.12 Go [2026-07-18 17:39:26] step 910450/1000000 | loss=2.0971 vocab=0.9188 token=1.3939 | lr=2.51e-05 | 6.11 steps/s | VRAM=26.16 Go [2026-07-18 17:39:34] step 910500/1000000 | loss=2.0781 vocab=0.6566 token=0.8115 | lr=2.51e-05 | 6.14 steps/s | VRAM=32.51 Go [2026-07-18 17:39:43] step 910550/1000000 | loss=2.0587 vocab=0.6625 token=1.0338 | lr=2.51e-05 | 6.10 steps/s | VRAM=26.53 Go [2026-07-18 17:39:51] step 910600/1000000 | loss=2.1084 vocab=0.7451 token=0.9980 | lr=2.50e-05 | 6.14 steps/s | VRAM=32.52 Go [2026-07-18 17:39:59] step 910650/1000000 | loss=2.0160 vocab=1.0248 token=0.9114 | lr=2.50e-05 | 6.13 steps/s | VRAM=25.95 Go [2026-07-18 17:40:07] step 910700/1000000 | loss=2.0636 vocab=0.5600 token=0.8457 | lr=2.50e-05 | 6.11 steps/s | VRAM=26.13 Go [2026-07-18 17:40:15] step 910750/1000000 | loss=2.0631 vocab=0.4421 token=0.8618 | lr=2.50e-05 | 6.14 steps/s | VRAM=28.00 Go [2026-07-18 17:40:23] step 910800/1000000 | loss=2.0299 vocab=0.8648 token=1.1746 | lr=2.49e-05 | 6.11 steps/s | VRAM=25.95 Go [2026-07-18 17:40:32] step 910850/1000000 | loss=2.0177 vocab=0.6166 token=1.1083 | lr=2.49e-05 | 6.13 steps/s | VRAM=26.26 Go [2026-07-18 17:40:40] step 910900/1000000 | loss=2.0695 vocab=1.1848 token=1.5042 | lr=2.49e-05 | 6.10 steps/s | VRAM=26.27 Go [2026-07-18 17:40:48] step 910950/1000000 | loss=2.1522 vocab=0.7233 token=1.2050 | lr=2.49e-05 | 6.12 steps/s | VRAM=28.74 Go [2026-07-18 17:40:56] step 911000/1000000 | loss=2.0425 vocab=1.0591 token=1.0349 | lr=2.49e-05 | 6.12 steps/s | VRAM=31.54 Go [2026-07-18 17:41:04] step 911050/1000000 | loss=2.0829 vocab=0.9085 token=1.3067 | lr=2.48e-05 | 6.15 steps/s | VRAM=31.96 Go [2026-07-18 17:41:12] step 911100/1000000 | loss=2.0638 vocab=0.8493 token=1.2312 | lr=2.48e-05 | 6.15 steps/s | VRAM=31.70 Go [2026-07-18 17:41:21] step 911150/1000000 | loss=2.1096 vocab=1.2512 token=1.4519 | lr=2.48e-05 | 6.08 steps/s | VRAM=26.54 Go [2026-07-18 17:41:29] step 911200/1000000 | loss=1.9928 vocab=1.1699 token=1.3488 | lr=2.48e-05 | 6.13 steps/s | VRAM=26.80 Go [2026-07-18 17:41:37] step 911250/1000000 | loss=1.9339 vocab=1.2163 token=1.1114 | lr=2.47e-05 | 6.16 steps/s | VRAM=31.72 Go [2026-07-18 17:41:45] step 911300/1000000 | loss=2.1072 vocab=0.9384 token=1.2408 | lr=2.47e-05 | 6.13 steps/s | VRAM=31.31 Go [2026-07-18 17:41:53] step 911350/1000000 | loss=2.0828 vocab=1.0723 token=0.9532 | lr=2.47e-05 | 6.16 steps/s | VRAM=28.79 Go [2026-07-18 17:42:01] step 911400/1000000 | loss=2.0297 vocab=0.9437 token=1.1137 | lr=2.47e-05 | 6.15 steps/s | VRAM=25.19 Go [2026-07-18 17:42:09] step 911450/1000000 | loss=2.0486 vocab=0.7807 token=1.2486 | lr=2.46e-05 | 6.13 steps/s | VRAM=24.90 Go [2026-07-18 17:42:18] step 911500/1000000 | loss=2.0043 vocab=0.8625 token=1.1561 | lr=2.46e-05 | 6.12 steps/s | VRAM=25.84 Go [2026-07-18 17:42:26] step 911550/1000000 | loss=2.0287 vocab=0.6719 token=1.0417 | lr=2.46e-05 | 6.15 steps/s | VRAM=32.07 Go [2026-07-18 17:42:34] step 911600/1000000 | loss=1.9940 vocab=1.0832 token=1.0774 | lr=2.46e-05 | 6.14 steps/s | VRAM=28.90 Go [2026-07-18 17:42:42] step 911650/1000000 | loss=2.0768 vocab=0.6616 token=1.2760 | lr=2.45e-05 | 6.14 steps/s | VRAM=31.81 Go [2026-07-18 17:42:50] step 911700/1000000 | loss=2.1322 vocab=0.9986 token=0.8184 | lr=2.45e-05 | 6.13 steps/s | VRAM=32.54 Go [2026-07-18 17:42:58] step 911750/1000000 | loss=2.0956 vocab=0.6209 token=0.8713 | lr=2.45e-05 | 6.15 steps/s | VRAM=25.20 Go [2026-07-18 17:43:06] step 911800/1000000 | loss=1.9883 vocab=0.6563 token=0.7962 | lr=2.45e-05 | 6.13 steps/s | VRAM=29.06 Go [2026-07-18 17:43:15] step 911850/1000000 | loss=2.0312 vocab=0.5165 token=0.8251 | lr=2.44e-05 | 6.12 steps/s | VRAM=28.37 Go [2026-07-18 17:43:23] step 911900/1000000 | loss=2.1158 vocab=1.3056 token=1.4799 | lr=2.44e-05 | 6.13 steps/s | VRAM=25.67 Go [2026-07-18 17:43:31] step 911950/1000000 | loss=2.0150 vocab=0.9454 token=1.3071 | lr=2.44e-05 | 6.14 steps/s | VRAM=32.11 Go [2026-07-18 17:43:39] step 912000/1000000 | loss=2.0277 vocab=1.4164 token=1.6459 | lr=2.44e-05 | 6.15 steps/s | VRAM=31.04 Go [2026-07-18 17:43:47] step 912050/1000000 | loss=2.0141 vocab=0.9459 token=1.2721 | lr=2.43e-05 | 6.14 steps/s | VRAM=31.39 Go [2026-07-18 17:43:55] step 912100/1000000 | loss=1.9754 vocab=0.7371 token=0.8689 | lr=2.43e-05 | 6.11 steps/s | VRAM=26.57 Go [2026-07-18 17:44:03] step 912150/1000000 | loss=2.0540 vocab=0.7168 token=1.1017 | lr=2.43e-05 | 6.12 steps/s | VRAM=29.14 Go [2026-07-18 17:44:12] step 912200/1000000 | loss=2.0238 vocab=0.6762 token=0.9564 | lr=2.43e-05 | 6.11 steps/s | VRAM=31.77 Go [2026-07-18 17:44:20] step 912250/1000000 | loss=2.0045 vocab=1.0181 token=1.2765 | lr=2.42e-05 | 6.12 steps/s | VRAM=29.07 Go [2026-07-18 17:44:28] step 912300/1000000 | loss=2.0660 vocab=0.8215 token=1.0484 | lr=2.42e-05 | 6.12 steps/s | VRAM=27.70 Go [2026-07-18 17:44:36] step 912350/1000000 | loss=2.0822 vocab=1.1586 token=1.0870 | lr=2.42e-05 | 6.16 steps/s | VRAM=31.56 Go [2026-07-18 17:44:44] step 912400/1000000 | loss=2.0621 vocab=0.8722 token=1.2915 | lr=2.42e-05 | 6.11 steps/s | VRAM=27.98 Go [2026-07-18 17:44:52] step 912450/1000000 | loss=2.1164 vocab=1.0650 token=1.2520 | lr=2.41e-05 | 6.16 steps/s | VRAM=31.78 Go [2026-07-18 17:45:01] step 912500/1000000 | loss=2.0934 vocab=1.0654 token=0.8374 | lr=2.41e-05 | 6.13 steps/s | VRAM=31.08 Go [2026-07-18 17:45:09] step 912550/1000000 | loss=2.0654 vocab=0.4626 token=0.7179 | lr=2.41e-05 | 6.10 steps/s | VRAM=28.68 Go [2026-07-18 17:45:17] step 912600/1000000 | loss=1.9210 vocab=0.9241 token=0.5475 | lr=2.41e-05 | 6.15 steps/s | VRAM=30.53 Go [2026-07-18 17:45:25] step 912650/1000000 | loss=2.1394 vocab=0.7988 token=1.2042 | lr=2.41e-05 | 6.15 steps/s | VRAM=25.84 Go [2026-07-18 17:45:33] step 912700/1000000 | loss=2.0996 vocab=0.6884 token=1.1406 | lr=2.40e-05 | 6.16 steps/s | VRAM=31.80 Go [2026-07-18 17:45:41] step 912750/1000000 | loss=2.1289 vocab=1.3752 token=1.9297 | lr=2.40e-05 | 6.13 steps/s | VRAM=29.41 Go [2026-07-18 17:45:49] step 912800/1000000 | loss=2.1086 vocab=0.6987 token=0.9317 | lr=2.40e-05 | 6.13 steps/s | VRAM=29.45 Go [2026-07-18 17:45:58] step 912850/1000000 | loss=2.1126 vocab=1.3512 token=1.6417 | lr=2.40e-05 | 6.10 steps/s | VRAM=26.21 Go [2026-07-18 17:46:06] step 912900/1000000 | loss=2.1804 vocab=1.4941 token=1.2310 | lr=2.39e-05 | 6.15 steps/s | VRAM=28.19 Go [2026-07-18 17:46:14] step 912950/1000000 | loss=2.0410 vocab=1.1779 token=1.6833 | lr=2.39e-05 | 6.12 steps/s | VRAM=32.32 Go [2026-07-18 17:46:22] step 913000/1000000 | loss=2.0217 vocab=0.6766 token=1.1706 | lr=2.39e-05 | 6.13 steps/s | VRAM=28.61 Go [2026-07-18 17:46:30] step 913050/1000000 | loss=2.0220 vocab=0.8051 token=0.8283 | lr=2.39e-05 | 6.08 steps/s | VRAM=28.09 Go [2026-07-18 17:46:39] step 913100/1000000 | loss=2.0190 vocab=0.7283 token=1.3203 | lr=2.38e-05 | 6.14 steps/s | VRAM=25.84 Go [2026-07-18 17:46:47] step 913150/1000000 | loss=1.9895 vocab=0.8019 token=1.1063 | lr=2.38e-05 | 6.14 steps/s | VRAM=31.85 Go [2026-07-18 17:46:55] step 913200/1000000 | loss=2.0376 vocab=1.1527 token=1.4336 | lr=2.38e-05 | 6.11 steps/s | VRAM=29.09 Go [2026-07-18 17:47:03] step 913250/1000000 | loss=2.1364 vocab=1.1804 token=1.2789 | lr=2.38e-05 | 6.12 steps/s | VRAM=26.84 Go [2026-07-18 17:47:11] step 913300/1000000 | loss=2.0511 vocab=0.5706 token=1.2797 | lr=2.37e-05 | 6.13 steps/s | VRAM=31.57 Go [2026-07-18 17:47:19] step 913350/1000000 | loss=2.0792 vocab=0.7643 token=1.1473 | lr=2.37e-05 | 6.10 steps/s | VRAM=25.81 Go [2026-07-18 17:47:28] step 913400/1000000 | loss=2.0846 vocab=1.5230 token=1.6237 | lr=2.37e-05 | 6.13 steps/s | VRAM=32.81 Go [2026-07-18 17:47:36] step 913450/1000000 | loss=1.9924 vocab=0.8600 token=1.0017 | lr=2.37e-05 | 6.10 steps/s | VRAM=28.83 Go [2026-07-18 17:47:44] step 913500/1000000 | loss=2.0693 vocab=0.6375 token=1.0401 | lr=2.36e-05 | 6.14 steps/s | VRAM=28.91 Go [2026-07-18 17:47:52] step 913550/1000000 | loss=2.0271 vocab=0.9229 token=1.1788 | lr=2.36e-05 | 6.17 steps/s | VRAM=24.62 Go [2026-07-18 17:48:00] step 913600/1000000 | loss=2.0308 vocab=0.9336 token=1.3596 | lr=2.36e-05 | 6.16 steps/s | VRAM=28.33 Go [2026-07-18 17:48:08] step 913650/1000000 | loss=2.0435 vocab=1.1249 token=1.3406 | lr=2.36e-05 | 6.13 steps/s | VRAM=32.71 Go [2026-07-18 17:48:16] step 913700/1000000 | loss=2.0617 vocab=0.8027 token=0.9129 | lr=2.35e-05 | 6.15 steps/s | VRAM=28.96 Go [2026-07-18 17:48:25] step 913750/1000000 | loss=2.1132 vocab=0.7254 token=1.0126 | lr=2.35e-05 | 6.13 steps/s | VRAM=32.65 Go [2026-07-18 17:48:33] step 913800/1000000 | loss=2.1177 vocab=1.0807 token=1.1143 | lr=2.35e-05 | 6.18 steps/s | VRAM=31.69 Go [2026-07-18 17:48:41] step 913850/1000000 | loss=2.0775 vocab=0.6648 token=1.2252 | lr=2.35e-05 | 6.15 steps/s | VRAM=32.07 Go [2026-07-18 17:48:49] step 913900/1000000 | loss=2.0720 vocab=0.8862 token=1.5330 | lr=2.35e-05 | 6.14 steps/s | VRAM=32.64 Go [2026-07-18 17:48:57] step 913950/1000000 | loss=1.9612 vocab=0.7724 token=1.3696 | lr=2.34e-05 | 6.13 steps/s | VRAM=29.09 Go [2026-07-18 17:49:05] step 914000/1000000 | loss=2.0434 vocab=1.4218 token=1.4799 | lr=2.34e-05 | 6.14 steps/s | VRAM=29.03 Go [2026-07-18 17:49:13] step 914050/1000000 | loss=2.0493 vocab=0.7483 token=1.3232 | lr=2.34e-05 | 6.12 steps/s | VRAM=26.87 Go [2026-07-18 17:49:21] step 914100/1000000 | loss=1.9960 vocab=0.6899 token=0.9632 | lr=2.34e-05 | 6.17 steps/s | VRAM=31.31 Go [2026-07-18 17:49:30] step 914150/1000000 | loss=2.0361 vocab=0.9693 token=1.4122 | lr=2.33e-05 | 6.14 steps/s | VRAM=25.56 Go [2026-07-18 17:49:38] step 914200/1000000 | loss=2.0629 vocab=1.0897 token=1.2122 | lr=2.33e-05 | 6.12 steps/s | VRAM=29.33 Go [2026-07-18 17:49:46] step 914250/1000000 | loss=1.9269 vocab=0.7469 token=1.2892 | lr=2.33e-05 | 6.14 steps/s | VRAM=28.21 Go [2026-07-18 17:49:54] step 914300/1000000 | loss=1.9901 vocab=1.5434 token=1.4927 | lr=2.33e-05 | 6.11 steps/s | VRAM=28.65 Go [2026-07-18 17:50:02] step 914350/1000000 | loss=2.1555 vocab=1.4483 token=1.4143 | lr=2.32e-05 | 6.14 steps/s | VRAM=25.99 Go [2026-07-18 17:50:10] step 914400/1000000 | loss=2.0214 vocab=0.8322 token=1.0910 | lr=2.32e-05 | 6.15 steps/s | VRAM=32.84 Go [2026-07-18 17:50:19] step 914450/1000000 | loss=1.9867 vocab=0.7513 token=1.1408 | lr=2.32e-05 | 6.11 steps/s | VRAM=26.55 Go [2026-07-18 17:50:27] step 914500/1000000 | loss=2.0548 vocab=0.7651 token=0.8822 | lr=2.32e-05 | 6.14 steps/s | VRAM=32.64 Go [2026-07-18 17:50:35] step 914550/1000000 | loss=2.0906 vocab=1.2781 token=1.3595 | lr=2.31e-05 | 6.13 steps/s | VRAM=28.74 Go [2026-07-18 17:50:43] step 914600/1000000 | loss=2.0160 vocab=0.7589 token=1.2208 | lr=2.31e-05 | 6.13 steps/s | VRAM=29.04 Go [2026-07-18 17:50:51] step 914650/1000000 | loss=2.0871 vocab=0.6884 token=1.0525 | lr=2.31e-05 | 6.13 steps/s | VRAM=28.98 Go [2026-07-18 17:50:59] step 914700/1000000 | loss=1.9546 vocab=1.1897 token=1.1374 | lr=2.31e-05 | 6.13 steps/s | VRAM=32.18 Go [2026-07-18 17:51:07] step 914750/1000000 | loss=2.0600 vocab=1.1848 token=1.3902 | lr=2.30e-05 | 6.15 steps/s | VRAM=31.49 Go [2026-07-18 17:51:16] step 914800/1000000 | loss=2.1230 vocab=1.0466 token=1.1634 | lr=2.30e-05 | 6.13 steps/s | VRAM=26.56 Go [2026-07-18 17:51:24] step 914850/1000000 | loss=2.0101 vocab=0.6422 token=0.9358 | lr=2.30e-05 | 6.16 steps/s | VRAM=28.11 Go [2026-07-18 17:51:32] step 914900/1000000 | loss=2.1417 vocab=0.2603 token=0.7816 | lr=2.30e-05 | 6.17 steps/s | VRAM=31.65 Go [2026-07-18 17:51:40] step 914950/1000000 | loss=2.0435 vocab=0.8943 token=1.2371 | lr=2.30e-05 | 6.13 steps/s | VRAM=32.58 Go [2026-07-18 17:51:48] step 915000/1000000 | loss=1.9422 vocab=0.8186 token=0.8061 | lr=2.29e-05 | 6.10 steps/s | VRAM=28.59 Go [2026-07-18 17:51:56] step 915050/1000000 | loss=2.0058 vocab=0.7499 token=0.9934 | lr=2.29e-05 | 6.14 steps/s | VRAM=28.56 Go [2026-07-18 17:52:04] step 915100/1000000 | loss=2.0505 vocab=0.9816 token=1.4261 | lr=2.29e-05 | 6.15 steps/s | VRAM=32.96 Go [2026-07-18 17:52:13] step 915150/1000000 | loss=2.0477 vocab=0.9228 token=1.2504 | lr=2.29e-05 | 6.11 steps/s | VRAM=26.53 Go [2026-07-18 17:52:21] step 915200/1000000 | loss=2.1443 vocab=1.3254 token=1.2632 | lr=2.28e-05 | 6.15 steps/s | VRAM=32.76 Go [2026-07-18 17:52:29] step 915250/1000000 | loss=1.9230 vocab=0.9057 token=0.5930 | lr=2.28e-05 | 6.15 steps/s | VRAM=31.55 Go [2026-07-18 17:52:37] step 915300/1000000 | loss=1.9907 vocab=0.9286 token=1.4468 | lr=2.28e-05 | 6.14 steps/s | VRAM=33.02 Go [2026-07-18 17:52:45] step 915350/1000000 | loss=2.0679 vocab=0.9127 token=1.1452 | lr=2.28e-05 | 6.13 steps/s | VRAM=29.22 Go [2026-07-18 17:52:53] step 915400/1000000 | loss=2.1301 vocab=0.5760 token=1.0509 | lr=2.27e-05 | 6.14 steps/s | VRAM=25.30 Go [2026-07-18 17:53:02] step 915450/1000000 | loss=2.0050 vocab=0.9139 token=1.3448 | lr=2.27e-05 | 6.14 steps/s | VRAM=24.48 Go [2026-07-18 17:53:10] step 915500/1000000 | loss=2.0264 vocab=0.6954 token=1.1323 | lr=2.27e-05 | 6.15 steps/s | VRAM=32.50 Go [2026-07-18 17:53:18] step 915550/1000000 | loss=1.9650 vocab=1.0455 token=0.8066 | lr=2.27e-05 | 6.12 steps/s | VRAM=29.29 Go [2026-07-18 17:53:26] step 915600/1000000 | loss=1.9804 vocab=1.2012 token=1.2874 | lr=2.26e-05 | 6.16 steps/s | VRAM=33.02 Go [2026-07-18 17:53:34] step 915650/1000000 | loss=2.1021 vocab=0.6886 token=1.3381 | lr=2.26e-05 | 6.15 steps/s | VRAM=31.76 Go [2026-07-18 17:53:42] step 915700/1000000 | loss=2.0715 vocab=1.0846 token=1.4869 | lr=2.26e-05 | 6.17 steps/s | VRAM=31.71 Go [2026-07-18 17:53:50] step 915750/1000000 | loss=2.0152 vocab=1.6277 token=1.0906 | lr=2.26e-05 | 6.17 steps/s | VRAM=31.71 Go [2026-07-18 17:53:58] step 915800/1000000 | loss=2.0780 vocab=0.6459 token=1.0717 | lr=2.25e-05 | 6.14 steps/s | VRAM=30.81 Go [2026-07-18 17:54:07] step 915850/1000000 | loss=2.0914 vocab=0.7506 token=1.0639 | lr=2.25e-05 | 6.13 steps/s | VRAM=24.12 Go [2026-07-18 17:54:15] step 915900/1000000 | loss=2.1362 vocab=1.0647 token=1.4718 | lr=2.25e-05 | 6.16 steps/s | VRAM=29.13 Go [2026-07-18 17:54:23] step 915950/1000000 | loss=1.8973 vocab=0.7781 token=1.2188 | lr=2.25e-05 | 6.19 steps/s | VRAM=32.39 Go [2026-07-18 17:54:31] step 916000/1000000 | loss=2.1082 vocab=0.7573 token=1.2043 | lr=2.25e-05 | 6.14 steps/s | VRAM=28.24 Go [2026-07-18 17:54:39] step 916050/1000000 | loss=1.9907 vocab=0.6056 token=0.8847 | lr=2.24e-05 | 6.15 steps/s | VRAM=28.47 Go [2026-07-18 17:54:47] step 916100/1000000 | loss=1.9751 vocab=0.4126 token=0.7424 | lr=2.24e-05 | 6.15 steps/s | VRAM=26.00 Go [2026-07-18 17:54:55] step 916150/1000000 | loss=1.9817 vocab=0.3518 token=0.8961 | lr=2.24e-05 | 6.16 steps/s | VRAM=32.32 Go [2026-07-18 17:55:03] step 916200/1000000 | loss=2.0377 vocab=1.0301 token=1.1171 | lr=2.24e-05 | 6.10 steps/s | VRAM=31.91 Go [2026-07-18 17:55:12] step 916250/1000000 | loss=2.1632 vocab=1.0265 token=1.2732 | lr=2.23e-05 | 6.13 steps/s | VRAM=28.35 Go [2026-07-18 17:55:20] step 916300/1000000 | loss=1.9795 vocab=0.5914 token=1.2966 | lr=2.23e-05 | 6.15 steps/s | VRAM=30.94 Go [2026-07-18 17:55:28] step 916350/1000000 | loss=1.9957 vocab=1.1111 token=1.1979 | lr=2.23e-05 | 6.12 steps/s | VRAM=30.23 Go [2026-07-18 17:55:36] step 916400/1000000 | loss=2.0419 vocab=0.9662 token=0.9723 | lr=2.23e-05 | 6.16 steps/s | VRAM=27.92 Go [2026-07-18 17:55:44] step 916450/1000000 | loss=2.1225 vocab=0.5370 token=0.8510 | lr=2.22e-05 | 6.13 steps/s | VRAM=32.85 Go [2026-07-18 17:55:52] step 916500/1000000 | loss=2.0376 vocab=0.9793 token=1.1558 | lr=2.22e-05 | 6.12 steps/s | VRAM=28.57 Go [2026-07-18 17:56:01] step 916550/1000000 | loss=1.9813 vocab=0.5853 token=0.8300 | lr=2.22e-05 | 6.13 steps/s | VRAM=28.05 Go [2026-07-18 17:56:09] step 916600/1000000 | loss=2.1697 vocab=0.5797 token=1.0588 | lr=2.22e-05 | 6.10 steps/s | VRAM=26.32 Go [2026-07-18 17:56:17] step 916650/1000000 | loss=2.0742 vocab=0.6099 token=0.8694 | lr=2.22e-05 | 6.15 steps/s | VRAM=27.52 Go [2026-07-18 17:56:25] step 916700/1000000 | loss=1.9550 vocab=0.8098 token=0.9279 | lr=2.21e-05 | 6.09 steps/s | VRAM=27.43 Go [2026-07-18 17:56:33] step 916750/1000000 | loss=2.0326 vocab=1.0779 token=1.5071 | lr=2.21e-05 | 6.07 steps/s | VRAM=28.95 Go [2026-07-18 17:56:41] step 916800/1000000 | loss=2.0655 vocab=1.3226 token=1.3660 | lr=2.21e-05 | 6.14 steps/s | VRAM=29.02 Go [2026-07-18 17:56:50] step 916850/1000000 | loss=2.0551 vocab=1.0594 token=0.9512 | lr=2.21e-05 | 6.13 steps/s | VRAM=26.33 Go [2026-07-18 17:56:58] step 916900/1000000 | loss=2.1143 vocab=1.4593 token=1.2481 | lr=2.20e-05 | 6.17 steps/s | VRAM=30.94 Go [2026-07-18 17:57:06] step 916950/1000000 | loss=1.9417 vocab=0.7607 token=0.9797 | lr=2.20e-05 | 6.14 steps/s | VRAM=27.36 Go [2026-07-18 17:57:14] step 917000/1000000 | loss=2.0751 vocab=0.7573 token=0.9727 | lr=2.20e-05 | 6.12 steps/s | VRAM=28.50 Go [2026-07-18 17:57:22] step 917050/1000000 | loss=1.9661 vocab=1.1528 token=1.6002 | lr=2.20e-05 | 6.15 steps/s | VRAM=32.05 Go [2026-07-18 17:57:30] step 917100/1000000 | loss=1.9551 vocab=1.1113 token=1.4999 | lr=2.19e-05 | 6.14 steps/s | VRAM=32.92 Go [2026-07-18 17:57:38] step 917150/1000000 | loss=2.0313 vocab=0.6979 token=0.8815 | lr=2.19e-05 | 6.17 steps/s | VRAM=32.10 Go [2026-07-18 17:57:47] step 917200/1000000 | loss=2.0726 vocab=0.8762 token=1.1260 | lr=2.19e-05 | 6.13 steps/s | VRAM=32.12 Go [2026-07-18 17:57:55] step 917250/1000000 | loss=2.0189 vocab=0.6099 token=1.2620 | lr=2.19e-05 | 6.13 steps/s | VRAM=28.17 Go [2026-07-18 17:58:03] step 917300/1000000 | loss=2.0136 vocab=0.8156 token=1.1081 | lr=2.18e-05 | 6.13 steps/s | VRAM=29.06 Go [2026-07-18 17:58:11] step 917350/1000000 | loss=2.1372 vocab=1.0770 token=1.0103 | lr=2.18e-05 | 6.12 steps/s | VRAM=28.69 Go [2026-07-18 17:58:19] step 917400/1000000 | loss=2.0746 vocab=0.9828 token=1.3137 | lr=2.18e-05 | 6.14 steps/s | VRAM=32.96 Go [2026-07-18 17:58:27] step 917450/1000000 | loss=2.0754 vocab=0.7352 token=1.2030 | lr=2.18e-05 | 6.12 steps/s | VRAM=25.98 Go [2026-07-18 17:58:36] step 917500/1000000 | loss=2.0384 vocab=0.8940 token=0.8267 | lr=2.18e-05 | 6.13 steps/s | VRAM=32.16 Go [2026-07-18 17:58:44] step 917550/1000000 | loss=2.0991 vocab=0.6506 token=1.0894 | lr=2.17e-05 | 6.13 steps/s | VRAM=32.59 Go [2026-07-18 17:58:52] step 917600/1000000 | loss=2.0484 vocab=0.7303 token=0.9711 | lr=2.17e-05 | 6.16 steps/s | VRAM=32.55 Go [2026-07-18 17:59:00] step 917650/1000000 | loss=1.9967 vocab=0.8607 token=0.9540 | lr=2.17e-05 | 6.12 steps/s | VRAM=31.47 Go [2026-07-18 17:59:08] step 917700/1000000 | loss=2.0292 vocab=0.7928 token=1.1982 | lr=2.17e-05 | 6.12 steps/s | VRAM=27.39 Go [2026-07-18 17:59:16] step 917750/1000000 | loss=2.0265 vocab=0.7432 token=1.1797 | lr=2.16e-05 | 6.14 steps/s | VRAM=32.45 Go [2026-07-18 17:59:24] step 917800/1000000 | loss=2.1552 vocab=1.0454 token=1.1543 | lr=2.16e-05 | 6.14 steps/s | VRAM=26.02 Go [2026-07-18 17:59:33] step 917850/1000000 | loss=2.1619 vocab=0.8207 token=1.1656 | lr=2.16e-05 | 6.14 steps/s | VRAM=32.85 Go [2026-07-18 17:59:41] step 917900/1000000 | loss=2.0494 vocab=0.4698 token=0.8798 | lr=2.16e-05 | 6.14 steps/s | VRAM=28.23 Go [2026-07-18 17:59:49] step 917950/1000000 | loss=2.0366 vocab=0.8011 token=1.1284 | lr=2.15e-05 | 6.14 steps/s | VRAM=32.02 Go [2026-07-18 17:59:57] step 918000/1000000 | loss=2.1141 vocab=0.7455 token=1.1284 | lr=2.15e-05 | 6.15 steps/s | VRAM=28.24 Go [2026-07-18 18:00:05] step 918050/1000000 | loss=2.0796 vocab=1.2255 token=1.5255 | lr=2.15e-05 | 6.13 steps/s | VRAM=25.21 Go [2026-07-18 18:00:13] step 918100/1000000 | loss=2.2070 vocab=0.5203 token=1.0138 | lr=2.15e-05 | 6.16 steps/s | VRAM=32.20 Go [2026-07-18 18:00:21] step 918150/1000000 | loss=2.0628 vocab=1.0977 token=0.8440 | lr=2.15e-05 | 6.11 steps/s | VRAM=29.65 Go [2026-07-18 18:00:30] step 918200/1000000 | loss=1.9927 vocab=1.1219 token=1.3049 | lr=2.14e-05 | 6.10 steps/s | VRAM=25.24 Go [2026-07-18 18:00:38] step 918250/1000000 | loss=1.9574 vocab=0.5946 token=1.0537 | lr=2.14e-05 | 6.14 steps/s | VRAM=29.33 Go [2026-07-18 18:00:46] step 918300/1000000 | loss=1.9868 vocab=1.1922 token=1.1614 | lr=2.14e-05 | 6.15 steps/s | VRAM=24.71 Go [2026-07-18 18:00:54] step 918350/1000000 | loss=2.0183 vocab=0.5961 token=0.9864 | lr=2.14e-05 | 6.11 steps/s | VRAM=29.55 Go [2026-07-18 18:01:02] step 918400/1000000 | loss=2.1038 vocab=0.7529 token=1.6970 | lr=2.13e-05 | 6.14 steps/s | VRAM=32.16 Go [2026-07-18 18:01:10] step 918450/1000000 | loss=2.0732 vocab=0.8812 token=1.0500 | lr=2.13e-05 | 6.13 steps/s | VRAM=32.65 Go [2026-07-18 18:01:19] step 918500/1000000 | loss=1.9952 vocab=0.4490 token=1.0475 | lr=2.13e-05 | 6.11 steps/s | VRAM=32.85 Go [2026-07-18 18:01:27] step 918550/1000000 | loss=1.9718 vocab=0.7878 token=1.3589 | lr=2.13e-05 | 6.16 steps/s | VRAM=32.38 Go [2026-07-18 18:01:35] step 918600/1000000 | loss=1.9624 vocab=0.7393 token=1.1357 | lr=2.12e-05 | 6.12 steps/s | VRAM=32.51 Go [2026-07-18 18:01:43] step 918650/1000000 | loss=2.0883 vocab=0.6499 token=1.1576 | lr=2.12e-05 | 6.10 steps/s | VRAM=28.57 Go [2026-07-18 18:01:51] step 918700/1000000 | loss=2.1145 vocab=0.5552 token=1.0809 | lr=2.12e-05 | 6.10 steps/s | VRAM=26.72 Go [2026-07-18 18:01:59] step 918750/1000000 | loss=2.0027 vocab=0.5911 token=1.0760 | lr=2.12e-05 | 6.13 steps/s | VRAM=28.43 Go [2026-07-18 18:02:08] step 918800/1000000 | loss=2.0760 vocab=0.7802 token=1.0670 | lr=2.12e-05 | 6.12 steps/s | VRAM=26.22 Go [2026-07-18 18:02:16] step 918850/1000000 | loss=1.9153 vocab=0.4851 token=0.7408 | lr=2.11e-05 | 6.11 steps/s | VRAM=28.55 Go [2026-07-18 18:02:24] step 918900/1000000 | loss=2.1827 vocab=0.7276 token=1.2693 | lr=2.11e-05 | 6.12 steps/s | VRAM=23.86 Go [2026-07-18 18:02:32] step 918950/1000000 | loss=2.0866 vocab=0.3739 token=0.6855 | lr=2.11e-05 | 6.14 steps/s | VRAM=25.61 Go [2026-07-18 18:02:40] step 919000/1000000 | loss=2.0464 vocab=0.7934 token=1.3573 | lr=2.11e-05 | 6.14 steps/s | VRAM=32.17 Go [2026-07-18 18:02:48] step 919050/1000000 | loss=2.0539 vocab=1.0441 token=1.2021 | lr=2.10e-05 | 6.13 steps/s | VRAM=30.53 Go [2026-07-18 18:02:57] step 919100/1000000 | loss=2.1230 vocab=0.8995 token=0.9994 | lr=2.10e-05 | 6.14 steps/s | VRAM=32.09 Go [2026-07-18 18:03:05] step 919150/1000000 | loss=2.0578 vocab=1.1977 token=1.4493 | lr=2.10e-05 | 6.13 steps/s | VRAM=31.78 Go [2026-07-18 18:03:13] step 919200/1000000 | loss=2.0605 vocab=1.0767 token=1.0873 | lr=2.10e-05 | 6.14 steps/s | VRAM=29.22 Go [2026-07-18 18:03:21] step 919250/1000000 | loss=2.1050 vocab=1.3505 token=1.2023 | lr=2.09e-05 | 6.18 steps/s | VRAM=30.83 Go [2026-07-18 18:03:29] step 919300/1000000 | loss=2.1060 vocab=0.8458 token=1.3318 | lr=2.09e-05 | 6.14 steps/s | VRAM=31.96 Go [2026-07-18 18:03:37] step 919350/1000000 | loss=2.0533 vocab=0.9612 token=1.5036 | lr=2.09e-05 | 6.13 steps/s | VRAM=26.02 Go [2026-07-18 18:03:45] step 919400/1000000 | loss=2.0624 vocab=0.6006 token=1.3665 | lr=2.09e-05 | 6.10 steps/s | VRAM=32.31 Go [2026-07-18 18:03:54] step 919450/1000000 | loss=2.0874 vocab=1.0319 token=1.3354 | lr=2.09e-05 | 6.12 steps/s | VRAM=29.51 Go [2026-07-18 18:04:02] step 919500/1000000 | loss=2.1298 vocab=0.8187 token=1.7449 | lr=2.08e-05 | 6.13 steps/s | VRAM=32.20 Go [2026-07-18 18:04:10] step 919550/1000000 | loss=2.0208 vocab=0.8608 token=1.3635 | lr=2.08e-05 | 6.14 steps/s | VRAM=32.58 Go [2026-07-18 18:04:18] step 919600/1000000 | loss=2.0992 vocab=0.9606 token=1.3827 | lr=2.08e-05 | 6.14 steps/s | VRAM=27.85 Go [2026-07-18 18:04:26] step 919650/1000000 | loss=2.0373 vocab=0.9896 token=1.3779 | lr=2.08e-05 | 6.14 steps/s | VRAM=29.20 Go [2026-07-18 18:04:34] step 919700/1000000 | loss=2.0734 vocab=0.9092 token=1.2446 | lr=2.07e-05 | 6.14 steps/s | VRAM=27.59 Go [2026-07-18 18:04:42] step 919750/1000000 | loss=2.0346 vocab=0.5386 token=0.6204 | lr=2.07e-05 | 6.12 steps/s | VRAM=26.12 Go [2026-07-18 18:04:51] step 919800/1000000 | loss=2.0891 vocab=1.0389 token=1.2513 | lr=2.07e-05 | 6.17 steps/s | VRAM=24.56 Go [2026-07-18 18:04:59] step 919850/1000000 | loss=2.0642 vocab=0.8357 token=1.0447 | lr=2.07e-05 | 6.13 steps/s | VRAM=29.81 Go [2026-07-18 18:05:07] step 919900/1000000 | loss=2.0806 vocab=0.6425 token=0.9543 | lr=2.07e-05 | 6.15 steps/s | VRAM=32.23 Go [2026-07-18 18:05:15] step 919950/1000000 | loss=2.0566 vocab=0.6003 token=1.0378 | lr=2.06e-05 | 6.17 steps/s | VRAM=30.55 Go [2026-07-18 18:05:23] step 920000/1000000 | loss=2.1122 vocab=0.8963 token=1.1232 | lr=2.06e-05 | 6.15 steps/s | VRAM=27.80 Go [2026-07-18 18:05:31] step 920050/1000000 | loss=1.9989 vocab=0.5969 token=1.2308 | lr=2.06e-05 | 6.10 steps/s | VRAM=29.58 Go [2026-07-18 18:05:39] step 920100/1000000 | loss=1.9901 vocab=0.8975 token=1.2320 | lr=2.06e-05 | 6.13 steps/s | VRAM=29.11 Go [2026-07-18 18:05:48] step 920150/1000000 | loss=2.1150 vocab=1.4643 token=1.3421 | lr=2.05e-05 | 6.14 steps/s | VRAM=28.19 Go [2026-07-18 18:05:56] step 920200/1000000 | loss=2.0615 vocab=1.2297 token=1.5070 | lr=2.05e-05 | 6.14 steps/s | VRAM=29.16 Go [2026-07-18 18:06:04] step 920250/1000000 | loss=1.9364 vocab=0.6291 token=1.1982 | lr=2.05e-05 | 6.13 steps/s | VRAM=28.97 Go [2026-07-18 18:06:12] step 920300/1000000 | loss=2.0276 vocab=0.5763 token=0.9355 | lr=2.05e-05 | 6.14 steps/s | VRAM=32.60 Go [2026-07-18 18:06:20] step 920350/1000000 | loss=2.0432 vocab=1.1945 token=1.0513 | lr=2.04e-05 | 6.11 steps/s | VRAM=28.62 Go [2026-07-18 18:06:28] step 920400/1000000 | loss=2.1179 vocab=0.9062 token=1.4615 | lr=2.04e-05 | 6.15 steps/s | VRAM=31.63 Go [2026-07-18 18:06:36] step 920450/1000000 | loss=2.0238 vocab=1.1331 token=1.4221 | lr=2.04e-05 | 6.15 steps/s | VRAM=28.33 Go [2026-07-18 18:06:45] step 920500/1000000 | loss=2.1147 vocab=0.9365 token=1.1944 | lr=2.04e-05 | 6.14 steps/s | VRAM=28.14 Go [2026-07-18 18:06:53] step 920550/1000000 | loss=2.0359 vocab=1.2379 token=1.2657 | lr=2.04e-05 | 6.14 steps/s | VRAM=26.28 Go [2026-07-18 18:07:01] step 920600/1000000 | loss=1.9569 vocab=1.2224 token=1.5996 | lr=2.03e-05 | 6.14 steps/s | VRAM=32.20 Go [2026-07-18 18:07:09] step 920650/1000000 | loss=2.0586 vocab=1.1457 token=1.0981 | lr=2.03e-05 | 6.12 steps/s | VRAM=28.03 Go [2026-07-18 18:07:17] step 920700/1000000 | loss=2.0234 vocab=1.2824 token=1.1789 | lr=2.03e-05 | 6.11 steps/s | VRAM=29.58 Go [2026-07-18 18:07:25] step 920750/1000000 | loss=2.0983 vocab=0.4143 token=0.8771 | lr=2.03e-05 | 6.11 steps/s | VRAM=28.62 Go [2026-07-18 18:07:34] step 920800/1000000 | loss=2.0523 vocab=0.5125 token=0.7419 | lr=2.02e-05 | 6.14 steps/s | VRAM=32.77 Go [2026-07-18 18:07:42] step 920850/1000000 | loss=2.0072 vocab=0.7226 token=0.9550 | lr=2.02e-05 | 6.16 steps/s | VRAM=31.44 Go [2026-07-18 18:07:50] step 920900/1000000 | loss=2.0518 vocab=0.4152 token=1.2033 | lr=2.02e-05 | 6.09 steps/s | VRAM=28.26 Go [2026-07-18 18:07:58] step 920950/1000000 | loss=2.1668 vocab=1.5117 token=1.4782 | lr=2.02e-05 | 6.15 steps/s | VRAM=26.19 Go [2026-07-18 18:08:06] step 921000/1000000 | loss=2.0144 vocab=0.5539 token=1.0199 | lr=2.02e-05 | 6.14 steps/s | VRAM=28.43 Go [2026-07-18 18:08:14] step 921050/1000000 | loss=2.0016 vocab=0.6976 token=1.1596 | lr=2.01e-05 | 6.14 steps/s | VRAM=32.95 Go [2026-07-18 18:08:22] step 921100/1000000 | loss=2.0909 vocab=1.4272 token=1.2871 | lr=2.01e-05 | 6.14 steps/s | VRAM=29.50 Go [2026-07-18 18:08:31] step 921150/1000000 | loss=2.0158 vocab=0.4221 token=0.8411 | lr=2.01e-05 | 6.11 steps/s | VRAM=28.41 Go [2026-07-18 18:08:39] step 921200/1000000 | loss=2.0132 vocab=0.9719 token=1.5264 | lr=2.01e-05 | 6.14 steps/s | VRAM=28.90 Go [2026-07-18 18:08:47] step 921250/1000000 | loss=2.0219 vocab=0.6576 token=0.8701 | lr=2.00e-05 | 6.15 steps/s | VRAM=29.14 Go [2026-07-18 18:08:55] step 921300/1000000 | loss=1.9655 vocab=0.5600 token=0.7982 | lr=2.00e-05 | 6.14 steps/s | VRAM=27.43 Go [2026-07-18 18:09:03] step 921350/1000000 | loss=2.0657 vocab=1.1806 token=1.2249 | lr=2.00e-05 | 6.15 steps/s | VRAM=28.03 Go [2026-07-18 18:09:11] step 921400/1000000 | loss=2.0065 vocab=1.3435 token=1.2843 | lr=2.00e-05 | 6.11 steps/s | VRAM=26.87 Go [2026-07-18 18:09:19] step 921450/1000000 | loss=1.9852 vocab=0.5180 token=1.1580 | lr=1.99e-05 | 6.15 steps/s | VRAM=28.41 Go [2026-07-18 18:09:28] step 921500/1000000 | loss=2.0805 vocab=0.7086 token=1.2142 | lr=1.99e-05 | 6.18 steps/s | VRAM=23.46 Go [2026-07-18 18:09:36] step 921550/1000000 | loss=2.1164 vocab=0.6892 token=1.3096 | lr=1.99e-05 | 6.12 steps/s | VRAM=29.29 Go [2026-07-18 18:09:44] step 921600/1000000 | loss=1.9715 vocab=0.6482 token=1.1552 | lr=1.99e-05 | 6.14 steps/s | VRAM=29.99 Go [2026-07-18 18:09:52] step 921650/1000000 | loss=2.0340 vocab=1.3906 token=1.0476 | lr=1.99e-05 | 6.14 steps/s | VRAM=31.29 Go [2026-07-18 18:10:00] step 921700/1000000 | loss=1.9515 vocab=0.5261 token=0.9456 | lr=1.98e-05 | 6.13 steps/s | VRAM=32.19 Go [2026-07-18 18:10:08] step 921750/1000000 | loss=2.1041 vocab=0.7402 token=1.4039 | lr=1.98e-05 | 6.13 steps/s | VRAM=25.93 Go [2026-07-18 18:10:16] step 921800/1000000 | loss=2.0159 vocab=1.1073 token=1.5709 | lr=1.98e-05 | 6.14 steps/s | VRAM=28.89 Go [2026-07-18 18:10:25] step 921850/1000000 | loss=2.0061 vocab=0.6754 token=0.7892 | lr=1.98e-05 | 6.15 steps/s | VRAM=29.03 Go [2026-07-18 18:10:33] step 921900/1000000 | loss=1.9826 vocab=0.8783 token=1.0143 | lr=1.97e-05 | 6.13 steps/s | VRAM=28.23 Go [2026-07-18 18:10:41] step 921950/1000000 | loss=2.1093 vocab=0.2806 token=0.9219 | lr=1.97e-05 | 6.15 steps/s | VRAM=32.12 Go [2026-07-18 18:10:49] step 922000/1000000 | loss=2.0206 vocab=0.8153 token=0.9371 | lr=1.97e-05 | 6.14 steps/s | VRAM=29.57 Go [2026-07-18 18:10:57] step 922050/1000000 | loss=2.0250 vocab=0.6742 token=1.0113 | lr=1.97e-05 | 6.12 steps/s | VRAM=28.81 Go [2026-07-18 18:11:05] step 922100/1000000 | loss=2.0055 vocab=0.6906 token=1.0039 | lr=1.97e-05 | 6.14 steps/s | VRAM=28.61 Go [2026-07-18 18:11:13] step 922150/1000000 | loss=2.0051 vocab=0.8360 token=0.8831 | lr=1.96e-05 | 6.17 steps/s | VRAM=31.26 Go [2026-07-18 18:11:22] step 922200/1000000 | loss=2.0610 vocab=0.8345 token=0.8940 | lr=1.96e-05 | 6.20 steps/s | VRAM=31.46 Go [2026-07-18 18:11:30] step 922250/1000000 | loss=2.1091 vocab=1.2141 token=1.3337 | lr=1.96e-05 | 6.13 steps/s | VRAM=27.82 Go [2026-07-18 18:11:38] step 922300/1000000 | loss=2.1254 vocab=0.5504 token=1.2190 | lr=1.96e-05 | 6.16 steps/s | VRAM=29.99 Go [2026-07-18 18:11:46] step 922350/1000000 | loss=2.0571 vocab=0.6731 token=1.0563 | lr=1.95e-05 | 6.16 steps/s | VRAM=24.68 Go [2026-07-18 18:11:54] step 922400/1000000 | loss=2.0922 vocab=1.1267 token=1.6639 | lr=1.95e-05 | 6.14 steps/s | VRAM=25.15 Go [2026-07-18 18:12:02] step 922450/1000000 | loss=1.9483 vocab=1.1100 token=1.0916 | lr=1.95e-05 | 6.14 steps/s | VRAM=32.42 Go [2026-07-18 18:12:10] step 922500/1000000 | loss=2.0890 vocab=1.6603 token=1.5396 | lr=1.95e-05 | 6.13 steps/s | VRAM=26.47 Go [2026-07-18 18:12:18] step 922550/1000000 | loss=1.9772 vocab=0.7381 token=1.2180 | lr=1.95e-05 | 6.16 steps/s | VRAM=24.80 Go [2026-07-18 18:12:27] step 922600/1000000 | loss=2.1054 vocab=0.8497 token=1.0704 | lr=1.94e-05 | 6.15 steps/s | VRAM=26.71 Go [2026-07-18 18:12:35] step 922650/1000000 | loss=2.0792 vocab=1.0074 token=1.5462 | lr=1.94e-05 | 6.15 steps/s | VRAM=28.54 Go [2026-07-18 18:12:43] step 922700/1000000 | loss=2.0838 vocab=0.8226 token=1.3723 | lr=1.94e-05 | 6.12 steps/s | VRAM=25.24 Go [2026-07-18 18:12:51] step 922750/1000000 | loss=2.0752 vocab=0.7228 token=1.0347 | lr=1.94e-05 | 6.14 steps/s | VRAM=24.77 Go [2026-07-18 18:12:59] step 922800/1000000 | loss=2.0503 vocab=0.7586 token=1.2254 | lr=1.93e-05 | 6.14 steps/s | VRAM=25.82 Go [2026-07-18 18:13:07] step 922850/1000000 | loss=2.0348 vocab=0.5151 token=1.1622 | lr=1.93e-05 | 6.13 steps/s | VRAM=26.29 Go [2026-07-18 18:13:16] step 922900/1000000 | loss=2.0556 vocab=0.8313 token=0.8706 | lr=1.93e-05 | 6.12 steps/s | VRAM=32.83 Go [2026-07-18 18:13:24] step 922950/1000000 | loss=2.1071 vocab=1.1857 token=1.3418 | lr=1.93e-05 | 6.17 steps/s | VRAM=31.27 Go [2026-07-18 18:13:32] step 923000/1000000 | loss=2.1395 vocab=0.9980 token=1.0735 | lr=1.93e-05 | 6.14 steps/s | VRAM=28.09 Go [2026-07-18 18:13:40] step 923050/1000000 | loss=2.0991 vocab=1.3865 token=1.2079 | lr=1.92e-05 | 6.12 steps/s | VRAM=28.19 Go [2026-07-18 18:13:48] step 923100/1000000 | loss=2.0738 vocab=0.8912 token=1.3546 | lr=1.92e-05 | 6.16 steps/s | VRAM=31.77 Go [2026-07-18 18:13:56] step 923150/1000000 | loss=1.9823 vocab=0.5262 token=0.9656 | lr=1.92e-05 | 6.15 steps/s | VRAM=32.43 Go [2026-07-18 18:14:04] step 923200/1000000 | loss=2.1153 vocab=0.8473 token=1.2916 | lr=1.92e-05 | 6.14 steps/s | VRAM=31.78 Go [2026-07-18 18:14:13] step 923250/1000000 | loss=1.9781 vocab=0.7999 token=0.8833 | lr=1.91e-05 | 6.11 steps/s | VRAM=25.77 Go [2026-07-18 18:14:21] step 923300/1000000 | loss=1.9896 vocab=0.8415 token=1.1023 | lr=1.91e-05 | 6.14 steps/s | VRAM=32.51 Go [2026-07-18 18:14:29] step 923350/1000000 | loss=2.0448 vocab=0.9103 token=1.3120 | lr=1.91e-05 | 6.18 steps/s | VRAM=32.58 Go [2026-07-18 18:14:37] step 923400/1000000 | loss=1.9190 vocab=0.7750 token=1.1771 | lr=1.91e-05 | 6.14 steps/s | VRAM=28.32 Go [2026-07-18 18:14:45] step 923450/1000000 | loss=2.2120 vocab=0.9858 token=1.5636 | lr=1.91e-05 | 6.13 steps/s | VRAM=28.88 Go [2026-07-18 18:14:53] step 923500/1000000 | loss=2.1733 vocab=1.1151 token=1.3278 | lr=1.90e-05 | 6.15 steps/s | VRAM=29.92 Go [2026-07-18 18:15:01] step 923550/1000000 | loss=2.0644 vocab=0.6540 token=0.9533 | lr=1.90e-05 | 6.15 steps/s | VRAM=31.49 Go [2026-07-18 18:15:09] step 923600/1000000 | loss=2.0743 vocab=0.5887 token=0.7672 | lr=1.90e-05 | 6.14 steps/s | VRAM=32.00 Go [2026-07-18 18:15:18] step 923650/1000000 | loss=1.9759 vocab=0.8360 token=1.0206 | lr=1.90e-05 | 6.15 steps/s | VRAM=33.02 Go [2026-07-18 18:15:26] step 923700/1000000 | loss=1.9727 vocab=0.9812 token=1.2093 | lr=1.89e-05 | 6.14 steps/s | VRAM=25.93 Go [2026-07-18 18:15:34] step 923750/1000000 | loss=2.0797 vocab=0.9011 token=1.2144 | lr=1.89e-05 | 6.15 steps/s | VRAM=32.06 Go [2026-07-18 18:15:42] step 923800/1000000 | loss=1.9068 vocab=0.5003 token=0.7436 | lr=1.89e-05 | 6.16 steps/s | VRAM=29.38 Go [2026-07-18 18:15:50] step 923850/1000000 | loss=2.1630 vocab=0.7956 token=0.9476 | lr=1.89e-05 | 6.11 steps/s | VRAM=26.69 Go [2026-07-18 18:15:58] step 923900/1000000 | loss=2.0511 vocab=0.7178 token=1.5371 | lr=1.89e-05 | 6.15 steps/s | VRAM=28.59 Go [2026-07-18 18:16:06] step 923950/1000000 | loss=2.0404 vocab=0.6947 token=0.9731 | lr=1.88e-05 | 6.16 steps/s | VRAM=32.26 Go [2026-07-18 18:16:15] step 924000/1000000 | loss=2.0105 vocab=1.2635 token=1.0981 | lr=1.88e-05 | 6.15 steps/s | VRAM=31.38 Go [2026-07-18 18:16:23] step 924050/1000000 | loss=2.1482 vocab=1.1039 token=1.5779 | lr=1.88e-05 | 6.14 steps/s | VRAM=31.34 Go [2026-07-18 18:16:31] step 924100/1000000 | loss=2.1455 vocab=0.3963 token=0.8499 | lr=1.88e-05 | 6.12 steps/s | VRAM=26.41 Go [2026-07-18 18:16:39] step 924150/1000000 | loss=2.1414 vocab=0.6972 token=1.0785 | lr=1.88e-05 | 6.14 steps/s | VRAM=29.18 Go [2026-07-18 18:16:47] step 924200/1000000 | loss=2.0338 vocab=0.5125 token=1.1100 | lr=1.87e-05 | 6.13 steps/s | VRAM=29.45 Go [2026-07-18 18:16:55] step 924250/1000000 | loss=1.9648 vocab=0.8693 token=1.1118 | lr=1.87e-05 | 6.09 steps/s | VRAM=29.48 Go [2026-07-18 18:17:04] step 924300/1000000 | loss=2.0320 vocab=0.7534 token=1.2012 | lr=1.87e-05 | 6.15 steps/s | VRAM=28.37 Go [2026-07-18 18:17:12] step 924350/1000000 | loss=2.0656 vocab=1.0226 token=1.5599 | lr=1.87e-05 | 6.15 steps/s | VRAM=24.86 Go [2026-07-18 18:17:20] step 924400/1000000 | loss=2.0600 vocab=1.3970 token=1.3529 | lr=1.86e-05 | 6.13 steps/s | VRAM=31.75 Go [2026-07-18 18:17:28] step 924450/1000000 | loss=1.9558 vocab=0.4389 token=0.8756 | lr=1.86e-05 | 6.16 steps/s | VRAM=24.90 Go [2026-07-18 18:17:36] step 924500/1000000 | loss=2.1587 vocab=0.9933 token=1.4915 | lr=1.86e-05 | 6.15 steps/s | VRAM=28.44 Go [2026-07-18 18:17:44] step 924550/1000000 | loss=2.0358 vocab=0.9812 token=1.2403 | lr=1.86e-05 | 6.14 steps/s | VRAM=29.04 Go [2026-07-18 18:17:52] step 924600/1000000 | loss=2.0329 vocab=0.8735 token=1.2900 | lr=1.86e-05 | 6.14 steps/s | VRAM=29.60 Go [2026-07-18 18:18:01] step 924650/1000000 | loss=1.9856 vocab=1.0310 token=1.2287 | lr=1.85e-05 | 6.13 steps/s | VRAM=28.70 Go [2026-07-18 18:18:09] step 924700/1000000 | loss=2.1106 vocab=1.0615 token=1.4813 | lr=1.85e-05 | 6.13 steps/s | VRAM=29.75 Go [2026-07-18 18:18:17] step 924750/1000000 | loss=1.9760 vocab=0.8961 token=1.3299 | lr=1.85e-05 | 6.16 steps/s | VRAM=31.54 Go [2026-07-18 18:18:25] step 924800/1000000 | loss=2.1061 vocab=1.2279 token=1.2236 | lr=1.85e-05 | 6.16 steps/s | VRAM=28.50 Go [2026-07-18 18:18:33] step 924850/1000000 | loss=2.0733 vocab=0.7891 token=1.5953 | lr=1.84e-05 | 6.15 steps/s | VRAM=26.68 Go [2026-07-18 18:18:41] step 924900/1000000 | loss=2.2002 vocab=0.7760 token=1.1373 | lr=1.84e-05 | 6.11 steps/s | VRAM=29.50 Go [2026-07-18 18:18:49] step 924950/1000000 | loss=2.0319 vocab=0.6265 token=0.8000 | lr=1.84e-05 | 6.12 steps/s | VRAM=26.11 Go [2026-07-18 18:18:58] step 925000/1000000 | loss=2.0345 vocab=0.8421 token=1.1381 | lr=1.84e-05 | 6.13 steps/s | VRAM=28.19 Go [2026-07-18 18:19:06] step 925050/1000000 | loss=2.0992 vocab=0.7727 token=0.8322 | lr=1.84e-05 | 6.10 steps/s | VRAM=24.58 Go [2026-07-18 18:19:14] step 925100/1000000 | loss=2.0016 vocab=0.7021 token=1.0462 | lr=1.83e-05 | 6.17 steps/s | VRAM=31.86 Go [2026-07-18 18:19:22] step 925150/1000000 | loss=2.0761 vocab=0.4678 token=0.9601 | lr=1.83e-05 | 6.07 steps/s | VRAM=26.14 Go [2026-07-18 18:19:30] step 925200/1000000 | loss=2.0863 vocab=0.5981 token=1.0809 | lr=1.83e-05 | 6.16 steps/s | VRAM=28.71 Go [2026-07-18 18:19:38] step 925250/1000000 | loss=2.0605 vocab=0.9315 token=1.5375 | lr=1.83e-05 | 6.13 steps/s | VRAM=28.27 Go [2026-07-18 18:19:46] step 925300/1000000 | loss=2.0570 vocab=1.0294 token=1.1299 | lr=1.83e-05 | 6.16 steps/s | VRAM=28.82 Go [2026-07-18 18:19:55] step 925350/1000000 | loss=1.9351 vocab=0.6950 token=1.1117 | lr=1.82e-05 | 6.14 steps/s | VRAM=26.19 Go [2026-07-18 18:20:03] step 925400/1000000 | loss=2.1087 vocab=0.8204 token=1.2878 | lr=1.82e-05 | 6.16 steps/s | VRAM=28.15 Go [2026-07-18 18:20:11] step 925450/1000000 | loss=2.0215 vocab=0.4274 token=1.1462 | lr=1.82e-05 | 6.15 steps/s | VRAM=28.61 Go [2026-07-18 18:20:19] step 925500/1000000 | loss=1.9944 vocab=0.8034 token=1.1487 | lr=1.82e-05 | 6.17 steps/s | VRAM=32.66 Go [2026-07-18 18:20:27] step 925550/1000000 | loss=2.0244 vocab=0.7044 token=0.8635 | lr=1.81e-05 | 6.13 steps/s | VRAM=32.30 Go [2026-07-18 18:20:35] step 925600/1000000 | loss=1.9579 vocab=1.0564 token=1.5040 | lr=1.81e-05 | 6.14 steps/s | VRAM=28.75 Go [2026-07-18 18:20:43] step 925650/1000000 | loss=1.9512 vocab=0.6842 token=1.0485 | lr=1.81e-05 | 6.11 steps/s | VRAM=29.73 Go [2026-07-18 18:20:52] step 925700/1000000 | loss=2.0242 vocab=0.4023 token=0.8856 | lr=1.81e-05 | 6.17 steps/s | VRAM=25.94 Go [2026-07-18 18:21:00] step 925750/1000000 | loss=1.9481 vocab=0.6856 token=1.1305 | lr=1.81e-05 | 6.16 steps/s | VRAM=28.00 Go [2026-07-18 18:21:08] step 925800/1000000 | loss=2.1035 vocab=0.4549 token=1.2183 | lr=1.80e-05 | 6.13 steps/s | VRAM=29.13 Go [2026-07-18 18:21:16] step 925850/1000000 | loss=2.1427 vocab=1.0420 token=1.1781 | lr=1.80e-05 | 6.18 steps/s | VRAM=31.25 Go [2026-07-18 18:21:24] step 925900/1000000 | loss=2.0873 vocab=1.2633 token=1.5839 | lr=1.80e-05 | 6.17 steps/s | VRAM=28.45 Go [2026-07-18 18:21:32] step 925950/1000000 | loss=1.9557 vocab=1.0360 token=1.3089 | lr=1.80e-05 | 6.16 steps/s | VRAM=28.41 Go [2026-07-18 18:21:40] step 926000/1000000 | loss=2.0007 vocab=0.6607 token=1.1205 | lr=1.79e-05 | 6.17 steps/s | VRAM=25.55 Go [2026-07-18 18:21:48] step 926050/1000000 | loss=2.0198 vocab=0.7112 token=1.0129 | lr=1.79e-05 | 6.14 steps/s | VRAM=31.86 Go [2026-07-18 18:21:57] step 926100/1000000 | loss=2.1537 vocab=0.8988 token=1.0628 | lr=1.79e-05 | 6.14 steps/s | VRAM=31.46 Go [2026-07-18 18:22:05] step 926150/1000000 | loss=1.9250 vocab=1.1087 token=1.3051 | lr=1.79e-05 | 6.18 steps/s | VRAM=31.12 Go [2026-07-18 18:22:13] step 926200/1000000 | loss=2.0070 vocab=0.6215 token=1.2042 | lr=1.79e-05 | 6.14 steps/s | VRAM=32.45 Go [2026-07-18 18:22:21] step 926250/1000000 | loss=2.0236 vocab=0.9239 token=1.1542 | lr=1.78e-05 | 6.15 steps/s | VRAM=31.98 Go [2026-07-18 18:22:29] step 926300/1000000 | loss=2.0173 vocab=1.1034 token=1.2083 | lr=1.78e-05 | 6.15 steps/s | VRAM=28.90 Go [2026-07-18 18:22:37] step 926350/1000000 | loss=2.0140 vocab=0.8957 token=1.2667 | lr=1.78e-05 | 6.14 steps/s | VRAM=29.27 Go [2026-07-18 18:22:45] step 926400/1000000 | loss=2.0287 vocab=0.5102 token=1.3270 | lr=1.78e-05 | 6.18 steps/s | VRAM=31.56 Go [2026-07-18 18:22:53] step 926450/1000000 | loss=2.1138 vocab=0.6810 token=0.9073 | lr=1.78e-05 | 6.12 steps/s | VRAM=29.02 Go [2026-07-18 18:23:02] step 926500/1000000 | loss=2.0103 vocab=0.5168 token=0.8667 | lr=1.77e-05 | 6.14 steps/s | VRAM=29.28 Go [2026-07-18 18:23:10] step 926550/1000000 | loss=2.0366 vocab=1.4046 token=1.2909 | lr=1.77e-05 | 6.16 steps/s | VRAM=31.84 Go [2026-07-18 18:23:18] step 926600/1000000 | loss=2.0419 vocab=0.8918 token=1.1066 | lr=1.77e-05 | 6.17 steps/s | VRAM=32.05 Go [2026-07-18 18:23:26] step 926650/1000000 | loss=2.0092 vocab=0.9909 token=1.1806 | lr=1.77e-05 | 6.15 steps/s | VRAM=28.53 Go [2026-07-18 18:23:34] step 926700/1000000 | loss=1.9431 vocab=0.9428 token=1.2603 | lr=1.76e-05 | 6.16 steps/s | VRAM=32.61 Go [2026-07-18 18:23:42] step 926750/1000000 | loss=1.9460 vocab=0.6727 token=1.1875 | lr=1.76e-05 | 6.17 steps/s | VRAM=30.84 Go [2026-07-18 18:23:50] step 926800/1000000 | loss=2.1251 vocab=1.0904 token=1.4086 | lr=1.76e-05 | 6.14 steps/s | VRAM=27.90 Go [2026-07-18 18:23:58] step 926850/1000000 | loss=2.0126 vocab=0.6042 token=1.1547 | lr=1.76e-05 | 6.13 steps/s | VRAM=29.64 Go [2026-07-18 18:24:07] step 926900/1000000 | loss=1.9581 vocab=1.0363 token=1.5094 | lr=1.76e-05 | 6.12 steps/s | VRAM=28.20 Go [2026-07-18 18:24:15] step 926950/1000000 | loss=2.0106 vocab=0.9261 token=0.7800 | lr=1.75e-05 | 6.15 steps/s | VRAM=28.71 Go [2026-07-18 18:24:23] step 927000/1000000 | loss=2.1035 vocab=0.8041 token=1.1569 | lr=1.75e-05 | 6.16 steps/s | VRAM=28.17 Go [2026-07-18 18:24:31] step 927050/1000000 | loss=1.9444 vocab=0.8648 token=1.1143 | lr=1.75e-05 | 6.13 steps/s | VRAM=25.81 Go [2026-07-18 18:24:39] step 927100/1000000 | loss=2.0604 vocab=0.5978 token=0.9638 | lr=1.75e-05 | 6.14 steps/s | VRAM=28.80 Go [2026-07-18 18:24:47] step 927150/1000000 | loss=2.0812 vocab=0.5991 token=1.0220 | lr=1.75e-05 | 6.17 steps/s | VRAM=31.29 Go [2026-07-18 18:24:55] step 927200/1000000 | loss=2.1295 vocab=0.6996 token=0.9381 | lr=1.74e-05 | 6.15 steps/s | VRAM=28.29 Go [2026-07-18 18:25:03] step 927250/1000000 | loss=2.0591 vocab=0.6364 token=1.2879 | lr=1.74e-05 | 6.19 steps/s | VRAM=31.70 Go [2026-07-18 18:25:12] step 927300/1000000 | loss=2.0904 vocab=0.8214 token=1.5222 | lr=1.74e-05 | 6.14 steps/s | VRAM=29.15 Go [2026-07-18 18:25:20] step 927350/1000000 | loss=2.0861 vocab=1.4508 token=1.2864 | lr=1.74e-05 | 6.14 steps/s | VRAM=31.12 Go [2026-07-18 18:25:28] step 927400/1000000 | loss=2.1953 vocab=0.8904 token=0.9847 | lr=1.74e-05 | 6.13 steps/s | VRAM=25.20 Go [2026-07-18 18:25:36] step 927450/1000000 | loss=2.0347 vocab=0.9392 token=1.3443 | lr=1.73e-05 | 6.12 steps/s | VRAM=26.08 Go [2026-07-18 18:25:44] step 927500/1000000 | loss=1.9862 vocab=0.7291 token=1.2402 | lr=1.73e-05 | 6.15 steps/s | VRAM=28.35 Go [2026-07-18 18:25:52] step 927550/1000000 | loss=2.0242 vocab=1.1922 token=1.0992 | lr=1.73e-05 | 6.13 steps/s | VRAM=28.48 Go [2026-07-18 18:26:01] step 927600/1000000 | loss=2.0549 vocab=0.6432 token=0.9821 | lr=1.73e-05 | 6.11 steps/s | VRAM=29.92 Go [2026-07-18 18:26:09] step 927650/1000000 | loss=1.9558 vocab=0.8945 token=1.2189 | lr=1.72e-05 | 6.15 steps/s | VRAM=28.32 Go [2026-07-18 18:26:17] step 927700/1000000 | loss=2.0160 vocab=0.8674 token=1.2672 | lr=1.72e-05 | 6.17 steps/s | VRAM=32.24 Go [2026-07-18 18:26:25] step 927750/1000000 | loss=2.1033 vocab=0.9414 token=1.2999 | lr=1.72e-05 | 6.14 steps/s | VRAM=26.30 Go [2026-07-18 18:26:33] step 927800/1000000 | loss=2.0422 vocab=0.6450 token=0.7624 | lr=1.72e-05 | 6.13 steps/s | VRAM=28.67 Go [2026-07-18 18:26:41] step 927850/1000000 | loss=1.9938 vocab=0.7727 token=0.9905 | lr=1.72e-05 | 6.15 steps/s | VRAM=29.80 Go [2026-07-18 18:26:49] step 927900/1000000 | loss=1.9958 vocab=1.2966 token=1.1045 | lr=1.71e-05 | 6.18 steps/s | VRAM=27.58 Go [2026-07-18 18:26:57] step 927950/1000000 | loss=2.0768 vocab=0.9594 token=1.2858 | lr=1.71e-05 | 6.14 steps/s | VRAM=27.09 Go [2026-07-18 18:27:06] step 928000/1000000 | loss=1.9940 vocab=0.8896 token=0.6062 | lr=1.71e-05 | 6.15 steps/s | VRAM=27.97 Go [2026-07-18 18:27:14] step 928050/1000000 | loss=2.0072 vocab=0.8788 token=1.1848 | lr=1.71e-05 | 6.12 steps/s | VRAM=28.87 Go [2026-07-18 18:27:22] step 928100/1000000 | loss=1.9776 vocab=0.8581 token=1.3197 | lr=1.71e-05 | 6.15 steps/s | VRAM=28.44 Go [2026-07-18 18:27:30] step 928150/1000000 | loss=2.0187 vocab=1.1572 token=1.3340 | lr=1.70e-05 | 6.12 steps/s | VRAM=29.06 Go [2026-07-18 18:27:38] step 928200/1000000 | loss=2.0196 vocab=0.5432 token=1.0641 | lr=1.70e-05 | 6.17 steps/s | VRAM=31.35 Go [2026-07-18 18:27:46] step 928250/1000000 | loss=2.0148 vocab=0.9159 token=1.3500 | lr=1.70e-05 | 6.13 steps/s | VRAM=27.69 Go [2026-07-18 18:27:54] step 928300/1000000 | loss=2.0743 vocab=0.9279 token=1.2885 | lr=1.70e-05 | 6.14 steps/s | VRAM=28.59 Go [2026-07-18 18:28:03] step 928350/1000000 | loss=2.1106 vocab=0.9864 token=1.0046 | lr=1.70e-05 | 6.13 steps/s | VRAM=29.31 Go [2026-07-18 18:28:11] step 928400/1000000 | loss=1.9503 vocab=0.3675 token=0.6853 | lr=1.69e-05 | 6.15 steps/s | VRAM=29.32 Go [2026-07-18 18:28:19] step 928450/1000000 | loss=2.1198 vocab=0.7535 token=0.8519 | lr=1.69e-05 | 6.15 steps/s | VRAM=29.45 Go [2026-07-18 18:28:27] step 928500/1000000 | loss=1.9667 vocab=1.3309 token=1.0812 | lr=1.69e-05 | 6.14 steps/s | VRAM=31.88 Go [2026-07-18 18:28:35] step 928550/1000000 | loss=2.0594 vocab=1.0026 token=1.1718 | lr=1.69e-05 | 6.15 steps/s | VRAM=28.85 Go [2026-07-18 18:28:43] step 928600/1000000 | loss=2.0656 vocab=0.6712 token=1.2075 | lr=1.68e-05 | 6.13 steps/s | VRAM=25.18 Go [2026-07-18 18:28:52] step 928650/1000000 | loss=2.0851 vocab=1.1719 token=1.0656 | lr=1.68e-05 | 6.11 steps/s | VRAM=29.12 Go [2026-07-18 18:29:00] step 928700/1000000 | loss=1.9666 vocab=1.2512 token=1.4319 | lr=1.68e-05 | 6.14 steps/s | VRAM=32.41 Go [2026-07-18 18:29:08] step 928750/1000000 | loss=1.9516 vocab=0.9774 token=1.3200 | lr=1.68e-05 | 6.11 steps/s | VRAM=27.11 Go [2026-07-18 18:29:16] step 928800/1000000 | loss=2.0697 vocab=0.8859 token=0.9015 | lr=1.68e-05 | 6.12 steps/s | VRAM=26.13 Go [2026-07-18 18:29:24] step 928850/1000000 | loss=2.0105 vocab=0.3483 token=0.7065 | lr=1.67e-05 | 6.17 steps/s | VRAM=28.11 Go [2026-07-18 18:29:32] step 928900/1000000 | loss=2.0082 vocab=1.1093 token=1.4782 | lr=1.67e-05 | 6.13 steps/s | VRAM=28.85 Go [2026-07-18 18:29:40] step 928950/1000000 | loss=2.0647 vocab=0.5675 token=1.0945 | lr=1.67e-05 | 6.16 steps/s | VRAM=31.81 Go [2026-07-18 18:29:49] step 929000/1000000 | loss=2.0035 vocab=0.6384 token=0.7745 | lr=1.67e-05 | 6.15 steps/s | VRAM=32.92 Go [2026-07-18 18:29:57] step 929050/1000000 | loss=2.1246 vocab=1.1645 token=1.3038 | lr=1.67e-05 | 6.16 steps/s | VRAM=28.35 Go [2026-07-18 18:30:05] step 929100/1000000 | loss=1.9911 vocab=0.9338 token=1.4477 | lr=1.66e-05 | 6.13 steps/s | VRAM=29.64 Go [2026-07-18 18:30:13] step 929150/1000000 | loss=2.0816 vocab=0.6474 token=1.0685 | lr=1.66e-05 | 6.12 steps/s | VRAM=26.59 Go [2026-07-18 18:30:21] step 929200/1000000 | loss=2.0836 vocab=0.9461 token=1.1706 | lr=1.66e-05 | 6.14 steps/s | VRAM=28.28 Go [2026-07-18 18:30:29] step 929250/1000000 | loss=2.0531 vocab=0.4819 token=0.6504 | lr=1.66e-05 | 6.11 steps/s | VRAM=26.60 Go [2026-07-18 18:30:37] step 929300/1000000 | loss=2.0207 vocab=0.7384 token=1.3291 | lr=1.66e-05 | 6.15 steps/s | VRAM=28.56 Go [2026-07-18 18:30:46] step 929350/1000000 | loss=2.0591 vocab=0.8772 token=0.6870 | lr=1.65e-05 | 6.17 steps/s | VRAM=31.40 Go [2026-07-18 18:30:54] step 929400/1000000 | loss=2.1184 vocab=0.6495 token=0.8881 | lr=1.65e-05 | 6.11 steps/s | VRAM=26.17 Go [2026-07-18 18:31:02] step 929450/1000000 | loss=2.0803 vocab=0.7152 token=1.0478 | lr=1.65e-05 | 6.12 steps/s | VRAM=29.09 Go [2026-07-18 18:31:10] step 929500/1000000 | loss=2.0694 vocab=0.8260 token=1.2338 | lr=1.65e-05 | 6.15 steps/s | VRAM=28.81 Go [2026-07-18 18:31:18] step 929550/1000000 | loss=2.0941 vocab=1.3701 token=1.3061 | lr=1.65e-05 | 6.14 steps/s | VRAM=26.47 Go [2026-07-18 18:31:26] step 929600/1000000 | loss=2.0681 vocab=0.8538 token=1.0714 | lr=1.64e-05 | 6.14 steps/s | VRAM=26.90 Go [2026-07-18 18:31:35] step 929650/1000000 | loss=2.0006 vocab=0.5195 token=1.1440 | lr=1.64e-05 | 6.13 steps/s | VRAM=29.02 Go [2026-07-18 18:31:43] step 929700/1000000 | loss=1.9611 vocab=0.7451 token=1.3144 | lr=1.64e-05 | 6.12 steps/s | VRAM=26.48 Go [2026-07-18 18:31:51] step 929750/1000000 | loss=2.1437 vocab=1.0214 token=1.4191 | lr=1.64e-05 | 6.14 steps/s | VRAM=29.52 Go [2026-07-18 18:31:59] step 929800/1000000 | loss=1.9658 vocab=0.5819 token=1.1238 | lr=1.63e-05 | 6.16 steps/s | VRAM=32.58 Go [2026-07-18 18:32:07] step 929850/1000000 | loss=2.0325 vocab=1.0577 token=0.7530 | lr=1.63e-05 | 6.11 steps/s | VRAM=26.37 Go [2026-07-18 18:32:15] step 929900/1000000 | loss=1.9283 vocab=0.7415 token=0.8791 | lr=1.63e-05 | 6.13 steps/s | VRAM=31.20 Go [2026-07-18 18:32:23] step 929950/1000000 | loss=2.0637 vocab=0.4436 token=1.2947 | lr=1.63e-05 | 6.14 steps/s | VRAM=28.42 Go [2026-07-18 18:32:32] step 930000/1000000 | loss=2.1034 vocab=1.2139 token=1.8476 | lr=1.63e-05 | 6.12 steps/s | VRAM=28.85 Go [2026-07-18 18:32:40] step 930050/1000000 | loss=2.1586 vocab=0.7760 token=0.9434 | lr=1.62e-05 | 6.11 steps/s | VRAM=31.60 Go [2026-07-18 18:32:48] step 930100/1000000 | loss=2.1334 vocab=1.1387 token=1.0062 | lr=1.62e-05 | 6.13 steps/s | VRAM=25.56 Go [2026-07-18 18:32:56] step 930150/1000000 | loss=2.1404 vocab=0.7658 token=1.3168 | lr=1.62e-05 | 6.17 steps/s | VRAM=28.27 Go [2026-07-18 18:33:04] step 930200/1000000 | loss=1.9110 vocab=0.5903 token=1.3196 | lr=1.62e-05 | 6.15 steps/s | VRAM=31.69 Go [2026-07-18 18:33:12] step 930250/1000000 | loss=2.0741 vocab=0.5776 token=1.0388 | lr=1.62e-05 | 6.12 steps/s | VRAM=30.46 Go [2026-07-18 18:33:20] step 930300/1000000 | loss=2.0882 vocab=1.3318 token=1.0774 | lr=1.61e-05 | 6.17 steps/s | VRAM=32.34 Go [2026-07-18 18:33:29] step 930350/1000000 | loss=2.0085 vocab=0.6090 token=1.0330 | lr=1.61e-05 | 6.14 steps/s | VRAM=29.13 Go [2026-07-18 18:33:37] step 930400/1000000 | loss=2.0677 vocab=0.9151 token=1.1415 | lr=1.61e-05 | 6.14 steps/s | VRAM=29.20 Go [2026-07-18 18:33:45] step 930450/1000000 | loss=2.1321 vocab=0.6006 token=1.0834 | lr=1.61e-05 | 6.14 steps/s | VRAM=28.81 Go [2026-07-18 18:33:53] step 930500/1000000 | loss=2.0922 vocab=1.0209 token=1.2943 | lr=1.61e-05 | 6.12 steps/s | VRAM=28.95 Go [2026-07-18 18:34:01] step 930550/1000000 | loss=2.0595 vocab=1.0791 token=0.8332 | lr=1.60e-05 | 6.14 steps/s | VRAM=28.17 Go [2026-07-18 18:34:09] step 930600/1000000 | loss=1.9949 vocab=1.0927 token=1.0343 | lr=1.60e-05 | 6.13 steps/s | VRAM=26.85 Go [2026-07-18 18:34:17] step 930650/1000000 | loss=1.9665 vocab=0.9902 token=1.3667 | lr=1.60e-05 | 6.16 steps/s | VRAM=28.36 Go [2026-07-18 18:34:26] step 930700/1000000 | loss=2.0656 vocab=0.6403 token=0.7540 | lr=1.60e-05 | 6.12 steps/s | VRAM=28.40 Go [2026-07-18 18:34:34] step 930750/1000000 | loss=2.1812 vocab=0.7897 token=1.4616 | lr=1.60e-05 | 6.13 steps/s | VRAM=32.57 Go [2026-07-18 18:34:42] step 930800/1000000 | loss=2.1425 vocab=0.9320 token=1.2209 | lr=1.59e-05 | 6.12 steps/s | VRAM=29.54 Go [2026-07-18 18:34:50] step 930850/1000000 | loss=1.9921 vocab=0.8313 token=1.1258 | lr=1.59e-05 | 6.16 steps/s | VRAM=29.27 Go [2026-07-18 18:34:58] step 930900/1000000 | loss=1.9936 vocab=0.8405 token=1.1070 | lr=1.59e-05 | 6.18 steps/s | VRAM=30.52 Go [2026-07-18 18:35:06] step 930950/1000000 | loss=2.0577 vocab=0.9744 token=1.3896 | lr=1.59e-05 | 6.16 steps/s | VRAM=25.55 Go [2026-07-18 18:35:14] step 931000/1000000 | loss=2.0354 vocab=0.6254 token=0.8556 | lr=1.59e-05 | 6.17 steps/s | VRAM=28.52 Go [2026-07-18 18:35:23] step 931050/1000000 | loss=1.9931 vocab=0.3348 token=1.0208 | lr=1.58e-05 | 6.13 steps/s | VRAM=26.22 Go [2026-07-18 18:35:31] step 931100/1000000 | loss=2.0515 vocab=1.1834 token=1.3867 | lr=1.58e-05 | 6.13 steps/s | VRAM=29.76 Go [2026-07-18 18:35:39] step 931150/1000000 | loss=2.0744 vocab=1.5047 token=1.4650 | lr=1.58e-05 | 6.13 steps/s | VRAM=31.44 Go [2026-07-18 18:35:47] step 931200/1000000 | loss=2.0964 vocab=1.1741 token=0.7115 | lr=1.58e-05 | 6.15 steps/s | VRAM=29.10 Go [2026-07-18 18:35:55] step 931250/1000000 | loss=1.9410 vocab=0.6283 token=0.5783 | lr=1.58e-05 | 6.15 steps/s | VRAM=28.50 Go [2026-07-18 18:36:03] step 931300/1000000 | loss=2.0509 vocab=0.7296 token=1.3804 | lr=1.57e-05 | 6.14 steps/s | VRAM=28.38 Go [2026-07-18 18:36:11] step 931350/1000000 | loss=2.0728 vocab=1.2352 token=1.1193 | lr=1.57e-05 | 6.15 steps/s | VRAM=25.82 Go [2026-07-18 18:36:20] step 931400/1000000 | loss=2.0314 vocab=0.9245 token=0.9727 | lr=1.57e-05 | 6.14 steps/s | VRAM=29.71 Go [2026-07-18 18:36:28] step 931450/1000000 | loss=1.9942 vocab=0.8308 token=1.1374 | lr=1.57e-05 | 6.15 steps/s | VRAM=31.85 Go [2026-07-18 18:36:36] step 931500/1000000 | loss=1.9712 vocab=1.0744 token=1.3638 | lr=1.57e-05 | 6.14 steps/s | VRAM=26.11 Go [2026-07-18 18:36:44] step 931550/1000000 | loss=2.1023 vocab=1.3044 token=1.2495 | lr=1.56e-05 | 6.13 steps/s | VRAM=29.54 Go [2026-07-18 18:36:52] step 931600/1000000 | loss=1.9966 vocab=0.4081 token=1.0328 | lr=1.56e-05 | 6.12 steps/s | VRAM=26.29 Go [2026-07-18 18:37:00] step 931650/1000000 | loss=2.0933 vocab=0.5600 token=1.2477 | lr=1.56e-05 | 6.15 steps/s | VRAM=28.69 Go [2026-07-18 18:37:08] step 931700/1000000 | loss=2.1172 vocab=0.6484 token=1.1067 | lr=1.56e-05 | 6.10 steps/s | VRAM=28.14 Go [2026-07-18 18:37:17] step 931750/1000000 | loss=1.9957 vocab=0.6502 token=1.0058 | lr=1.56e-05 | 6.13 steps/s | VRAM=26.00 Go [2026-07-18 18:37:25] step 931800/1000000 | loss=2.1211 vocab=1.1929 token=1.4339 | lr=1.55e-05 | 6.15 steps/s | VRAM=28.29 Go [2026-07-18 18:37:33] step 931850/1000000 | loss=2.1175 vocab=0.7502 token=1.4984 | lr=1.55e-05 | 6.15 steps/s | VRAM=31.76 Go [2026-07-18 18:37:41] step 931900/1000000 | loss=2.0999 vocab=0.6355 token=1.2366 | lr=1.55e-05 | 6.14 steps/s | VRAM=29.22 Go [2026-07-18 18:37:49] step 931950/1000000 | loss=2.1695 vocab=0.5849 token=1.0915 | lr=1.55e-05 | 6.13 steps/s | VRAM=32.08 Go [2026-07-18 18:37:57] step 932000/1000000 | loss=2.1567 vocab=0.7500 token=1.0434 | lr=1.54e-05 | 6.12 steps/s | VRAM=31.60 Go [2026-07-18 18:38:05] step 932050/1000000 | loss=2.1226 vocab=1.2008 token=1.2722 | lr=1.54e-05 | 6.17 steps/s | VRAM=31.17 Go [2026-07-18 18:38:14] step 932100/1000000 | loss=2.0348 vocab=1.1452 token=1.1948 | lr=1.54e-05 | 6.14 steps/s | VRAM=32.50 Go [2026-07-18 18:38:22] step 932150/1000000 | loss=2.0215 vocab=0.7810 token=1.1889 | lr=1.54e-05 | 6.17 steps/s | VRAM=31.39 Go [2026-07-18 18:38:30] step 932200/1000000 | loss=2.1471 vocab=0.5641 token=0.9906 | lr=1.54e-05 | 6.12 steps/s | VRAM=26.05 Go [2026-07-18 18:38:38] step 932250/1000000 | loss=2.0368 vocab=0.8288 token=1.3270 | lr=1.53e-05 | 6.13 steps/s | VRAM=29.47 Go [2026-07-18 18:38:46] step 932300/1000000 | loss=2.0366 vocab=0.6330 token=1.1866 | lr=1.53e-05 | 6.15 steps/s | VRAM=24.95 Go [2026-07-18 18:38:54] step 932350/1000000 | loss=2.0865 vocab=0.8505 token=1.4158 | lr=1.53e-05 | 6.14 steps/s | VRAM=32.62 Go [2026-07-18 18:39:02] step 932400/1000000 | loss=2.1388 vocab=1.2998 token=1.4253 | lr=1.53e-05 | 6.11 steps/s | VRAM=28.94 Go [2026-07-18 18:39:11] step 932450/1000000 | loss=2.0141 vocab=0.6802 token=1.3696 | lr=1.53e-05 | 6.15 steps/s | VRAM=23.71 Go [2026-07-18 18:39:19] step 932500/1000000 | loss=1.9652 vocab=0.9397 token=1.1818 | lr=1.52e-05 | 6.10 steps/s | VRAM=32.77 Go [2026-07-18 18:39:27] step 932550/1000000 | loss=1.9568 vocab=0.2292 token=0.3812 | lr=1.52e-05 | 6.14 steps/s | VRAM=29.79 Go [2026-07-18 18:39:35] step 932600/1000000 | loss=2.0781 vocab=0.9719 token=1.0503 | lr=1.52e-05 | 6.16 steps/s | VRAM=28.45 Go [2026-07-18 18:39:43] step 932650/1000000 | loss=1.9594 vocab=0.5326 token=1.3021 | lr=1.52e-05 | 6.14 steps/s | VRAM=32.22 Go [2026-07-18 18:39:51] step 932700/1000000 | loss=2.1317 vocab=0.9747 token=1.1884 | lr=1.52e-05 | 6.12 steps/s | VRAM=28.26 Go [2026-07-18 18:39:59] step 932750/1000000 | loss=2.0725 vocab=0.6816 token=1.1348 | lr=1.51e-05 | 6.17 steps/s | VRAM=29.12 Go [2026-07-18 18:40:08] step 932800/1000000 | loss=2.0695 vocab=0.8929 token=0.7336 | lr=1.51e-05 | 6.11 steps/s | VRAM=32.05 Go [2026-07-18 18:40:16] step 932850/1000000 | loss=2.0191 vocab=0.9824 token=1.1232 | lr=1.51e-05 | 6.15 steps/s | VRAM=26.21 Go [2026-07-18 18:40:24] step 932900/1000000 | loss=2.0350 vocab=1.2910 token=1.0544 | lr=1.51e-05 | 6.15 steps/s | VRAM=32.56 Go [2026-07-18 18:40:32] step 932950/1000000 | loss=1.9723 vocab=0.7879 token=1.5041 | lr=1.51e-05 | 6.15 steps/s | VRAM=28.78 Go [2026-07-18 18:40:40] step 933000/1000000 | loss=2.0388 vocab=0.8054 token=1.3022 | lr=1.50e-05 | 6.16 steps/s | VRAM=31.14 Go [2026-07-18 18:40:48] step 933050/1000000 | loss=2.0078 vocab=0.6383 token=0.7067 | lr=1.50e-05 | 6.15 steps/s | VRAM=25.51 Go [2026-07-18 18:40:56] step 933100/1000000 | loss=2.0362 vocab=0.9518 token=1.3633 | lr=1.50e-05 | 6.10 steps/s | VRAM=29.05 Go [2026-07-18 18:41:05] step 933150/1000000 | loss=2.0811 vocab=0.6929 token=0.8160 | lr=1.50e-05 | 6.13 steps/s | VRAM=26.71 Go [2026-07-18 18:41:13] step 933200/1000000 | loss=2.0345 vocab=0.8993 token=1.3837 | lr=1.50e-05 | 6.13 steps/s | VRAM=26.60 Go [2026-07-18 18:41:21] step 933250/1000000 | loss=2.2228 vocab=1.1405 token=1.5533 | lr=1.49e-05 | 6.14 steps/s | VRAM=28.44 Go [2026-07-18 18:41:29] step 933300/1000000 | loss=1.8602 vocab=0.9734 token=1.1355 | lr=1.49e-05 | 6.14 steps/s | VRAM=29.55 Go [2026-07-18 18:41:37] step 933350/1000000 | loss=2.1487 vocab=0.6182 token=1.0433 | lr=1.49e-05 | 6.15 steps/s | VRAM=28.94 Go [2026-07-18 18:41:45] step 933400/1000000 | loss=2.1678 vocab=0.8840 token=1.2542 | lr=1.49e-05 | 6.13 steps/s | VRAM=26.59 Go [2026-07-18 18:41:54] step 933450/1000000 | loss=1.9533 vocab=0.6736 token=0.7477 | lr=1.49e-05 | 6.15 steps/s | VRAM=32.07 Go [2026-07-18 18:42:02] step 933500/1000000 | loss=1.9862 vocab=1.0942 token=1.3148 | lr=1.49e-05 | 6.15 steps/s | VRAM=28.95 Go [2026-07-18 18:42:10] step 933550/1000000 | loss=2.0396 vocab=0.9284 token=1.2926 | lr=1.48e-05 | 6.14 steps/s | VRAM=29.14 Go [2026-07-18 18:42:18] step 933600/1000000 | loss=2.1265 vocab=1.0723 token=1.3307 | lr=1.48e-05 | 6.16 steps/s | VRAM=25.43 Go [2026-07-18 18:42:26] step 933650/1000000 | loss=2.1247 vocab=0.8307 token=1.1859 | lr=1.48e-05 | 6.14 steps/s | VRAM=28.47 Go [2026-07-18 18:42:34] step 933700/1000000 | loss=1.9644 vocab=1.0129 token=0.8796 | lr=1.48e-05 | 6.14 steps/s | VRAM=27.41 Go [2026-07-18 18:42:42] step 933750/1000000 | loss=1.9465 vocab=0.7853 token=1.0272 | lr=1.48e-05 | 6.16 steps/s | VRAM=28.38 Go [2026-07-18 18:42:50] step 933800/1000000 | loss=1.9459 vocab=0.8305 token=1.2541 | lr=1.47e-05 | 6.15 steps/s | VRAM=28.82 Go [2026-07-18 18:42:59] step 933850/1000000 | loss=2.0147 vocab=0.9212 token=1.3416 | lr=1.47e-05 | 6.12 steps/s | VRAM=26.50 Go [2026-07-18 18:43:07] step 933900/1000000 | loss=1.9387 vocab=0.9721 token=1.5254 | lr=1.47e-05 | 6.14 steps/s | VRAM=32.85 Go [2026-07-18 18:43:15] step 933950/1000000 | loss=2.0526 vocab=0.9093 token=1.2905 | lr=1.47e-05 | 6.16 steps/s | VRAM=28.57 Go [2026-07-18 18:43:23] step 934000/1000000 | loss=2.0021 vocab=0.7216 token=1.0288 | lr=1.47e-05 | 6.10 steps/s | VRAM=26.20 Go [2026-07-18 18:43:31] step 934050/1000000 | loss=2.0668 vocab=1.1499 token=1.1598 | lr=1.46e-05 | 6.16 steps/s | VRAM=32.82 Go [2026-07-18 18:43:39] step 934100/1000000 | loss=2.0635 vocab=1.3079 token=1.3985 | lr=1.46e-05 | 6.13 steps/s | VRAM=32.07 Go [2026-07-18 18:43:47] step 934150/1000000 | loss=1.9696 vocab=1.1915 token=0.8371 | lr=1.46e-05 | 6.14 steps/s | VRAM=32.28 Go [2026-07-18 18:43:56] step 934200/1000000 | loss=2.0263 vocab=0.3971 token=0.7027 | lr=1.46e-05 | 6.12 steps/s | VRAM=25.73 Go [2026-07-18 18:44:04] step 934250/1000000 | loss=2.0588 vocab=0.9481 token=1.4050 | lr=1.46e-05 | 6.15 steps/s | VRAM=31.85 Go [2026-07-18 18:44:12] step 934300/1000000 | loss=2.0378 vocab=0.8317 token=0.9398 | lr=1.45e-05 | 6.16 steps/s | VRAM=28.98 Go [2026-07-18 18:44:20] step 934350/1000000 | loss=2.0260 vocab=0.5431 token=0.8876 | lr=1.45e-05 | 6.07 steps/s | VRAM=26.58 Go [2026-07-18 18:44:28] step 934400/1000000 | loss=1.9829 vocab=1.0699 token=1.3228 | lr=1.45e-05 | 6.12 steps/s | VRAM=28.64 Go [2026-07-18 18:44:36] step 934450/1000000 | loss=2.0045 vocab=0.7192 token=1.0986 | lr=1.45e-05 | 6.15 steps/s | VRAM=31.92 Go [2026-07-18 18:44:45] step 934500/1000000 | loss=2.1203 vocab=0.5100 token=0.9344 | lr=1.45e-05 | 6.12 steps/s | VRAM=29.42 Go [2026-07-18 18:44:53] step 934550/1000000 | loss=2.0068 vocab=0.8981 token=0.8179 | lr=1.44e-05 | 6.16 steps/s | VRAM=28.28 Go [2026-07-18 18:45:01] step 934600/1000000 | loss=2.0429 vocab=1.6257 token=1.3230 | lr=1.44e-05 | 6.17 steps/s | VRAM=32.19 Go [2026-07-18 18:45:09] step 934650/1000000 | loss=2.0905 vocab=1.3298 token=1.1215 | lr=1.44e-05 | 6.15 steps/s | VRAM=32.89 Go [2026-07-18 18:45:17] step 934700/1000000 | loss=2.0346 vocab=1.5551 token=1.6091 | lr=1.44e-05 | 6.15 steps/s | VRAM=27.13 Go [2026-07-18 18:45:25] step 934750/1000000 | loss=2.1104 vocab=1.2971 token=1.2336 | lr=1.44e-05 | 6.14 steps/s | VRAM=29.50 Go [2026-07-18 18:45:33] step 934800/1000000 | loss=2.0581 vocab=1.4511 token=1.4203 | lr=1.43e-05 | 6.15 steps/s | VRAM=24.90 Go [2026-07-18 18:45:42] step 934850/1000000 | loss=2.0593 vocab=0.5945 token=1.2863 | lr=1.43e-05 | 6.12 steps/s | VRAM=29.47 Go [2026-07-18 18:45:50] step 934900/1000000 | loss=1.9547 vocab=0.7064 token=1.2390 | lr=1.43e-05 | 6.13 steps/s | VRAM=28.69 Go [2026-07-18 18:45:58] step 934950/1000000 | loss=1.9899 vocab=0.8400 token=1.0978 | lr=1.43e-05 | 6.15 steps/s | VRAM=29.54 Go [2026-07-18 18:46:06] step 935000/1000000 | loss=2.1598 vocab=1.3346 token=1.5366 | lr=1.43e-05 | 6.10 steps/s | VRAM=25.79 Go [2026-07-18 18:46:14] step 935050/1000000 | loss=1.9667 vocab=0.9612 token=1.0757 | lr=1.42e-05 | 6.11 steps/s | VRAM=28.96 Go [2026-07-18 18:46:22] step 935100/1000000 | loss=2.2371 vocab=0.8467 token=1.4091 | lr=1.42e-05 | 6.16 steps/s | VRAM=31.72 Go [2026-07-18 18:46:30] step 935150/1000000 | loss=2.0085 vocab=0.7587 token=1.5793 | lr=1.42e-05 | 6.14 steps/s | VRAM=26.22 Go [2026-07-18 18:46:39] step 935200/1000000 | loss=2.0683 vocab=0.6563 token=1.4037 | lr=1.42e-05 | 6.11 steps/s | VRAM=26.32 Go [2026-07-18 18:46:47] step 935250/1000000 | loss=2.0413 vocab=0.9372 token=1.2646 | lr=1.42e-05 | 6.13 steps/s | VRAM=32.83 Go [2026-07-18 18:46:55] step 935300/1000000 | loss=2.0648 vocab=0.8698 token=1.5321 | lr=1.41e-05 | 6.11 steps/s | VRAM=29.39 Go [2026-07-18 18:47:03] step 935350/1000000 | loss=1.9197 vocab=0.3995 token=0.7365 | lr=1.41e-05 | 6.12 steps/s | VRAM=25.89 Go [2026-07-18 18:47:11] step 935400/1000000 | loss=2.0661 vocab=0.5666 token=1.0400 | lr=1.41e-05 | 6.17 steps/s | VRAM=32.21 Go [2026-07-18 18:47:19] step 935450/1000000 | loss=2.0210 vocab=0.9541 token=1.3975 | lr=1.41e-05 | 6.12 steps/s | VRAM=32.68 Go [2026-07-18 18:47:28] step 935500/1000000 | loss=2.0741 vocab=0.4927 token=1.2640 | lr=1.41e-05 | 6.10 steps/s | VRAM=28.46 Go [2026-07-18 18:47:36] step 935550/1000000 | loss=2.1226 vocab=1.0245 token=1.5619 | lr=1.40e-05 | 6.13 steps/s | VRAM=25.24 Go [2026-07-18 18:47:44] step 935600/1000000 | loss=2.0060 vocab=1.1526 token=1.5799 | lr=1.40e-05 | 6.15 steps/s | VRAM=28.22 Go [2026-07-18 18:47:52] step 935650/1000000 | loss=2.0749 vocab=1.5191 token=1.1312 | lr=1.40e-05 | 6.14 steps/s | VRAM=29.22 Go [2026-07-18 18:48:00] step 935700/1000000 | loss=2.0529 vocab=1.3672 token=1.5030 | lr=1.40e-05 | 6.13 steps/s | VRAM=28.93 Go [2026-07-18 18:48:08] step 935750/1000000 | loss=2.1077 vocab=1.3972 token=1.3895 | lr=1.40e-05 | 6.13 steps/s | VRAM=25.71 Go [2026-07-18 18:48:17] step 935800/1000000 | loss=2.0548 vocab=1.0695 token=1.0568 | lr=1.40e-05 | 6.14 steps/s | VRAM=28.28 Go [2026-07-18 18:48:25] step 935850/1000000 | loss=2.0064 vocab=1.0010 token=1.3603 | lr=1.39e-05 | 6.15 steps/s | VRAM=28.64 Go [2026-07-18 18:48:33] step 935900/1000000 | loss=2.1833 vocab=0.6621 token=1.1350 | lr=1.39e-05 | 6.16 steps/s | VRAM=25.65 Go [2026-07-18 18:48:41] step 935950/1000000 | loss=1.9950 vocab=0.8413 token=1.3495 | lr=1.39e-05 | 6.16 steps/s | VRAM=28.96 Go [2026-07-18 18:48:49] step 936000/1000000 | loss=1.9956 vocab=0.6365 token=1.3911 | lr=1.39e-05 | 6.16 steps/s | VRAM=26.14 Go [2026-07-18 18:48:57] step 936050/1000000 | loss=2.0113 vocab=0.6237 token=1.2332 | lr=1.39e-05 | 6.13 steps/s | VRAM=29.50 Go [2026-07-18 18:49:05] step 936100/1000000 | loss=1.9430 vocab=1.7996 token=1.0143 | lr=1.38e-05 | 6.16 steps/s | VRAM=32.89 Go [2026-07-18 18:49:13] step 936150/1000000 | loss=1.9644 vocab=0.5662 token=1.2437 | lr=1.38e-05 | 6.13 steps/s | VRAM=28.50 Go [2026-07-18 18:49:22] step 936200/1000000 | loss=1.9829 vocab=0.7907 token=1.2623 | lr=1.38e-05 | 6.12 steps/s | VRAM=26.43 Go [2026-07-18 18:49:30] step 936250/1000000 | loss=2.0326 vocab=0.4644 token=1.1020 | lr=1.38e-05 | 6.14 steps/s | VRAM=25.61 Go [2026-07-18 18:49:38] step 936300/1000000 | loss=1.9916 vocab=0.9816 token=1.2140 | lr=1.38e-05 | 6.18 steps/s | VRAM=31.84 Go [2026-07-18 18:49:46] step 936350/1000000 | loss=2.0000 vocab=0.9588 token=1.1440 | lr=1.37e-05 | 6.13 steps/s | VRAM=26.70 Go [2026-07-18 18:49:54] step 936400/1000000 | loss=1.9691 vocab=0.6047 token=1.1225 | lr=1.37e-05 | 6.15 steps/s | VRAM=28.40 Go [2026-07-18 18:50:02] step 936450/1000000 | loss=1.9968 vocab=0.5579 token=1.1791 | lr=1.37e-05 | 6.16 steps/s | VRAM=24.81 Go [2026-07-18 18:50:10] step 936500/1000000 | loss=2.0487 vocab=0.6848 token=1.5482 | lr=1.37e-05 | 6.16 steps/s | VRAM=28.93 Go [2026-07-18 18:50:18] step 936550/1000000 | loss=1.9378 vocab=0.5498 token=0.9291 | lr=1.37e-05 | 6.15 steps/s | VRAM=29.43 Go [2026-07-18 18:50:27] step 936600/1000000 | loss=2.0325 vocab=1.1790 token=1.0213 | lr=1.36e-05 | 6.16 steps/s | VRAM=32.39 Go [2026-07-18 18:50:35] step 936650/1000000 | loss=2.0168 vocab=1.1480 token=1.3183 | lr=1.36e-05 | 6.13 steps/s | VRAM=29.30 Go [2026-07-18 18:50:43] step 936700/1000000 | loss=1.9696 vocab=0.6838 token=1.1877 | lr=1.36e-05 | 6.15 steps/s | VRAM=28.78 Go [2026-07-18 18:50:51] step 936750/1000000 | loss=1.9446 vocab=0.5531 token=0.7053 | lr=1.36e-05 | 6.14 steps/s | VRAM=29.15 Go [2026-07-18 18:50:59] step 936800/1000000 | loss=1.9544 vocab=0.3699 token=0.6497 | lr=1.36e-05 | 6.11 steps/s | VRAM=28.24 Go [2026-07-18 18:51:07] step 936850/1000000 | loss=1.8777 vocab=0.3575 token=0.9213 | lr=1.36e-05 | 6.15 steps/s | VRAM=27.99 Go [2026-07-18 18:51:15] step 936900/1000000 | loss=2.0434 vocab=1.0168 token=1.3680 | lr=1.35e-05 | 6.14 steps/s | VRAM=29.55 Go [2026-07-18 18:51:24] step 936950/1000000 | loss=2.0747 vocab=1.0372 token=1.3974 | lr=1.35e-05 | 6.15 steps/s | VRAM=25.51 Go [2026-07-18 18:51:32] step 937000/1000000 | loss=2.0965 vocab=0.7496 token=0.7645 | lr=1.35e-05 | 6.14 steps/s | VRAM=28.82 Go [2026-07-18 18:51:40] step 937050/1000000 | loss=1.9849 vocab=0.9780 token=1.3950 | lr=1.35e-05 | 6.14 steps/s | VRAM=29.86 Go [2026-07-18 18:51:48] step 937100/1000000 | loss=2.0290 vocab=0.8322 token=0.9692 | lr=1.35e-05 | 6.13 steps/s | VRAM=24.96 Go [2026-07-18 18:51:56] step 937150/1000000 | loss=2.0627 vocab=0.8287 token=1.1810 | lr=1.34e-05 | 6.19 steps/s | VRAM=32.57 Go [2026-07-18 18:52:04] step 937200/1000000 | loss=2.0530 vocab=0.5754 token=1.2262 | lr=1.34e-05 | 6.17 steps/s | VRAM=31.16 Go [2026-07-18 18:52:12] step 937250/1000000 | loss=2.0048 vocab=1.2255 token=1.0275 | lr=1.34e-05 | 6.16 steps/s | VRAM=29.92 Go [2026-07-18 18:52:20] step 937300/1000000 | loss=1.9800 vocab=0.7540 token=0.9243 | lr=1.34e-05 | 6.16 steps/s | VRAM=28.54 Go [2026-07-18 18:52:29] step 937350/1000000 | loss=2.1000 vocab=0.4211 token=0.8805 | lr=1.34e-05 | 6.14 steps/s | VRAM=25.46 Go [2026-07-18 18:52:37] step 937400/1000000 | loss=2.0607 vocab=0.6597 token=1.4201 | lr=1.33e-05 | 6.14 steps/s | VRAM=32.67 Go [2026-07-18 18:52:45] step 937450/1000000 | loss=2.0853 vocab=1.2432 token=1.5667 | lr=1.33e-05 | 6.15 steps/s | VRAM=25.98 Go [2026-07-18 18:52:53] step 937500/1000000 | loss=1.9180 vocab=0.8372 token=0.8101 | lr=1.33e-05 | 6.15 steps/s | VRAM=25.46 Go [2026-07-18 18:53:01] step 937550/1000000 | loss=2.1178 vocab=0.8596 token=1.3350 | lr=1.33e-05 | 6.16 steps/s | VRAM=28.76 Go [2026-07-18 18:53:09] step 937600/1000000 | loss=2.2048 vocab=0.9560 token=1.1456 | lr=1.33e-05 | 6.14 steps/s | VRAM=26.32 Go [2026-07-18 18:53:17] step 937650/1000000 | loss=2.1348 vocab=1.0166 token=1.2502 | lr=1.32e-05 | 6.17 steps/s | VRAM=32.58 Go [2026-07-18 18:53:25] step 937700/1000000 | loss=1.9947 vocab=1.2004 token=1.3276 | lr=1.32e-05 | 6.15 steps/s | VRAM=29.33 Go [2026-07-18 18:53:34] step 937750/1000000 | loss=2.0137 vocab=0.6602 token=1.0299 | lr=1.32e-05 | 6.19 steps/s | VRAM=31.55 Go [2026-07-18 18:53:42] step 937800/1000000 | loss=2.0365 vocab=0.8655 token=1.1860 | lr=1.32e-05 | 6.15 steps/s | VRAM=32.38 Go [2026-07-18 18:53:50] step 937850/1000000 | loss=2.0448 vocab=0.6170 token=1.1936 | lr=1.32e-05 | 6.13 steps/s | VRAM=26.79 Go [2026-07-18 18:53:58] step 937900/1000000 | loss=1.9486 vocab=1.0387 token=1.5010 | lr=1.32e-05 | 6.15 steps/s | VRAM=28.72 Go [2026-07-18 18:54:06] step 937950/1000000 | loss=1.9925 vocab=0.7782 token=1.1710 | lr=1.31e-05 | 6.13 steps/s | VRAM=29.12 Go [2026-07-18 18:54:14] step 938000/1000000 | loss=1.9820 vocab=0.8155 token=1.0102 | lr=1.31e-05 | 6.12 steps/s | VRAM=28.65 Go [2026-07-18 18:54:22] step 938050/1000000 | loss=2.0012 vocab=0.6671 token=0.9100 | lr=1.31e-05 | 6.11 steps/s | VRAM=24.72 Go [2026-07-18 18:54:31] step 938100/1000000 | loss=2.0471 vocab=0.6109 token=1.5158 | lr=1.31e-05 | 6.12 steps/s | VRAM=25.94 Go [2026-07-18 18:54:39] step 938150/1000000 | loss=2.0912 vocab=1.0961 token=1.1696 | lr=1.31e-05 | 6.17 steps/s | VRAM=24.84 Go [2026-07-18 18:54:47] step 938200/1000000 | loss=2.0093 vocab=0.5267 token=0.7987 | lr=1.30e-05 | 6.15 steps/s | VRAM=32.42 Go [2026-07-18 18:54:55] step 938250/1000000 | loss=2.0933 vocab=0.4929 token=1.1217 | lr=1.30e-05 | 6.14 steps/s | VRAM=31.65 Go [2026-07-18 18:55:03] step 938300/1000000 | loss=2.0764 vocab=1.2749 token=1.5538 | lr=1.30e-05 | 6.15 steps/s | VRAM=32.67 Go [2026-07-18 18:55:11] step 938350/1000000 | loss=1.9630 vocab=0.7188 token=0.9682 | lr=1.30e-05 | 6.16 steps/s | VRAM=31.61 Go [2026-07-18 18:55:19] step 938400/1000000 | loss=1.9325 vocab=0.4511 token=1.0537 | lr=1.30e-05 | 6.14 steps/s | VRAM=26.01 Go [2026-07-18 18:55:28] step 938450/1000000 | loss=2.0616 vocab=0.8220 token=1.0697 | lr=1.29e-05 | 6.14 steps/s | VRAM=28.11 Go [2026-07-18 18:55:36] step 938500/1000000 | loss=2.0531 vocab=1.4975 token=1.4029 | lr=1.29e-05 | 6.12 steps/s | VRAM=26.02 Go [2026-07-18 18:55:44] step 938550/1000000 | loss=2.0164 vocab=0.8836 token=1.0166 | lr=1.29e-05 | 6.11 steps/s | VRAM=26.15 Go [2026-07-18 18:55:52] step 938600/1000000 | loss=2.1800 vocab=0.9759 token=1.4465 | lr=1.29e-05 | 6.11 steps/s | VRAM=28.36 Go [2026-07-18 18:56:00] step 938650/1000000 | loss=2.0995 vocab=0.7298 token=1.0862 | lr=1.29e-05 | 6.14 steps/s | VRAM=28.49 Go [2026-07-18 18:56:08] step 938700/1000000 | loss=2.0393 vocab=0.4884 token=1.1702 | lr=1.29e-05 | 6.17 steps/s | VRAM=32.52 Go [2026-07-18 18:56:17] step 938750/1000000 | loss=2.1066 vocab=0.4647 token=0.5861 | lr=1.28e-05 | 6.13 steps/s | VRAM=32.31 Go [2026-07-18 18:56:25] step 938800/1000000 | loss=2.0409 vocab=0.8066 token=0.9707 | lr=1.28e-05 | 6.12 steps/s | VRAM=26.39 Go [2026-07-18 18:56:33] step 938850/1000000 | loss=2.0402 vocab=0.8713 token=1.0135 | lr=1.28e-05 | 6.16 steps/s | VRAM=31.22 Go [2026-07-18 18:56:41] step 938900/1000000 | loss=2.0517 vocab=0.9859 token=1.3274 | lr=1.28e-05 | 6.13 steps/s | VRAM=28.33 Go [2026-07-18 18:56:49] step 938950/1000000 | loss=1.9759 vocab=0.6032 token=0.9492 | lr=1.28e-05 | 6.15 steps/s | VRAM=32.29 Go [2026-07-18 18:56:57] step 939000/1000000 | loss=2.0820 vocab=0.8700 token=1.4413 | lr=1.27e-05 | 6.11 steps/s | VRAM=28.05 Go [2026-07-18 18:57:05] step 939050/1000000 | loss=1.9182 vocab=0.8649 token=1.1305 | lr=1.27e-05 | 6.13 steps/s | VRAM=27.34 Go [2026-07-18 18:57:14] step 939100/1000000 | loss=2.0863 vocab=0.6640 token=1.3200 | lr=1.27e-05 | 6.16 steps/s | VRAM=32.14 Go [2026-07-18 18:57:22] step 939150/1000000 | loss=1.9942 vocab=0.8863 token=0.7708 | lr=1.27e-05 | 6.14 steps/s | VRAM=32.28 Go [2026-07-18 18:57:30] step 939200/1000000 | loss=2.0272 vocab=0.9710 token=0.8761 | lr=1.27e-05 | 6.15 steps/s | VRAM=28.86 Go [2026-07-18 18:57:38] step 939250/1000000 | loss=2.0270 vocab=0.5810 token=0.9373 | lr=1.27e-05 | 6.10 steps/s | VRAM=29.63 Go [2026-07-18 18:57:46] step 939300/1000000 | loss=2.1423 vocab=0.3091 token=0.9734 | lr=1.26e-05 | 6.14 steps/s | VRAM=28.85 Go [2026-07-18 18:57:54] step 939350/1000000 | loss=2.0891 vocab=1.0475 token=1.5295 | lr=1.26e-05 | 6.13 steps/s | VRAM=29.07 Go [2026-07-18 18:58:02] step 939400/1000000 | loss=1.9594 vocab=0.6773 token=1.3269 | lr=1.26e-05 | 6.14 steps/s | VRAM=25.67 Go [2026-07-18 18:58:11] step 939450/1000000 | loss=2.0830 vocab=0.6529 token=0.9148 | lr=1.26e-05 | 6.14 steps/s | VRAM=28.59 Go [2026-07-18 18:58:19] step 939500/1000000 | loss=2.0766 vocab=1.0526 token=1.3729 | lr=1.26e-05 | 6.09 steps/s | VRAM=27.43 Go [2026-07-18 18:58:27] step 939550/1000000 | loss=1.9786 vocab=1.0593 token=1.0735 | lr=1.25e-05 | 6.15 steps/s | VRAM=28.04 Go [2026-07-18 18:58:35] step 939600/1000000 | loss=2.1237 vocab=0.6873 token=0.7257 | lr=1.25e-05 | 6.11 steps/s | VRAM=26.81 Go [2026-07-18 18:58:43] step 939650/1000000 | loss=2.0320 vocab=0.7751 token=1.0032 | lr=1.25e-05 | 6.12 steps/s | VRAM=24.53 Go [2026-07-18 18:58:51] step 939700/1000000 | loss=2.0667 vocab=0.9611 token=1.4051 | lr=1.25e-05 | 6.15 steps/s | VRAM=28.89 Go [2026-07-18 18:59:00] step 939750/1000000 | loss=2.0865 vocab=0.8569 token=1.2522 | lr=1.25e-05 | 6.15 steps/s | VRAM=33.01 Go [2026-07-18 18:59:08] step 939800/1000000 | loss=2.0213 vocab=1.6334 token=1.5359 | lr=1.25e-05 | 6.16 steps/s | VRAM=25.15 Go [2026-07-18 18:59:16] step 939850/1000000 | loss=2.0591 vocab=1.9379 token=1.0735 | lr=1.24e-05 | 6.16 steps/s | VRAM=26.22 Go [2026-07-18 18:59:24] step 939900/1000000 | loss=2.1445 vocab=0.6386 token=1.5492 | lr=1.24e-05 | 6.15 steps/s | VRAM=31.97 Go [2026-07-18 18:59:32] step 939950/1000000 | loss=2.1462 vocab=1.0908 token=1.3935 | lr=1.24e-05 | 6.13 steps/s | VRAM=30.08 Go [2026-07-18 18:59:40] step 940000/1000000 | loss=1.9061 vocab=0.5874 token=1.3593 | lr=1.24e-05 | 6.16 steps/s | VRAM=28.10 Go [2026-07-18 18:59:48] step 940050/1000000 | loss=2.1107 vocab=0.9160 token=1.2693 | lr=1.24e-05 | 6.13 steps/s | VRAM=26.04 Go [2026-07-18 18:59:57] step 940100/1000000 | loss=1.9922 vocab=0.6458 token=0.8765 | lr=1.23e-05 | 6.13 steps/s | VRAM=28.97 Go [2026-07-18 19:00:05] step 940150/1000000 | loss=2.0333 vocab=1.4201 token=0.7009 | lr=1.23e-05 | 6.15 steps/s | VRAM=31.73 Go [2026-07-18 19:00:13] step 940200/1000000 | loss=2.0681 vocab=0.5654 token=0.8480 | lr=1.23e-05 | 6.14 steps/s | VRAM=32.76 Go [2026-07-18 19:00:21] step 940250/1000000 | loss=2.1097 vocab=0.6851 token=1.2608 | lr=1.23e-05 | 6.12 steps/s | VRAM=31.42 Go [2026-07-18 19:00:29] step 940300/1000000 | loss=2.0906 vocab=0.9136 token=1.5649 | lr=1.23e-05 | 6.15 steps/s | VRAM=25.66 Go [2026-07-18 19:00:37] step 940350/1000000 | loss=1.8713 vocab=0.9393 token=0.9318 | lr=1.23e-05 | 6.17 steps/s | VRAM=25.26 Go [2026-07-18 19:00:45] step 940400/1000000 | loss=2.0623 vocab=0.6926 token=1.1432 | lr=1.22e-05 | 6.16 steps/s | VRAM=31.57 Go [2026-07-18 19:00:53] step 940450/1000000 | loss=1.9939 vocab=0.6959 token=0.8812 | lr=1.22e-05 | 6.17 steps/s | VRAM=29.44 Go [2026-07-18 19:01:02] step 940500/1000000 | loss=2.0157 vocab=0.9880 token=1.7794 | lr=1.22e-05 | 6.16 steps/s | VRAM=28.31 Go [2026-07-18 19:01:10] step 940550/1000000 | loss=2.0548 vocab=1.0650 token=1.3290 | lr=1.22e-05 | 6.13 steps/s | VRAM=25.65 Go [2026-07-18 19:01:18] step 940600/1000000 | loss=2.0185 vocab=1.4503 token=1.7480 | lr=1.22e-05 | 6.13 steps/s | VRAM=32.75 Go [2026-07-18 19:01:26] step 940650/1000000 | loss=2.0347 vocab=0.6384 token=0.9189 | lr=1.21e-05 | 6.13 steps/s | VRAM=29.74 Go [2026-07-18 19:01:34] step 940700/1000000 | loss=1.9745 vocab=0.7559 token=1.2111 | lr=1.21e-05 | 6.14 steps/s | VRAM=32.13 Go [2026-07-18 19:01:42] step 940750/1000000 | loss=2.0438 vocab=0.9670 token=1.2515 | lr=1.21e-05 | 6.14 steps/s | VRAM=31.92 Go [2026-07-18 19:01:50] step 940800/1000000 | loss=1.9909 vocab=1.2601 token=1.0596 | lr=1.21e-05 | 6.14 steps/s | VRAM=28.17 Go [2026-07-18 19:01:59] step 940850/1000000 | loss=2.0240 vocab=0.8963 token=1.3198 | lr=1.21e-05 | 6.14 steps/s | VRAM=29.05 Go [2026-07-18 19:02:07] step 940900/1000000 | loss=2.0376 vocab=1.0059 token=1.2830 | lr=1.21e-05 | 6.14 steps/s | VRAM=27.64 Go [2026-07-18 19:02:15] step 940950/1000000 | loss=2.0228 vocab=0.8346 token=0.8030 | lr=1.20e-05 | 6.12 steps/s | VRAM=30.26 Go [2026-07-18 19:02:23] step 941000/1000000 | loss=2.0082 vocab=0.8153 token=1.0311 | lr=1.20e-05 | 6.15 steps/s | VRAM=31.58 Go [2026-07-18 19:02:31] step 941050/1000000 | loss=2.1643 vocab=1.2515 token=1.2327 | lr=1.20e-05 | 6.14 steps/s | VRAM=25.01 Go [2026-07-18 19:02:39] step 941100/1000000 | loss=1.9440 vocab=0.6370 token=0.8786 | lr=1.20e-05 | 6.18 steps/s | VRAM=32.02 Go [2026-07-18 19:02:47] step 941150/1000000 | loss=2.0713 vocab=0.7964 token=1.0544 | lr=1.20e-05 | 6.14 steps/s | VRAM=29.05 Go [2026-07-18 19:02:56] step 941200/1000000 | loss=2.0300 vocab=0.8583 token=1.2420 | lr=1.19e-05 | 6.15 steps/s | VRAM=29.49 Go [2026-07-18 19:03:04] step 941250/1000000 | loss=1.9809 vocab=0.7419 token=1.3780 | lr=1.19e-05 | 6.13 steps/s | VRAM=26.40 Go [2026-07-18 19:03:12] step 941300/1000000 | loss=2.0397 vocab=0.6625 token=1.2091 | lr=1.19e-05 | 6.12 steps/s | VRAM=27.22 Go [2026-07-18 19:03:20] step 941350/1000000 | loss=2.1011 vocab=0.7082 token=1.1192 | lr=1.19e-05 | 6.17 steps/s | VRAM=31.31 Go [2026-07-18 19:03:28] step 941400/1000000 | loss=1.9058 vocab=0.7102 token=1.1068 | lr=1.19e-05 | 6.14 steps/s | VRAM=25.80 Go [2026-07-18 19:03:36] step 941450/1000000 | loss=2.0847 vocab=0.6201 token=1.3035 | lr=1.19e-05 | 6.12 steps/s | VRAM=29.37 Go [2026-07-18 19:03:44] step 941500/1000000 | loss=1.8737 vocab=0.6604 token=1.1531 | lr=1.18e-05 | 6.14 steps/s | VRAM=29.37 Go [2026-07-18 19:03:53] step 941550/1000000 | loss=2.1016 vocab=0.8800 token=1.1784 | lr=1.18e-05 | 6.13 steps/s | VRAM=28.92 Go [2026-07-18 19:04:01] step 941600/1000000 | loss=1.9531 vocab=0.4626 token=1.1111 | lr=1.18e-05 | 6.10 steps/s | VRAM=28.71 Go [2026-07-18 19:04:09] step 941650/1000000 | loss=1.9522 vocab=1.0692 token=1.0385 | lr=1.18e-05 | 6.09 steps/s | VRAM=26.55 Go [2026-07-18 19:04:17] step 941700/1000000 | loss=2.0939 vocab=0.8672 token=1.1252 | lr=1.18e-05 | 6.12 steps/s | VRAM=24.80 Go [2026-07-18 19:04:25] step 941750/1000000 | loss=2.0548 vocab=0.7113 token=1.3638 | lr=1.17e-05 | 6.13 steps/s | VRAM=29.78 Go [2026-07-18 19:04:33] step 941800/1000000 | loss=1.9971 vocab=0.4114 token=0.9901 | lr=1.17e-05 | 6.16 steps/s | VRAM=29.07 Go [2026-07-18 19:04:42] step 941850/1000000 | loss=1.9910 vocab=0.5887 token=1.6400 | lr=1.17e-05 | 6.14 steps/s | VRAM=28.65 Go [2026-07-18 19:04:50] step 941900/1000000 | loss=2.0162 vocab=0.7838 token=0.8979 | lr=1.17e-05 | 6.13 steps/s | VRAM=32.10 Go [2026-07-18 19:04:58] step 941950/1000000 | loss=2.0898 vocab=1.6154 token=1.2353 | lr=1.17e-05 | 6.15 steps/s | VRAM=28.60 Go [2026-07-18 19:05:06] step 942000/1000000 | loss=1.9957 vocab=0.6309 token=0.7753 | lr=1.17e-05 | 6.13 steps/s | VRAM=28.97 Go [2026-07-18 19:05:14] step 942050/1000000 | loss=2.0313 vocab=0.6003 token=0.7797 | lr=1.16e-05 | 6.08 steps/s | VRAM=26.44 Go [2026-07-18 19:05:22] step 942100/1000000 | loss=1.9667 vocab=1.0594 token=1.3648 | lr=1.16e-05 | 6.14 steps/s | VRAM=27.86 Go [2026-07-18 19:05:31] step 942150/1000000 | loss=2.0498 vocab=0.7845 token=1.4219 | lr=1.16e-05 | 6.14 steps/s | VRAM=31.34 Go [2026-07-18 19:05:39] step 942200/1000000 | loss=2.1604 vocab=0.9566 token=1.5305 | lr=1.16e-05 | 6.12 steps/s | VRAM=25.57 Go [2026-07-18 19:05:47] step 942250/1000000 | loss=2.0696 vocab=0.5210 token=0.5801 | lr=1.16e-05 | 6.14 steps/s | VRAM=32.27 Go [2026-07-18 19:05:55] step 942300/1000000 | loss=1.9976 vocab=0.6688 token=1.2762 | lr=1.16e-05 | 6.12 steps/s | VRAM=29.92 Go [2026-07-18 19:06:03] step 942350/1000000 | loss=1.9447 vocab=0.5691 token=0.9132 | lr=1.15e-05 | 6.10 steps/s | VRAM=25.73 Go [2026-07-18 19:06:11] step 942400/1000000 | loss=1.9728 vocab=1.1367 token=1.2754 | lr=1.15e-05 | 6.15 steps/s | VRAM=28.83 Go [2026-07-18 19:06:20] step 942450/1000000 | loss=2.0120 vocab=1.3102 token=1.3585 | lr=1.15e-05 | 6.12 steps/s | VRAM=25.34 Go [2026-07-18 19:06:28] step 942500/1000000 | loss=2.0593 vocab=0.7474 token=1.0640 | lr=1.15e-05 | 6.14 steps/s | VRAM=27.16 Go [2026-07-18 19:06:36] step 942550/1000000 | loss=1.9982 vocab=1.2396 token=1.2587 | lr=1.15e-05 | 6.13 steps/s | VRAM=25.03 Go [2026-07-18 19:06:44] step 942600/1000000 | loss=1.9197 vocab=0.6684 token=1.3760 | lr=1.14e-05 | 6.12 steps/s | VRAM=29.30 Go [2026-07-18 19:06:52] step 942650/1000000 | loss=2.0741 vocab=1.0392 token=0.9366 | lr=1.14e-05 | 6.07 steps/s | VRAM=27.46 Go [2026-07-18 19:07:00] step 942700/1000000 | loss=1.9223 vocab=0.4015 token=0.8303 | lr=1.14e-05 | 6.16 steps/s | VRAM=31.81 Go [2026-07-18 19:07:09] step 942750/1000000 | loss=1.9824 vocab=0.8348 token=0.9443 | lr=1.14e-05 | 6.13 steps/s | VRAM=28.70 Go [2026-07-18 19:07:17] step 942800/1000000 | loss=2.0153 vocab=0.9508 token=1.0628 | lr=1.14e-05 | 6.14 steps/s | VRAM=32.64 Go [2026-07-18 19:07:25] step 942850/1000000 | loss=2.0624 vocab=0.9387 token=1.3562 | lr=1.14e-05 | 6.09 steps/s | VRAM=29.29 Go [2026-07-18 19:07:33] step 942900/1000000 | loss=2.0745 vocab=1.0480 token=1.1941 | lr=1.13e-05 | 6.13 steps/s | VRAM=24.69 Go [2026-07-18 19:07:41] step 942950/1000000 | loss=2.0886 vocab=0.8307 token=0.9431 | lr=1.13e-05 | 6.14 steps/s | VRAM=27.62 Go [2026-07-18 19:07:49] step 943000/1000000 | loss=2.0962 vocab=0.7856 token=1.4159 | lr=1.13e-05 | 6.11 steps/s | VRAM=26.38 Go [2026-07-18 19:07:58] step 943050/1000000 | loss=2.0514 vocab=1.3338 token=1.3706 | lr=1.13e-05 | 6.12 steps/s | VRAM=26.92 Go [2026-07-18 19:08:06] step 943100/1000000 | loss=1.9448 vocab=0.8668 token=1.0020 | lr=1.13e-05 | 6.12 steps/s | VRAM=29.12 Go [2026-07-18 19:08:14] step 943150/1000000 | loss=2.0583 vocab=1.2171 token=1.5589 | lr=1.13e-05 | 6.13 steps/s | VRAM=29.59 Go [2026-07-18 19:08:22] step 943200/1000000 | loss=1.9536 vocab=0.5163 token=1.4114 | lr=1.12e-05 | 6.10 steps/s | VRAM=27.79 Go [2026-07-18 19:08:30] step 943250/1000000 | loss=1.9991 vocab=0.4309 token=1.0656 | lr=1.12e-05 | 6.15 steps/s | VRAM=31.55 Go [2026-07-18 19:08:38] step 943300/1000000 | loss=2.1169 vocab=1.0646 token=1.5566 | lr=1.12e-05 | 6.15 steps/s | VRAM=26.32 Go [2026-07-18 19:08:46] step 943350/1000000 | loss=2.0586 vocab=1.2874 token=1.5310 | lr=1.12e-05 | 6.14 steps/s | VRAM=29.53 Go [2026-07-18 19:08:55] step 943400/1000000 | loss=1.9842 vocab=0.7716 token=1.2413 | lr=1.12e-05 | 6.15 steps/s | VRAM=27.35 Go [2026-07-18 19:09:03] step 943450/1000000 | loss=2.1527 vocab=0.7496 token=1.3021 | lr=1.12e-05 | 6.12 steps/s | VRAM=26.46 Go [2026-07-18 19:09:11] step 943500/1000000 | loss=2.1385 vocab=0.9994 token=1.3363 | lr=1.11e-05 | 6.15 steps/s | VRAM=31.54 Go [2026-07-18 19:09:19] step 943550/1000000 | loss=2.1178 vocab=0.8521 token=1.1176 | lr=1.11e-05 | 6.16 steps/s | VRAM=32.60 Go [2026-07-18 19:09:27] step 943600/1000000 | loss=1.9797 vocab=0.7978 token=1.0806 | lr=1.11e-05 | 6.13 steps/s | VRAM=25.36 Go [2026-07-18 19:09:35] step 943650/1000000 | loss=2.0363 vocab=1.1338 token=1.0938 | lr=1.11e-05 | 6.12 steps/s | VRAM=31.85 Go [2026-07-18 19:09:44] step 943700/1000000 | loss=2.1229 vocab=1.2437 token=1.2185 | lr=1.11e-05 | 6.08 steps/s | VRAM=30.49 Go [2026-07-18 19:09:52] step 943750/1000000 | loss=2.0179 vocab=0.5688 token=0.9155 | lr=1.10e-05 | 6.11 steps/s | VRAM=28.60 Go [2026-07-18 19:10:00] step 943800/1000000 | loss=2.0583 vocab=0.8861 token=1.2521 | lr=1.10e-05 | 6.08 steps/s | VRAM=30.66 Go [2026-07-18 19:10:08] step 943850/1000000 | loss=2.0095 vocab=0.9001 token=1.0409 | lr=1.10e-05 | 6.14 steps/s | VRAM=31.23 Go [2026-07-18 19:10:16] step 943900/1000000 | loss=2.0261 vocab=0.5851 token=1.2015 | lr=1.10e-05 | 6.14 steps/s | VRAM=28.49 Go [2026-07-18 19:10:24] step 943950/1000000 | loss=1.9717 vocab=1.1812 token=1.3266 | lr=1.10e-05 | 6.14 steps/s | VRAM=31.81 Go [2026-07-18 19:10:33] step 944000/1000000 | loss=2.0520 vocab=0.9289 token=1.0848 | lr=1.10e-05 | 6.13 steps/s | VRAM=28.71 Go [2026-07-18 19:10:41] step 944050/1000000 | loss=2.1114 vocab=0.4123 token=0.8791 | lr=1.09e-05 | 6.13 steps/s | VRAM=32.74 Go [2026-07-18 19:10:49] step 944100/1000000 | loss=1.9506 vocab=1.1078 token=1.3857 | lr=1.09e-05 | 6.10 steps/s | VRAM=26.15 Go [2026-07-18 19:10:57] step 944150/1000000 | loss=1.9902 vocab=0.7277 token=1.1763 | lr=1.09e-05 | 6.12 steps/s | VRAM=27.96 Go [2026-07-18 19:11:05] step 944200/1000000 | loss=2.0932 vocab=0.7501 token=1.0932 | lr=1.09e-05 | 6.15 steps/s | VRAM=31.76 Go [2026-07-18 19:11:13] step 944250/1000000 | loss=1.9581 vocab=0.6673 token=1.3306 | lr=1.09e-05 | 6.09 steps/s | VRAM=24.96 Go [2026-07-18 19:11:22] step 944300/1000000 | loss=1.8917 vocab=0.2925 token=0.6919 | lr=1.09e-05 | 6.10 steps/s | VRAM=27.61 Go [2026-07-18 19:11:30] step 944350/1000000 | loss=1.9647 vocab=0.7978 token=0.9177 | lr=1.08e-05 | 6.12 steps/s | VRAM=28.35 Go [2026-07-18 19:11:38] step 944400/1000000 | loss=2.1340 vocab=0.9287 token=1.5110 | lr=1.08e-05 | 6.12 steps/s | VRAM=29.19 Go [2026-07-18 19:11:46] step 944450/1000000 | loss=1.9350 vocab=0.9000 token=1.5024 | lr=1.08e-05 | 6.15 steps/s | VRAM=27.50 Go [2026-07-18 19:11:54] step 944500/1000000 | loss=2.1057 vocab=0.7041 token=0.8846 | lr=1.08e-05 | 6.14 steps/s | VRAM=28.65 Go [2026-07-18 19:12:02] step 944550/1000000 | loss=2.0378 vocab=0.8683 token=0.9666 | lr=1.08e-05 | 6.12 steps/s | VRAM=26.25 Go [2026-07-18 19:12:11] step 944600/1000000 | loss=2.0742 vocab=0.6769 token=0.8059 | lr=1.08e-05 | 6.12 steps/s | VRAM=28.54 Go [2026-07-18 19:12:19] step 944650/1000000 | loss=2.1687 vocab=0.5866 token=0.8950 | lr=1.07e-05 | 6.14 steps/s | VRAM=28.41 Go [2026-07-18 19:12:27] step 944700/1000000 | loss=2.0796 vocab=1.2590 token=1.0108 | lr=1.07e-05 | 6.13 steps/s | VRAM=26.53 Go [2026-07-18 19:12:35] step 944750/1000000 | loss=2.0380 vocab=0.5329 token=1.1919 | lr=1.07e-05 | 6.10 steps/s | VRAM=28.93 Go [2026-07-18 19:12:43] step 944800/1000000 | loss=2.0072 vocab=1.2888 token=1.4920 | lr=1.07e-05 | 6.12 steps/s | VRAM=28.86 Go [2026-07-18 19:12:51] step 944850/1000000 | loss=2.1300 vocab=0.7927 token=1.2011 | lr=1.07e-05 | 6.10 steps/s | VRAM=26.55 Go [2026-07-18 19:13:00] step 944900/1000000 | loss=1.9292 vocab=0.5994 token=0.6726 | lr=1.07e-05 | 6.14 steps/s | VRAM=24.82 Go [2026-07-18 19:13:08] step 944950/1000000 | loss=2.0853 vocab=0.7979 token=0.9584 | lr=1.06e-05 | 6.17 steps/s | VRAM=32.48 Go [2026-07-18 19:13:16] step 945000/1000000 | loss=2.0381 vocab=0.9141 token=1.2307 | lr=1.06e-05 | 6.10 steps/s | VRAM=32.01 Go [2026-07-18 19:13:24] step 945050/1000000 | loss=1.9908 vocab=1.3757 token=1.5360 | lr=1.06e-05 | 6.14 steps/s | VRAM=24.45 Go [2026-07-18 19:13:32] step 945100/1000000 | loss=1.9931 vocab=0.5793 token=0.9229 | lr=1.06e-05 | 6.13 steps/s | VRAM=30.42 Go [2026-07-18 19:13:40] step 945150/1000000 | loss=2.0823 vocab=0.3592 token=0.8062 | lr=1.06e-05 | 6.14 steps/s | VRAM=24.45 Go [2026-07-18 19:13:48] step 945200/1000000 | loss=2.0807 vocab=0.8049 token=1.1616 | lr=1.06e-05 | 6.15 steps/s | VRAM=27.19 Go [2026-07-18 19:13:57] step 945250/1000000 | loss=2.0301 vocab=0.7498 token=1.5473 | lr=1.05e-05 | 6.11 steps/s | VRAM=28.28 Go [2026-07-18 19:14:05] step 945300/1000000 | loss=2.0109 vocab=1.1182 token=1.1776 | lr=1.05e-05 | 6.11 steps/s | VRAM=27.20 Go [2026-07-18 19:14:13] step 945350/1000000 | loss=2.0908 vocab=0.7359 token=1.2830 | lr=1.05e-05 | 6.16 steps/s | VRAM=30.99 Go [2026-07-18 19:14:21] step 945400/1000000 | loss=1.9855 vocab=1.1914 token=1.5024 | lr=1.05e-05 | 6.11 steps/s | VRAM=28.22 Go [2026-07-18 19:14:29] step 945450/1000000 | loss=2.0709 vocab=0.9001 token=1.5293 | lr=1.05e-05 | 6.12 steps/s | VRAM=28.66 Go [2026-07-18 19:14:38] step 945500/1000000 | loss=2.0024 vocab=0.9074 token=0.9509 | lr=1.05e-05 | 6.09 steps/s | VRAM=29.24 Go [2026-07-18 19:14:46] step 945550/1000000 | loss=2.0385 vocab=0.6906 token=0.8671 | lr=1.04e-05 | 6.13 steps/s | VRAM=27.99 Go [2026-07-18 19:14:54] step 945600/1000000 | loss=1.9791 vocab=0.4963 token=1.1802 | lr=1.04e-05 | 6.12 steps/s | VRAM=28.40 Go [2026-07-18 19:15:02] step 945650/1000000 | loss=2.1154 vocab=0.8496 token=1.1230 | lr=1.04e-05 | 6.12 steps/s | VRAM=27.91 Go [2026-07-18 19:15:10] step 945700/1000000 | loss=2.0227 vocab=1.6917 token=1.3900 | lr=1.04e-05 | 6.13 steps/s | VRAM=27.16 Go [2026-07-18 19:15:18] step 945750/1000000 | loss=2.0211 vocab=0.6014 token=0.9888 | lr=1.04e-05 | 6.14 steps/s | VRAM=25.68 Go [2026-07-18 19:15:26] step 945800/1000000 | loss=2.0259 vocab=0.3977 token=1.0556 | lr=1.04e-05 | 6.14 steps/s | VRAM=30.23 Go [2026-07-18 19:15:35] step 945850/1000000 | loss=2.0885 vocab=0.4233 token=1.0242 | lr=1.03e-05 | 6.13 steps/s | VRAM=28.63 Go [2026-07-18 19:15:43] step 945900/1000000 | loss=2.0035 vocab=1.1746 token=1.0760 | lr=1.03e-05 | 6.13 steps/s | VRAM=26.52 Go [2026-07-18 19:15:51] step 945950/1000000 | loss=1.9708 vocab=0.6870 token=1.1307 | lr=1.03e-05 | 6.07 steps/s | VRAM=26.29 Go [2026-07-18 19:15:59] step 946000/1000000 | loss=2.0080 vocab=0.6039 token=1.2413 | lr=1.03e-05 | 6.14 steps/s | VRAM=30.99 Go [2026-07-18 19:16:07] step 946050/1000000 | loss=2.0900 vocab=0.8555 token=0.9283 | lr=1.03e-05 | 6.12 steps/s | VRAM=26.02 Go [2026-07-18 19:16:15] step 946100/1000000 | loss=2.0547 vocab=0.8631 token=0.5993 | lr=1.03e-05 | 6.13 steps/s | VRAM=26.44 Go [2026-07-18 19:16:24] step 946150/1000000 | loss=2.0233 vocab=0.7119 token=0.9818 | lr=1.02e-05 | 6.11 steps/s | VRAM=28.24 Go [2026-07-18 19:16:32] step 946200/1000000 | loss=1.9748 vocab=0.6597 token=1.2691 | lr=1.02e-05 | 6.11 steps/s | VRAM=23.54 Go [2026-07-18 19:16:40] step 946250/1000000 | loss=1.9086 vocab=0.4921 token=1.0501 | lr=1.02e-05 | 6.18 steps/s | VRAM=32.33 Go [2026-07-18 19:16:48] step 946300/1000000 | loss=2.1385 vocab=1.1304 token=1.4161 | lr=1.02e-05 | 6.14 steps/s | VRAM=24.59 Go [2026-07-18 19:16:56] step 946350/1000000 | loss=2.0310 vocab=0.9255 token=1.2546 | lr=1.02e-05 | 6.14 steps/s | VRAM=28.79 Go [2026-07-18 19:17:04] step 946400/1000000 | loss=2.0831 vocab=0.8955 token=1.3435 | lr=1.02e-05 | 6.12 steps/s | VRAM=26.19 Go [2026-07-18 19:17:13] step 946450/1000000 | loss=2.1420 vocab=0.5604 token=0.7664 | lr=1.01e-05 | 6.12 steps/s | VRAM=28.54 Go [2026-07-18 19:17:21] step 946500/1000000 | loss=1.9857 vocab=0.3787 token=1.0300 | lr=1.01e-05 | 6.10 steps/s | VRAM=29.82 Go [2026-07-18 19:17:29] step 946550/1000000 | loss=2.1409 vocab=0.9326 token=0.9358 | lr=1.01e-05 | 6.15 steps/s | VRAM=30.31 Go [2026-07-18 19:17:37] step 946600/1000000 | loss=1.9679 vocab=0.5175 token=1.3596 | lr=1.01e-05 | 6.15 steps/s | VRAM=28.98 Go [2026-07-18 19:17:45] step 946650/1000000 | loss=2.1142 vocab=1.1361 token=1.3138 | lr=1.01e-05 | 6.13 steps/s | VRAM=28.30 Go [2026-07-18 19:17:53] step 946700/1000000 | loss=1.9314 vocab=0.4698 token=0.8603 | lr=1.01e-05 | 6.09 steps/s | VRAM=28.94 Go [2026-07-18 19:18:02] step 946750/1000000 | loss=1.9545 vocab=0.7206 token=1.2200 | lr=1.00e-05 | 6.12 steps/s | VRAM=29.52 Go [2026-07-18 19:18:10] step 946800/1000000 | loss=2.0200 vocab=0.8348 token=1.0111 | lr=1.00e-05 | 6.12 steps/s | VRAM=28.00 Go [2026-07-18 19:18:18] step 946850/1000000 | loss=2.1778 vocab=1.1151 token=1.5412 | lr=1.00e-05 | 6.14 steps/s | VRAM=29.88 Go [2026-07-18 19:18:26] step 946900/1000000 | loss=2.0272 vocab=0.9140 token=0.9628 | lr=9.99e-06 | 6.16 steps/s | VRAM=24.70 Go [2026-07-18 19:18:34] step 946950/1000000 | loss=2.0286 vocab=0.9746 token=1.4220 | lr=9.97e-06 | 6.15 steps/s | VRAM=30.01 Go [2026-07-18 19:18:42] step 947000/1000000 | loss=2.0482 vocab=0.6845 token=0.7219 | lr=9.96e-06 | 6.14 steps/s | VRAM=25.87 Go [2026-07-18 19:18:50] step 947050/1000000 | loss=2.0117 vocab=0.5523 token=0.9810 | lr=9.94e-06 | 6.11 steps/s | VRAM=28.39 Go [2026-07-18 19:18:59] step 947100/1000000 | loss=1.9109 vocab=0.8375 token=1.4124 | lr=9.92e-06 | 6.10 steps/s | VRAM=29.30 Go [2026-07-18 19:19:07] step 947150/1000000 | loss=1.9871 vocab=1.2319 token=0.9676 | lr=9.91e-06 | 6.13 steps/s | VRAM=28.50 Go [2026-07-18 19:19:15] step 947200/1000000 | loss=2.0694 vocab=0.6074 token=1.5419 | lr=9.89e-06 | 6.11 steps/s | VRAM=25.83 Go [2026-07-18 19:19:23] step 947250/1000000 | loss=2.0218 vocab=0.8828 token=0.8024 | lr=9.88e-06 | 6.12 steps/s | VRAM=25.89 Go [2026-07-18 19:19:31] step 947300/1000000 | loss=1.9663 vocab=0.5549 token=0.7585 | lr=9.86e-06 | 6.09 steps/s | VRAM=28.93 Go [2026-07-18 19:19:40] step 947350/1000000 | loss=2.1162 vocab=0.8186 token=1.1422 | lr=9.84e-06 | 6.11 steps/s | VRAM=26.34 Go [2026-07-18 19:19:48] step 947400/1000000 | loss=2.0847 vocab=1.2209 token=1.1825 | lr=9.83e-06 | 6.14 steps/s | VRAM=25.94 Go [2026-07-18 19:19:56] step 947450/1000000 | loss=2.0525 vocab=0.6262 token=0.9926 | lr=9.81e-06 | 6.12 steps/s | VRAM=26.46 Go [2026-07-18 19:20:04] step 947500/1000000 | loss=1.9419 vocab=0.8583 token=1.2443 | lr=9.79e-06 | 6.10 steps/s | VRAM=24.31 Go [2026-07-18 19:20:12] step 947550/1000000 | loss=1.8805 vocab=0.7011 token=0.9833 | lr=9.78e-06 | 6.16 steps/s | VRAM=32.09 Go [2026-07-18 19:20:20] step 947600/1000000 | loss=2.0888 vocab=0.6096 token=1.1335 | lr=9.76e-06 | 6.12 steps/s | VRAM=29.05 Go [2026-07-18 19:20:29] step 947650/1000000 | loss=1.9965 vocab=0.9356 token=0.9779 | lr=9.75e-06 | 6.11 steps/s | VRAM=28.93 Go [2026-07-18 19:20:37] step 947700/1000000 | loss=2.1324 vocab=0.9542 token=1.4455 | lr=9.73e-06 | 6.12 steps/s | VRAM=26.54 Go [2026-07-18 19:20:45] step 947750/1000000 | loss=2.0271 vocab=1.1878 token=1.3829 | lr=9.71e-06 | 6.13 steps/s | VRAM=30.82 Go [2026-07-18 19:20:53] step 947800/1000000 | loss=1.8681 vocab=0.7817 token=1.2945 | lr=9.70e-06 | 6.14 steps/s | VRAM=32.90 Go [2026-07-18 19:21:01] step 947850/1000000 | loss=2.0437 vocab=1.0036 token=1.1706 | lr=9.68e-06 | 6.14 steps/s | VRAM=29.58 Go [2026-07-18 19:21:09] step 947900/1000000 | loss=1.9860 vocab=0.8827 token=1.1392 | lr=9.66e-06 | 6.14 steps/s | VRAM=28.32 Go [2026-07-18 19:21:17] step 947950/1000000 | loss=2.0798 vocab=1.0303 token=1.5820 | lr=9.65e-06 | 6.14 steps/s | VRAM=27.99 Go [2026-07-18 19:21:26] step 948000/1000000 | loss=1.9672 vocab=0.6765 token=1.0734 | lr=9.63e-06 | 6.14 steps/s | VRAM=25.51 Go [2026-07-18 19:21:34] step 948050/1000000 | loss=2.0044 vocab=0.6670 token=1.2733 | lr=9.62e-06 | 6.12 steps/s | VRAM=29.54 Go [2026-07-18 19:21:42] step 948100/1000000 | loss=1.9225 vocab=1.1016 token=1.1554 | lr=9.60e-06 | 6.14 steps/s | VRAM=25.10 Go [2026-07-18 19:21:50] step 948150/1000000 | loss=2.0247 vocab=0.3824 token=0.9183 | lr=9.58e-06 | 6.15 steps/s | VRAM=31.83 Go [2026-07-18 19:21:58] step 948200/1000000 | loss=2.0594 vocab=0.9082 token=1.0220 | lr=9.57e-06 | 6.14 steps/s | VRAM=28.94 Go [2026-07-18 19:22:06] step 948250/1000000 | loss=2.0445 vocab=0.8495 token=1.0207 | lr=9.55e-06 | 6.13 steps/s | VRAM=29.10 Go [2026-07-18 19:22:14] step 948300/1000000 | loss=2.0683 vocab=1.1420 token=1.0024 | lr=9.54e-06 | 6.14 steps/s | VRAM=32.20 Go [2026-07-18 19:22:23] step 948350/1000000 | loss=2.0986 vocab=0.8255 token=1.3229 | lr=9.52e-06 | 6.13 steps/s | VRAM=29.43 Go [2026-07-18 19:22:31] step 948400/1000000 | loss=1.9937 vocab=0.8363 token=1.3325 | lr=9.50e-06 | 6.16 steps/s | VRAM=32.26 Go [2026-07-18 19:22:39] step 948450/1000000 | loss=2.0910 vocab=0.9295 token=1.2844 | lr=9.49e-06 | 6.14 steps/s | VRAM=29.87 Go [2026-07-18 19:22:47] step 948500/1000000 | loss=1.9632 vocab=0.7391 token=0.8707 | lr=9.47e-06 | 6.14 steps/s | VRAM=28.78 Go [2026-07-18 19:22:55] step 948550/1000000 | loss=2.1190 vocab=0.7153 token=1.2422 | lr=9.46e-06 | 6.15 steps/s | VRAM=25.65 Go [2026-07-18 19:23:03] step 948600/1000000 | loss=2.0802 vocab=1.2309 token=1.6725 | lr=9.44e-06 | 6.14 steps/s | VRAM=32.34 Go [2026-07-18 19:23:11] step 948650/1000000 | loss=2.1019 vocab=0.7770 token=0.9730 | lr=9.42e-06 | 6.12 steps/s | VRAM=26.76 Go [2026-07-18 19:23:20] step 948700/1000000 | loss=2.0846 vocab=0.6848 token=1.1592 | lr=9.41e-06 | 6.13 steps/s | VRAM=28.16 Go [2026-07-18 19:23:28] step 948750/1000000 | loss=1.9014 vocab=0.7256 token=1.1001 | lr=9.39e-06 | 6.15 steps/s | VRAM=27.90 Go [2026-07-18 19:23:36] step 948800/1000000 | loss=2.1025 vocab=1.0903 token=1.1229 | lr=9.38e-06 | 6.13 steps/s | VRAM=27.30 Go [2026-07-18 19:23:44] step 948850/1000000 | loss=1.9747 vocab=0.5908 token=1.1159 | lr=9.36e-06 | 6.17 steps/s | VRAM=27.27 Go [2026-07-18 19:23:52] step 948900/1000000 | loss=2.1263 vocab=1.3790 token=1.3315 | lr=9.35e-06 | 6.14 steps/s | VRAM=29.06 Go [2026-07-18 19:24:00] step 948950/1000000 | loss=2.0336 vocab=0.8221 token=1.1431 | lr=9.33e-06 | 6.11 steps/s | VRAM=26.87 Go [2026-07-18 19:24:08] step 949000/1000000 | loss=1.9575 vocab=0.9540 token=1.5409 | lr=9.31e-06 | 6.14 steps/s | VRAM=28.40 Go [2026-07-18 19:24:17] step 949050/1000000 | loss=2.0011 vocab=1.2200 token=1.5267 | lr=9.30e-06 | 6.17 steps/s | VRAM=28.22 Go [2026-07-18 19:24:25] step 949100/1000000 | loss=1.9817 vocab=1.0259 token=1.7936 | lr=9.28e-06 | 6.12 steps/s | VRAM=27.75 Go [2026-07-18 19:24:33] step 949150/1000000 | loss=1.9903 vocab=0.8686 token=1.1079 | lr=9.27e-06 | 6.14 steps/s | VRAM=24.50 Go [2026-07-18 19:24:41] step 949200/1000000 | loss=2.0071 vocab=0.4327 token=1.0062 | lr=9.25e-06 | 6.15 steps/s | VRAM=32.50 Go [2026-07-18 19:24:49] step 949250/1000000 | loss=2.0236 vocab=0.5366 token=1.0502 | lr=9.23e-06 | 6.11 steps/s | VRAM=26.19 Go [2026-07-18 19:24:57] step 949300/1000000 | loss=2.0063 vocab=0.4343 token=1.1029 | lr=9.22e-06 | 6.15 steps/s | VRAM=27.92 Go [2026-07-18 19:25:05] step 949350/1000000 | loss=1.9769 vocab=1.0311 token=1.2808 | lr=9.20e-06 | 6.15 steps/s | VRAM=25.30 Go [2026-07-18 19:25:14] step 949400/1000000 | loss=1.9616 vocab=1.1715 token=1.6348 | lr=9.19e-06 | 6.12 steps/s | VRAM=25.94 Go [2026-07-18 19:25:22] step 949450/1000000 | loss=2.0744 vocab=0.4637 token=1.1774 | lr=9.17e-06 | 6.14 steps/s | VRAM=27.91 Go [2026-07-18 19:25:30] step 949500/1000000 | loss=1.9405 vocab=0.7046 token=1.3496 | lr=9.16e-06 | 6.11 steps/s | VRAM=28.58 Go [2026-07-18 19:25:38] step 949550/1000000 | loss=1.9514 vocab=0.6017 token=1.0226 | lr=9.14e-06 | 6.14 steps/s | VRAM=29.17 Go [2026-07-18 19:25:46] step 949600/1000000 | loss=2.0619 vocab=1.2069 token=1.1482 | lr=9.12e-06 | 6.14 steps/s | VRAM=28.26 Go [2026-07-18 19:25:54] step 949650/1000000 | loss=2.1147 vocab=0.5161 token=1.3711 | lr=9.11e-06 | 6.14 steps/s | VRAM=25.89 Go [2026-07-18 19:26:03] step 949700/1000000 | loss=1.9588 vocab=0.7665 token=1.1857 | lr=9.09e-06 | 6.15 steps/s | VRAM=29.21 Go [2026-07-18 19:26:11] step 949750/1000000 | loss=2.0023 vocab=0.9748 token=1.0673 | lr=9.08e-06 | 6.17 steps/s | VRAM=32.24 Go [2026-07-18 19:26:19] step 949800/1000000 | loss=2.1427 vocab=1.0188 token=1.1240 | lr=9.06e-06 | 6.15 steps/s | VRAM=25.01 Go [2026-07-18 19:26:27] step 949850/1000000 | loss=2.0695 vocab=1.0373 token=1.3501 | lr=9.05e-06 | 6.16 steps/s | VRAM=31.18 Go [2026-07-18 19:26:35] step 949900/1000000 | loss=2.0251 vocab=0.7676 token=1.6694 | lr=9.03e-06 | 6.14 steps/s | VRAM=30.61 Go [2026-07-18 19:26:43] step 949950/1000000 | loss=2.1035 vocab=1.1029 token=1.4751 | lr=9.02e-06 | 6.13 steps/s | VRAM=26.54 Go [2026-07-18 19:26:51] step 950000/1000000 | loss=2.0891 vocab=0.9470 token=1.1882 | lr=9.00e-06 | 6.14 steps/s | VRAM=28.84 Go [accuracy] step 950000 : 66.25% (log: /workspace/plbert_train/precision_log.csv) [push] step 950000 OK (loss_token=1.1882, accuracy=66.25%) [2026-07-18 19:28:19] step 950050/1000000 | loss=1.9974 vocab=0.8207 token=0.8994 | lr=8.98e-06 | 0.57 steps/s | VRAM=31.19 Go [2026-07-18 19:28:27] step 950100/1000000 | loss=2.1366 vocab=0.8314 token=1.1004 | lr=8.97e-06 | 6.23 steps/s | VRAM=25.23 Go [2026-07-18 19:28:35] step 950150/1000000 | loss=2.1183 vocab=0.5994 token=0.9187 | lr=8.95e-06 | 6.21 steps/s | VRAM=24.37 Go [2026-07-18 19:28:43] step 950200/1000000 | loss=2.0525 vocab=1.0880 token=1.1229 | lr=8.94e-06 | 6.19 steps/s | VRAM=26.63 Go [2026-07-18 19:28:51] step 950250/1000000 | loss=2.0275 vocab=0.9578 token=1.5703 | lr=8.92e-06 | 6.17 steps/s | VRAM=27.99 Go [2026-07-18 19:28:59] step 950300/1000000 | loss=2.0442 vocab=0.5511 token=0.8759 | lr=8.91e-06 | 6.17 steps/s | VRAM=25.60 Go [2026-07-18 19:29:08] step 950350/1000000 | loss=2.0525 vocab=0.6229 token=1.0542 | lr=8.89e-06 | 6.15 steps/s | VRAM=28.60 Go [2026-07-18 19:29:16] step 950400/1000000 | loss=2.0533 vocab=1.0017 token=1.0151 | lr=8.88e-06 | 6.12 steps/s | VRAM=25.48 Go [2026-07-18 19:29:24] step 950450/1000000 | loss=2.0356 vocab=0.5969 token=0.8830 | lr=8.86e-06 | 6.19 steps/s | VRAM=29.18 Go [2026-07-18 19:29:32] step 950500/1000000 | loss=2.0441 vocab=1.0354 token=1.1795 | lr=8.84e-06 | 6.13 steps/s | VRAM=28.63 Go [2026-07-18 19:29:40] step 950550/1000000 | loss=2.0510 vocab=1.0521 token=1.1417 | lr=8.83e-06 | 6.15 steps/s | VRAM=29.00 Go [2026-07-18 19:29:48] step 950600/1000000 | loss=2.0214 vocab=1.4669 token=1.4899 | lr=8.81e-06 | 6.15 steps/s | VRAM=31.30 Go [2026-07-18 19:29:56] step 950650/1000000 | loss=2.0892 vocab=0.8341 token=1.5018 | lr=8.80e-06 | 6.12 steps/s | VRAM=26.44 Go [2026-07-18 19:30:05] step 950700/1000000 | loss=2.1335 vocab=0.7487 token=1.3135 | lr=8.78e-06 | 6.13 steps/s | VRAM=29.26 Go [2026-07-18 19:30:13] step 950750/1000000 | loss=2.0598 vocab=0.7586 token=1.0131 | lr=8.77e-06 | 6.12 steps/s | VRAM=26.58 Go [2026-07-18 19:30:21] step 950800/1000000 | loss=2.0826 vocab=0.7936 token=1.2408 | lr=8.75e-06 | 6.10 steps/s | VRAM=28.88 Go [2026-07-18 19:30:29] step 950850/1000000 | loss=2.0144 vocab=0.4785 token=1.0496 | lr=8.74e-06 | 6.12 steps/s | VRAM=28.98 Go [2026-07-18 19:30:37] step 950900/1000000 | loss=2.0820 vocab=1.1339 token=1.2734 | lr=8.72e-06 | 6.10 steps/s | VRAM=29.10 Go [2026-07-18 19:30:45] step 950950/1000000 | loss=2.0926 vocab=1.0157 token=1.1405 | lr=8.71e-06 | 6.14 steps/s | VRAM=28.10 Go [2026-07-18 19:30:54] step 951000/1000000 | loss=2.0563 vocab=0.9815 token=1.4036 | lr=8.69e-06 | 6.13 steps/s | VRAM=31.57 Go [2026-07-18 19:31:02] step 951050/1000000 | loss=2.0789 vocab=0.8962 token=1.1268 | lr=8.68e-06 | 6.10 steps/s | VRAM=26.41 Go [2026-07-18 19:31:10] step 951100/1000000 | loss=2.0951 vocab=1.3841 token=1.5031 | lr=8.66e-06 | 6.10 steps/s | VRAM=29.17 Go [2026-07-18 19:31:18] step 951150/1000000 | loss=1.9873 vocab=0.9900 token=1.0960 | lr=8.65e-06 | 6.11 steps/s | VRAM=28.74 Go [2026-07-18 19:31:26] step 951200/1000000 | loss=2.0622 vocab=0.7365 token=1.0709 | lr=8.63e-06 | 6.10 steps/s | VRAM=32.57 Go [2026-07-18 19:31:34] step 951250/1000000 | loss=1.8686 vocab=0.8245 token=1.4199 | lr=8.62e-06 | 6.15 steps/s | VRAM=28.66 Go [2026-07-18 19:31:43] step 951300/1000000 | loss=1.9889 vocab=0.9265 token=1.2476 | lr=8.60e-06 | 6.14 steps/s | VRAM=32.39 Go [2026-07-18 19:31:51] step 951350/1000000 | loss=1.9517 vocab=0.8415 token=1.0266 | lr=8.58e-06 | 6.09 steps/s | VRAM=26.22 Go [2026-07-18 19:31:59] step 951400/1000000 | loss=1.9871 vocab=0.6575 token=0.9518 | lr=8.57e-06 | 6.14 steps/s | VRAM=31.73 Go [2026-07-18 19:32:07] step 951450/1000000 | loss=1.9626 vocab=0.8282 token=1.1254 | lr=8.55e-06 | 6.13 steps/s | VRAM=27.97 Go [2026-07-18 19:32:15] step 951500/1000000 | loss=2.1229 vocab=1.4381 token=0.9936 | lr=8.54e-06 | 6.12 steps/s | VRAM=29.42 Go [2026-07-18 19:32:23] step 951550/1000000 | loss=2.0227 vocab=1.1224 token=1.2209 | lr=8.52e-06 | 6.13 steps/s | VRAM=29.07 Go [2026-07-18 19:32:32] step 951600/1000000 | loss=2.0165 vocab=0.8734 token=1.3549 | lr=8.51e-06 | 6.14 steps/s | VRAM=32.82 Go [2026-07-18 19:32:40] step 951650/1000000 | loss=2.0419 vocab=0.7940 token=1.1140 | lr=8.49e-06 | 6.14 steps/s | VRAM=29.48 Go [2026-07-18 19:32:48] step 951700/1000000 | loss=2.0359 vocab=0.4303 token=1.0419 | lr=8.48e-06 | 6.12 steps/s | VRAM=29.06 Go [2026-07-18 19:32:56] step 951750/1000000 | loss=1.9749 vocab=0.5336 token=1.1178 | lr=8.46e-06 | 6.14 steps/s | VRAM=29.01 Go [2026-07-18 19:33:04] step 951800/1000000 | loss=2.0103 vocab=0.9559 token=1.3205 | lr=8.45e-06 | 6.13 steps/s | VRAM=28.21 Go [2026-07-18 19:33:12] step 951850/1000000 | loss=2.0827 vocab=0.7822 token=1.0110 | lr=8.43e-06 | 6.11 steps/s | VRAM=25.13 Go [2026-07-18 19:33:21] step 951900/1000000 | loss=2.0252 vocab=0.7646 token=0.9592 | lr=8.42e-06 | 6.10 steps/s | VRAM=27.25 Go [2026-07-18 19:33:29] step 951950/1000000 | loss=1.9254 vocab=0.4495 token=0.9648 | lr=8.40e-06 | 6.10 steps/s | VRAM=28.38 Go [2026-07-18 19:33:37] step 952000/1000000 | loss=2.0404 vocab=0.3546 token=0.8215 | lr=8.39e-06 | 6.13 steps/s | VRAM=32.28 Go [2026-07-18 19:33:45] step 952050/1000000 | loss=1.9746 vocab=0.6584 token=1.2564 | lr=8.37e-06 | 6.13 steps/s | VRAM=25.23 Go [2026-07-18 19:33:53] step 952100/1000000 | loss=1.9910 vocab=0.9633 token=1.2555 | lr=8.36e-06 | 6.13 steps/s | VRAM=30.13 Go [2026-07-18 19:34:01] step 952150/1000000 | loss=1.9824 vocab=0.7622 token=1.3857 | lr=8.34e-06 | 6.13 steps/s | VRAM=32.01 Go [2026-07-18 19:34:10] step 952200/1000000 | loss=2.0957 vocab=0.9687 token=1.2548 | lr=8.33e-06 | 6.12 steps/s | VRAM=26.30 Go [2026-07-18 19:34:18] step 952250/1000000 | loss=1.9834 vocab=0.8638 token=1.0494 | lr=8.31e-06 | 6.15 steps/s | VRAM=28.11 Go [2026-07-18 19:34:26] step 952300/1000000 | loss=2.0387 vocab=0.6831 token=0.7681 | lr=8.30e-06 | 6.13 steps/s | VRAM=32.68 Go [2026-07-18 19:34:34] step 952350/1000000 | loss=1.9501 vocab=0.4189 token=0.8563 | lr=8.28e-06 | 6.12 steps/s | VRAM=28.35 Go [2026-07-18 19:34:42] step 952400/1000000 | loss=2.0189 vocab=1.5236 token=1.2073 | lr=8.27e-06 | 6.12 steps/s | VRAM=25.60 Go [2026-07-18 19:34:50] step 952450/1000000 | loss=1.9851 vocab=0.6189 token=1.1178 | lr=8.25e-06 | 6.11 steps/s | VRAM=28.69 Go [2026-07-18 19:34:59] step 952500/1000000 | loss=2.0690 vocab=0.7411 token=1.2425 | lr=8.24e-06 | 6.14 steps/s | VRAM=24.67 Go [2026-07-18 19:35:07] step 952550/1000000 | loss=2.1621 vocab=1.2356 token=1.2503 | lr=8.22e-06 | 6.13 steps/s | VRAM=32.37 Go [2026-07-18 19:35:15] step 952600/1000000 | loss=2.1625 vocab=0.7379 token=1.1490 | lr=8.21e-06 | 6.12 steps/s | VRAM=26.38 Go [2026-07-18 19:35:23] step 952650/1000000 | loss=2.0022 vocab=0.5978 token=0.6739 | lr=8.19e-06 | 6.12 steps/s | VRAM=28.88 Go [2026-07-18 19:35:31] step 952700/1000000 | loss=2.0603 vocab=0.9612 token=1.5520 | lr=8.18e-06 | 6.10 steps/s | VRAM=29.82 Go [2026-07-18 19:35:39] step 952750/1000000 | loss=1.9319 vocab=0.9351 token=1.2019 | lr=8.17e-06 | 6.14 steps/s | VRAM=28.49 Go [2026-07-18 19:35:48] step 952800/1000000 | loss=2.0218 vocab=0.9573 token=1.2030 | lr=8.15e-06 | 6.12 steps/s | VRAM=32.16 Go [2026-07-18 19:35:56] step 952850/1000000 | loss=1.9405 vocab=0.5755 token=1.2576 | lr=8.14e-06 | 6.07 steps/s | VRAM=25.74 Go [2026-07-18 19:36:04] step 952900/1000000 | loss=2.0776 vocab=0.8964 token=1.8129 | lr=8.12e-06 | 6.09 steps/s | VRAM=29.92 Go [2026-07-18 19:36:12] step 952950/1000000 | loss=1.9773 vocab=0.7770 token=1.2601 | lr=8.11e-06 | 6.13 steps/s | VRAM=32.38 Go [2026-07-18 19:36:20] step 953000/1000000 | loss=2.0709 vocab=0.7234 token=1.0133 | lr=8.09e-06 | 6.12 steps/s | VRAM=28.58 Go [2026-07-18 19:36:29] step 953050/1000000 | loss=1.9629 vocab=0.9219 token=1.1585 | lr=8.08e-06 | 6.14 steps/s | VRAM=25.39 Go [2026-07-18 19:36:37] step 953100/1000000 | loss=1.9922 vocab=0.9442 token=1.2139 | lr=8.06e-06 | 6.13 steps/s | VRAM=28.86 Go [2026-07-18 19:36:45] step 953150/1000000 | loss=1.9944 vocab=0.5944 token=0.9930 | lr=8.05e-06 | 6.13 steps/s | VRAM=28.66 Go [2026-07-18 19:36:53] step 953200/1000000 | loss=1.9646 vocab=0.6179 token=1.0800 | lr=8.03e-06 | 6.12 steps/s | VRAM=28.49 Go [2026-07-18 19:37:01] step 953250/1000000 | loss=2.0270 vocab=1.0524 token=1.0517 | lr=8.02e-06 | 6.13 steps/s | VRAM=29.28 Go [2026-07-18 19:37:09] step 953300/1000000 | loss=2.0451 vocab=0.4434 token=0.8933 | lr=8.00e-06 | 6.12 steps/s | VRAM=28.13 Go [2026-07-18 19:37:17] step 953350/1000000 | loss=1.9749 vocab=0.5585 token=1.2282 | lr=7.99e-06 | 6.15 steps/s | VRAM=32.17 Go [2026-07-18 19:37:26] step 953400/1000000 | loss=1.9877 vocab=0.7176 token=1.1607 | lr=7.97e-06 | 6.14 steps/s | VRAM=29.37 Go [2026-07-18 19:37:34] step 953450/1000000 | loss=2.0465 vocab=0.6221 token=1.4562 | lr=7.96e-06 | 6.13 steps/s | VRAM=30.74 Go [2026-07-18 19:37:42] step 953500/1000000 | loss=1.9497 vocab=0.5396 token=1.0749 | lr=7.94e-06 | 6.12 steps/s | VRAM=27.17 Go [2026-07-18 19:37:50] step 953550/1000000 | loss=2.0656 vocab=0.6420 token=1.1356 | lr=7.93e-06 | 6.13 steps/s | VRAM=28.95 Go [2026-07-18 19:37:58] step 953600/1000000 | loss=2.0086 vocab=1.0911 token=1.6094 | lr=7.92e-06 | 6.09 steps/s | VRAM=26.35 Go [2026-07-18 19:38:06] step 953650/1000000 | loss=2.1069 vocab=0.9140 token=1.0211 | lr=7.90e-06 | 6.11 steps/s | VRAM=28.56 Go [2026-07-18 19:38:15] step 953700/1000000 | loss=2.1295 vocab=1.0075 token=1.5315 | lr=7.89e-06 | 6.10 steps/s | VRAM=32.42 Go [2026-07-18 19:38:23] step 953750/1000000 | loss=1.9531 vocab=0.7036 token=0.9489 | lr=7.87e-06 | 6.12 steps/s | VRAM=32.11 Go [2026-07-18 19:38:31] step 953800/1000000 | loss=2.0408 vocab=0.3629 token=1.1094 | lr=7.86e-06 | 6.13 steps/s | VRAM=28.78 Go [2026-07-18 19:38:39] step 953850/1000000 | loss=2.0105 vocab=0.7392 token=1.2836 | lr=7.84e-06 | 6.11 steps/s | VRAM=29.04 Go [2026-07-18 19:38:47] step 953900/1000000 | loss=2.0530 vocab=0.5757 token=1.1132 | lr=7.83e-06 | 6.13 steps/s | VRAM=26.27 Go [2026-07-18 19:38:55] step 953950/1000000 | loss=1.9870 vocab=0.8268 token=1.3090 | lr=7.81e-06 | 6.11 steps/s | VRAM=26.16 Go [2026-07-18 19:39:04] step 954000/1000000 | loss=2.0455 vocab=0.7593 token=1.0256 | lr=7.80e-06 | 6.13 steps/s | VRAM=29.31 Go [2026-07-18 19:39:12] step 954050/1000000 | loss=2.0082 vocab=0.8643 token=1.5057 | lr=7.79e-06 | 6.13 steps/s | VRAM=24.91 Go [2026-07-18 19:39:20] step 954100/1000000 | loss=2.0048 vocab=0.7303 token=1.2882 | lr=7.77e-06 | 6.10 steps/s | VRAM=29.15 Go [2026-07-18 19:39:28] step 954150/1000000 | loss=2.1654 vocab=0.5730 token=1.1067 | lr=7.76e-06 | 6.14 steps/s | VRAM=28.85 Go [2026-07-18 19:39:36] step 954200/1000000 | loss=1.9605 vocab=0.7602 token=0.8757 | lr=7.74e-06 | 6.11 steps/s | VRAM=28.06 Go [2026-07-18 19:39:45] step 954250/1000000 | loss=2.0119 vocab=0.4874 token=1.0684 | lr=7.73e-06 | 6.11 steps/s | VRAM=26.57 Go [2026-07-18 19:39:53] step 954300/1000000 | loss=1.8907 vocab=0.9382 token=1.1527 | lr=7.71e-06 | 6.13 steps/s | VRAM=23.23 Go [2026-07-18 19:40:01] step 954350/1000000 | loss=2.0585 vocab=0.9191 token=1.1555 | lr=7.70e-06 | 6.17 steps/s | VRAM=28.39 Go [2026-07-18 19:40:09] step 954400/1000000 | loss=1.8828 vocab=0.4630 token=0.9834 | lr=7.68e-06 | 6.12 steps/s | VRAM=28.99 Go [2026-07-18 19:40:17] step 954450/1000000 | loss=2.0305 vocab=0.9448 token=1.2803 | lr=7.67e-06 | 6.13 steps/s | VRAM=28.47 Go [2026-07-18 19:40:25] step 954500/1000000 | loss=1.9795 vocab=0.6518 token=1.1825 | lr=7.66e-06 | 6.18 steps/s | VRAM=32.16 Go [2026-07-18 19:40:33] step 954550/1000000 | loss=2.1329 vocab=0.5919 token=0.8522 | lr=7.64e-06 | 6.17 steps/s | VRAM=24.86 Go [2026-07-18 19:40:41] step 954600/1000000 | loss=2.0323 vocab=0.9444 token=1.0745 | lr=7.63e-06 | 6.13 steps/s | VRAM=28.72 Go [2026-07-18 19:40:50] step 954650/1000000 | loss=1.9338 vocab=1.0919 token=0.8032 | lr=7.61e-06 | 6.11 steps/s | VRAM=25.86 Go [2026-07-18 19:40:58] step 954700/1000000 | loss=2.0762 vocab=0.9949 token=1.3525 | lr=7.60e-06 | 6.14 steps/s | VRAM=28.83 Go [2026-07-18 19:41:06] step 954750/1000000 | loss=2.0529 vocab=1.7174 token=1.4761 | lr=7.58e-06 | 6.14 steps/s | VRAM=31.48 Go [2026-07-18 19:41:14] step 954800/1000000 | loss=2.0197 vocab=0.4735 token=1.0623 | lr=7.57e-06 | 6.15 steps/s | VRAM=31.80 Go [2026-07-18 19:41:22] step 954850/1000000 | loss=2.0299 vocab=1.0196 token=1.5982 | lr=7.56e-06 | 6.12 steps/s | VRAM=29.26 Go [2026-07-18 19:41:30] step 954900/1000000 | loss=1.8455 vocab=0.5615 token=1.0825 | lr=7.54e-06 | 6.12 steps/s | VRAM=27.13 Go [2026-07-18 19:41:39] step 954950/1000000 | loss=1.9677 vocab=0.7146 token=0.9396 | lr=7.53e-06 | 6.12 steps/s | VRAM=26.73 Go [2026-07-18 19:41:47] step 955000/1000000 | loss=2.0212 vocab=0.6106 token=1.2950 | lr=7.51e-06 | 6.15 steps/s | VRAM=32.19 Go [2026-07-18 19:41:55] step 955050/1000000 | loss=2.0279 vocab=0.7216 token=1.2441 | lr=7.50e-06 | 6.14 steps/s | VRAM=28.29 Go [2026-07-18 19:42:03] step 955100/1000000 | loss=2.0742 vocab=0.5188 token=0.7709 | lr=7.49e-06 | 6.11 steps/s | VRAM=28.97 Go [2026-07-18 19:42:11] step 955150/1000000 | loss=2.0514 vocab=0.9061 token=1.4090 | lr=7.47e-06 | 6.18 steps/s | VRAM=31.62 Go [2026-07-18 19:42:19] step 955200/1000000 | loss=2.0384 vocab=0.8114 token=1.0599 | lr=7.46e-06 | 6.13 steps/s | VRAM=32.78 Go [2026-07-18 19:42:27] step 955250/1000000 | loss=1.9538 vocab=0.7357 token=1.2179 | lr=7.44e-06 | 6.13 steps/s | VRAM=29.69 Go [2026-07-18 19:42:36] step 955300/1000000 | loss=2.1733 vocab=0.8866 token=1.1670 | lr=7.43e-06 | 6.12 steps/s | VRAM=28.56 Go [2026-07-18 19:42:44] step 955350/1000000 | loss=2.1092 vocab=0.5268 token=0.7753 | lr=7.42e-06 | 6.14 steps/s | VRAM=28.78 Go [2026-07-18 19:42:52] step 955400/1000000 | loss=1.9574 vocab=0.4767 token=0.9353 | lr=7.40e-06 | 6.11 steps/s | VRAM=28.32 Go [2026-07-18 19:43:00] step 955450/1000000 | loss=2.0018 vocab=0.7459 token=0.6828 | lr=7.39e-06 | 6.14 steps/s | VRAM=28.57 Go [2026-07-18 19:43:08] step 955500/1000000 | loss=2.1423 vocab=1.3196 token=0.9156 | lr=7.37e-06 | 6.14 steps/s | VRAM=27.51 Go [2026-07-18 19:43:16] step 955550/1000000 | loss=1.9977 vocab=0.9351 token=1.0595 | lr=7.36e-06 | 6.16 steps/s | VRAM=31.94 Go [2026-07-18 19:43:24] step 955600/1000000 | loss=1.9898 vocab=0.4741 token=0.8190 | lr=7.35e-06 | 6.15 steps/s | VRAM=29.19 Go [2026-07-18 19:43:33] step 955650/1000000 | loss=2.0561 vocab=0.8511 token=1.2613 | lr=7.33e-06 | 6.12 steps/s | VRAM=32.42 Go [2026-07-18 19:43:41] step 955700/1000000 | loss=2.0820 vocab=1.2236 token=1.2382 | lr=7.32e-06 | 6.09 steps/s | VRAM=29.24 Go [2026-07-18 19:43:49] step 955750/1000000 | loss=1.9698 vocab=0.7635 token=1.1575 | lr=7.30e-06 | 6.14 steps/s | VRAM=32.38 Go [2026-07-18 19:43:57] step 955800/1000000 | loss=2.0082 vocab=0.5808 token=1.1370 | lr=7.29e-06 | 6.14 steps/s | VRAM=26.25 Go [2026-07-18 19:44:05] step 955850/1000000 | loss=1.9569 vocab=0.5278 token=0.9497 | lr=7.28e-06 | 6.15 steps/s | VRAM=32.08 Go [2026-07-18 19:44:13] step 955900/1000000 | loss=1.8997 vocab=1.4433 token=1.0316 | lr=7.26e-06 | 6.13 steps/s | VRAM=25.76 Go [2026-07-18 19:44:22] step 955950/1000000 | loss=1.9327 vocab=1.1888 token=1.3376 | lr=7.25e-06 | 6.15 steps/s | VRAM=28.33 Go [2026-07-18 19:44:30] step 956000/1000000 | loss=1.9113 vocab=1.0107 token=1.2039 | lr=7.23e-06 | 6.13 steps/s | VRAM=28.12 Go [2026-07-18 19:44:38] step 956050/1000000 | loss=2.0019 vocab=0.5959 token=0.9936 | lr=7.22e-06 | 6.14 steps/s | VRAM=28.58 Go [2026-07-18 19:44:46] step 956100/1000000 | loss=1.9524 vocab=0.6103 token=1.2679 | lr=7.21e-06 | 6.15 steps/s | VRAM=29.09 Go [2026-07-18 19:44:54] step 956150/1000000 | loss=1.9744 vocab=0.7789 token=0.9993 | lr=7.19e-06 | 6.13 steps/s | VRAM=28.01 Go [2026-07-18 19:45:02] step 956200/1000000 | loss=2.0385 vocab=1.0797 token=1.2466 | lr=7.18e-06 | 6.16 steps/s | VRAM=27.96 Go [2026-07-18 19:45:10] step 956250/1000000 | loss=1.9255 vocab=1.1159 token=1.4113 | lr=7.16e-06 | 6.17 steps/s | VRAM=27.89 Go [2026-07-18 19:45:18] step 956300/1000000 | loss=2.0234 vocab=0.4603 token=1.0419 | lr=7.15e-06 | 6.14 steps/s | VRAM=32.46 Go [2026-07-18 19:45:27] step 956350/1000000 | loss=2.0625 vocab=1.2264 token=1.2619 | lr=7.14e-06 | 6.14 steps/s | VRAM=32.26 Go [2026-07-18 19:45:35] step 956400/1000000 | loss=1.9201 vocab=0.3866 token=1.1295 | lr=7.12e-06 | 6.15 steps/s | VRAM=26.02 Go [2026-07-18 19:45:43] step 956450/1000000 | loss=1.9545 vocab=0.6157 token=0.9580 | lr=7.11e-06 | 6.15 steps/s | VRAM=29.85 Go [2026-07-18 19:45:51] step 956500/1000000 | loss=1.9569 vocab=0.7510 token=1.2830 | lr=7.10e-06 | 6.16 steps/s | VRAM=32.33 Go [2026-07-18 19:45:59] step 956550/1000000 | loss=1.9575 vocab=0.4811 token=1.1606 | lr=7.08e-06 | 6.15 steps/s | VRAM=25.05 Go [2026-07-18 19:46:07] step 956600/1000000 | loss=2.0124 vocab=0.8797 token=0.8200 | lr=7.07e-06 | 6.15 steps/s | VRAM=31.85 Go [2026-07-18 19:46:15] step 956650/1000000 | loss=2.0273 vocab=0.8457 token=1.2999 | lr=7.05e-06 | 6.15 steps/s | VRAM=26.39 Go [2026-07-18 19:46:23] step 956700/1000000 | loss=2.0207 vocab=1.5070 token=1.5815 | lr=7.04e-06 | 6.19 steps/s | VRAM=31.21 Go [2026-07-18 19:46:32] step 956750/1000000 | loss=2.0122 vocab=1.2874 token=1.0526 | lr=7.03e-06 | 6.15 steps/s | VRAM=32.46 Go [2026-07-18 19:46:40] step 956800/1000000 | loss=1.9509 vocab=0.7491 token=0.8983 | lr=7.01e-06 | 6.15 steps/s | VRAM=32.58 Go [2026-07-18 19:46:48] step 956850/1000000 | loss=2.0762 vocab=0.8948 token=0.8551 | lr=7.00e-06 | 6.18 steps/s | VRAM=24.70 Go [2026-07-18 19:46:56] step 956900/1000000 | loss=2.0311 vocab=0.9684 token=1.3913 | lr=6.99e-06 | 6.14 steps/s | VRAM=32.73 Go [2026-07-18 19:47:04] step 956950/1000000 | loss=2.0322 vocab=0.9475 token=1.1156 | lr=6.97e-06 | 6.14 steps/s | VRAM=31.47 Go [2026-07-18 19:47:12] step 957000/1000000 | loss=2.0097 vocab=0.8930 token=1.2547 | lr=6.96e-06 | 6.14 steps/s | VRAM=25.10 Go [2026-07-18 19:47:20] step 957050/1000000 | loss=2.0337 vocab=0.6952 token=1.0456 | lr=6.95e-06 | 6.15 steps/s | VRAM=27.81 Go [2026-07-18 19:47:29] step 957100/1000000 | loss=1.9883 vocab=1.0929 token=1.1573 | lr=6.93e-06 | 6.16 steps/s | VRAM=29.53 Go [2026-07-18 19:47:37] step 957150/1000000 | loss=2.0389 vocab=0.9012 token=1.0482 | lr=6.92e-06 | 6.14 steps/s | VRAM=24.83 Go [2026-07-18 19:47:45] step 957200/1000000 | loss=2.0538 vocab=0.9813 token=1.3824 | lr=6.91e-06 | 6.17 steps/s | VRAM=31.73 Go [2026-07-18 19:47:53] step 957250/1000000 | loss=2.0167 vocab=1.0149 token=1.4572 | lr=6.89e-06 | 6.16 steps/s | VRAM=31.46 Go [2026-07-18 19:48:01] step 957300/1000000 | loss=2.0557 vocab=0.6364 token=1.2949 | lr=6.88e-06 | 6.17 steps/s | VRAM=25.62 Go [2026-07-18 19:48:09] step 957350/1000000 | loss=1.9710 vocab=0.9832 token=1.2808 | lr=6.86e-06 | 6.11 steps/s | VRAM=28.59 Go [2026-07-18 19:48:17] step 957400/1000000 | loss=1.9738 vocab=0.7095 token=1.2187 | lr=6.85e-06 | 6.11 steps/s | VRAM=25.88 Go [2026-07-18 19:48:25] step 957450/1000000 | loss=1.9842 vocab=0.9940 token=1.2766 | lr=6.84e-06 | 6.19 steps/s | VRAM=29.69 Go [2026-07-18 19:48:34] step 957500/1000000 | loss=2.0273 vocab=0.8677 token=1.3265 | lr=6.82e-06 | 6.17 steps/s | VRAM=31.10 Go [2026-07-18 19:48:42] step 957550/1000000 | loss=2.0329 vocab=0.4601 token=1.0765 | lr=6.81e-06 | 6.16 steps/s | VRAM=28.68 Go [2026-07-18 19:48:50] step 957600/1000000 | loss=2.1424 vocab=0.7360 token=0.9492 | lr=6.80e-06 | 6.18 steps/s | VRAM=28.55 Go [2026-07-18 19:48:58] step 957650/1000000 | loss=1.9845 vocab=1.6267 token=1.8075 | lr=6.78e-06 | 6.14 steps/s | VRAM=27.84 Go [2026-07-18 19:49:06] step 957700/1000000 | loss=2.0822 vocab=0.8926 token=0.7507 | lr=6.77e-06 | 6.12 steps/s | VRAM=29.34 Go [2026-07-18 19:49:14] step 957750/1000000 | loss=2.0198 vocab=0.5179 token=0.7822 | lr=6.76e-06 | 6.16 steps/s | VRAM=26.51 Go [2026-07-18 19:49:22] step 957800/1000000 | loss=1.9815 vocab=1.0456 token=1.2571 | lr=6.74e-06 | 6.18 steps/s | VRAM=32.29 Go [2026-07-18 19:49:30] step 957850/1000000 | loss=1.9847 vocab=0.5856 token=1.3695 | lr=6.73e-06 | 6.11 steps/s | VRAM=29.50 Go [2026-07-18 19:49:39] step 957900/1000000 | loss=2.0710 vocab=0.8251 token=1.0236 | lr=6.72e-06 | 6.16 steps/s | VRAM=25.80 Go [2026-07-18 19:49:47] step 957950/1000000 | loss=1.9474 vocab=0.9370 token=1.1042 | lr=6.70e-06 | 6.13 steps/s | VRAM=28.50 Go [2026-07-18 19:49:55] step 958000/1000000 | loss=2.0277 vocab=0.6784 token=1.0940 | lr=6.69e-06 | 6.13 steps/s | VRAM=26.31 Go [2026-07-18 19:50:03] step 958050/1000000 | loss=1.9704 vocab=0.5944 token=1.1005 | lr=6.68e-06 | 6.15 steps/s | VRAM=29.08 Go [2026-07-18 19:50:11] step 958100/1000000 | loss=1.9100 vocab=0.8088 token=1.0840 | lr=6.66e-06 | 6.15 steps/s | VRAM=28.84 Go [2026-07-18 19:50:19] step 958150/1000000 | loss=2.0773 vocab=0.5472 token=1.1598 | lr=6.65e-06 | 6.16 steps/s | VRAM=29.00 Go [2026-07-18 19:50:27] step 958200/1000000 | loss=2.0719 vocab=0.8828 token=1.1248 | lr=6.64e-06 | 6.15 steps/s | VRAM=32.75 Go [2026-07-18 19:50:35] step 958250/1000000 | loss=1.9925 vocab=0.4925 token=0.8730 | lr=6.62e-06 | 6.14 steps/s | VRAM=25.95 Go [2026-07-18 19:50:44] step 958300/1000000 | loss=1.9647 vocab=0.8593 token=1.4150 | lr=6.61e-06 | 6.15 steps/s | VRAM=29.21 Go [2026-07-18 19:50:52] step 958350/1000000 | loss=1.9168 vocab=0.5770 token=1.0866 | lr=6.60e-06 | 6.14 steps/s | VRAM=28.74 Go [2026-07-18 19:51:00] step 958400/1000000 | loss=1.9805 vocab=0.7366 token=0.9876 | lr=6.59e-06 | 6.11 steps/s | VRAM=32.68 Go [2026-07-18 19:51:08] step 958450/1000000 | loss=2.0069 vocab=0.3855 token=0.6569 | lr=6.57e-06 | 6.14 steps/s | VRAM=26.19 Go [2026-07-18 19:51:16] step 958500/1000000 | loss=1.9713 vocab=0.9001 token=1.2140 | lr=6.56e-06 | 6.18 steps/s | VRAM=29.77 Go [2026-07-18 19:51:24] step 958550/1000000 | loss=1.9842 vocab=1.1072 token=1.3744 | lr=6.55e-06 | 6.17 steps/s | VRAM=29.43 Go [2026-07-18 19:51:32] step 958600/1000000 | loss=2.1134 vocab=1.1923 token=1.5445 | lr=6.53e-06 | 6.16 steps/s | VRAM=28.68 Go [2026-07-18 19:51:41] step 958650/1000000 | loss=2.0926 vocab=0.9864 token=1.4384 | lr=6.52e-06 | 6.12 steps/s | VRAM=29.19 Go [2026-07-18 19:51:49] step 958700/1000000 | loss=2.0367 vocab=0.6132 token=0.8683 | lr=6.51e-06 | 6.13 steps/s | VRAM=26.19 Go [2026-07-18 19:51:57] step 958750/1000000 | loss=1.9943 vocab=0.7905 token=1.2658 | lr=6.49e-06 | 6.18 steps/s | VRAM=32.82 Go [2026-07-18 19:52:05] step 958800/1000000 | loss=1.9345 vocab=0.7068 token=0.9720 | lr=6.48e-06 | 6.14 steps/s | VRAM=28.71 Go [2026-07-18 19:52:13] step 958850/1000000 | loss=1.9319 vocab=0.9327 token=1.0177 | lr=6.47e-06 | 6.16 steps/s | VRAM=28.09 Go [2026-07-18 19:52:21] step 958900/1000000 | loss=2.0615 vocab=0.5393 token=0.8099 | lr=6.45e-06 | 6.12 steps/s | VRAM=32.90 Go [2026-07-18 19:52:29] step 958950/1000000 | loss=2.0000 vocab=1.0636 token=1.0232 | lr=6.44e-06 | 6.16 steps/s | VRAM=31.17 Go [2026-07-18 19:52:38] step 959000/1000000 | loss=1.9632 vocab=1.2789 token=1.1690 | lr=6.43e-06 | 6.14 steps/s | VRAM=29.43 Go [2026-07-18 19:52:46] step 959050/1000000 | loss=2.1002 vocab=0.9518 token=1.5482 | lr=6.42e-06 | 6.15 steps/s | VRAM=26.00 Go [2026-07-18 19:52:54] step 959100/1000000 | loss=1.9657 vocab=0.5523 token=1.0462 | lr=6.40e-06 | 6.15 steps/s | VRAM=28.49 Go [2026-07-18 19:53:02] step 959150/1000000 | loss=1.9916 vocab=1.3049 token=1.4787 | lr=6.39e-06 | 6.13 steps/s | VRAM=26.35 Go [2026-07-18 19:53:10] step 959200/1000000 | loss=2.0188 vocab=0.8579 token=1.1501 | lr=6.38e-06 | 6.13 steps/s | VRAM=29.05 Go [2026-07-18 19:53:18] step 959250/1000000 | loss=1.9588 vocab=0.9446 token=1.0875 | lr=6.36e-06 | 6.13 steps/s | VRAM=26.35 Go [2026-07-18 19:53:26] step 959300/1000000 | loss=1.9835 vocab=0.7463 token=0.6185 | lr=6.35e-06 | 6.14 steps/s | VRAM=28.09 Go [2026-07-18 19:53:35] step 959350/1000000 | loss=1.9917 vocab=0.6400 token=0.9407 | lr=6.34e-06 | 6.12 steps/s | VRAM=31.59 Go [2026-07-18 19:53:43] step 959400/1000000 | loss=1.9606 vocab=0.7534 token=1.0717 | lr=6.32e-06 | 6.12 steps/s | VRAM=26.08 Go [2026-07-18 19:53:51] step 959450/1000000 | loss=1.9993 vocab=0.7447 token=0.9096 | lr=6.31e-06 | 6.17 steps/s | VRAM=32.95 Go [2026-07-18 19:53:59] step 959500/1000000 | loss=2.1072 vocab=0.8394 token=1.0233 | lr=6.30e-06 | 6.14 steps/s | VRAM=31.91 Go [2026-07-18 19:54:07] step 959550/1000000 | loss=1.9590 vocab=1.1492 token=1.2953 | lr=6.29e-06 | 6.14 steps/s | VRAM=31.96 Go [2026-07-18 19:54:15] step 959600/1000000 | loss=1.9107 vocab=1.2174 token=1.1230 | lr=6.27e-06 | 6.15 steps/s | VRAM=32.73 Go [2026-07-18 19:54:23] step 959650/1000000 | loss=2.0498 vocab=0.4981 token=0.9522 | lr=6.26e-06 | 6.15 steps/s | VRAM=29.17 Go [2026-07-18 19:54:32] step 959700/1000000 | loss=2.1016 vocab=0.8624 token=1.3404 | lr=6.25e-06 | 6.14 steps/s | VRAM=29.59 Go [2026-07-18 19:54:40] step 959750/1000000 | loss=2.0833 vocab=0.9872 token=0.9671 | lr=6.23e-06 | 6.11 steps/s | VRAM=29.66 Go [2026-07-18 19:54:48] step 959800/1000000 | loss=1.9825 vocab=0.6290 token=1.0492 | lr=6.22e-06 | 6.16 steps/s | VRAM=28.49 Go [2026-07-18 19:54:56] step 959850/1000000 | loss=2.0721 vocab=0.8871 token=1.1613 | lr=6.21e-06 | 6.16 steps/s | VRAM=30.97 Go [2026-07-18 19:55:04] step 959900/1000000 | loss=2.0216 vocab=0.3692 token=0.9441 | lr=6.20e-06 | 6.15 steps/s | VRAM=28.78 Go [2026-07-18 19:55:12] step 959950/1000000 | loss=2.0015 vocab=0.4885 token=0.8592 | lr=6.18e-06 | 6.15 steps/s | VRAM=25.83 Go [2026-07-18 19:55:20] step 960000/1000000 | loss=1.9337 vocab=0.6859 token=1.0635 | lr=6.17e-06 | 6.15 steps/s | VRAM=28.58 Go [2026-07-18 19:55:28] step 960050/1000000 | loss=2.0295 vocab=0.9696 token=1.1247 | lr=6.16e-06 | 6.17 steps/s | VRAM=29.66 Go [2026-07-18 19:55:37] step 960100/1000000 | loss=2.0338 vocab=0.7347 token=1.2990 | lr=6.15e-06 | 6.15 steps/s | VRAM=28.51 Go [2026-07-18 19:55:45] step 960150/1000000 | loss=2.0435 vocab=1.0717 token=1.2747 | lr=6.13e-06 | 6.13 steps/s | VRAM=32.44 Go [2026-07-18 19:55:53] step 960200/1000000 | loss=1.9670 vocab=0.9413 token=1.6256 | lr=6.12e-06 | 6.15 steps/s | VRAM=25.75 Go [2026-07-18 19:56:01] step 960250/1000000 | loss=1.9771 vocab=0.8985 token=0.9791 | lr=6.11e-06 | 6.11 steps/s | VRAM=26.41 Go [2026-07-18 19:56:09] step 960300/1000000 | loss=2.0142 vocab=0.4804 token=1.1390 | lr=6.10e-06 | 6.13 steps/s | VRAM=29.12 Go [2026-07-18 19:56:17] step 960350/1000000 | loss=2.0197 vocab=0.8387 token=1.4477 | lr=6.08e-06 | 6.15 steps/s | VRAM=27.83 Go [2026-07-18 19:56:25] step 960400/1000000 | loss=2.0043 vocab=1.4940 token=1.6001 | lr=6.07e-06 | 6.18 steps/s | VRAM=28.08 Go [2026-07-18 19:56:34] step 960450/1000000 | loss=2.0031 vocab=1.3049 token=1.6391 | lr=6.06e-06 | 6.14 steps/s | VRAM=29.86 Go [2026-07-18 19:56:42] step 960500/1000000 | loss=1.9465 vocab=0.3341 token=0.8821 | lr=6.05e-06 | 6.16 steps/s | VRAM=28.36 Go [2026-07-18 19:56:50] step 960550/1000000 | loss=1.9937 vocab=1.2013 token=1.4802 | lr=6.03e-06 | 6.17 steps/s | VRAM=29.38 Go [2026-07-18 19:56:58] step 960600/1000000 | loss=2.0289 vocab=0.4445 token=0.8309 | lr=6.02e-06 | 6.14 steps/s | VRAM=26.25 Go [2026-07-18 19:57:06] step 960650/1000000 | loss=1.9713 vocab=0.6974 token=1.2972 | lr=6.01e-06 | 6.15 steps/s | VRAM=28.80 Go [2026-07-18 19:57:14] step 960700/1000000 | loss=2.0440 vocab=0.8685 token=1.1222 | lr=5.99e-06 | 6.10 steps/s | VRAM=28.42 Go [2026-07-18 19:57:22] step 960750/1000000 | loss=2.0379 vocab=1.0393 token=1.2221 | lr=5.98e-06 | 6.17 steps/s | VRAM=31.94 Go [2026-07-18 19:57:31] step 960800/1000000 | loss=2.0160 vocab=0.7910 token=1.2357 | lr=5.97e-06 | 6.12 steps/s | VRAM=26.19 Go [2026-07-18 19:57:39] step 960850/1000000 | loss=1.8634 vocab=0.5644 token=0.8720 | lr=5.96e-06 | 6.15 steps/s | VRAM=29.43 Go [2026-07-18 19:57:47] step 960900/1000000 | loss=1.8788 vocab=1.1477 token=1.5198 | lr=5.95e-06 | 6.14 steps/s | VRAM=28.29 Go [2026-07-18 19:57:55] step 960950/1000000 | loss=1.9573 vocab=0.9560 token=1.3884 | lr=5.93e-06 | 6.15 steps/s | VRAM=24.48 Go [2026-07-18 19:58:03] step 961000/1000000 | loss=1.9713 vocab=0.7050 token=1.2539 | lr=5.92e-06 | 6.12 steps/s | VRAM=26.02 Go [2026-07-18 19:58:11] step 961050/1000000 | loss=2.0625 vocab=1.0720 token=1.1359 | lr=5.91e-06 | 6.13 steps/s | VRAM=30.22 Go [2026-07-18 19:58:19] step 961100/1000000 | loss=2.0624 vocab=0.4913 token=0.8557 | lr=5.90e-06 | 6.15 steps/s | VRAM=32.96 Go [2026-07-18 19:58:28] step 961150/1000000 | loss=2.0746 vocab=0.6894 token=1.0675 | lr=5.88e-06 | 6.15 steps/s | VRAM=29.79 Go [2026-07-18 19:58:36] step 961200/1000000 | loss=1.9049 vocab=0.9163 token=1.1939 | lr=5.87e-06 | 6.18 steps/s | VRAM=30.84 Go [2026-07-18 19:58:44] step 961250/1000000 | loss=2.0347 vocab=1.0166 token=1.1819 | lr=5.86e-06 | 6.14 steps/s | VRAM=32.65 Go [2026-07-18 19:58:52] step 961300/1000000 | loss=2.0902 vocab=0.9424 token=1.2637 | lr=5.85e-06 | 6.09 steps/s | VRAM=25.59 Go [2026-07-18 19:59:00] step 961350/1000000 | loss=2.0584 vocab=1.1340 token=1.2456 | lr=5.83e-06 | 6.14 steps/s | VRAM=28.94 Go [2026-07-18 19:59:08] step 961400/1000000 | loss=2.0451 vocab=1.6799 token=1.1997 | lr=5.82e-06 | 6.14 steps/s | VRAM=24.15 Go [2026-07-18 19:59:16] step 961450/1000000 | loss=2.0305 vocab=0.6976 token=0.8502 | lr=5.81e-06 | 6.13 steps/s | VRAM=27.45 Go [2026-07-18 19:59:25] step 961500/1000000 | loss=2.0125 vocab=0.8792 token=1.4510 | lr=5.80e-06 | 6.14 steps/s | VRAM=28.25 Go [2026-07-18 19:59:33] step 961550/1000000 | loss=1.9180 vocab=0.8953 token=1.2599 | lr=5.78e-06 | 6.15 steps/s | VRAM=29.29 Go [2026-07-18 19:59:41] step 961600/1000000 | loss=2.0094 vocab=1.0891 token=1.1726 | lr=5.77e-06 | 6.16 steps/s | VRAM=25.11 Go [2026-07-18 19:59:49] step 961650/1000000 | loss=1.9949 vocab=1.5483 token=1.9112 | lr=5.76e-06 | 6.15 steps/s | VRAM=28.57 Go [2026-07-18 19:59:57] step 961700/1000000 | loss=2.1019 vocab=0.9173 token=1.3075 | lr=5.75e-06 | 6.13 steps/s | VRAM=29.76 Go [2026-07-18 20:00:05] step 961750/1000000 | loss=2.0231 vocab=0.7018 token=1.1312 | lr=5.74e-06 | 6.14 steps/s | VRAM=32.12 Go [2026-07-18 20:00:13] step 961800/1000000 | loss=2.0132 vocab=0.9103 token=1.4109 | lr=5.72e-06 | 6.16 steps/s | VRAM=26.51 Go [2026-07-18 20:00:21] step 961850/1000000 | loss=1.9713 vocab=0.4706 token=1.1806 | lr=5.71e-06 | 6.18 steps/s | VRAM=28.75 Go [2026-07-18 20:00:30] step 961900/1000000 | loss=1.9940 vocab=0.5817 token=1.0500 | lr=5.70e-06 | 6.15 steps/s | VRAM=29.72 Go [2026-07-18 20:00:38] step 961950/1000000 | loss=1.9799 vocab=0.7582 token=1.0052 | lr=5.69e-06 | 6.14 steps/s | VRAM=28.24 Go [2026-07-18 20:00:46] step 962000/1000000 | loss=1.9798 vocab=1.1196 token=1.6013 | lr=5.68e-06 | 6.17 steps/s | VRAM=31.65 Go [2026-07-18 20:00:54] step 962050/1000000 | loss=1.9640 vocab=1.0102 token=0.8707 | lr=5.66e-06 | 6.17 steps/s | VRAM=27.99 Go [2026-07-18 20:01:02] step 962100/1000000 | loss=2.0912 vocab=0.6193 token=1.3741 | lr=5.65e-06 | 6.16 steps/s | VRAM=31.66 Go [2026-07-18 20:01:10] step 962150/1000000 | loss=1.9820 vocab=0.7879 token=1.3332 | lr=5.64e-06 | 6.16 steps/s | VRAM=28.70 Go [2026-07-18 20:01:18] step 962200/1000000 | loss=1.9699 vocab=0.8383 token=0.9777 | lr=5.63e-06 | 6.10 steps/s | VRAM=26.71 Go [2026-07-18 20:01:26] step 962250/1000000 | loss=2.0366 vocab=0.3988 token=0.6432 | lr=5.61e-06 | 6.15 steps/s | VRAM=32.56 Go [2026-07-18 20:01:35] step 962300/1000000 | loss=1.9576 vocab=1.1545 token=0.9725 | lr=5.60e-06 | 6.14 steps/s | VRAM=28.17 Go [2026-07-18 20:01:43] step 962350/1000000 | loss=2.0556 vocab=1.2885 token=1.7170 | lr=5.59e-06 | 6.14 steps/s | VRAM=32.59 Go [2026-07-18 20:01:51] step 962400/1000000 | loss=2.0117 vocab=1.2214 token=0.9232 | lr=5.58e-06 | 6.15 steps/s | VRAM=25.58 Go [2026-07-18 20:01:59] step 962450/1000000 | loss=2.1489 vocab=0.6549 token=1.2647 | lr=5.57e-06 | 6.15 steps/s | VRAM=28.61 Go [2026-07-18 20:02:07] step 962500/1000000 | loss=1.9175 vocab=0.5609 token=1.3689 | lr=5.55e-06 | 6.15 steps/s | VRAM=32.21 Go [2026-07-18 20:02:15] step 962550/1000000 | loss=1.9868 vocab=0.9489 token=0.6715 | lr=5.54e-06 | 6.16 steps/s | VRAM=31.13 Go [2026-07-18 20:02:23] step 962600/1000000 | loss=1.9509 vocab=0.8149 token=1.2306 | lr=5.53e-06 | 6.17 steps/s | VRAM=24.64 Go [2026-07-18 20:02:31] step 962650/1000000 | loss=2.1025 vocab=0.5774 token=0.6017 | lr=5.52e-06 | 6.13 steps/s | VRAM=28.81 Go [2026-07-18 20:02:40] step 962700/1000000 | loss=2.0218 vocab=0.5655 token=1.3988 | lr=5.51e-06 | 6.14 steps/s | VRAM=26.27 Go [2026-07-18 20:02:48] step 962750/1000000 | loss=2.1055 vocab=0.9481 token=1.2314 | lr=5.50e-06 | 6.14 steps/s | VRAM=29.45 Go [2026-07-18 20:02:56] step 962800/1000000 | loss=2.0712 vocab=0.9531 token=1.2503 | lr=5.48e-06 | 6.14 steps/s | VRAM=31.87 Go [2026-07-18 20:03:04] step 962850/1000000 | loss=1.9520 vocab=1.0024 token=1.1519 | lr=5.47e-06 | 6.12 steps/s | VRAM=28.51 Go [2026-07-18 20:03:12] step 962900/1000000 | loss=1.9078 vocab=0.5066 token=1.2268 | lr=5.46e-06 | 6.12 steps/s | VRAM=26.55 Go [2026-07-18 20:03:20] step 962950/1000000 | loss=2.0134 vocab=0.9133 token=1.0079 | lr=5.45e-06 | 6.14 steps/s | VRAM=24.55 Go [2026-07-18 20:03:29] step 963000/1000000 | loss=1.9009 vocab=0.8179 token=0.9759 | lr=5.44e-06 | 6.13 steps/s | VRAM=29.31 Go [2026-07-18 20:03:37] step 963050/1000000 | loss=1.9413 vocab=0.5589 token=1.0027 | lr=5.42e-06 | 6.18 steps/s | VRAM=31.29 Go [2026-07-18 20:03:45] step 963100/1000000 | loss=2.0149 vocab=0.7909 token=1.2339 | lr=5.41e-06 | 6.15 steps/s | VRAM=29.28 Go [2026-07-18 20:03:53] step 963150/1000000 | loss=1.9539 vocab=0.6533 token=1.2739 | lr=5.40e-06 | 6.12 steps/s | VRAM=28.83 Go [2026-07-18 20:04:01] step 963200/1000000 | loss=1.9400 vocab=1.1793 token=1.3643 | lr=5.39e-06 | 6.13 steps/s | VRAM=29.53 Go [2026-07-18 20:04:09] step 963250/1000000 | loss=2.0457 vocab=1.0795 token=1.3963 | lr=5.38e-06 | 6.17 steps/s | VRAM=28.13 Go [2026-07-18 20:04:17] step 963300/1000000 | loss=2.0077 vocab=0.8899 token=1.4128 | lr=5.37e-06 | 6.18 steps/s | VRAM=28.44 Go [2026-07-18 20:04:25] step 963350/1000000 | loss=2.0846 vocab=0.7458 token=1.3237 | lr=5.35e-06 | 6.15 steps/s | VRAM=28.62 Go [2026-07-18 20:04:34] step 963400/1000000 | loss=1.9514 vocab=0.2958 token=0.6472 | lr=5.34e-06 | 6.13 steps/s | VRAM=28.78 Go [2026-07-18 20:04:42] step 963450/1000000 | loss=2.0269 vocab=0.8721 token=1.2939 | lr=5.33e-06 | 6.15 steps/s | VRAM=28.42 Go [2026-07-18 20:04:50] step 963500/1000000 | loss=2.1341 vocab=1.0798 token=1.3207 | lr=5.32e-06 | 6.13 steps/s | VRAM=28.48 Go [2026-07-18 20:04:58] step 963550/1000000 | loss=2.0373 vocab=0.5957 token=1.0221 | lr=5.31e-06 | 6.16 steps/s | VRAM=27.96 Go [2026-07-18 20:05:06] step 963600/1000000 | loss=2.0681 vocab=0.9829 token=1.1162 | lr=5.30e-06 | 6.13 steps/s | VRAM=30.71 Go [2026-07-18 20:05:14] step 963650/1000000 | loss=1.9875 vocab=0.8489 token=0.7829 | lr=5.28e-06 | 6.15 steps/s | VRAM=27.53 Go [2026-07-18 20:05:22] step 963700/1000000 | loss=2.0808 vocab=0.4725 token=0.7697 | lr=5.27e-06 | 6.15 steps/s | VRAM=32.20 Go [2026-07-18 20:05:31] step 963750/1000000 | loss=2.0365 vocab=1.2463 token=1.4337 | lr=5.26e-06 | 6.09 steps/s | VRAM=23.41 Go [2026-07-18 20:05:39] step 963800/1000000 | loss=2.0275 vocab=1.2527 token=1.2582 | lr=5.25e-06 | 6.15 steps/s | VRAM=28.71 Go [2026-07-18 20:05:47] step 963850/1000000 | loss=2.0204 vocab=0.8087 token=1.1745 | lr=5.24e-06 | 6.16 steps/s | VRAM=28.81 Go [2026-07-18 20:05:55] step 963900/1000000 | loss=1.9873 vocab=0.7411 token=0.9908 | lr=5.23e-06 | 6.15 steps/s | VRAM=28.99 Go [2026-07-18 20:06:03] step 963950/1000000 | loss=2.0279 vocab=1.3662 token=1.4788 | lr=5.21e-06 | 6.12 steps/s | VRAM=28.76 Go [2026-07-18 20:06:11] step 964000/1000000 | loss=1.9328 vocab=0.5924 token=1.0860 | lr=5.20e-06 | 6.12 steps/s | VRAM=28.18 Go [2026-07-18 20:06:19] step 964050/1000000 | loss=2.0159 vocab=0.6342 token=0.9386 | lr=5.19e-06 | 6.12 steps/s | VRAM=26.52 Go [2026-07-18 20:06:28] step 964100/1000000 | loss=1.8841 vocab=0.5596 token=1.0976 | lr=5.18e-06 | 6.12 steps/s | VRAM=32.59 Go [2026-07-18 20:06:36] step 964150/1000000 | loss=2.0370 vocab=0.8862 token=1.3352 | lr=5.17e-06 | 6.11 steps/s | VRAM=28.37 Go [2026-07-18 20:06:44] step 964200/1000000 | loss=2.0063 vocab=0.6394 token=1.1370 | lr=5.16e-06 | 6.15 steps/s | VRAM=32.29 Go [2026-07-18 20:06:52] step 964250/1000000 | loss=1.9614 vocab=0.8838 token=0.9528 | lr=5.15e-06 | 6.11 steps/s | VRAM=28.41 Go [2026-07-18 20:07:00] step 964300/1000000 | loss=2.0407 vocab=1.1703 token=0.9905 | lr=5.13e-06 | 6.14 steps/s | VRAM=26.29 Go [2026-07-18 20:07:08] step 964350/1000000 | loss=2.0171 vocab=0.8373 token=1.2101 | lr=5.12e-06 | 6.13 steps/s | VRAM=29.63 Go [2026-07-18 20:07:17] step 964400/1000000 | loss=2.0190 vocab=0.9064 token=0.9793 | lr=5.11e-06 | 6.15 steps/s | VRAM=24.99 Go [2026-07-18 20:07:25] step 964450/1000000 | loss=1.9283 vocab=0.8187 token=1.3869 | lr=5.10e-06 | 6.17 steps/s | VRAM=25.35 Go [2026-07-18 20:07:33] step 964500/1000000 | loss=2.1697 vocab=0.8993 token=1.4273 | lr=5.09e-06 | 6.12 steps/s | VRAM=26.48 Go [2026-07-18 20:07:41] step 964550/1000000 | loss=2.0545 vocab=1.0233 token=1.4578 | lr=5.08e-06 | 6.17 steps/s | VRAM=27.75 Go [2026-07-18 20:07:49] step 964600/1000000 | loss=2.0211 vocab=0.4547 token=1.0752 | lr=5.07e-06 | 6.12 steps/s | VRAM=27.08 Go [2026-07-18 20:07:57] step 964650/1000000 | loss=1.9386 vocab=0.5812 token=0.9765 | lr=5.05e-06 | 6.14 steps/s | VRAM=30.17 Go [2026-07-18 20:08:05] step 964700/1000000 | loss=2.0010 vocab=1.2061 token=1.2524 | lr=5.04e-06 | 6.15 steps/s | VRAM=26.19 Go [2026-07-18 20:08:14] step 964750/1000000 | loss=2.0753 vocab=0.5392 token=0.7544 | lr=5.03e-06 | 6.16 steps/s | VRAM=32.66 Go [2026-07-18 20:08:22] step 964800/1000000 | loss=2.0348 vocab=0.5720 token=0.8928 | lr=5.02e-06 | 6.15 steps/s | VRAM=27.97 Go [2026-07-18 20:08:30] step 964850/1000000 | loss=2.1116 vocab=1.2520 token=1.2457 | lr=5.01e-06 | 6.15 steps/s | VRAM=32.33 Go [2026-07-18 20:08:38] step 964900/1000000 | loss=1.9333 vocab=0.7766 token=1.0621 | lr=5.00e-06 | 6.13 steps/s | VRAM=28.95 Go [2026-07-18 20:08:46] step 964950/1000000 | loss=1.8860 vocab=0.8322 token=1.1299 | lr=4.99e-06 | 6.13 steps/s | VRAM=31.96 Go [2026-07-18 20:08:54] step 965000/1000000 | loss=2.0632 vocab=0.7303 token=1.0921 | lr=4.98e-06 | 6.14 steps/s | VRAM=26.00 Go [2026-07-18 20:09:02] step 965050/1000000 | loss=1.9223 vocab=1.4948 token=0.9223 | lr=4.96e-06 | 6.16 steps/s | VRAM=26.14 Go [2026-07-18 20:09:10] step 965100/1000000 | loss=1.9837 vocab=0.8323 token=1.3770 | lr=4.95e-06 | 6.13 steps/s | VRAM=31.70 Go [2026-07-18 20:09:19] step 965150/1000000 | loss=1.9369 vocab=0.8553 token=0.9613 | lr=4.94e-06 | 6.15 steps/s | VRAM=28.18 Go [2026-07-18 20:09:27] step 965200/1000000 | loss=2.0364 vocab=1.0816 token=1.4795 | lr=4.93e-06 | 6.18 steps/s | VRAM=31.80 Go [2026-07-18 20:09:35] step 965250/1000000 | loss=2.0810 vocab=1.1005 token=1.5034 | lr=4.92e-06 | 6.14 steps/s | VRAM=28.59 Go [2026-07-18 20:09:43] step 965300/1000000 | loss=2.0455 vocab=0.7884 token=1.0877 | lr=4.91e-06 | 6.16 steps/s | VRAM=32.65 Go [2026-07-18 20:09:51] step 965350/1000000 | loss=2.0804 vocab=1.1976 token=1.6339 | lr=4.90e-06 | 6.15 steps/s | VRAM=28.21 Go [2026-07-18 20:09:59] step 965400/1000000 | loss=2.0271 vocab=0.4493 token=1.3403 | lr=4.89e-06 | 6.15 steps/s | VRAM=28.81 Go [2026-07-18 20:10:07] step 965450/1000000 | loss=2.0523 vocab=0.8236 token=1.2980 | lr=4.88e-06 | 6.14 steps/s | VRAM=31.70 Go [2026-07-18 20:10:16] step 965500/1000000 | loss=1.9247 vocab=0.7963 token=0.9501 | lr=4.86e-06 | 6.11 steps/s | VRAM=29.04 Go [2026-07-18 20:10:24] step 965550/1000000 | loss=2.0114 vocab=0.9206 token=1.1853 | lr=4.85e-06 | 6.15 steps/s | VRAM=25.65 Go [2026-07-18 20:10:32] step 965600/1000000 | loss=2.0068 vocab=0.4302 token=1.0074 | lr=4.84e-06 | 6.14 steps/s | VRAM=31.73 Go [2026-07-18 20:10:40] step 965650/1000000 | loss=2.0322 vocab=1.0948 token=1.1999 | lr=4.83e-06 | 6.14 steps/s | VRAM=29.05 Go [2026-07-18 20:10:48] step 965700/1000000 | loss=1.9917 vocab=1.2594 token=1.5581 | lr=4.82e-06 | 6.14 steps/s | VRAM=30.31 Go [2026-07-18 20:10:56] step 965750/1000000 | loss=2.0617 vocab=0.9040 token=1.4303 | lr=4.81e-06 | 6.11 steps/s | VRAM=28.33 Go [2026-07-18 20:11:04] step 965800/1000000 | loss=2.1155 vocab=0.5813 token=1.2820 | lr=4.80e-06 | 6.14 steps/s | VRAM=31.83 Go [2026-07-18 20:11:13] step 965850/1000000 | loss=2.0005 vocab=1.1895 token=1.3039 | lr=4.79e-06 | 6.15 steps/s | VRAM=29.26 Go [2026-07-18 20:11:21] step 965900/1000000 | loss=2.0375 vocab=0.8266 token=1.2953 | lr=4.78e-06 | 6.12 steps/s | VRAM=28.96 Go [2026-07-18 20:11:29] step 965950/1000000 | loss=2.0298 vocab=0.7488 token=1.1505 | lr=4.77e-06 | 6.10 steps/s | VRAM=26.07 Go [2026-07-18 20:11:37] step 966000/1000000 | loss=2.0726 vocab=0.7631 token=1.0586 | lr=4.75e-06 | 6.15 steps/s | VRAM=28.24 Go [2026-07-18 20:11:45] step 966050/1000000 | loss=2.0858 vocab=0.8852 token=1.2660 | lr=4.74e-06 | 6.11 steps/s | VRAM=25.89 Go [2026-07-18 20:11:53] step 966100/1000000 | loss=2.0004 vocab=1.6641 token=1.5217 | lr=4.73e-06 | 6.16 steps/s | VRAM=32.48 Go [2026-07-18 20:12:02] step 966150/1000000 | loss=1.9921 vocab=0.8579 token=1.2858 | lr=4.72e-06 | 6.12 steps/s | VRAM=26.64 Go [2026-07-18 20:12:10] step 966200/1000000 | loss=2.0412 vocab=0.5459 token=1.1902 | lr=4.71e-06 | 6.13 steps/s | VRAM=33.00 Go [2026-07-18 20:12:18] step 966250/1000000 | loss=1.9710 vocab=0.4719 token=1.0868 | lr=4.70e-06 | 6.13 steps/s | VRAM=26.47 Go [2026-07-18 20:12:26] step 966300/1000000 | loss=2.0299 vocab=1.0356 token=1.3013 | lr=4.69e-06 | 6.12 steps/s | VRAM=28.64 Go [2026-07-18 20:12:34] step 966350/1000000 | loss=1.9977 vocab=0.6763 token=0.6728 | lr=4.68e-06 | 6.12 steps/s | VRAM=26.55 Go [2026-07-18 20:12:42] step 966400/1000000 | loss=1.9989 vocab=0.7040 token=0.9801 | lr=4.67e-06 | 6.16 steps/s | VRAM=29.19 Go [2026-07-18 20:12:50] step 966450/1000000 | loss=1.9947 vocab=0.9730 token=1.3675 | lr=4.66e-06 | 6.16 steps/s | VRAM=28.07 Go [2026-07-18 20:12:59] step 966500/1000000 | loss=2.1178 vocab=1.4538 token=1.3507 | lr=4.65e-06 | 6.16 steps/s | VRAM=32.05 Go [2026-07-18 20:13:07] step 966550/1000000 | loss=1.9643 vocab=1.0197 token=1.0189 | lr=4.64e-06 | 6.12 steps/s | VRAM=26.32 Go [2026-07-18 20:13:15] step 966600/1000000 | loss=1.9362 vocab=1.0755 token=1.6560 | lr=4.62e-06 | 6.16 steps/s | VRAM=27.98 Go [2026-07-18 20:13:23] step 966650/1000000 | loss=2.1165 vocab=0.8646 token=0.9048 | lr=4.61e-06 | 6.12 steps/s | VRAM=32.81 Go [2026-07-18 20:13:31] step 966700/1000000 | loss=1.9879 vocab=0.9770 token=1.3190 | lr=4.60e-06 | 6.13 steps/s | VRAM=25.35 Go [2026-07-18 20:13:39] step 966750/1000000 | loss=1.9852 vocab=1.6862 token=1.1434 | lr=4.59e-06 | 6.14 steps/s | VRAM=24.73 Go [2026-07-18 20:13:47] step 966800/1000000 | loss=2.0527 vocab=0.9402 token=1.1583 | lr=4.58e-06 | 6.12 steps/s | VRAM=32.71 Go [2026-07-18 20:13:56] step 966850/1000000 | loss=2.0688 vocab=0.6243 token=1.1584 | lr=4.57e-06 | 6.15 steps/s | VRAM=31.05 Go [2026-07-18 20:14:04] step 966900/1000000 | loss=1.9980 vocab=0.9331 token=1.5937 | lr=4.56e-06 | 6.16 steps/s | VRAM=32.12 Go [2026-07-18 20:14:12] step 966950/1000000 | loss=2.0130 vocab=0.9874 token=1.3941 | lr=4.55e-06 | 6.17 steps/s | VRAM=31.85 Go [2026-07-18 20:14:20] step 967000/1000000 | loss=2.0089 vocab=0.6740 token=1.2847 | lr=4.54e-06 | 6.13 steps/s | VRAM=25.73 Go [2026-07-18 20:14:28] step 967050/1000000 | loss=2.0013 vocab=1.0016 token=1.7022 | lr=4.53e-06 | 6.16 steps/s | VRAM=31.34 Go [2026-07-18 20:14:36] step 967100/1000000 | loss=2.0029 vocab=1.0401 token=1.2257 | lr=4.52e-06 | 6.13 steps/s | VRAM=26.12 Go [2026-07-18 20:14:44] step 967150/1000000 | loss=2.0283 vocab=0.9707 token=1.5473 | lr=4.51e-06 | 6.11 steps/s | VRAM=29.16 Go [2026-07-18 20:14:53] step 967200/1000000 | loss=2.0427 vocab=0.8476 token=1.3939 | lr=4.50e-06 | 6.14 steps/s | VRAM=29.25 Go [2026-07-18 20:15:01] step 967250/1000000 | loss=2.0526 vocab=0.9248 token=1.3709 | lr=4.49e-06 | 6.14 steps/s | VRAM=32.13 Go [2026-07-18 20:15:09] step 967300/1000000 | loss=2.0602 vocab=1.2978 token=1.2992 | lr=4.48e-06 | 6.14 steps/s | VRAM=31.12 Go [2026-07-18 20:15:17] step 967350/1000000 | loss=1.9890 vocab=0.9822 token=1.3104 | lr=4.47e-06 | 6.14 steps/s | VRAM=31.85 Go [2026-07-18 20:15:25] step 967400/1000000 | loss=1.9148 vocab=0.7254 token=0.9880 | lr=4.46e-06 | 6.10 steps/s | VRAM=25.91 Go [2026-07-18 20:15:33] step 967450/1000000 | loss=2.0627 vocab=0.7725 token=0.8957 | lr=4.44e-06 | 6.12 steps/s | VRAM=26.76 Go [2026-07-18 20:15:42] step 967500/1000000 | loss=1.9872 vocab=1.1493 token=1.2587 | lr=4.43e-06 | 6.14 steps/s | VRAM=28.39 Go [2026-07-18 20:15:50] step 967550/1000000 | loss=2.1575 vocab=0.9826 token=1.2679 | lr=4.42e-06 | 6.16 steps/s | VRAM=32.57 Go [2026-07-18 20:15:58] step 967600/1000000 | loss=2.0646 vocab=0.9648 token=0.8605 | lr=4.41e-06 | 6.15 steps/s | VRAM=31.58 Go [2026-07-18 20:16:06] step 967650/1000000 | loss=1.9160 vocab=0.6435 token=1.1898 | lr=4.40e-06 | 6.12 steps/s | VRAM=30.66 Go [2026-07-18 20:16:14] step 967700/1000000 | loss=2.0660 vocab=0.6727 token=0.9166 | lr=4.39e-06 | 6.13 steps/s | VRAM=32.32 Go [2026-07-18 20:16:22] step 967750/1000000 | loss=2.0460 vocab=0.7214 token=1.2637 | lr=4.38e-06 | 6.13 steps/s | VRAM=28.59 Go [2026-07-18 20:16:30] step 967800/1000000 | loss=2.0771 vocab=0.6754 token=1.3439 | lr=4.37e-06 | 6.13 steps/s | VRAM=29.17 Go [2026-07-18 20:16:39] step 967850/1000000 | loss=1.9513 vocab=1.0889 token=1.5877 | lr=4.36e-06 | 6.13 steps/s | VRAM=25.02 Go [2026-07-18 20:16:47] step 967900/1000000 | loss=2.0255 vocab=0.4896 token=0.9241 | lr=4.35e-06 | 6.14 steps/s | VRAM=31.98 Go [2026-07-18 20:16:55] step 967950/1000000 | loss=2.0656 vocab=1.1241 token=1.5782 | lr=4.34e-06 | 6.14 steps/s | VRAM=31.60 Go [2026-07-18 20:17:03] step 968000/1000000 | loss=2.0132 vocab=0.8556 token=1.2519 | lr=4.33e-06 | 6.13 steps/s | VRAM=26.08 Go [2026-07-18 20:17:11] step 968050/1000000 | loss=2.0135 vocab=0.6035 token=0.6909 | lr=4.32e-06 | 6.15 steps/s | VRAM=28.38 Go [2026-07-18 20:17:19] step 968100/1000000 | loss=1.9096 vocab=0.5672 token=1.4359 | lr=4.31e-06 | 6.14 steps/s | VRAM=31.29 Go [2026-07-18 20:17:27] step 968150/1000000 | loss=2.0206 vocab=0.8727 token=1.0779 | lr=4.30e-06 | 6.16 steps/s | VRAM=24.93 Go [2026-07-18 20:17:36] step 968200/1000000 | loss=2.1026 vocab=1.0152 token=0.7455 | lr=4.29e-06 | 6.17 steps/s | VRAM=32.09 Go [2026-07-18 20:17:44] step 968250/1000000 | loss=2.0075 vocab=1.0753 token=1.4198 | lr=4.28e-06 | 6.15 steps/s | VRAM=31.56 Go [2026-07-18 20:17:52] step 968300/1000000 | loss=2.0307 vocab=0.8354 token=1.3452 | lr=4.27e-06 | 6.14 steps/s | VRAM=28.55 Go [2026-07-18 20:18:00] step 968350/1000000 | loss=1.9701 vocab=0.8643 token=1.1064 | lr=4.26e-06 | 6.14 steps/s | VRAM=28.41 Go [2026-07-18 20:18:08] step 968400/1000000 | loss=1.9442 vocab=0.5706 token=0.9708 | lr=4.25e-06 | 6.15 steps/s | VRAM=28.21 Go [2026-07-18 20:18:16] step 968450/1000000 | loss=1.9958 vocab=0.7307 token=1.1586 | lr=4.24e-06 | 6.15 steps/s | VRAM=31.78 Go [2026-07-18 20:18:24] step 968500/1000000 | loss=1.9416 vocab=0.4688 token=0.9161 | lr=4.23e-06 | 6.14 steps/s | VRAM=28.84 Go [2026-07-18 20:18:33] step 968550/1000000 | loss=2.0684 vocab=0.9515 token=1.2790 | lr=4.22e-06 | 6.13 steps/s | VRAM=29.11 Go [2026-07-18 20:18:41] step 968600/1000000 | loss=2.0251 vocab=0.8072 token=1.1694 | lr=4.21e-06 | 6.15 steps/s | VRAM=29.16 Go [2026-07-18 20:18:49] step 968650/1000000 | loss=1.9478 vocab=0.5097 token=1.1441 | lr=4.20e-06 | 6.15 steps/s | VRAM=28.56 Go [2026-07-18 20:18:57] step 968700/1000000 | loss=2.0366 vocab=1.0701 token=1.1777 | lr=4.19e-06 | 6.15 steps/s | VRAM=25.26 Go [2026-07-18 20:19:05] step 968750/1000000 | loss=2.0556 vocab=1.0207 token=1.2543 | lr=4.18e-06 | 6.14 steps/s | VRAM=32.55 Go [2026-07-18 20:19:13] step 968800/1000000 | loss=2.0878 vocab=0.9382 token=1.1661 | lr=4.17e-06 | 6.16 steps/s | VRAM=25.49 Go [2026-07-18 20:19:21] step 968850/1000000 | loss=1.9756 vocab=1.0696 token=1.2480 | lr=4.16e-06 | 6.12 steps/s | VRAM=25.69 Go [2026-07-18 20:19:29] step 968900/1000000 | loss=2.0607 vocab=0.9744 token=0.9355 | lr=4.15e-06 | 6.14 steps/s | VRAM=31.57 Go [2026-07-18 20:19:38] step 968950/1000000 | loss=2.0692 vocab=0.7366 token=0.9479 | lr=4.14e-06 | 6.15 steps/s | VRAM=30.86 Go [2026-07-18 20:19:46] step 969000/1000000 | loss=1.9061 vocab=1.0892 token=1.3010 | lr=4.13e-06 | 6.11 steps/s | VRAM=29.18 Go [2026-07-18 20:19:54] step 969050/1000000 | loss=1.9944 vocab=1.5245 token=1.1740 | lr=4.12e-06 | 6.11 steps/s | VRAM=26.73 Go [2026-07-18 20:20:02] step 969100/1000000 | loss=1.9855 vocab=0.5790 token=1.3880 | lr=4.11e-06 | 6.14 steps/s | VRAM=28.02 Go [2026-07-18 20:20:10] step 969150/1000000 | loss=2.0486 vocab=1.9249 token=1.0961 | lr=4.10e-06 | 6.16 steps/s | VRAM=30.71 Go [2026-07-18 20:20:18] step 969200/1000000 | loss=2.0646 vocab=0.7379 token=1.1983 | lr=4.09e-06 | 6.14 steps/s | VRAM=29.17 Go [2026-07-18 20:20:27] step 969250/1000000 | loss=1.9731 vocab=0.7881 token=1.4364 | lr=4.08e-06 | 6.10 steps/s | VRAM=28.93 Go [2026-07-18 20:20:35] step 969300/1000000 | loss=2.0126 vocab=1.0119 token=1.1855 | lr=4.07e-06 | 6.14 steps/s | VRAM=25.75 Go [2026-07-18 20:20:43] step 969350/1000000 | loss=2.1475 vocab=0.8678 token=1.4410 | lr=4.06e-06 | 6.14 steps/s | VRAM=24.67 Go [2026-07-18 20:20:51] step 969400/1000000 | loss=2.0272 vocab=0.9551 token=1.0400 | lr=4.05e-06 | 6.16 steps/s | VRAM=24.88 Go [2026-07-18 20:20:59] step 969450/1000000 | loss=2.0058 vocab=0.8930 token=0.9575 | lr=4.04e-06 | 6.12 steps/s | VRAM=32.58 Go [2026-07-18 20:21:07] step 969500/1000000 | loss=1.9857 vocab=0.7148 token=1.0285 | lr=4.03e-06 | 6.16 steps/s | VRAM=28.33 Go [2026-07-18 20:21:15] step 969550/1000000 | loss=2.0043 vocab=0.9441 token=1.4078 | lr=4.02e-06 | 6.11 steps/s | VRAM=28.01 Go [2026-07-18 20:21:24] step 969600/1000000 | loss=2.0028 vocab=1.1751 token=1.2228 | lr=4.01e-06 | 6.15 steps/s | VRAM=32.44 Go [2026-07-18 20:21:32] step 969650/1000000 | loss=2.0602 vocab=0.4940 token=1.0605 | lr=4.00e-06 | 6.18 steps/s | VRAM=29.36 Go [2026-07-18 20:21:40] step 969700/1000000 | loss=1.9796 vocab=0.7777 token=0.9226 | lr=3.99e-06 | 6.17 steps/s | VRAM=29.36 Go [2026-07-18 20:21:48] step 969750/1000000 | loss=2.0012 vocab=0.8708 token=0.8988 | lr=3.98e-06 | 6.18 steps/s | VRAM=31.76 Go [2026-07-18 20:21:56] step 969800/1000000 | loss=1.9696 vocab=0.6692 token=1.1496 | lr=3.97e-06 | 6.13 steps/s | VRAM=28.92 Go [2026-07-18 20:22:04] step 969850/1000000 | loss=2.0425 vocab=0.9749 token=1.0656 | lr=3.96e-06 | 6.14 steps/s | VRAM=29.07 Go [2026-07-18 20:22:12] step 969900/1000000 | loss=2.0617 vocab=1.0356 token=1.4426 | lr=3.95e-06 | 6.13 steps/s | VRAM=22.74 Go [2026-07-18 20:22:20] step 969950/1000000 | loss=2.0437 vocab=0.9804 token=1.1710 | lr=3.94e-06 | 6.13 steps/s | VRAM=24.74 Go [2026-07-18 20:22:29] step 970000/1000000 | loss=2.0508 vocab=1.0189 token=1.1127 | lr=3.93e-06 | 6.14 steps/s | VRAM=32.33 Go [2026-07-18 20:22:37] step 970050/1000000 | loss=2.0399 vocab=0.9323 token=1.4484 | lr=3.92e-06 | 6.17 steps/s | VRAM=25.50 Go [2026-07-18 20:22:45] step 970100/1000000 | loss=1.9920 vocab=0.8448 token=1.3522 | lr=3.91e-06 | 6.14 steps/s | VRAM=29.31 Go [2026-07-18 20:22:53] step 970150/1000000 | loss=2.0460 vocab=0.9821 token=1.2334 | lr=3.90e-06 | 6.15 steps/s | VRAM=28.58 Go [2026-07-18 20:23:01] step 970200/1000000 | loss=2.0081 vocab=0.8424 token=1.1654 | lr=3.89e-06 | 6.17 steps/s | VRAM=28.61 Go [2026-07-18 20:23:09] step 970250/1000000 | loss=2.0031 vocab=0.7414 token=1.4806 | lr=3.88e-06 | 6.15 steps/s | VRAM=28.39 Go [2026-07-18 20:23:17] step 970300/1000000 | loss=2.0270 vocab=0.4879 token=0.7517 | lr=3.87e-06 | 6.13 steps/s | VRAM=26.65 Go [2026-07-18 20:23:26] step 970350/1000000 | loss=2.0527 vocab=1.4298 token=1.2123 | lr=3.86e-06 | 6.12 steps/s | VRAM=29.10 Go [2026-07-18 20:23:34] step 970400/1000000 | loss=1.9910 vocab=0.9405 token=1.4398 | lr=3.85e-06 | 6.13 steps/s | VRAM=32.29 Go [2026-07-18 20:23:42] step 970450/1000000 | loss=1.9792 vocab=0.9536 token=1.2752 | lr=3.85e-06 | 6.15 steps/s | VRAM=26.33 Go [2026-07-18 20:23:50] step 970500/1000000 | loss=2.0047 vocab=0.8144 token=1.3735 | lr=3.84e-06 | 6.15 steps/s | VRAM=29.21 Go [2026-07-18 20:23:58] step 970550/1000000 | loss=1.9989 vocab=0.8675 token=1.1352 | lr=3.83e-06 | 6.15 steps/s | VRAM=32.21 Go [2026-07-18 20:24:06] step 970600/1000000 | loss=1.9546 vocab=0.8872 token=1.1058 | lr=3.82e-06 | 6.16 steps/s | VRAM=31.87 Go [2026-07-18 20:24:14] step 970650/1000000 | loss=1.9669 vocab=1.3776 token=1.3684 | lr=3.81e-06 | 6.15 steps/s | VRAM=31.26 Go [2026-07-18 20:24:22] step 970700/1000000 | loss=2.0647 vocab=0.5447 token=1.1124 | lr=3.80e-06 | 6.13 steps/s | VRAM=28.12 Go [2026-07-18 20:24:31] step 970750/1000000 | loss=1.9984 vocab=0.5060 token=1.1097 | lr=3.79e-06 | 6.16 steps/s | VRAM=25.44 Go [2026-07-18 20:24:39] step 970800/1000000 | loss=2.1429 vocab=0.7204 token=1.0430 | lr=3.78e-06 | 6.12 steps/s | VRAM=26.00 Go [2026-07-18 20:24:47] step 970850/1000000 | loss=1.9799 vocab=0.8735 token=1.2082 | lr=3.77e-06 | 6.14 steps/s | VRAM=26.46 Go [2026-07-18 20:24:55] step 970900/1000000 | loss=2.0845 vocab=0.8256 token=1.3280 | lr=3.76e-06 | 6.13 steps/s | VRAM=26.35 Go [2026-07-18 20:25:03] step 970950/1000000 | loss=1.9989 vocab=0.6303 token=1.1276 | lr=3.75e-06 | 6.14 steps/s | VRAM=28.99 Go [2026-07-18 20:25:11] step 971000/1000000 | loss=2.1093 vocab=0.8677 token=1.0937 | lr=3.74e-06 | 6.16 steps/s | VRAM=28.85 Go [2026-07-18 20:25:19] step 971050/1000000 | loss=2.0073 vocab=0.8874 token=1.3243 | lr=3.73e-06 | 6.14 steps/s | VRAM=25.91 Go [2026-07-18 20:25:28] step 971100/1000000 | loss=2.0469 vocab=0.8169 token=1.1563 | lr=3.72e-06 | 6.15 steps/s | VRAM=26.36 Go [2026-07-18 20:25:36] step 971150/1000000 | loss=2.0407 vocab=0.4806 token=0.8522 | lr=3.71e-06 | 6.15 steps/s | VRAM=28.85 Go [2026-07-18 20:25:44] step 971200/1000000 | loss=2.1099 vocab=0.6819 token=0.8621 | lr=3.70e-06 | 6.13 steps/s | VRAM=28.04 Go [2026-07-18 20:25:52] step 971250/1000000 | loss=2.1170 vocab=0.8932 token=1.2947 | lr=3.69e-06 | 6.14 steps/s | VRAM=28.30 Go [2026-07-18 20:26:00] step 971300/1000000 | loss=2.0164 vocab=0.5438 token=1.2836 | lr=3.69e-06 | 6.12 steps/s | VRAM=27.25 Go [2026-07-18 20:26:08] step 971350/1000000 | loss=2.0653 vocab=0.6494 token=1.0325 | lr=3.68e-06 | 6.13 steps/s | VRAM=29.60 Go [2026-07-18 20:26:16] step 971400/1000000 | loss=2.0992 vocab=0.8856 token=1.1424 | lr=3.67e-06 | 6.18 steps/s | VRAM=31.53 Go [2026-07-18 20:26:25] step 971450/1000000 | loss=1.9879 vocab=0.5994 token=1.2802 | lr=3.66e-06 | 6.15 steps/s | VRAM=32.60 Go [2026-07-18 20:26:33] step 971500/1000000 | loss=1.9177 vocab=0.9708 token=1.2706 | lr=3.65e-06 | 6.10 steps/s | VRAM=30.14 Go [2026-07-18 20:26:41] step 971550/1000000 | loss=2.0397 vocab=1.1454 token=1.3402 | lr=3.64e-06 | 6.14 steps/s | VRAM=28.76 Go [2026-07-18 20:26:49] step 971600/1000000 | loss=2.1252 vocab=0.7466 token=1.2954 | lr=3.63e-06 | 6.15 steps/s | VRAM=28.58 Go [2026-07-18 20:26:57] step 971650/1000000 | loss=2.0625 vocab=1.4104 token=1.7644 | lr=3.62e-06 | 6.14 steps/s | VRAM=28.33 Go [2026-07-18 20:27:05] step 971700/1000000 | loss=2.1274 vocab=1.2877 token=1.4854 | lr=3.61e-06 | 6.16 steps/s | VRAM=28.17 Go [2026-07-18 20:27:13] step 971750/1000000 | loss=2.0087 vocab=0.6722 token=0.9726 | lr=3.60e-06 | 6.11 steps/s | VRAM=29.41 Go [2026-07-18 20:27:22] step 971800/1000000 | loss=1.9381 vocab=0.4084 token=1.2112 | lr=3.59e-06 | 6.17 steps/s | VRAM=31.10 Go [2026-07-18 20:27:30] step 971850/1000000 | loss=2.0526 vocab=0.5104 token=1.1665 | lr=3.58e-06 | 6.14 steps/s | VRAM=28.59 Go [2026-07-18 20:27:38] step 971900/1000000 | loss=2.0331 vocab=0.6629 token=0.9627 | lr=3.58e-06 | 6.12 steps/s | VRAM=28.11 Go [2026-07-18 20:27:46] step 971950/1000000 | loss=2.0027 vocab=0.9249 token=0.9356 | lr=3.57e-06 | 6.10 steps/s | VRAM=28.87 Go [2026-07-18 20:27:54] step 972000/1000000 | loss=1.9592 vocab=0.7897 token=1.2811 | lr=3.56e-06 | 6.13 steps/s | VRAM=31.92 Go [2026-07-18 20:28:02] step 972050/1000000 | loss=2.0062 vocab=0.9537 token=0.8720 | lr=3.55e-06 | 6.12 steps/s | VRAM=28.87 Go [2026-07-18 20:28:11] step 972100/1000000 | loss=2.0174 vocab=0.6883 token=1.2951 | lr=3.54e-06 | 6.16 steps/s | VRAM=28.48 Go [2026-07-18 20:28:19] step 972150/1000000 | loss=1.9165 vocab=0.7822 token=0.9205 | lr=3.53e-06 | 6.14 steps/s | VRAM=29.23 Go [2026-07-18 20:28:27] step 972200/1000000 | loss=2.1230 vocab=1.4707 token=1.8466 | lr=3.52e-06 | 6.13 steps/s | VRAM=25.60 Go [2026-07-18 20:28:35] step 972250/1000000 | loss=1.8948 vocab=0.6539 token=0.8296 | lr=3.51e-06 | 6.14 steps/s | VRAM=27.81 Go [2026-07-18 20:28:43] step 972300/1000000 | loss=2.1214 vocab=0.9295 token=0.7547 | lr=3.50e-06 | 6.11 steps/s | VRAM=26.70 Go [2026-07-18 20:28:51] step 972350/1000000 | loss=1.9958 vocab=0.9566 token=1.0665 | lr=3.49e-06 | 6.16 steps/s | VRAM=28.49 Go [2026-07-18 20:28:59] step 972400/1000000 | loss=1.9487 vocab=1.5290 token=1.2822 | lr=3.49e-06 | 6.14 steps/s | VRAM=28.68 Go [2026-07-18 20:29:08] step 972450/1000000 | loss=2.0814 vocab=0.8632 token=1.0065 | lr=3.48e-06 | 6.11 steps/s | VRAM=28.94 Go [2026-07-18 20:29:16] step 972500/1000000 | loss=2.0724 vocab=0.8255 token=1.1198 | lr=3.47e-06 | 6.14 steps/s | VRAM=27.10 Go [2026-07-18 20:29:24] step 972550/1000000 | loss=1.9901 vocab=1.7382 token=1.6828 | lr=3.46e-06 | 6.17 steps/s | VRAM=31.12 Go [2026-07-18 20:29:32] step 972600/1000000 | loss=2.0511 vocab=0.6145 token=1.4579 | lr=3.45e-06 | 6.13 steps/s | VRAM=29.05 Go [2026-07-18 20:29:40] step 972650/1000000 | loss=1.9901 vocab=0.7965 token=1.0687 | lr=3.44e-06 | 6.10 steps/s | VRAM=24.13 Go [2026-07-18 20:29:48] step 972700/1000000 | loss=1.9684 vocab=0.7741 token=0.8846 | lr=3.43e-06 | 6.16 steps/s | VRAM=28.21 Go [2026-07-18 20:29:57] step 972750/1000000 | loss=2.0535 vocab=1.3074 token=1.0758 | lr=3.42e-06 | 6.12 steps/s | VRAM=32.53 Go [2026-07-18 20:30:05] step 972800/1000000 | loss=1.9269 vocab=1.0736 token=1.5667 | lr=3.41e-06 | 6.12 steps/s | VRAM=28.39 Go [2026-07-18 20:30:13] step 972850/1000000 | loss=1.9492 vocab=0.7705 token=1.1349 | lr=3.41e-06 | 6.12 steps/s | VRAM=26.21 Go [2026-07-18 20:30:21] step 972900/1000000 | loss=2.0483 vocab=0.7937 token=1.1767 | lr=3.40e-06 | 6.10 steps/s | VRAM=25.42 Go [2026-07-18 20:30:29] step 972950/1000000 | loss=2.0026 vocab=1.3095 token=1.1836 | lr=3.39e-06 | 6.12 steps/s | VRAM=28.61 Go [2026-07-18 20:30:37] step 973000/1000000 | loss=1.9805 vocab=0.5175 token=1.1676 | lr=3.38e-06 | 6.14 steps/s | VRAM=28.94 Go [2026-07-18 20:30:45] step 973050/1000000 | loss=1.9362 vocab=0.5280 token=0.7998 | lr=3.37e-06 | 6.14 steps/s | VRAM=28.83 Go [2026-07-18 20:30:54] step 973100/1000000 | loss=1.9382 vocab=0.6761 token=1.1485 | lr=3.36e-06 | 6.14 steps/s | VRAM=29.42 Go [2026-07-18 20:31:02] step 973150/1000000 | loss=2.0027 vocab=0.6420 token=0.7264 | lr=3.35e-06 | 6.13 steps/s | VRAM=32.83 Go [2026-07-18 20:31:10] step 973200/1000000 | loss=1.9691 vocab=1.2035 token=1.1657 | lr=3.34e-06 | 6.13 steps/s | VRAM=27.98 Go [2026-07-18 20:31:18] step 973250/1000000 | loss=2.0607 vocab=0.9952 token=1.1149 | lr=3.34e-06 | 6.12 steps/s | VRAM=28.64 Go [2026-07-18 20:31:26] step 973300/1000000 | loss=2.0871 vocab=0.3920 token=1.4090 | lr=3.33e-06 | 6.11 steps/s | VRAM=26.49 Go [2026-07-18 20:31:34] step 973350/1000000 | loss=2.0628 vocab=0.9687 token=1.3534 | lr=3.32e-06 | 6.14 steps/s | VRAM=29.31 Go [2026-07-18 20:31:43] step 973400/1000000 | loss=1.9023 vocab=0.7594 token=1.2475 | lr=3.31e-06 | 6.12 steps/s | VRAM=26.47 Go [2026-07-18 20:31:51] step 973450/1000000 | loss=2.0454 vocab=0.3241 token=0.8607 | lr=3.30e-06 | 6.13 steps/s | VRAM=32.71 Go [2026-07-18 20:31:59] step 973500/1000000 | loss=1.9270 vocab=1.0690 token=1.2197 | lr=3.29e-06 | 6.12 steps/s | VRAM=30.54 Go [2026-07-18 20:32:07] step 973550/1000000 | loss=1.9725 vocab=0.2211 token=0.6609 | lr=3.28e-06 | 6.14 steps/s | VRAM=32.43 Go [2026-07-18 20:32:15] step 973600/1000000 | loss=1.9497 vocab=0.8164 token=1.5559 | lr=3.28e-06 | 6.13 steps/s | VRAM=31.89 Go [2026-07-18 20:32:23] step 973650/1000000 | loss=2.0859 vocab=1.1121 token=1.5291 | lr=3.27e-06 | 6.14 steps/s | VRAM=29.20 Go [2026-07-18 20:32:32] step 973700/1000000 | loss=2.0442 vocab=1.7649 token=1.1915 | lr=3.26e-06 | 6.15 steps/s | VRAM=26.37 Go [2026-07-18 20:32:40] step 973750/1000000 | loss=1.9465 vocab=0.9062 token=1.3748 | lr=3.25e-06 | 5.72 steps/s | VRAM=26.80 Go [2026-07-18 20:32:48] step 973800/1000000 | loss=2.0287 vocab=0.8748 token=1.2983 | lr=3.24e-06 | 6.15 steps/s | VRAM=32.17 Go [2026-07-18 20:32:57] step 973850/1000000 | loss=1.9586 vocab=0.7991 token=1.3448 | lr=3.23e-06 | 6.14 steps/s | VRAM=28.32 Go [2026-07-18 20:33:05] step 973900/1000000 | loss=1.9970 vocab=0.8813 token=1.4621 | lr=3.22e-06 | 6.12 steps/s | VRAM=26.01 Go [2026-07-18 20:33:13] step 973950/1000000 | loss=1.9929 vocab=0.5878 token=0.8767 | lr=3.22e-06 | 6.19 steps/s | VRAM=32.14 Go [2026-07-18 20:33:21] step 974000/1000000 | loss=1.9628 vocab=0.6431 token=1.2444 | lr=3.21e-06 | 6.13 steps/s | VRAM=31.71 Go [2026-07-18 20:33:29] step 974050/1000000 | loss=1.8756 vocab=0.7307 token=1.0755 | lr=3.20e-06 | 6.11 steps/s | VRAM=29.42 Go [2026-07-18 20:33:37] step 974100/1000000 | loss=1.9714 vocab=1.3541 token=1.4616 | lr=3.19e-06 | 6.15 steps/s | VRAM=31.78 Go [2026-07-18 20:33:45] step 974150/1000000 | loss=2.0012 vocab=0.6960 token=1.0364 | lr=3.18e-06 | 6.14 steps/s | VRAM=24.35 Go [2026-07-18 20:33:54] step 974200/1000000 | loss=1.8885 vocab=0.7522 token=0.6935 | lr=3.17e-06 | 6.14 steps/s | VRAM=32.80 Go [2026-07-18 20:34:02] step 974250/1000000 | loss=2.0432 vocab=0.8342 token=1.1154 | lr=3.17e-06 | 6.14 steps/s | VRAM=26.09 Go [2026-07-18 20:34:10] step 974300/1000000 | loss=2.0081 vocab=0.5960 token=1.1097 | lr=3.16e-06 | 6.15 steps/s | VRAM=26.14 Go [2026-07-18 20:34:18] step 974350/1000000 | loss=2.0001 vocab=0.8227 token=1.0964 | lr=3.15e-06 | 6.16 steps/s | VRAM=27.36 Go [2026-07-18 20:34:26] step 974400/1000000 | loss=1.9834 vocab=1.0800 token=1.4672 | lr=3.14e-06 | 6.16 steps/s | VRAM=28.94 Go [2026-07-18 20:34:34] step 974450/1000000 | loss=1.9742 vocab=0.5621 token=0.9860 | lr=3.13e-06 | 6.14 steps/s | VRAM=28.79 Go [2026-07-18 20:34:42] step 974500/1000000 | loss=1.9777 vocab=1.1518 token=1.6814 | lr=3.12e-06 | 6.15 steps/s | VRAM=27.98 Go [2026-07-18 20:34:50] step 974550/1000000 | loss=2.0785 vocab=0.9131 token=1.1337 | lr=3.12e-06 | 6.17 steps/s | VRAM=28.50 Go [2026-07-18 20:34:59] step 974600/1000000 | loss=2.0251 vocab=0.3418 token=0.7545 | lr=3.11e-06 | 6.13 steps/s | VRAM=29.19 Go [2026-07-18 20:35:07] step 974650/1000000 | loss=1.9665 vocab=1.0124 token=1.1930 | lr=3.10e-06 | 6.17 steps/s | VRAM=32.33 Go [2026-07-18 20:35:15] step 974700/1000000 | loss=1.9516 vocab=0.7007 token=0.9574 | lr=3.09e-06 | 6.15 steps/s | VRAM=28.41 Go [2026-07-18 20:35:23] step 974750/1000000 | loss=1.8902 vocab=1.0515 token=1.4345 | lr=3.08e-06 | 6.14 steps/s | VRAM=28.31 Go [2026-07-18 20:35:31] step 974800/1000000 | loss=1.9909 vocab=0.6809 token=1.2130 | lr=3.07e-06 | 6.14 steps/s | VRAM=29.01 Go [2026-07-18 20:35:39] step 974850/1000000 | loss=2.0055 vocab=1.0379 token=0.9415 | lr=3.07e-06 | 6.13 steps/s | VRAM=28.42 Go [2026-07-18 20:35:47] step 974900/1000000 | loss=1.9978 vocab=0.7490 token=0.8838 | lr=3.06e-06 | 6.16 steps/s | VRAM=31.92 Go [2026-07-18 20:35:56] step 974950/1000000 | loss=2.1356 vocab=1.0315 token=1.4599 | lr=3.05e-06 | 6.17 steps/s | VRAM=32.13 Go [2026-07-18 20:36:04] step 975000/1000000 | loss=2.0626 vocab=0.7842 token=1.0215 | lr=3.04e-06 | 6.11 steps/s | VRAM=26.59 Go [2026-07-18 20:36:12] step 975050/1000000 | loss=1.9361 vocab=0.5800 token=1.0413 | lr=3.03e-06 | 6.13 steps/s | VRAM=28.68 Go [2026-07-18 20:36:20] step 975100/1000000 | loss=1.9166 vocab=1.4768 token=1.6105 | lr=3.03e-06 | 6.14 steps/s | VRAM=26.03 Go [2026-07-18 20:36:28] step 975150/1000000 | loss=2.0463 vocab=0.7730 token=1.1105 | lr=3.02e-06 | 6.14 steps/s | VRAM=26.35 Go [2026-07-18 20:36:36] step 975200/1000000 | loss=2.0327 vocab=0.7034 token=1.4090 | lr=3.01e-06 | 6.12 steps/s | VRAM=28.82 Go [2026-07-18 20:36:44] step 975250/1000000 | loss=2.0226 vocab=1.0239 token=0.9881 | lr=3.00e-06 | 6.16 steps/s | VRAM=32.15 Go [2026-07-18 20:36:53] step 975300/1000000 | loss=1.9307 vocab=0.5389 token=0.9581 | lr=2.99e-06 | 6.14 steps/s | VRAM=28.73 Go [2026-07-18 20:37:01] step 975350/1000000 | loss=2.0477 vocab=0.3180 token=0.9674 | lr=2.99e-06 | 6.16 steps/s | VRAM=31.60 Go [2026-07-18 20:37:09] step 975400/1000000 | loss=2.0828 vocab=1.0142 token=1.6660 | lr=2.98e-06 | 6.15 steps/s | VRAM=25.78 Go [2026-07-18 20:37:17] step 975450/1000000 | loss=1.9285 vocab=1.0361 token=1.1168 | lr=2.97e-06 | 6.14 steps/s | VRAM=29.12 Go [2026-07-18 20:37:25] step 975500/1000000 | loss=1.9461 vocab=1.0335 token=0.9172 | lr=2.96e-06 | 6.12 steps/s | VRAM=30.64 Go [2026-07-18 20:37:33] step 975550/1000000 | loss=2.0888 vocab=0.8558 token=1.1976 | lr=2.95e-06 | 6.16 steps/s | VRAM=28.96 Go [2026-07-18 20:37:41] step 975600/1000000 | loss=2.0368 vocab=0.5038 token=1.1070 | lr=2.95e-06 | 6.16 steps/s | VRAM=32.33 Go [2026-07-18 20:37:49] step 975650/1000000 | loss=1.9834 vocab=0.6314 token=1.1998 | lr=2.94e-06 | 6.14 steps/s | VRAM=32.47 Go [2026-07-18 20:37:58] step 975700/1000000 | loss=2.0330 vocab=0.6331 token=0.6765 | lr=2.93e-06 | 6.15 steps/s | VRAM=25.43 Go [2026-07-18 20:38:06] step 975750/1000000 | loss=2.0571 vocab=0.6975 token=1.1859 | lr=2.92e-06 | 6.17 steps/s | VRAM=28.65 Go [2026-07-18 20:38:14] step 975800/1000000 | loss=1.9611 vocab=0.4884 token=0.9420 | lr=2.91e-06 | 6.10 steps/s | VRAM=26.23 Go [2026-07-18 20:38:22] step 975850/1000000 | loss=1.9397 vocab=1.0366 token=1.3199 | lr=2.91e-06 | 6.14 steps/s | VRAM=26.85 Go [2026-07-18 20:38:30] step 975900/1000000 | loss=2.0259 vocab=0.4051 token=0.6058 | lr=2.90e-06 | 6.17 steps/s | VRAM=31.07 Go [2026-07-18 20:38:38] step 975950/1000000 | loss=1.9295 vocab=0.5047 token=0.8345 | lr=2.89e-06 | 6.13 steps/s | VRAM=26.01 Go [2026-07-18 20:38:46] step 976000/1000000 | loss=2.0089 vocab=1.3459 token=1.3349 | lr=2.88e-06 | 6.13 steps/s | VRAM=29.54 Go [2026-07-18 20:38:55] step 976050/1000000 | loss=2.0108 vocab=0.6859 token=0.9917 | lr=2.88e-06 | 6.15 steps/s | VRAM=31.57 Go [2026-07-18 20:39:03] step 976100/1000000 | loss=1.9256 vocab=1.2251 token=1.4169 | lr=2.87e-06 | 6.15 steps/s | VRAM=31.85 Go [2026-07-18 20:39:11] step 976150/1000000 | loss=2.0920 vocab=1.2308 token=1.3546 | lr=2.86e-06 | 6.12 steps/s | VRAM=28.51 Go [2026-07-18 20:39:19] step 976200/1000000 | loss=1.9570 vocab=1.2638 token=1.5597 | lr=2.85e-06 | 6.16 steps/s | VRAM=32.06 Go [2026-07-18 20:39:27] step 976250/1000000 | loss=2.0852 vocab=1.3644 token=1.2814 | lr=2.84e-06 | 6.12 steps/s | VRAM=25.71 Go [2026-07-18 20:39:35] step 976300/1000000 | loss=1.8746 vocab=0.4048 token=1.1288 | lr=2.84e-06 | 6.16 steps/s | VRAM=26.56 Go [2026-07-18 20:39:43] step 976350/1000000 | loss=1.9304 vocab=0.9921 token=1.1932 | lr=2.83e-06 | 6.18 steps/s | VRAM=25.12 Go [2026-07-18 20:39:52] step 976400/1000000 | loss=2.0514 vocab=0.5190 token=1.1649 | lr=2.82e-06 | 6.13 steps/s | VRAM=28.51 Go [2026-07-18 20:40:00] step 976450/1000000 | loss=2.0134 vocab=0.5957 token=1.1934 | lr=2.81e-06 | 6.10 steps/s | VRAM=30.15 Go [2026-07-18 20:40:08] step 976500/1000000 | loss=2.1309 vocab=0.9706 token=1.2644 | lr=2.81e-06 | 6.18 steps/s | VRAM=31.70 Go [2026-07-18 20:40:16] step 976550/1000000 | loss=1.9777 vocab=0.8071 token=0.8371 | lr=2.80e-06 | 6.13 steps/s | VRAM=28.89 Go [2026-07-18 20:40:24] step 976600/1000000 | loss=1.8588 vocab=0.4507 token=0.7705 | lr=2.79e-06 | 6.14 steps/s | VRAM=29.36 Go [2026-07-18 20:40:32] step 976650/1000000 | loss=1.9246 vocab=0.4753 token=0.9281 | lr=2.78e-06 | 6.15 steps/s | VRAM=31.86 Go [2026-07-18 20:40:40] step 976700/1000000 | loss=2.1642 vocab=0.7727 token=1.3860 | lr=2.78e-06 | 6.13 steps/s | VRAM=29.35 Go [2026-07-18 20:40:49] step 976750/1000000 | loss=1.9840 vocab=0.7426 token=0.9570 | lr=2.77e-06 | 6.13 steps/s | VRAM=29.07 Go [2026-07-18 20:40:57] step 976800/1000000 | loss=1.9130 vocab=0.8524 token=1.3624 | lr=2.76e-06 | 6.14 steps/s | VRAM=27.98 Go [2026-07-18 20:41:05] step 976850/1000000 | loss=2.0158 vocab=0.7156 token=0.9993 | lr=2.75e-06 | 6.13 steps/s | VRAM=25.29 Go [2026-07-18 20:41:13] step 976900/1000000 | loss=2.0844 vocab=0.9237 token=1.1322 | lr=2.75e-06 | 6.14 steps/s | VRAM=26.68 Go [2026-07-18 20:41:21] step 976950/1000000 | loss=2.0099 vocab=0.8765 token=1.4609 | lr=2.74e-06 | 6.16 steps/s | VRAM=28.46 Go [2026-07-18 20:41:29] step 977000/1000000 | loss=1.9834 vocab=0.8483 token=1.3065 | lr=2.73e-06 | 6.14 steps/s | VRAM=28.83 Go [2026-07-18 20:41:37] step 977050/1000000 | loss=2.0074 vocab=0.7010 token=0.8308 | lr=2.72e-06 | 6.16 steps/s | VRAM=31.20 Go [2026-07-18 20:41:46] step 977100/1000000 | loss=1.9919 vocab=0.6986 token=1.1555 | lr=2.72e-06 | 6.14 steps/s | VRAM=28.62 Go [2026-07-18 20:41:54] step 977150/1000000 | loss=1.9890 vocab=0.9369 token=1.2550 | lr=2.71e-06 | 6.14 steps/s | VRAM=29.01 Go [2026-07-18 20:42:02] step 977200/1000000 | loss=1.9873 vocab=0.8059 token=1.1125 | lr=2.70e-06 | 6.17 steps/s | VRAM=30.99 Go [2026-07-18 20:42:10] step 977250/1000000 | loss=1.8594 vocab=0.5723 token=1.5642 | lr=2.69e-06 | 6.15 steps/s | VRAM=29.25 Go [2026-07-18 20:42:18] step 977300/1000000 | loss=1.9781 vocab=0.4232 token=0.8347 | lr=2.69e-06 | 6.14 steps/s | VRAM=28.70 Go [2026-07-18 20:42:26] step 977350/1000000 | loss=2.0383 vocab=0.5912 token=1.0676 | lr=2.68e-06 | 6.16 steps/s | VRAM=31.49 Go [2026-07-18 20:42:34] step 977400/1000000 | loss=2.0515 vocab=0.6714 token=1.1365 | lr=2.67e-06 | 6.14 steps/s | VRAM=28.57 Go [2026-07-18 20:42:42] step 977450/1000000 | loss=2.0373 vocab=0.6871 token=0.9193 | lr=2.66e-06 | 6.14 steps/s | VRAM=29.27 Go [2026-07-18 20:42:51] step 977500/1000000 | loss=1.9257 vocab=0.6116 token=1.1153 | lr=2.66e-06 | 6.15 steps/s | VRAM=25.50 Go [2026-07-18 20:42:59] step 977550/1000000 | loss=2.0248 vocab=0.4276 token=1.1034 | lr=2.65e-06 | 6.15 steps/s | VRAM=28.31 Go [2026-07-18 20:43:07] step 977600/1000000 | loss=2.0433 vocab=1.3766 token=1.1740 | lr=2.64e-06 | 6.13 steps/s | VRAM=32.93 Go [2026-07-18 20:43:15] step 977650/1000000 | loss=2.0768 vocab=0.7677 token=1.0582 | lr=2.63e-06 | 6.16 steps/s | VRAM=31.31 Go [2026-07-18 20:43:23] step 977700/1000000 | loss=1.9625 vocab=0.7685 token=0.8508 | lr=2.63e-06 | 6.15 steps/s | VRAM=29.50 Go [2026-07-18 20:43:31] step 977750/1000000 | loss=1.9335 vocab=0.5931 token=1.0642 | lr=2.62e-06 | 6.15 steps/s | VRAM=32.18 Go [2026-07-18 20:43:39] step 977800/1000000 | loss=1.9567 vocab=0.8318 token=1.3999 | lr=2.61e-06 | 6.14 steps/s | VRAM=27.92 Go [2026-07-18 20:43:48] step 977850/1000000 | loss=1.9999 vocab=0.6651 token=1.2875 | lr=2.61e-06 | 6.11 steps/s | VRAM=26.88 Go [2026-07-18 20:43:56] step 977900/1000000 | loss=2.0131 vocab=0.6875 token=1.1570 | lr=2.60e-06 | 6.14 steps/s | VRAM=27.82 Go [2026-07-18 20:44:04] step 977950/1000000 | loss=1.9035 vocab=0.6029 token=0.8619 | lr=2.59e-06 | 6.14 steps/s | VRAM=25.88 Go [2026-07-18 20:44:12] step 978000/1000000 | loss=2.0269 vocab=0.9494 token=1.2979 | lr=2.58e-06 | 6.17 steps/s | VRAM=31.51 Go [2026-07-18 20:44:20] step 978050/1000000 | loss=1.9775 vocab=0.6620 token=1.1601 | lr=2.58e-06 | 6.13 steps/s | VRAM=29.45 Go [2026-07-18 20:44:28] step 978100/1000000 | loss=1.8905 vocab=0.6529 token=0.9595 | lr=2.57e-06 | 6.14 steps/s | VRAM=26.77 Go [2026-07-18 20:44:36] step 978150/1000000 | loss=2.0464 vocab=0.9837 token=1.1106 | lr=2.56e-06 | 6.15 steps/s | VRAM=32.06 Go [2026-07-18 20:44:45] step 978200/1000000 | loss=1.9560 vocab=0.6821 token=0.9964 | lr=2.56e-06 | 6.15 steps/s | VRAM=28.29 Go [2026-07-18 20:44:53] step 978250/1000000 | loss=1.9483 vocab=0.8966 token=1.0407 | lr=2.55e-06 | 6.15 steps/s | VRAM=25.70 Go [2026-07-18 20:45:01] step 978300/1000000 | loss=2.0211 vocab=0.4364 token=0.9685 | lr=2.54e-06 | 6.15 steps/s | VRAM=28.95 Go [2026-07-18 20:45:09] step 978350/1000000 | loss=1.9500 vocab=0.5814 token=0.5607 | lr=2.53e-06 | 6.15 steps/s | VRAM=28.66 Go [2026-07-18 20:45:17] step 978400/1000000 | loss=1.9798 vocab=0.5864 token=0.8058 | lr=2.53e-06 | 6.16 steps/s | VRAM=31.62 Go [2026-07-18 20:45:25] step 978450/1000000 | loss=2.0867 vocab=0.8666 token=1.1670 | lr=2.52e-06 | 6.12 steps/s | VRAM=28.12 Go [2026-07-18 20:45:33] step 978500/1000000 | loss=2.0271 vocab=0.9162 token=1.2558 | lr=2.51e-06 | 6.13 steps/s | VRAM=24.74 Go [2026-07-18 20:45:42] step 978550/1000000 | loss=2.0544 vocab=1.0506 token=1.2695 | lr=2.51e-06 | 6.12 steps/s | VRAM=26.27 Go [2026-07-18 20:45:50] step 978600/1000000 | loss=2.0592 vocab=1.1578 token=1.1995 | lr=2.50e-06 | 6.15 steps/s | VRAM=28.68 Go [2026-07-18 20:45:58] step 978650/1000000 | loss=1.9161 vocab=1.2424 token=0.9040 | lr=2.49e-06 | 6.13 steps/s | VRAM=24.79 Go [2026-07-18 20:46:06] step 978700/1000000 | loss=2.0296 vocab=0.6058 token=1.1682 | lr=2.49e-06 | 6.13 steps/s | VRAM=25.74 Go [2026-07-18 20:46:14] step 978750/1000000 | loss=1.9723 vocab=0.6083 token=0.9501 | lr=2.48e-06 | 6.14 steps/s | VRAM=29.25 Go [2026-07-18 20:46:22] step 978800/1000000 | loss=2.0001 vocab=0.9964 token=1.3538 | lr=2.47e-06 | 6.13 steps/s | VRAM=29.29 Go [2026-07-18 20:46:30] step 978850/1000000 | loss=1.9906 vocab=0.3065 token=0.7497 | lr=2.46e-06 | 6.14 steps/s | VRAM=31.74 Go [2026-07-18 20:46:39] step 978900/1000000 | loss=2.0585 vocab=1.0039 token=1.5114 | lr=2.46e-06 | 6.14 steps/s | VRAM=30.44 Go [2026-07-18 20:46:47] step 978950/1000000 | loss=1.9151 vocab=0.8734 token=1.2135 | lr=2.45e-06 | 6.16 steps/s | VRAM=31.19 Go [2026-07-18 20:46:55] step 979000/1000000 | loss=1.9973 vocab=1.0197 token=1.1843 | lr=2.44e-06 | 6.18 steps/s | VRAM=31.45 Go [2026-07-18 20:47:03] step 979050/1000000 | loss=2.0634 vocab=0.7748 token=1.2148 | lr=2.44e-06 | 6.14 steps/s | VRAM=32.53 Go [2026-07-18 20:47:11] step 979100/1000000 | loss=1.9248 vocab=1.1956 token=1.0750 | lr=2.43e-06 | 6.12 steps/s | VRAM=28.96 Go [2026-07-18 20:47:19] step 979150/1000000 | loss=2.0288 vocab=0.7077 token=0.6619 | lr=2.42e-06 | 6.14 steps/s | VRAM=26.36 Go [2026-07-18 20:47:27] step 979200/1000000 | loss=2.0957 vocab=0.6180 token=0.9050 | lr=2.42e-06 | 6.13 steps/s | VRAM=33.00 Go [2026-07-18 20:47:36] step 979250/1000000 | loss=2.0129 vocab=0.8007 token=1.1569 | lr=2.41e-06 | 6.16 steps/s | VRAM=28.28 Go [2026-07-18 20:47:44] step 979300/1000000 | loss=1.9562 vocab=0.8717 token=1.2569 | lr=2.40e-06 | 6.13 steps/s | VRAM=26.43 Go [2026-07-18 20:47:52] step 979350/1000000 | loss=1.8932 vocab=1.6486 token=1.2604 | lr=2.40e-06 | 6.16 steps/s | VRAM=25.33 Go [2026-07-18 20:48:00] step 979400/1000000 | loss=2.0438 vocab=0.7782 token=0.9252 | lr=2.39e-06 | 6.16 steps/s | VRAM=31.78 Go [2026-07-18 20:48:08] step 979450/1000000 | loss=2.1007 vocab=0.6017 token=1.4888 | lr=2.38e-06 | 6.13 steps/s | VRAM=32.73 Go [2026-07-18 20:48:16] step 979500/1000000 | loss=2.0048 vocab=0.9305 token=1.0663 | lr=2.38e-06 | 6.14 steps/s | VRAM=32.81 Go [2026-07-18 20:48:24] step 979550/1000000 | loss=1.9815 vocab=0.6705 token=1.1277 | lr=2.37e-06 | 6.16 steps/s | VRAM=32.84 Go [2026-07-18 20:48:32] step 979600/1000000 | loss=2.0368 vocab=1.1411 token=1.4969 | lr=2.36e-06 | 6.14 steps/s | VRAM=28.99 Go [2026-07-18 20:48:41] step 979650/1000000 | loss=2.0131 vocab=0.5722 token=1.1249 | lr=2.36e-06 | 6.12 steps/s | VRAM=29.54 Go [2026-07-18 20:48:49] step 979700/1000000 | loss=2.1200 vocab=0.7844 token=0.9782 | lr=2.35e-06 | 6.17 steps/s | VRAM=29.67 Go [2026-07-18 20:48:57] step 979750/1000000 | loss=1.8068 vocab=0.5916 token=0.8272 | lr=2.34e-06 | 6.16 steps/s | VRAM=25.11 Go [2026-07-18 20:49:05] step 979800/1000000 | loss=2.0175 vocab=1.1472 token=1.5252 | lr=2.34e-06 | 6.15 steps/s | VRAM=32.06 Go [2026-07-18 20:49:13] step 979850/1000000 | loss=1.9588 vocab=0.6076 token=0.6669 | lr=2.33e-06 | 6.16 steps/s | VRAM=31.85 Go [2026-07-18 20:49:21] step 979900/1000000 | loss=1.9703 vocab=0.8479 token=1.1312 | lr=2.32e-06 | 6.14 steps/s | VRAM=29.68 Go [2026-07-18 20:49:29] step 979950/1000000 | loss=2.0110 vocab=0.8925 token=1.2419 | lr=2.32e-06 | 6.17 steps/s | VRAM=27.99 Go [2026-07-18 20:49:37] step 980000/1000000 | loss=2.0494 vocab=1.1634 token=1.1788 | lr=2.31e-06 | 6.17 steps/s | VRAM=25.03 Go [2026-07-18 20:49:46] step 980050/1000000 | loss=1.9238 vocab=0.7648 token=1.0586 | lr=2.30e-06 | 6.13 steps/s | VRAM=28.63 Go [2026-07-18 20:49:54] step 980100/1000000 | loss=1.9853 vocab=0.7474 token=1.5753 | lr=2.30e-06 | 6.15 steps/s | VRAM=31.86 Go [2026-07-18 20:50:02] step 980150/1000000 | loss=1.9161 vocab=0.5609 token=1.0015 | lr=2.29e-06 | 6.15 steps/s | VRAM=28.18 Go [2026-07-18 20:50:10] step 980200/1000000 | loss=1.9246 vocab=0.8882 token=1.0910 | lr=2.28e-06 | 6.16 steps/s | VRAM=28.88 Go [2026-07-18 20:50:18] step 980250/1000000 | loss=1.8792 vocab=1.1463 token=1.1637 | lr=2.28e-06 | 6.16 steps/s | VRAM=31.29 Go [2026-07-18 20:50:26] step 980300/1000000 | loss=1.9195 vocab=0.7343 token=1.0458 | lr=2.27e-06 | 6.12 steps/s | VRAM=26.80 Go [2026-07-18 20:50:34] step 980350/1000000 | loss=1.9620 vocab=1.3033 token=1.4588 | lr=2.26e-06 | 6.16 steps/s | VRAM=28.41 Go [2026-07-18 20:50:43] step 980400/1000000 | loss=2.0137 vocab=0.8171 token=1.4583 | lr=2.26e-06 | 6.15 steps/s | VRAM=26.86 Go [2026-07-18 20:50:51] step 980450/1000000 | loss=2.0192 vocab=1.0604 token=1.1564 | lr=2.25e-06 | 6.16 steps/s | VRAM=27.89 Go [2026-07-18 20:50:59] step 980500/1000000 | loss=2.0651 vocab=0.6088 token=1.0700 | lr=2.25e-06 | 6.17 steps/s | VRAM=32.14 Go [2026-07-18 20:51:07] step 980550/1000000 | loss=1.9826 vocab=0.5749 token=0.8114 | lr=2.24e-06 | 6.14 steps/s | VRAM=26.28 Go [2026-07-18 20:51:15] step 980600/1000000 | loss=1.9536 vocab=0.6508 token=1.3457 | lr=2.23e-06 | 6.15 steps/s | VRAM=29.25 Go [2026-07-18 20:51:23] step 980650/1000000 | loss=2.0817 vocab=0.8779 token=1.0511 | lr=2.23e-06 | 6.12 steps/s | VRAM=28.82 Go [2026-07-18 20:51:31] step 980700/1000000 | loss=1.9866 vocab=0.5710 token=0.9385 | lr=2.22e-06 | 6.17 steps/s | VRAM=31.83 Go [2026-07-18 20:51:39] step 980750/1000000 | loss=2.0289 vocab=1.1212 token=1.2845 | lr=2.21e-06 | 6.15 steps/s | VRAM=28.91 Go [2026-07-18 20:51:48] step 980800/1000000 | loss=2.0997 vocab=1.0027 token=1.2201 | lr=2.21e-06 | 6.16 steps/s | VRAM=28.93 Go [2026-07-18 20:51:56] step 980850/1000000 | loss=1.9863 vocab=1.0424 token=1.1009 | lr=2.20e-06 | 6.17 steps/s | VRAM=24.34 Go [2026-07-18 20:52:04] step 980900/1000000 | loss=2.0143 vocab=0.4581 token=0.8344 | lr=2.20e-06 | 6.13 steps/s | VRAM=26.51 Go [2026-07-18 20:52:12] step 980950/1000000 | loss=1.9708 vocab=0.6541 token=1.2320 | lr=2.19e-06 | 6.15 steps/s | VRAM=27.96 Go [2026-07-18 20:52:20] step 981000/1000000 | loss=2.0839 vocab=0.3840 token=0.8557 | lr=2.18e-06 | 6.14 steps/s | VRAM=26.60 Go [2026-07-18 20:52:28] step 981050/1000000 | loss=1.8094 vocab=0.8680 token=0.8460 | lr=2.18e-06 | 6.16 steps/s | VRAM=28.52 Go [2026-07-18 20:52:36] step 981100/1000000 | loss=2.0284 vocab=1.1145 token=1.0573 | lr=2.17e-06 | 6.17 steps/s | VRAM=25.50 Go [2026-07-18 20:52:44] step 981150/1000000 | loss=2.0870 vocab=1.1814 token=1.1294 | lr=2.16e-06 | 6.17 steps/s | VRAM=26.80 Go [2026-07-18 20:52:53] step 981200/1000000 | loss=1.9855 vocab=0.8753 token=1.1638 | lr=2.16e-06 | 6.16 steps/s | VRAM=24.46 Go [2026-07-18 20:53:01] step 981250/1000000 | loss=2.0358 vocab=0.8175 token=0.9472 | lr=2.15e-06 | 6.17 steps/s | VRAM=27.79 Go [2026-07-18 20:53:09] step 981300/1000000 | loss=2.0895 vocab=0.3412 token=0.7339 | lr=2.15e-06 | 6.15 steps/s | VRAM=26.99 Go [2026-07-18 20:53:17] step 981350/1000000 | loss=2.0819 vocab=0.8842 token=1.1186 | lr=2.14e-06 | 6.13 steps/s | VRAM=28.66 Go [2026-07-18 20:53:25] step 981400/1000000 | loss=2.0082 vocab=1.2805 token=1.7986 | lr=2.13e-06 | 6.13 steps/s | VRAM=29.53 Go [2026-07-18 20:53:33] step 981450/1000000 | loss=2.0528 vocab=0.6026 token=0.6668 | lr=2.13e-06 | 6.13 steps/s | VRAM=29.26 Go [2026-07-18 20:53:41] step 981500/1000000 | loss=1.9709 vocab=0.5842 token=1.0415 | lr=2.12e-06 | 6.11 steps/s | VRAM=29.24 Go [2026-07-18 20:53:50] step 981550/1000000 | loss=1.9252 vocab=1.0571 token=0.8472 | lr=2.12e-06 | 6.13 steps/s | VRAM=28.15 Go [2026-07-18 20:53:58] step 981600/1000000 | loss=1.9227 vocab=0.9360 token=1.3662 | lr=2.11e-06 | 6.16 steps/s | VRAM=30.75 Go [2026-07-18 20:54:06] step 981650/1000000 | loss=1.9968 vocab=0.7729 token=1.0891 | lr=2.10e-06 | 6.15 steps/s | VRAM=26.61 Go [2026-07-18 20:54:14] step 981700/1000000 | loss=2.0133 vocab=0.9462 token=1.3507 | lr=2.10e-06 | 6.15 steps/s | VRAM=29.37 Go [2026-07-18 20:54:22] step 981750/1000000 | loss=1.9564 vocab=0.5772 token=0.7387 | lr=2.09e-06 | 6.15 steps/s | VRAM=32.96 Go [2026-07-18 20:54:30] step 981800/1000000 | loss=1.9906 vocab=0.5215 token=1.1989 | lr=2.09e-06 | 6.15 steps/s | VRAM=29.89 Go [2026-07-18 20:54:38] step 981850/1000000 | loss=2.0505 vocab=1.3207 token=1.5239 | lr=2.08e-06 | 6.14 steps/s | VRAM=29.31 Go [2026-07-18 20:54:46] step 981900/1000000 | loss=2.0780 vocab=1.2577 token=1.7012 | lr=2.07e-06 | 6.14 steps/s | VRAM=28.18 Go [2026-07-18 20:54:55] step 981950/1000000 | loss=2.0264 vocab=0.7659 token=0.9735 | lr=2.07e-06 | 6.15 steps/s | VRAM=26.29 Go [2026-07-18 20:55:03] step 982000/1000000 | loss=1.9335 vocab=0.7648 token=1.1141 | lr=2.06e-06 | 6.16 steps/s | VRAM=28.64 Go [2026-07-18 20:55:11] step 982050/1000000 | loss=1.9869 vocab=0.5094 token=1.0307 | lr=2.06e-06 | 6.14 steps/s | VRAM=32.36 Go [2026-07-18 20:55:19] step 982100/1000000 | loss=1.9443 vocab=0.8400 token=1.4098 | lr=2.05e-06 | 6.17 steps/s | VRAM=28.80 Go [2026-07-18 20:55:27] step 982150/1000000 | loss=1.9315 vocab=1.1715 token=1.3644 | lr=2.04e-06 | 6.14 steps/s | VRAM=24.91 Go [2026-07-18 20:55:35] step 982200/1000000 | loss=2.0811 vocab=1.1028 token=1.3870 | lr=2.04e-06 | 6.17 steps/s | VRAM=31.70 Go [2026-07-18 20:55:43] step 982250/1000000 | loss=1.9491 vocab=0.7499 token=1.1341 | lr=2.03e-06 | 6.14 steps/s | VRAM=25.75 Go [2026-07-18 20:55:51] step 982300/1000000 | loss=1.9829 vocab=0.8219 token=1.2307 | lr=2.03e-06 | 6.16 steps/s | VRAM=31.96 Go [2026-07-18 20:56:00] step 982350/1000000 | loss=2.0727 vocab=0.8754 token=1.1902 | lr=2.02e-06 | 6.14 steps/s | VRAM=28.41 Go [2026-07-18 20:56:08] step 982400/1000000 | loss=1.9677 vocab=0.4766 token=0.8238 | lr=2.02e-06 | 6.17 steps/s | VRAM=23.83 Go [2026-07-18 20:56:16] step 982450/1000000 | loss=1.8985 vocab=1.0428 token=1.3717 | lr=2.01e-06 | 6.17 steps/s | VRAM=28.33 Go [2026-07-18 20:56:24] step 982500/1000000 | loss=2.0436 vocab=0.4217 token=1.1069 | lr=2.00e-06 | 6.17 steps/s | VRAM=32.32 Go [2026-07-18 20:56:32] step 982550/1000000 | loss=2.0063 vocab=0.6445 token=1.1577 | lr=2.00e-06 | 6.15 steps/s | VRAM=28.19 Go [2026-07-18 20:56:40] step 982600/1000000 | loss=2.0859 vocab=0.6399 token=1.2357 | lr=1.99e-06 | 6.15 steps/s | VRAM=31.71 Go [2026-07-18 20:56:48] step 982650/1000000 | loss=2.0417 vocab=0.5536 token=0.9767 | lr=1.99e-06 | 6.16 steps/s | VRAM=28.95 Go [2026-07-18 20:56:56] step 982700/1000000 | loss=1.9736 vocab=0.6866 token=0.9370 | lr=1.98e-06 | 6.14 steps/s | VRAM=26.61 Go [2026-07-18 20:57:05] step 982750/1000000 | loss=1.9755 vocab=1.0767 token=1.2506 | lr=1.98e-06 | 6.17 steps/s | VRAM=28.89 Go [2026-07-18 20:57:13] step 982800/1000000 | loss=2.0077 vocab=0.5268 token=1.3532 | lr=1.97e-06 | 6.14 steps/s | VRAM=25.87 Go [2026-07-18 20:57:21] step 982850/1000000 | loss=2.0683 vocab=1.6990 token=1.4391 | lr=1.96e-06 | 6.13 steps/s | VRAM=28.94 Go [2026-07-18 20:57:29] step 982900/1000000 | loss=1.9664 vocab=1.0212 token=1.3459 | lr=1.96e-06 | 6.13 steps/s | VRAM=27.84 Go [2026-07-18 20:57:37] step 982950/1000000 | loss=1.9978 vocab=0.9987 token=1.1230 | lr=1.95e-06 | 6.18 steps/s | VRAM=27.60 Go [2026-07-18 20:57:45] step 983000/1000000 | loss=2.0346 vocab=0.9646 token=1.2366 | lr=1.95e-06 | 6.13 steps/s | VRAM=26.21 Go [2026-07-18 20:57:53] step 983050/1000000 | loss=1.9994 vocab=1.0171 token=1.2017 | lr=1.94e-06 | 6.14 steps/s | VRAM=26.77 Go [2026-07-18 20:58:02] step 983100/1000000 | loss=1.9931 vocab=1.3104 token=0.9543 | lr=1.94e-06 | 6.14 steps/s | VRAM=30.15 Go [2026-07-18 20:58:10] step 983150/1000000 | loss=1.9308 vocab=1.4812 token=1.9326 | lr=1.93e-06 | 6.13 steps/s | VRAM=28.32 Go [2026-07-18 20:58:18] step 983200/1000000 | loss=1.9975 vocab=0.6839 token=0.8551 | lr=1.93e-06 | 6.14 steps/s | VRAM=28.65 Go [2026-07-18 20:58:26] step 983250/1000000 | loss=1.8996 vocab=0.7232 token=0.9873 | lr=1.92e-06 | 6.15 steps/s | VRAM=27.96 Go [2026-07-18 20:58:34] step 983300/1000000 | loss=2.0438 vocab=0.9352 token=1.2911 | lr=1.91e-06 | 6.13 steps/s | VRAM=28.22 Go [2026-07-18 20:58:42] step 983350/1000000 | loss=1.9721 vocab=0.7871 token=1.4547 | lr=1.91e-06 | 6.14 steps/s | VRAM=29.41 Go [2026-07-18 20:58:50] step 983400/1000000 | loss=2.1037 vocab=1.0917 token=1.5044 | lr=1.90e-06 | 6.16 steps/s | VRAM=24.62 Go [2026-07-18 20:58:59] step 983450/1000000 | loss=1.9302 vocab=0.6216 token=1.1637 | lr=1.90e-06 | 6.16 steps/s | VRAM=25.45 Go [2026-07-18 20:59:07] step 983500/1000000 | loss=1.9855 vocab=0.9766 token=1.5887 | lr=1.89e-06 | 6.16 steps/s | VRAM=29.02 Go [2026-07-18 20:59:15] step 983550/1000000 | loss=2.0466 vocab=0.7340 token=1.1122 | lr=1.89e-06 | 6.13 steps/s | VRAM=31.77 Go [2026-07-18 20:59:23] step 983600/1000000 | loss=2.0462 vocab=0.6524 token=1.0139 | lr=1.88e-06 | 6.16 steps/s | VRAM=28.90 Go [2026-07-18 20:59:31] step 983650/1000000 | loss=2.0064 vocab=0.8679 token=0.7687 | lr=1.88e-06 | 6.13 steps/s | VRAM=28.96 Go [2026-07-18 20:59:39] step 983700/1000000 | loss=2.0062 vocab=0.4665 token=0.9241 | lr=1.87e-06 | 6.12 steps/s | VRAM=28.19 Go [2026-07-18 20:59:47] step 983750/1000000 | loss=1.9625 vocab=0.6929 token=1.0774 | lr=1.87e-06 | 6.13 steps/s | VRAM=32.67 Go [2026-07-18 20:59:56] step 983800/1000000 | loss=1.9785 vocab=0.8196 token=0.9114 | lr=1.86e-06 | 6.11 steps/s | VRAM=30.71 Go [2026-07-18 21:00:04] step 983850/1000000 | loss=1.9605 vocab=0.9551 token=1.5559 | lr=1.86e-06 | 6.13 steps/s | VRAM=28.82 Go [2026-07-18 21:00:12] step 983900/1000000 | loss=2.0679 vocab=0.7715 token=1.1801 | lr=1.85e-06 | 6.15 steps/s | VRAM=30.15 Go [2026-07-18 21:00:20] step 983950/1000000 | loss=1.9518 vocab=0.5233 token=1.0526 | lr=1.85e-06 | 6.13 steps/s | VRAM=29.26 Go [2026-07-18 21:00:28] step 984000/1000000 | loss=1.9717 vocab=0.5443 token=0.9371 | lr=1.84e-06 | 6.16 steps/s | VRAM=28.95 Go [2026-07-18 21:00:36] step 984050/1000000 | loss=1.9260 vocab=0.6552 token=0.9562 | lr=1.83e-06 | 6.14 steps/s | VRAM=28.84 Go [2026-07-18 21:00:44] step 984100/1000000 | loss=2.0423 vocab=0.6961 token=0.9191 | lr=1.83e-06 | 6.17 steps/s | VRAM=25.04 Go [2026-07-18 21:00:52] step 984150/1000000 | loss=1.9595 vocab=0.8179 token=1.0799 | lr=1.82e-06 | 6.16 steps/s | VRAM=28.37 Go [2026-07-18 21:01:01] step 984200/1000000 | loss=2.0867 vocab=0.7302 token=1.2298 | lr=1.82e-06 | 6.13 steps/s | VRAM=28.64 Go [2026-07-18 21:01:09] step 984250/1000000 | loss=1.9805 vocab=1.0917 token=1.0668 | lr=1.81e-06 | 6.13 steps/s | VRAM=32.48 Go [2026-07-18 21:01:17] step 984300/1000000 | loss=1.9545 vocab=0.4830 token=0.7392 | lr=1.81e-06 | 6.15 steps/s | VRAM=28.45 Go [2026-07-18 21:01:25] step 984350/1000000 | loss=2.0524 vocab=0.6720 token=1.1865 | lr=1.80e-06 | 6.18 steps/s | VRAM=32.14 Go [2026-07-18 21:01:33] step 984400/1000000 | loss=2.0375 vocab=0.6889 token=1.2085 | lr=1.80e-06 | 6.15 steps/s | VRAM=25.75 Go [2026-07-18 21:01:41] step 984450/1000000 | loss=1.9189 vocab=0.4138 token=1.0437 | lr=1.79e-06 | 6.13 steps/s | VRAM=30.11 Go [2026-07-18 21:01:49] step 984500/1000000 | loss=2.0075 vocab=0.7289 token=1.2910 | lr=1.79e-06 | 6.17 steps/s | VRAM=25.67 Go [2026-07-18 21:01:58] step 984550/1000000 | loss=1.9478 vocab=0.5530 token=1.3825 | lr=1.78e-06 | 6.12 steps/s | VRAM=25.73 Go [2026-07-18 21:02:06] step 984600/1000000 | loss=2.0989 vocab=0.8266 token=1.1549 | lr=1.78e-06 | 6.19 steps/s | VRAM=31.78 Go [2026-07-18 21:02:14] step 984650/1000000 | loss=1.9674 vocab=0.4011 token=1.1372 | lr=1.77e-06 | 6.18 steps/s | VRAM=25.54 Go [2026-07-18 21:02:22] step 984700/1000000 | loss=2.0499 vocab=1.0579 token=0.9109 | lr=1.77e-06 | 6.19 steps/s | VRAM=31.53 Go [2026-07-18 21:02:30] step 984750/1000000 | loss=1.9823 vocab=1.3955 token=1.2733 | lr=1.76e-06 | 6.17 steps/s | VRAM=27.92 Go [2026-07-18 21:02:38] step 984800/1000000 | loss=2.0743 vocab=0.9826 token=1.3024 | lr=1.76e-06 | 6.18 steps/s | VRAM=32.59 Go [2026-07-18 21:02:46] step 984850/1000000 | loss=2.0691 vocab=1.1517 token=0.9731 | lr=1.75e-06 | 6.17 steps/s | VRAM=29.27 Go [2026-07-18 21:02:54] step 984900/1000000 | loss=1.9126 vocab=0.7592 token=1.0780 | lr=1.75e-06 | 6.18 steps/s | VRAM=32.16 Go [2026-07-18 21:03:02] step 984950/1000000 | loss=1.8907 vocab=1.0155 token=1.5807 | lr=1.74e-06 | 6.12 steps/s | VRAM=32.66 Go [2026-07-18 21:03:11] step 985000/1000000 | loss=1.9502 vocab=0.6551 token=1.2346 | lr=1.74e-06 | 6.15 steps/s | VRAM=28.62 Go [2026-07-18 21:03:19] step 985050/1000000 | loss=2.1471 vocab=0.5764 token=0.7774 | lr=1.73e-06 | 6.13 steps/s | VRAM=29.83 Go [2026-07-18 21:03:27] step 985100/1000000 | loss=1.9979 vocab=0.6745 token=1.1622 | lr=1.73e-06 | 6.19 steps/s | VRAM=26.80 Go [2026-07-18 21:03:35] step 985150/1000000 | loss=1.9330 vocab=0.9335 token=1.2883 | lr=1.72e-06 | 6.15 steps/s | VRAM=26.22 Go [2026-07-18 21:03:43] step 985200/1000000 | loss=1.9936 vocab=0.8809 token=1.2994 | lr=1.72e-06 | 6.17 steps/s | VRAM=28.45 Go [2026-07-18 21:03:51] step 985250/1000000 | loss=2.0363 vocab=1.2054 token=1.0113 | lr=1.71e-06 | 6.16 steps/s | VRAM=28.72 Go [2026-07-18 21:03:59] step 985300/1000000 | loss=1.9885 vocab=0.7489 token=1.5470 | lr=1.71e-06 | 6.13 steps/s | VRAM=30.27 Go [2026-07-18 21:04:07] step 985350/1000000 | loss=2.0054 vocab=0.3156 token=1.1105 | lr=1.70e-06 | 6.18 steps/s | VRAM=31.83 Go [2026-07-18 21:04:15] step 985400/1000000 | loss=2.0000 vocab=0.7332 token=0.9244 | lr=1.70e-06 | 6.13 steps/s | VRAM=26.35 Go [2026-07-18 21:04:24] step 985450/1000000 | loss=1.9843 vocab=1.2085 token=1.3445 | lr=1.69e-06 | 6.12 steps/s | VRAM=29.02 Go [2026-07-18 21:04:32] step 985500/1000000 | loss=1.9328 vocab=1.1501 token=1.2146 | lr=1.69e-06 | 6.16 steps/s | VRAM=29.39 Go [2026-07-18 21:04:40] step 985550/1000000 | loss=2.0468 vocab=0.6405 token=1.1013 | lr=1.69e-06 | 6.13 steps/s | VRAM=26.28 Go [2026-07-18 21:04:48] step 985600/1000000 | loss=1.9174 vocab=1.2501 token=1.1717 | lr=1.68e-06 | 6.16 steps/s | VRAM=29.06 Go [2026-07-18 21:04:56] step 985650/1000000 | loss=1.9672 vocab=0.8087 token=1.0143 | lr=1.68e-06 | 6.17 steps/s | VRAM=32.29 Go [2026-07-18 21:05:04] step 985700/1000000 | loss=1.9883 vocab=0.7974 token=1.1597 | lr=1.67e-06 | 6.13 steps/s | VRAM=29.11 Go [2026-07-18 21:05:12] step 985750/1000000 | loss=1.9468 vocab=0.5671 token=0.5735 | lr=1.67e-06 | 6.15 steps/s | VRAM=26.04 Go [2026-07-18 21:05:21] step 985800/1000000 | loss=2.0611 vocab=0.7600 token=1.4706 | lr=1.66e-06 | 6.17 steps/s | VRAM=29.26 Go [2026-07-18 21:05:29] step 985850/1000000 | loss=2.0375 vocab=0.4510 token=1.0829 | lr=1.66e-06 | 6.16 steps/s | VRAM=28.93 Go [2026-07-18 21:05:37] step 985900/1000000 | loss=1.8736 vocab=1.2645 token=1.3244 | lr=1.65e-06 | 6.17 steps/s | VRAM=32.69 Go [2026-07-18 21:05:45] step 985950/1000000 | loss=1.9654 vocab=1.1703 token=1.0202 | lr=1.65e-06 | 6.17 steps/s | VRAM=29.25 Go [2026-07-18 21:05:53] step 986000/1000000 | loss=1.9699 vocab=0.9400 token=1.1293 | lr=1.64e-06 | 6.16 steps/s | VRAM=30.48 Go [2026-07-18 21:06:01] step 986050/1000000 | loss=2.0133 vocab=0.6603 token=0.9313 | lr=1.64e-06 | 6.10 steps/s | VRAM=27.88 Go [2026-07-18 21:06:09] step 986100/1000000 | loss=2.0173 vocab=1.0939 token=1.5530 | lr=1.63e-06 | 6.14 steps/s | VRAM=28.13 Go [2026-07-18 21:06:17] step 986150/1000000 | loss=1.9147 vocab=0.3275 token=0.9399 | lr=1.63e-06 | 6.16 steps/s | VRAM=28.49 Go [2026-07-18 21:06:26] step 986200/1000000 | loss=2.0773 vocab=0.8963 token=1.4702 | lr=1.63e-06 | 6.15 steps/s | VRAM=27.37 Go [2026-07-18 21:06:34] step 986250/1000000 | loss=2.0670 vocab=0.4759 token=0.8837 | lr=1.62e-06 | 6.14 steps/s | VRAM=28.99 Go [2026-07-18 21:06:42] step 986300/1000000 | loss=1.9909 vocab=1.4591 token=1.3446 | lr=1.62e-06 | 6.14 steps/s | VRAM=30.06 Go [2026-07-18 21:06:50] step 986350/1000000 | loss=1.9663 vocab=0.9826 token=1.3266 | lr=1.61e-06 | 6.12 steps/s | VRAM=29.06 Go [2026-07-18 21:06:58] step 986400/1000000 | loss=1.9512 vocab=0.7843 token=1.3348 | lr=1.61e-06 | 6.14 steps/s | VRAM=32.64 Go [2026-07-18 21:07:06] step 986450/1000000 | loss=2.1167 vocab=0.6420 token=0.9666 | lr=1.60e-06 | 6.07 steps/s | VRAM=27.58 Go [2026-07-18 21:07:15] step 986500/1000000 | loss=1.9497 vocab=1.0088 token=1.1678 | lr=1.60e-06 | 6.18 steps/s | VRAM=32.27 Go [2026-07-18 21:07:23] step 986550/1000000 | loss=1.9102 vocab=0.7411 token=1.1584 | lr=1.59e-06 | 6.18 steps/s | VRAM=32.94 Go [2026-07-18 21:07:31] step 986600/1000000 | loss=2.0385 vocab=1.0959 token=1.0183 | lr=1.59e-06 | 6.13 steps/s | VRAM=26.22 Go [2026-07-18 21:07:39] step 986650/1000000 | loss=1.9363 vocab=0.6986 token=0.9243 | lr=1.59e-06 | 6.16 steps/s | VRAM=32.72 Go [2026-07-18 21:07:47] step 986700/1000000 | loss=2.1384 vocab=0.6559 token=1.1772 | lr=1.58e-06 | 6.15 steps/s | VRAM=27.94 Go [2026-07-18 21:07:55] step 986750/1000000 | loss=1.9669 vocab=0.9499 token=1.1481 | lr=1.58e-06 | 6.16 steps/s | VRAM=32.13 Go [2026-07-18 21:08:03] step 986800/1000000 | loss=2.0206 vocab=0.9768 token=1.0429 | lr=1.57e-06 | 6.13 steps/s | VRAM=26.65 Go [2026-07-18 21:08:11] step 986850/1000000 | loss=1.9779 vocab=1.6170 token=1.6403 | lr=1.57e-06 | 6.15 steps/s | VRAM=32.64 Go [2026-07-18 21:08:19] step 986900/1000000 | loss=2.0110 vocab=0.9397 token=1.1655 | lr=1.56e-06 | 6.19 steps/s | VRAM=30.96 Go [2026-07-18 21:08:28] step 986950/1000000 | loss=2.0223 vocab=0.9586 token=1.3258 | lr=1.56e-06 | 6.13 steps/s | VRAM=30.32 Go [2026-07-18 21:08:36] step 987000/1000000 | loss=2.0432 vocab=0.7605 token=1.3317 | lr=1.55e-06 | 6.14 steps/s | VRAM=27.84 Go [2026-07-18 21:08:44] step 987050/1000000 | loss=1.9648 vocab=1.2288 token=1.1804 | lr=1.55e-06 | 6.17 steps/s | VRAM=28.54 Go [2026-07-18 21:08:52] step 987100/1000000 | loss=1.9272 vocab=0.9107 token=1.1621 | lr=1.55e-06 | 6.16 steps/s | VRAM=27.12 Go [2026-07-18 21:09:00] step 987150/1000000 | loss=1.9427 vocab=0.7605 token=0.8712 | lr=1.54e-06 | 6.16 steps/s | VRAM=28.39 Go [2026-07-18 21:09:08] step 987200/1000000 | loss=1.8906 vocab=0.8131 token=1.1852 | lr=1.54e-06 | 6.14 steps/s | VRAM=30.75 Go [2026-07-18 21:09:16] step 987250/1000000 | loss=2.0068 vocab=0.8252 token=1.0293 | lr=1.53e-06 | 6.14 steps/s | VRAM=26.12 Go [2026-07-18 21:09:25] step 987300/1000000 | loss=1.9599 vocab=0.6146 token=0.8575 | lr=1.53e-06 | 6.15 steps/s | VRAM=32.25 Go [2026-07-18 21:09:33] step 987350/1000000 | loss=1.8992 vocab=0.7317 token=1.0290 | lr=1.53e-06 | 6.14 steps/s | VRAM=29.21 Go [2026-07-18 21:09:41] step 987400/1000000 | loss=1.9733 vocab=0.5482 token=1.0394 | lr=1.52e-06 | 6.16 steps/s | VRAM=29.71 Go [2026-07-18 21:09:49] step 987450/1000000 | loss=2.0055 vocab=1.0353 token=1.0879 | lr=1.52e-06 | 6.12 steps/s | VRAM=27.03 Go [2026-07-18 21:09:57] step 987500/1000000 | loss=2.0693 vocab=0.8147 token=0.9735 | lr=1.51e-06 | 6.16 steps/s | VRAM=28.86 Go [2026-07-18 21:10:05] step 987550/1000000 | loss=1.9726 vocab=0.9668 token=1.5825 | lr=1.51e-06 | 6.19 steps/s | VRAM=30.93 Go [2026-07-18 21:10:13] step 987600/1000000 | loss=1.9856 vocab=1.0449 token=1.4617 | lr=1.50e-06 | 6.16 steps/s | VRAM=32.11 Go [2026-07-18 21:10:21] step 987650/1000000 | loss=1.9844 vocab=0.7843 token=1.2356 | lr=1.50e-06 | 6.16 steps/s | VRAM=22.84 Go [2026-07-18 21:10:30] step 987700/1000000 | loss=1.9614 vocab=0.9849 token=1.2314 | lr=1.50e-06 | 6.17 steps/s | VRAM=25.68 Go [2026-07-18 21:10:38] step 987750/1000000 | loss=1.9361 vocab=0.5455 token=0.8374 | lr=1.49e-06 | 6.15 steps/s | VRAM=24.51 Go [2026-07-18 21:10:46] step 987800/1000000 | loss=1.9441 vocab=0.8217 token=1.2292 | lr=1.49e-06 | 6.17 steps/s | VRAM=28.69 Go [2026-07-18 21:10:54] step 987850/1000000 | loss=1.9678 vocab=0.5430 token=1.2118 | lr=1.48e-06 | 6.13 steps/s | VRAM=23.42 Go [2026-07-18 21:11:02] step 987900/1000000 | loss=1.9457 vocab=0.7040 token=1.2509 | lr=1.48e-06 | 6.16 steps/s | VRAM=32.40 Go [2026-07-18 21:11:10] step 987950/1000000 | loss=1.9672 vocab=0.3548 token=0.7397 | lr=1.48e-06 | 6.15 steps/s | VRAM=28.61 Go [2026-07-18 21:11:18] step 988000/1000000 | loss=2.0062 vocab=0.8204 token=1.2176 | lr=1.47e-06 | 6.18 steps/s | VRAM=32.70 Go [2026-07-18 21:11:26] step 988050/1000000 | loss=1.9824 vocab=0.4982 token=0.8784 | lr=1.47e-06 | 6.12 steps/s | VRAM=29.30 Go [2026-07-18 21:11:35] step 988100/1000000 | loss=2.0256 vocab=0.4519 token=1.1503 | lr=1.47e-06 | 6.16 steps/s | VRAM=28.62 Go [2026-07-18 21:11:43] step 988150/1000000 | loss=1.9412 vocab=0.6335 token=1.3682 | lr=1.46e-06 | 6.16 steps/s | VRAM=28.82 Go [2026-07-18 21:11:51] step 988200/1000000 | loss=2.0402 vocab=0.7652 token=0.9351 | lr=1.46e-06 | 6.11 steps/s | VRAM=26.74 Go [2026-07-18 21:11:59] step 988250/1000000 | loss=1.9495 vocab=1.3949 token=1.3211 | lr=1.45e-06 | 6.18 steps/s | VRAM=27.69 Go [2026-07-18 21:12:07] step 988300/1000000 | loss=2.0212 vocab=0.8226 token=1.5219 | lr=1.45e-06 | 6.12 steps/s | VRAM=29.14 Go [2026-07-18 21:12:15] step 988350/1000000 | loss=1.9702 vocab=0.5655 token=0.9482 | lr=1.45e-06 | 6.14 steps/s | VRAM=29.85 Go [2026-07-18 21:12:23] step 988400/1000000 | loss=1.9370 vocab=0.5353 token=0.8069 | lr=1.44e-06 | 6.14 steps/s | VRAM=28.64 Go [2026-07-18 21:12:32] step 988450/1000000 | loss=1.9220 vocab=1.2209 token=1.2670 | lr=1.44e-06 | 6.14 steps/s | VRAM=28.76 Go [2026-07-18 21:12:40] step 988500/1000000 | loss=2.0335 vocab=1.4110 token=1.2034 | lr=1.43e-06 | 6.16 steps/s | VRAM=24.91 Go [2026-07-18 21:12:48] step 988550/1000000 | loss=1.9605 vocab=1.0760 token=0.8478 | lr=1.43e-06 | 6.15 steps/s | VRAM=30.84 Go [2026-07-18 21:12:56] step 988600/1000000 | loss=1.9332 vocab=0.5633 token=1.1912 | lr=1.43e-06 | 6.16 steps/s | VRAM=28.90 Go [2026-07-18 21:13:04] step 988650/1000000 | loss=2.0506 vocab=1.1773 token=1.3816 | lr=1.42e-06 | 6.17 steps/s | VRAM=32.07 Go [2026-07-18 21:13:12] step 988700/1000000 | loss=1.9248 vocab=1.0116 token=1.1232 | lr=1.42e-06 | 6.16 steps/s | VRAM=31.35 Go [2026-07-18 21:13:20] step 988750/1000000 | loss=2.0825 vocab=0.8188 token=1.0819 | lr=1.42e-06 | 6.13 steps/s | VRAM=32.07 Go [2026-07-18 21:13:28] step 988800/1000000 | loss=2.0638 vocab=1.0031 token=1.2235 | lr=1.41e-06 | 6.14 steps/s | VRAM=28.72 Go [2026-07-18 21:13:37] step 988850/1000000 | loss=1.9513 vocab=0.5008 token=1.0350 | lr=1.41e-06 | 6.15 steps/s | VRAM=29.01 Go [2026-07-18 21:13:45] step 988900/1000000 | loss=2.0822 vocab=1.3296 token=1.1673 | lr=1.40e-06 | 6.15 steps/s | VRAM=28.36 Go [2026-07-18 21:13:53] step 988950/1000000 | loss=2.1079 vocab=0.6933 token=1.1742 | lr=1.40e-06 | 6.14 steps/s | VRAM=29.27 Go [2026-07-18 21:14:01] step 989000/1000000 | loss=2.0637 vocab=0.9670 token=1.3143 | lr=1.40e-06 | 6.14 steps/s | VRAM=27.95 Go [2026-07-18 21:14:09] step 989050/1000000 | loss=1.9445 vocab=0.8766 token=1.5665 | lr=1.39e-06 | 6.14 steps/s | VRAM=28.45 Go [2026-07-18 21:14:17] step 989100/1000000 | loss=1.9289 vocab=0.7528 token=1.3182 | lr=1.39e-06 | 6.17 steps/s | VRAM=31.34 Go [2026-07-18 21:14:25] step 989150/1000000 | loss=1.9880 vocab=0.9759 token=1.6067 | lr=1.39e-06 | 6.14 steps/s | VRAM=26.19 Go [2026-07-18 21:14:33] step 989200/1000000 | loss=2.1199 vocab=1.0488 token=1.4188 | lr=1.38e-06 | 6.15 steps/s | VRAM=25.85 Go [2026-07-18 21:14:42] step 989250/1000000 | loss=2.1154 vocab=1.1947 token=1.3294 | lr=1.38e-06 | 6.15 steps/s | VRAM=31.99 Go [2026-07-18 21:14:50] step 989300/1000000 | loss=1.9217 vocab=0.9494 token=0.9781 | lr=1.38e-06 | 6.15 steps/s | VRAM=32.02 Go [2026-07-18 21:14:58] step 989350/1000000 | loss=1.9431 vocab=0.8440 token=1.1633 | lr=1.37e-06 | 6.15 steps/s | VRAM=32.70 Go [2026-07-18 21:15:06] step 989400/1000000 | loss=2.0624 vocab=1.7152 token=1.7937 | lr=1.37e-06 | 6.16 steps/s | VRAM=31.71 Go [2026-07-18 21:15:14] step 989450/1000000 | loss=2.0327 vocab=1.1795 token=1.2321 | lr=1.37e-06 | 6.14 steps/s | VRAM=29.04 Go [2026-07-18 21:15:22] step 989500/1000000 | loss=2.0192 vocab=0.4871 token=0.7728 | lr=1.36e-06 | 6.16 steps/s | VRAM=28.08 Go [2026-07-18 21:15:30] step 989550/1000000 | loss=2.0942 vocab=1.3749 token=1.2591 | lr=1.36e-06 | 6.15 steps/s | VRAM=32.91 Go [2026-07-18 21:15:38] step 989600/1000000 | loss=1.9659 vocab=0.7299 token=0.9989 | lr=1.36e-06 | 6.16 steps/s | VRAM=31.92 Go [2026-07-18 21:15:47] step 989650/1000000 | loss=2.0054 vocab=0.5861 token=0.8167 | lr=1.35e-06 | 6.14 steps/s | VRAM=26.68 Go [2026-07-18 21:15:55] step 989700/1000000 | loss=1.9001 vocab=0.5772 token=1.1498 | lr=1.35e-06 | 6.16 steps/s | VRAM=31.51 Go [2026-07-18 21:16:03] step 989750/1000000 | loss=1.9320 vocab=1.0416 token=1.2478 | lr=1.35e-06 | 6.20 steps/s | VRAM=31.33 Go [2026-07-18 21:16:11] step 989800/1000000 | loss=2.0834 vocab=0.9469 token=1.3080 | lr=1.34e-06 | 6.17 steps/s | VRAM=28.86 Go [2026-07-18 21:16:19] step 989850/1000000 | loss=1.8591 vocab=0.9935 token=0.8452 | lr=1.34e-06 | 6.15 steps/s | VRAM=26.06 Go [2026-07-18 21:16:27] step 989900/1000000 | loss=1.9801 vocab=1.0976 token=1.2172 | lr=1.34e-06 | 6.14 steps/s | VRAM=26.33 Go [2026-07-18 21:16:35] step 989950/1000000 | loss=1.9119 vocab=0.6398 token=1.2067 | lr=1.33e-06 | 6.18 steps/s | VRAM=31.57 Go [2026-07-18 21:16:43] step 990000/1000000 | loss=1.8864 vocab=0.8978 token=1.0529 | lr=1.33e-06 | 6.20 steps/s | VRAM=31.79 Go [2026-07-18 21:16:52] step 990050/1000000 | loss=1.8276 vocab=0.5926 token=0.9776 | lr=1.33e-06 | 6.13 steps/s | VRAM=28.19 Go [2026-07-18 21:17:00] step 990100/1000000 | loss=1.9556 vocab=0.8486 token=1.5533 | lr=1.32e-06 | 6.13 steps/s | VRAM=28.80 Go [2026-07-18 21:17:08] step 990150/1000000 | loss=1.9693 vocab=0.9669 token=1.5000 | lr=1.32e-06 | 6.13 steps/s | VRAM=26.79 Go [2026-07-18 21:17:16] step 990200/1000000 | loss=1.9535 vocab=0.4009 token=1.1038 | lr=1.32e-06 | 6.15 steps/s | VRAM=29.07 Go [2026-07-18 21:17:24] step 990250/1000000 | loss=2.0170 vocab=0.6281 token=1.0290 | lr=1.31e-06 | 6.15 steps/s | VRAM=28.10 Go [2026-07-18 21:17:32] step 990300/1000000 | loss=1.9896 vocab=0.7755 token=0.9237 | lr=1.31e-06 | 6.17 steps/s | VRAM=28.42 Go [2026-07-18 21:17:40] step 990350/1000000 | loss=2.0274 vocab=0.7996 token=1.5068 | lr=1.31e-06 | 6.18 steps/s | VRAM=31.08 Go [2026-07-18 21:17:48] step 990400/1000000 | loss=2.0598 vocab=1.3248 token=1.2355 | lr=1.30e-06 | 6.15 steps/s | VRAM=27.98 Go [2026-07-18 21:17:57] step 990450/1000000 | loss=1.9939 vocab=1.2625 token=1.4028 | lr=1.30e-06 | 6.14 steps/s | VRAM=25.80 Go [2026-07-18 21:18:05] step 990500/1000000 | loss=2.0219 vocab=0.3769 token=0.9481 | lr=1.30e-06 | 6.15 steps/s | VRAM=26.11 Go [2026-07-18 21:18:13] step 990550/1000000 | loss=1.9182 vocab=0.8726 token=1.3433 | lr=1.29e-06 | 6.16 steps/s | VRAM=32.12 Go [2026-07-18 21:18:21] step 990600/1000000 | loss=2.1000 vocab=0.5544 token=0.6977 | lr=1.29e-06 | 6.11 steps/s | VRAM=26.63 Go [2026-07-18 21:18:29] step 990650/1000000 | loss=1.9999 vocab=1.0257 token=1.6196 | lr=1.29e-06 | 6.18 steps/s | VRAM=25.41 Go [2026-07-18 21:18:37] step 990700/1000000 | loss=2.0543 vocab=0.5777 token=1.3062 | lr=1.28e-06 | 6.15 steps/s | VRAM=26.80 Go [2026-07-18 21:18:45] step 990750/1000000 | loss=1.8744 vocab=0.8152 token=1.2737 | lr=1.28e-06 | 6.14 steps/s | VRAM=28.21 Go [2026-07-18 21:18:53] step 990800/1000000 | loss=2.0464 vocab=1.0105 token=1.5104 | lr=1.28e-06 | 6.16 steps/s | VRAM=31.58 Go [2026-07-18 21:19:02] step 990850/1000000 | loss=1.9702 vocab=0.5966 token=1.2163 | lr=1.28e-06 | 6.15 steps/s | VRAM=29.25 Go [2026-07-18 21:19:10] step 990900/1000000 | loss=2.1352 vocab=1.1023 token=1.2194 | lr=1.27e-06 | 6.16 steps/s | VRAM=25.18 Go [2026-07-18 21:19:18] step 990950/1000000 | loss=1.9914 vocab=0.6596 token=0.9771 | lr=1.27e-06 | 6.13 steps/s | VRAM=24.40 Go [2026-07-18 21:19:26] step 991000/1000000 | loss=2.0426 vocab=0.9529 token=0.8615 | lr=1.27e-06 | 6.18 steps/s | VRAM=28.17 Go [2026-07-18 21:19:34] step 991050/1000000 | loss=1.9897 vocab=0.6078 token=1.3343 | lr=1.26e-06 | 6.17 steps/s | VRAM=27.93 Go [2026-07-18 21:19:42] step 991100/1000000 | loss=2.0461 vocab=0.4718 token=0.8644 | lr=1.26e-06 | 6.16 steps/s | VRAM=28.25 Go [2026-07-18 21:19:50] step 991150/1000000 | loss=2.0091 vocab=1.0229 token=1.2426 | lr=1.26e-06 | 6.19 steps/s | VRAM=28.54 Go [2026-07-18 21:19:58] step 991200/1000000 | loss=2.0171 vocab=0.7863 token=1.2072 | lr=1.25e-06 | 6.15 steps/s | VRAM=26.34 Go [2026-07-18 21:20:07] step 991250/1000000 | loss=2.0432 vocab=1.1412 token=1.7447 | lr=1.25e-06 | 6.15 steps/s | VRAM=28.83 Go [2026-07-18 21:20:15] step 991300/1000000 | loss=2.0059 vocab=0.7593 token=1.0902 | lr=1.25e-06 | 6.15 steps/s | VRAM=32.87 Go [2026-07-18 21:20:23] step 991350/1000000 | loss=2.1264 vocab=0.8997 token=1.2201 | lr=1.25e-06 | 6.15 steps/s | VRAM=31.87 Go [2026-07-18 21:20:31] step 991400/1000000 | loss=1.9394 vocab=0.4279 token=1.1484 | lr=1.24e-06 | 6.16 steps/s | VRAM=29.23 Go [2026-07-18 21:20:39] step 991450/1000000 | loss=2.0532 vocab=0.4376 token=1.2346 | lr=1.24e-06 | 6.16 steps/s | VRAM=26.51 Go [2026-07-18 21:20:47] step 991500/1000000 | loss=2.0544 vocab=0.5939 token=1.0508 | lr=1.24e-06 | 6.17 steps/s | VRAM=28.42 Go [2026-07-18 21:20:55] step 991550/1000000 | loss=1.9793 vocab=0.5632 token=0.9701 | lr=1.23e-06 | 6.13 steps/s | VRAM=25.95 Go [2026-07-18 21:21:03] step 991600/1000000 | loss=1.9546 vocab=1.0540 token=0.8085 | lr=1.23e-06 | 6.17 steps/s | VRAM=31.22 Go [2026-07-18 21:21:12] step 991650/1000000 | loss=2.0066 vocab=0.7033 token=1.2278 | lr=1.23e-06 | 6.16 steps/s | VRAM=26.31 Go [2026-07-18 21:21:20] step 991700/1000000 | loss=2.0131 vocab=0.6527 token=1.1405 | lr=1.23e-06 | 6.17 steps/s | VRAM=31.04 Go [2026-07-18 21:21:28] step 991750/1000000 | loss=1.8991 vocab=0.6072 token=0.8852 | lr=1.22e-06 | 6.12 steps/s | VRAM=25.99 Go [2026-07-18 21:21:36] step 991800/1000000 | loss=1.9856 vocab=1.3465 token=1.1011 | lr=1.22e-06 | 6.12 steps/s | VRAM=28.42 Go [2026-07-18 21:21:44] step 991850/1000000 | loss=2.1212 vocab=0.8936 token=1.0731 | lr=1.22e-06 | 6.13 steps/s | VRAM=26.04 Go [2026-07-18 21:21:52] step 991900/1000000 | loss=1.9847 vocab=1.2777 token=1.2115 | lr=1.22e-06 | 6.17 steps/s | VRAM=31.98 Go [2026-07-18 21:22:00] step 991950/1000000 | loss=2.0671 vocab=1.0803 token=1.5686 | lr=1.21e-06 | 6.17 steps/s | VRAM=31.94 Go [2026-07-18 21:22:08] step 992000/1000000 | loss=2.0327 vocab=0.5037 token=1.0662 | lr=1.21e-06 | 6.15 steps/s | VRAM=32.21 Go [2026-07-18 21:22:17] step 992050/1000000 | loss=1.9668 vocab=0.7148 token=1.3093 | lr=1.21e-06 | 6.12 steps/s | VRAM=28.18 Go [2026-07-18 21:22:25] step 992100/1000000 | loss=1.9428 vocab=0.6383 token=1.4678 | lr=1.21e-06 | 6.11 steps/s | VRAM=27.32 Go [2026-07-18 21:22:33] step 992150/1000000 | loss=1.9770 vocab=0.6396 token=1.3329 | lr=1.20e-06 | 6.16 steps/s | VRAM=28.15 Go [2026-07-18 21:22:41] step 992200/1000000 | loss=2.0080 vocab=0.9009 token=1.0584 | lr=1.20e-06 | 6.09 steps/s | VRAM=24.07 Go [2026-07-18 21:22:49] step 992250/1000000 | loss=2.1105 vocab=0.6957 token=1.4522 | lr=1.20e-06 | 6.16 steps/s | VRAM=28.59 Go [2026-07-18 21:22:57] step 992300/1000000 | loss=1.9268 vocab=1.0927 token=1.2319 | lr=1.19e-06 | 6.15 steps/s | VRAM=28.05 Go [2026-07-18 21:23:05] step 992350/1000000 | loss=2.0832 vocab=0.7244 token=1.0537 | lr=1.19e-06 | 6.14 steps/s | VRAM=28.38 Go [2026-07-18 21:23:14] step 992400/1000000 | loss=1.9545 vocab=0.5313 token=1.0125 | lr=1.19e-06 | 6.16 steps/s | VRAM=32.47 Go [2026-07-18 21:23:22] step 992450/1000000 | loss=1.9514 vocab=0.7255 token=0.6916 | lr=1.19e-06 | 6.15 steps/s | VRAM=29.20 Go [2026-07-18 21:23:30] step 992500/1000000 | loss=2.0162 vocab=0.7368 token=1.1011 | lr=1.18e-06 | 6.12 steps/s | VRAM=26.22 Go [2026-07-18 21:23:38] step 992550/1000000 | loss=2.0583 vocab=1.0548 token=1.3507 | lr=1.18e-06 | 6.17 steps/s | VRAM=28.04 Go [2026-07-18 21:23:46] step 992600/1000000 | loss=1.9053 vocab=0.7492 token=1.3335 | lr=1.18e-06 | 6.17 steps/s | VRAM=31.50 Go [2026-07-18 21:23:54] step 992650/1000000 | loss=1.9945 vocab=0.6276 token=1.1755 | lr=1.18e-06 | 6.15 steps/s | VRAM=32.38 Go [2026-07-18 21:24:02] step 992700/1000000 | loss=2.0073 vocab=0.8528 token=1.2758 | lr=1.18e-06 | 6.17 steps/s | VRAM=28.42 Go [2026-07-18 21:24:10] step 992750/1000000 | loss=1.9510 vocab=0.7076 token=1.0571 | lr=1.17e-06 | 6.17 steps/s | VRAM=31.32 Go [2026-07-18 21:24:19] step 992800/1000000 | loss=2.1063 vocab=0.9243 token=1.4611 | lr=1.17e-06 | 6.15 steps/s | VRAM=29.32 Go [2026-07-18 21:24:27] step 992850/1000000 | loss=2.0913 vocab=0.7815 token=1.9444 | lr=1.17e-06 | 6.16 steps/s | VRAM=31.59 Go [2026-07-18 21:24:35] step 992900/1000000 | loss=2.0206 vocab=0.7537 token=0.8419 | lr=1.17e-06 | 6.16 steps/s | VRAM=28.68 Go [2026-07-18 21:24:43] step 992950/1000000 | loss=1.9673 vocab=0.9983 token=1.2956 | lr=1.16e-06 | 6.15 steps/s | VRAM=29.31 Go [2026-07-18 21:24:51] step 993000/1000000 | loss=1.9245 vocab=0.7736 token=1.3292 | lr=1.16e-06 | 6.13 steps/s | VRAM=29.33 Go [2026-07-18 21:24:59] step 993050/1000000 | loss=1.9654 vocab=1.0602 token=1.2410 | lr=1.16e-06 | 6.13 steps/s | VRAM=28.19 Go [2026-07-18 21:25:07] step 993100/1000000 | loss=1.9771 vocab=0.5044 token=1.0542 | lr=1.16e-06 | 6.16 steps/s | VRAM=31.60 Go [2026-07-18 21:25:16] step 993150/1000000 | loss=2.0840 vocab=0.7568 token=1.1690 | lr=1.15e-06 | 6.12 steps/s | VRAM=29.07 Go [2026-07-18 21:25:24] step 993200/1000000 | loss=1.8909 vocab=0.4671 token=0.7947 | lr=1.15e-06 | 6.14 steps/s | VRAM=28.44 Go [2026-07-18 21:25:32] step 993250/1000000 | loss=2.0056 vocab=0.9706 token=1.5133 | lr=1.15e-06 | 6.12 steps/s | VRAM=23.32 Go [2026-07-18 21:25:40] step 993300/1000000 | loss=1.9638 vocab=0.5398 token=0.7574 | lr=1.15e-06 | 6.17 steps/s | VRAM=31.63 Go [2026-07-18 21:25:48] step 993350/1000000 | loss=1.9658 vocab=1.0808 token=1.0438 | lr=1.15e-06 | 6.13 steps/s | VRAM=28.71 Go [2026-07-18 21:25:56] step 993400/1000000 | loss=1.8532 vocab=0.9397 token=1.3217 | lr=1.14e-06 | 6.16 steps/s | VRAM=29.09 Go [2026-07-18 21:26:04] step 993450/1000000 | loss=1.9284 vocab=0.8729 token=1.0616 | lr=1.14e-06 | 6.16 steps/s | VRAM=32.77 Go [2026-07-18 21:26:12] step 993500/1000000 | loss=2.0454 vocab=0.8614 token=0.8573 | lr=1.14e-06 | 6.15 steps/s | VRAM=29.83 Go [2026-07-18 21:26:21] step 993550/1000000 | loss=2.0317 vocab=0.5939 token=0.8719 | lr=1.14e-06 | 6.14 steps/s | VRAM=32.48 Go [2026-07-18 21:26:29] step 993600/1000000 | loss=2.0593 vocab=0.8941 token=1.2102 | lr=1.13e-06 | 6.13 steps/s | VRAM=25.83 Go [2026-07-18 21:26:37] step 993650/1000000 | loss=1.9616 vocab=0.6767 token=0.8325 | lr=1.13e-06 | 6.15 steps/s | VRAM=28.88 Go [2026-07-18 21:26:45] step 993700/1000000 | loss=1.9936 vocab=0.5541 token=1.2076 | lr=1.13e-06 | 6.16 steps/s | VRAM=29.43 Go [2026-07-18 21:26:53] step 993750/1000000 | loss=1.9834 vocab=1.2663 token=1.4454 | lr=1.13e-06 | 6.16 steps/s | VRAM=26.68 Go [2026-07-18 21:27:01] step 993800/1000000 | loss=2.0004 vocab=0.4427 token=0.9701 | lr=1.13e-06 | 6.13 steps/s | VRAM=26.05 Go [2026-07-18 21:27:09] step 993850/1000000 | loss=1.9930 vocab=1.2004 token=1.3288 | lr=1.12e-06 | 6.18 steps/s | VRAM=31.28 Go [2026-07-18 21:27:18] step 993900/1000000 | loss=1.9723 vocab=0.8820 token=1.5626 | lr=1.12e-06 | 6.15 steps/s | VRAM=24.88 Go [2026-07-18 21:27:26] step 993950/1000000 | loss=1.9899 vocab=0.7673 token=0.9714 | lr=1.12e-06 | 6.12 steps/s | VRAM=26.82 Go [2026-07-18 21:27:34] step 994000/1000000 | loss=2.0023 vocab=1.1804 token=0.9769 | lr=1.12e-06 | 6.17 steps/s | VRAM=31.76 Go [2026-07-18 21:27:42] step 994050/1000000 | loss=2.0805 vocab=0.9045 token=0.9113 | lr=1.12e-06 | 6.12 steps/s | VRAM=25.78 Go [2026-07-18 21:27:50] step 994100/1000000 | loss=1.9164 vocab=0.4812 token=1.2135 | lr=1.11e-06 | 6.13 steps/s | VRAM=32.67 Go [2026-07-18 21:27:58] step 994150/1000000 | loss=1.9825 vocab=1.0321 token=1.3610 | lr=1.11e-06 | 6.16 steps/s | VRAM=28.42 Go [2026-07-18 21:28:06] step 994200/1000000 | loss=2.0506 vocab=1.3724 token=1.0978 | lr=1.11e-06 | 6.15 steps/s | VRAM=25.92 Go [2026-07-18 21:28:15] step 994250/1000000 | loss=1.9381 vocab=0.6514 token=1.3398 | lr=1.11e-06 | 6.13 steps/s | VRAM=23.85 Go [2026-07-18 21:28:23] step 994300/1000000 | loss=1.9757 vocab=0.6695 token=1.6024 | lr=1.11e-06 | 6.13 steps/s | VRAM=28.35 Go [2026-07-18 21:28:31] step 994350/1000000 | loss=2.0332 vocab=0.5105 token=0.9594 | lr=1.10e-06 | 6.16 steps/s | VRAM=32.18 Go [2026-07-18 21:28:39] step 994400/1000000 | loss=2.0876 vocab=0.4450 token=0.9931 | lr=1.10e-06 | 6.15 steps/s | VRAM=32.78 Go [2026-07-18 21:28:47] step 994450/1000000 | loss=2.0114 vocab=0.7706 token=0.7081 | lr=1.10e-06 | 6.15 steps/s | VRAM=25.82 Go [2026-07-18 21:28:55] step 994500/1000000 | loss=1.9406 vocab=0.7054 token=0.9722 | lr=1.10e-06 | 6.12 steps/s | VRAM=27.07 Go [2026-07-18 21:29:03] step 994550/1000000 | loss=1.9549 vocab=0.7784 token=0.9576 | lr=1.10e-06 | 6.16 steps/s | VRAM=29.80 Go [2026-07-18 21:29:12] step 994600/1000000 | loss=2.0129 vocab=0.3047 token=0.8500 | lr=1.10e-06 | 6.14 steps/s | VRAM=31.79 Go [2026-07-18 21:29:20] step 994650/1000000 | loss=2.0047 vocab=1.0834 token=1.0887 | lr=1.09e-06 | 6.18 steps/s | VRAM=29.04 Go [2026-07-18 21:29:28] step 994700/1000000 | loss=1.9622 vocab=1.1467 token=1.2275 | lr=1.09e-06 | 6.15 steps/s | VRAM=28.90 Go [2026-07-18 21:29:36] step 994750/1000000 | loss=2.0726 vocab=0.7483 token=1.4807 | lr=1.09e-06 | 6.16 steps/s | VRAM=32.45 Go [2026-07-18 21:29:44] step 994800/1000000 | loss=2.0643 vocab=0.5820 token=1.5525 | lr=1.09e-06 | 6.16 steps/s | VRAM=32.64 Go [2026-07-18 21:29:52] step 994850/1000000 | loss=2.0001 vocab=0.9020 token=1.1654 | lr=1.09e-06 | 6.12 steps/s | VRAM=29.03 Go [2026-07-18 21:30:00] step 994900/1000000 | loss=2.0612 vocab=0.8330 token=1.0788 | lr=1.09e-06 | 6.14 steps/s | VRAM=28.59 Go [2026-07-18 21:30:08] step 994950/1000000 | loss=1.9863 vocab=0.8883 token=1.1332 | lr=1.08e-06 | 6.17 steps/s | VRAM=31.72 Go [2026-07-18 21:30:17] step 995000/1000000 | loss=2.0572 vocab=0.8849 token=1.0655 | lr=1.08e-06 | 6.12 steps/s | VRAM=29.68 Go [2026-07-18 21:30:25] step 995050/1000000 | loss=2.0249 vocab=1.0183 token=1.1262 | lr=1.08e-06 | 6.15 steps/s | VRAM=29.35 Go [2026-07-18 21:30:33] step 995100/1000000 | loss=1.9548 vocab=0.6949 token=1.0667 | lr=1.08e-06 | 6.13 steps/s | VRAM=26.18 Go [2026-07-18 21:30:41] step 995150/1000000 | loss=1.9779 vocab=0.5147 token=0.9827 | lr=1.08e-06 | 6.12 steps/s | VRAM=29.75 Go [2026-07-18 21:30:49] step 995200/1000000 | loss=1.9572 vocab=0.7128 token=1.4212 | lr=1.08e-06 | 6.18 steps/s | VRAM=31.17 Go [2026-07-18 21:30:57] step 995250/1000000 | loss=2.0268 vocab=0.8451 token=1.2520 | lr=1.07e-06 | 6.14 steps/s | VRAM=29.50 Go [2026-07-18 21:31:05] step 995300/1000000 | loss=2.0242 vocab=1.1733 token=1.2014 | lr=1.07e-06 | 6.14 steps/s | VRAM=28.35 Go [2026-07-18 21:31:13] step 995350/1000000 | loss=1.9370 vocab=0.9179 token=0.9639 | lr=1.07e-06 | 6.17 steps/s | VRAM=24.42 Go [2026-07-18 21:31:22] step 995400/1000000 | loss=2.0974 vocab=1.1965 token=1.3406 | lr=1.07e-06 | 6.18 steps/s | VRAM=27.75 Go [2026-07-18 21:31:30] step 995450/1000000 | loss=1.8925 vocab=0.3269 token=0.8979 | lr=1.07e-06 | 6.15 steps/s | VRAM=31.84 Go [2026-07-18 21:31:38] step 995500/1000000 | loss=1.9860 vocab=0.7367 token=0.9692 | lr=1.07e-06 | 6.20 steps/s | VRAM=31.35 Go [2026-07-18 21:31:46] step 995550/1000000 | loss=2.0153 vocab=0.6779 token=0.5140 | lr=1.07e-06 | 6.19 steps/s | VRAM=25.83 Go [2026-07-18 21:31:54] step 995600/1000000 | loss=2.0340 vocab=0.6075 token=1.2286 | lr=1.06e-06 | 6.15 steps/s | VRAM=32.07 Go [2026-07-18 21:32:02] step 995650/1000000 | loss=2.0214 vocab=0.6062 token=0.9701 | lr=1.06e-06 | 6.15 steps/s | VRAM=27.02 Go [2026-07-18 21:32:10] step 995700/1000000 | loss=2.0232 vocab=1.2561 token=1.3163 | lr=1.06e-06 | 6.14 steps/s | VRAM=30.03 Go [2026-07-18 21:32:18] step 995750/1000000 | loss=1.9501 vocab=0.8137 token=1.2044 | lr=1.06e-06 | 6.17 steps/s | VRAM=26.88 Go [2026-07-18 21:32:26] step 995800/1000000 | loss=2.0114 vocab=1.0915 token=1.2234 | lr=1.06e-06 | 6.15 steps/s | VRAM=32.57 Go [2026-07-18 21:32:35] step 995850/1000000 | loss=2.0467 vocab=1.0703 token=1.6038 | lr=1.06e-06 | 6.17 steps/s | VRAM=25.39 Go [2026-07-18 21:32:43] step 995900/1000000 | loss=2.0778 vocab=0.8518 token=1.0172 | lr=1.06e-06 | 6.18 steps/s | VRAM=32.49 Go [2026-07-18 21:32:51] step 995950/1000000 | loss=2.0028 vocab=0.8600 token=0.9898 | lr=1.05e-06 | 6.16 steps/s | VRAM=32.23 Go [2026-07-18 21:32:59] step 996000/1000000 | loss=2.1073 vocab=1.3137 token=1.5451 | lr=1.05e-06 | 6.12 steps/s | VRAM=28.06 Go [2026-07-18 21:33:07] step 996050/1000000 | loss=2.0284 vocab=0.7529 token=1.2464 | lr=1.05e-06 | 6.17 steps/s | VRAM=30.96 Go [2026-07-18 21:33:15] step 996100/1000000 | loss=1.8661 vocab=1.0771 token=1.4683 | lr=1.05e-06 | 6.13 steps/s | VRAM=27.92 Go [2026-07-18 21:33:23] step 996150/1000000 | loss=1.9667 vocab=0.8713 token=1.3568 | lr=1.05e-06 | 6.16 steps/s | VRAM=32.11 Go [2026-07-18 21:33:32] step 996200/1000000 | loss=1.9822 vocab=1.1059 token=1.5290 | lr=1.05e-06 | 6.12 steps/s | VRAM=32.73 Go [2026-07-18 21:33:40] step 996250/1000000 | loss=2.0819 vocab=0.7061 token=1.0597 | lr=1.05e-06 | 6.15 steps/s | VRAM=28.95 Go [2026-07-18 21:33:48] step 996300/1000000 | loss=2.0424 vocab=1.1649 token=1.4397 | lr=1.05e-06 | 6.15 steps/s | VRAM=29.19 Go [2026-07-18 21:33:56] step 996350/1000000 | loss=1.9756 vocab=0.5872 token=0.7744 | lr=1.04e-06 | 6.14 steps/s | VRAM=26.87 Go [2026-07-18 21:34:04] step 996400/1000000 | loss=2.0453 vocab=0.8373 token=1.1336 | lr=1.04e-06 | 6.16 steps/s | VRAM=25.42 Go [2026-07-18 21:34:12] step 996450/1000000 | loss=2.0212 vocab=1.0035 token=0.5734 | lr=1.04e-06 | 6.14 steps/s | VRAM=28.53 Go [2026-07-18 21:34:20] step 996500/1000000 | loss=1.8196 vocab=0.7686 token=1.2490 | lr=1.04e-06 | 6.12 steps/s | VRAM=32.06 Go [2026-07-18 21:34:28] step 996550/1000000 | loss=1.8986 vocab=0.5660 token=0.9662 | lr=1.04e-06 | 6.16 steps/s | VRAM=26.18 Go [2026-07-18 21:34:37] step 996600/1000000 | loss=1.8878 vocab=0.7333 token=1.1814 | lr=1.04e-06 | 6.16 steps/s | VRAM=28.47 Go [2026-07-18 21:34:45] step 996650/1000000 | loss=1.9989 vocab=1.0021 token=1.2099 | lr=1.04e-06 | 6.12 steps/s | VRAM=28.23 Go [2026-07-18 21:34:53] step 996700/1000000 | loss=1.9860 vocab=0.8250 token=0.9733 | lr=1.04e-06 | 6.14 steps/s | VRAM=32.27 Go [2026-07-18 21:35:01] step 996750/1000000 | loss=1.9688 vocab=0.8395 token=1.3123 | lr=1.03e-06 | 6.14 steps/s | VRAM=29.32 Go [2026-07-18 21:35:09] step 996800/1000000 | loss=1.9239 vocab=1.0067 token=1.3009 | lr=1.03e-06 | 6.15 steps/s | VRAM=29.00 Go [2026-07-18 21:35:17] step 996850/1000000 | loss=1.9235 vocab=0.7378 token=0.9453 | lr=1.03e-06 | 6.15 steps/s | VRAM=28.80 Go [2026-07-18 21:35:25] step 996900/1000000 | loss=2.0830 vocab=0.8786 token=1.0941 | lr=1.03e-06 | 6.18 steps/s | VRAM=28.67 Go [2026-07-18 21:35:34] step 996950/1000000 | loss=2.0806 vocab=0.9890 token=0.8735 | lr=1.03e-06 | 6.12 steps/s | VRAM=26.45 Go [2026-07-18 21:35:42] step 997000/1000000 | loss=1.9704 vocab=0.9095 token=0.9254 | lr=1.03e-06 | 6.11 steps/s | VRAM=30.12 Go [2026-07-18 21:35:50] step 997050/1000000 | loss=2.0215 vocab=0.7330 token=1.1946 | lr=1.03e-06 | 6.15 steps/s | VRAM=32.13 Go [2026-07-18 21:35:58] step 997100/1000000 | loss=2.0059 vocab=1.1017 token=1.6289 | lr=1.03e-06 | 6.15 steps/s | VRAM=32.91 Go [2026-07-18 21:36:06] step 997150/1000000 | loss=2.0420 vocab=0.6299 token=1.3201 | lr=1.03e-06 | 6.14 steps/s | VRAM=28.66 Go [2026-07-18 21:36:14] step 997200/1000000 | loss=2.0039 vocab=0.2931 token=0.7943 | lr=1.03e-06 | 6.17 steps/s | VRAM=28.66 Go [2026-07-18 21:36:22] step 997250/1000000 | loss=2.0186 vocab=0.4698 token=0.8756 | lr=1.02e-06 | 6.13 steps/s | VRAM=28.53 Go [2026-07-18 21:36:31] step 997300/1000000 | loss=2.0229 vocab=0.8736 token=1.2245 | lr=1.02e-06 | 6.10 steps/s | VRAM=28.60 Go [2026-07-18 21:36:39] step 997350/1000000 | loss=1.8863 vocab=1.6899 token=1.2305 | lr=1.02e-06 | 6.12 steps/s | VRAM=29.88 Go [2026-07-18 21:36:47] step 997400/1000000 | loss=2.0112 vocab=0.4838 token=1.0116 | lr=1.02e-06 | 6.15 steps/s | VRAM=32.23 Go [2026-07-18 21:36:55] step 997450/1000000 | loss=1.8678 vocab=0.9393 token=0.8848 | lr=1.02e-06 | 6.10 steps/s | VRAM=28.62 Go [2026-07-18 21:37:03] step 997500/1000000 | loss=1.9203 vocab=0.4863 token=1.0076 | lr=1.02e-06 | 6.11 steps/s | VRAM=28.47 Go [2026-07-18 21:37:11] step 997550/1000000 | loss=2.0247 vocab=0.6879 token=1.1454 | lr=1.02e-06 | 6.17 steps/s | VRAM=28.54 Go [2026-07-18 21:37:20] step 997600/1000000 | loss=1.9367 vocab=0.2483 token=0.6335 | lr=1.02e-06 | 6.14 steps/s | VRAM=25.34 Go [2026-07-18 21:37:28] step 997650/1000000 | loss=2.0609 vocab=0.6434 token=1.0749 | lr=1.02e-06 | 6.14 steps/s | VRAM=29.63 Go [2026-07-18 21:37:36] step 997700/1000000 | loss=2.0370 vocab=0.7165 token=1.1137 | lr=1.02e-06 | 6.14 steps/s | VRAM=27.72 Go [2026-07-18 21:37:44] step 997750/1000000 | loss=1.9600 vocab=0.7941 token=1.1053 | lr=1.02e-06 | 6.16 steps/s | VRAM=31.47 Go [2026-07-18 21:37:52] step 997800/1000000 | loss=1.9950 vocab=0.6392 token=1.3401 | lr=1.02e-06 | 6.17 steps/s | VRAM=32.04 Go [2026-07-18 21:38:00] step 997850/1000000 | loss=1.9653 vocab=0.5551 token=0.8716 | lr=1.02e-06 | 6.11 steps/s | VRAM=26.61 Go [2026-07-18 21:38:08] step 997900/1000000 | loss=2.0339 vocab=0.4768 token=0.7182 | lr=1.01e-06 | 6.12 steps/s | VRAM=29.04 Go [2026-07-18 21:38:17] step 997950/1000000 | loss=2.0191 vocab=0.5406 token=0.7763 | lr=1.01e-06 | 6.14 steps/s | VRAM=28.07 Go [2026-07-18 21:38:25] step 998000/1000000 | loss=2.1100 vocab=0.6956 token=1.1124 | lr=1.01e-06 | 6.13 steps/s | VRAM=29.19 Go [2026-07-18 21:38:33] step 998050/1000000 | loss=2.0444 vocab=1.0133 token=1.3100 | lr=1.01e-06 | 6.12 steps/s | VRAM=27.02 Go [2026-07-18 21:38:41] step 998100/1000000 | loss=1.9432 vocab=1.0286 token=1.1504 | lr=1.01e-06 | 6.13 steps/s | VRAM=31.32 Go [2026-07-18 21:38:49] step 998150/1000000 | loss=1.9932 vocab=0.6539 token=1.0375 | lr=1.01e-06 | 6.11 steps/s | VRAM=28.71 Go [2026-07-18 21:38:57] step 998200/1000000 | loss=2.0102 vocab=0.9416 token=1.1600 | lr=1.01e-06 | 6.13 steps/s | VRAM=28.69 Go [2026-07-18 21:39:06] step 998250/1000000 | loss=1.9807 vocab=0.8136 token=1.0530 | lr=1.01e-06 | 6.14 steps/s | VRAM=25.59 Go [2026-07-18 21:39:14] step 998300/1000000 | loss=1.8939 vocab=0.6892 token=1.2933 | lr=1.01e-06 | 6.10 steps/s | VRAM=27.67 Go [2026-07-18 21:39:22] step 998350/1000000 | loss=2.0291 vocab=0.9632 token=1.1564 | lr=1.01e-06 | 6.11 steps/s | VRAM=28.98 Go [2026-07-18 21:39:30] step 998400/1000000 | loss=1.8826 vocab=0.9994 token=1.1125 | lr=1.01e-06 | 6.13 steps/s | VRAM=29.24 Go [2026-07-18 21:39:38] step 998450/1000000 | loss=2.0141 vocab=0.6328 token=1.2568 | lr=1.01e-06 | 6.12 steps/s | VRAM=29.23 Go [2026-07-18 21:39:46] step 998500/1000000 | loss=1.9840 vocab=1.4801 token=1.3619 | lr=1.01e-06 | 6.13 steps/s | VRAM=28.07 Go [2026-07-18 21:39:55] step 998550/1000000 | loss=1.9110 vocab=0.7058 token=0.9705 | lr=1.01e-06 | 6.09 steps/s | VRAM=28.94 Go [2026-07-18 21:40:03] step 998600/1000000 | loss=1.9914 vocab=0.5112 token=0.9391 | lr=1.01e-06 | 6.12 steps/s | VRAM=28.77 Go [2026-07-18 21:40:11] step 998650/1000000 | loss=2.0527 vocab=0.7880 token=1.3322 | lr=1.01e-06 | 6.14 steps/s | VRAM=28.04 Go [2026-07-18 21:40:19] step 998700/1000000 | loss=2.0022 vocab=0.9383 token=1.2584 | lr=1.01e-06 | 6.13 steps/s | VRAM=28.22 Go [2026-07-18 21:40:27] step 998750/1000000 | loss=2.0288 vocab=1.1315 token=1.2921 | lr=1.01e-06 | 6.14 steps/s | VRAM=29.57 Go [2026-07-18 21:40:35] step 998800/1000000 | loss=1.9494 vocab=0.5557 token=0.9227 | lr=1.00e-06 | 6.15 steps/s | VRAM=26.69 Go [2026-07-18 21:40:44] step 998850/1000000 | loss=1.9617 vocab=1.0464 token=1.2800 | lr=1.00e-06 | 6.12 steps/s | VRAM=26.38 Go [2026-07-18 21:40:52] step 998900/1000000 | loss=2.0349 vocab=0.8162 token=1.0888 | lr=1.00e-06 | 6.14 steps/s | VRAM=29.45 Go [2026-07-18 21:41:00] step 998950/1000000 | loss=2.0256 vocab=1.1366 token=1.0456 | lr=1.00e-06 | 6.13 steps/s | VRAM=28.50 Go [2026-07-18 21:41:08] step 999000/1000000 | loss=2.0931 vocab=1.0874 token=1.4191 | lr=1.00e-06 | 6.16 steps/s | VRAM=25.08 Go [2026-07-18 21:41:16] step 999050/1000000 | loss=2.0386 vocab=0.9557 token=1.2422 | lr=1.00e-06 | 6.14 steps/s | VRAM=30.04 Go [2026-07-18 21:41:24] step 999100/1000000 | loss=2.0030 vocab=0.8850 token=1.7499 | lr=1.00e-06 | 6.14 steps/s | VRAM=31.34 Go [2026-07-18 21:41:32] step 999150/1000000 | loss=1.9153 vocab=0.8404 token=1.5444 | lr=1.00e-06 | 6.13 steps/s | VRAM=28.59 Go [2026-07-18 21:41:41] step 999200/1000000 | loss=1.9496 vocab=1.1120 token=1.5386 | lr=1.00e-06 | 6.13 steps/s | VRAM=28.31 Go [2026-07-18 21:41:49] step 999250/1000000 | loss=2.0260 vocab=0.7516 token=1.1217 | lr=1.00e-06 | 6.12 steps/s | VRAM=29.14 Go [2026-07-18 21:41:57] step 999300/1000000 | loss=2.0046 vocab=0.7681 token=1.5841 | lr=1.00e-06 | 6.17 steps/s | VRAM=27.54 Go [2026-07-18 21:42:05] step 999350/1000000 | loss=2.0695 vocab=1.4201 token=1.4082 | lr=1.00e-06 | 6.13 steps/s | VRAM=27.38 Go [2026-07-18 21:42:13] step 999400/1000000 | loss=1.9333 vocab=0.6463 token=1.2136 | lr=1.00e-06 | 6.15 steps/s | VRAM=28.86 Go [2026-07-18 21:42:21] step 999450/1000000 | loss=2.0516 vocab=0.9033 token=0.7781 | lr=1.00e-06 | 6.18 steps/s | VRAM=28.72 Go [2026-07-18 21:42:29] step 999500/1000000 | loss=1.9434 vocab=0.8880 token=1.2194 | lr=1.00e-06 | 6.14 steps/s | VRAM=29.03 Go [2026-07-18 21:42:38] step 999550/1000000 | loss=1.9679 vocab=0.7371 token=1.0878 | lr=1.00e-06 | 6.12 steps/s | VRAM=32.85 Go [2026-07-18 21:42:46] step 999600/1000000 | loss=2.0239 vocab=1.0639 token=0.9813 | lr=1.00e-06 | 6.15 steps/s | VRAM=28.72 Go [2026-07-18 21:42:54] step 999650/1000000 | loss=1.9620 vocab=0.8450 token=1.1630 | lr=1.00e-06 | 6.11 steps/s | VRAM=28.50 Go [2026-07-18 21:43:02] step 999700/1000000 | loss=1.9259 vocab=1.2901 token=1.3311 | lr=1.00e-06 | 6.18 steps/s | VRAM=30.74 Go [2026-07-18 21:43:10] step 999750/1000000 | loss=1.9685 vocab=0.9107 token=1.1755 | lr=1.00e-06 | 6.16 steps/s | VRAM=31.56 Go [2026-07-18 21:43:18] step 999800/1000000 | loss=2.0124 vocab=1.1495 token=1.3557 | lr=1.00e-06 | 6.15 steps/s | VRAM=24.55 Go [2026-07-18 21:43:26] step 999850/1000000 | loss=2.0902 vocab=0.9363 token=1.0751 | lr=1.00e-06 | 6.13 steps/s | VRAM=26.00 Go [2026-07-18 21:43:34] step 999900/1000000 | loss=1.9914 vocab=0.4535 token=1.0393 | lr=1.00e-06 | 6.17 steps/s | VRAM=31.54 Go [2026-07-18 21:43:43] step 999950/1000000 | loss=1.9969 vocab=0.8660 token=1.0791 | lr=1.00e-06 | 6.13 steps/s | VRAM=25.54 Go [2026-07-18 21:43:51] step 1000000/1000000 | loss=1.8922 vocab=0.5753 token=1.2230 | lr=1.00e-06 | 6.16 steps/s | VRAM=31.58 Go [accuracy] step 1000000 : 65.52% (log: /workspace/plbert_train/precision_log.csv)