| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.4833252779120348, | |
| "eval_steps": 500, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 75.8236083984375, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 3.7148, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 52.23726272583008, | |
| "learning_rate": 9e-07, | |
| "loss": 3.2721, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 35.391998291015625, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 2.4262, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 15.225934028625488, | |
| "learning_rate": 1.9e-06, | |
| "loss": 1.8828, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 13.488654136657715, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 1.5398, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 8.773133277893066, | |
| "learning_rate": 2.8500000000000002e-06, | |
| "loss": 1.2772, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 6.736795902252197, | |
| "learning_rate": 3.3500000000000005e-06, | |
| "loss": 1.1159, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 2.9463019371032715, | |
| "learning_rate": 3.85e-06, | |
| "loss": 0.9883, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.3851969242095947, | |
| "learning_rate": 4.35e-06, | |
| "loss": 0.902, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.4642181396484375, | |
| "learning_rate": 4.85e-06, | |
| "loss": 0.8701, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.6514687538146973, | |
| "learning_rate": 5.3500000000000004e-06, | |
| "loss": 0.8548, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.4448617994785309, | |
| "learning_rate": 5.850000000000001e-06, | |
| "loss": 0.8038, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.5425733327865601, | |
| "learning_rate": 6.35e-06, | |
| "loss": 0.825, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.47352057695388794, | |
| "learning_rate": 6.8500000000000005e-06, | |
| "loss": 0.812, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.40401744842529297, | |
| "learning_rate": 7.35e-06, | |
| "loss": 0.7912, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.3112718164920807, | |
| "learning_rate": 7.850000000000001e-06, | |
| "loss": 0.8057, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.45351892709732056, | |
| "learning_rate": 8.350000000000001e-06, | |
| "loss": 0.7912, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.38080641627311707, | |
| "learning_rate": 8.85e-06, | |
| "loss": 0.7956, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.22949431836605072, | |
| "learning_rate": 9.35e-06, | |
| "loss": 0.7785, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.8331744074821472, | |
| "learning_rate": 9.85e-06, | |
| "loss": 0.7769, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.30213871598243713, | |
| "learning_rate": 1.035e-05, | |
| "loss": 0.7797, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.32015591859817505, | |
| "learning_rate": 1.0850000000000001e-05, | |
| "loss": 0.7941, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.22447015345096588, | |
| "learning_rate": 1.1350000000000001e-05, | |
| "loss": 0.8014, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.2546006143093109, | |
| "learning_rate": 1.185e-05, | |
| "loss": 0.771, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.23833389580249786, | |
| "learning_rate": 1.235e-05, | |
| "loss": 0.7956, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.1831669956445694, | |
| "learning_rate": 1.285e-05, | |
| "loss": 0.7833, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.19860319793224335, | |
| "learning_rate": 1.3350000000000001e-05, | |
| "loss": 0.7956, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.21220380067825317, | |
| "learning_rate": 1.3850000000000001e-05, | |
| "loss": 0.7942, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.26025038957595825, | |
| "learning_rate": 1.435e-05, | |
| "loss": 0.7877, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.2029470056295395, | |
| "learning_rate": 1.485e-05, | |
| "loss": 0.7637, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.21639183163642883, | |
| "learning_rate": 1.535e-05, | |
| "loss": 0.7885, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.23670387268066406, | |
| "learning_rate": 1.5850000000000002e-05, | |
| "loss": 0.7558, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.2855335772037506, | |
| "learning_rate": 1.635e-05, | |
| "loss": 0.7854, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.21515759825706482, | |
| "learning_rate": 1.6850000000000003e-05, | |
| "loss": 0.7577, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.18981383740901947, | |
| "learning_rate": 1.7349999999999998e-05, | |
| "loss": 0.7802, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.2674611508846283, | |
| "learning_rate": 1.785e-05, | |
| "loss": 0.7809, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.24821774661540985, | |
| "learning_rate": 1.8350000000000002e-05, | |
| "loss": 0.7965, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.3644929528236389, | |
| "learning_rate": 1.885e-05, | |
| "loss": 0.7957, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.3568972647190094, | |
| "learning_rate": 1.9350000000000003e-05, | |
| "loss": 0.7882, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.3236319422721863, | |
| "learning_rate": 1.985e-05, | |
| "loss": 0.7841, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.24553723633289337, | |
| "learning_rate": 2.035e-05, | |
| "loss": 0.7824, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.1855400949716568, | |
| "learning_rate": 2.085e-05, | |
| "loss": 0.7645, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.26212435960769653, | |
| "learning_rate": 2.135e-05, | |
| "loss": 0.7824, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.3167509138584137, | |
| "learning_rate": 2.1850000000000003e-05, | |
| "loss": 0.7908, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.21845023334026337, | |
| "learning_rate": 2.235e-05, | |
| "loss": 0.7888, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.30318892002105713, | |
| "learning_rate": 2.2850000000000003e-05, | |
| "loss": 0.7682, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.28828638792037964, | |
| "learning_rate": 2.3350000000000002e-05, | |
| "loss": 0.7588, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.2422240674495697, | |
| "learning_rate": 2.385e-05, | |
| "loss": 0.7787, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.3579326272010803, | |
| "learning_rate": 2.435e-05, | |
| "loss": 0.7932, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.3549540042877197, | |
| "learning_rate": 2.485e-05, | |
| "loss": 0.7634, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "eval_loss": 0.7744565606117249, | |
| "eval_runtime": 43.9626, | |
| "eval_samples_per_second": 45.493, | |
| "eval_steps_per_second": 0.364, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.2931281328201294, | |
| "learning_rate": 2.5350000000000003e-05, | |
| "loss": 0.7763, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.21228083968162537, | |
| "learning_rate": 2.585e-05, | |
| "loss": 0.7748, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.26342302560806274, | |
| "learning_rate": 2.6350000000000004e-05, | |
| "loss": 0.789, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.2840302288532257, | |
| "learning_rate": 2.6850000000000002e-05, | |
| "loss": 0.768, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.273771196603775, | |
| "learning_rate": 2.7350000000000004e-05, | |
| "loss": 0.7707, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 1.6022837162017822, | |
| "learning_rate": 2.7850000000000003e-05, | |
| "loss": 0.7705, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.3082919716835022, | |
| "learning_rate": 2.8349999999999998e-05, | |
| "loss": 0.7849, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.25341132283210754, | |
| "learning_rate": 2.885e-05, | |
| "loss": 0.7658, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.2250708043575287, | |
| "learning_rate": 2.935e-05, | |
| "loss": 0.756, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.23522531986236572, | |
| "learning_rate": 2.985e-05, | |
| "loss": 0.7692, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.2448827624320984, | |
| "learning_rate": 3.035e-05, | |
| "loss": 0.7635, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.19717586040496826, | |
| "learning_rate": 3.0850000000000004e-05, | |
| "loss": 0.7548, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.24299098551273346, | |
| "learning_rate": 3.135e-05, | |
| "loss": 0.7673, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.2555953562259674, | |
| "learning_rate": 3.185e-05, | |
| "loss": 0.7528, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.2102127969264984, | |
| "learning_rate": 3.235e-05, | |
| "loss": 0.7678, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.21767646074295044, | |
| "learning_rate": 3.2850000000000006e-05, | |
| "loss": 0.7564, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.3582068085670471, | |
| "learning_rate": 3.3350000000000004e-05, | |
| "loss": 0.7645, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.27804431319236755, | |
| "learning_rate": 3.385e-05, | |
| "loss": 0.7645, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.2464047521352768, | |
| "learning_rate": 3.435e-05, | |
| "loss": 0.7325, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.22323842346668243, | |
| "learning_rate": 3.485e-05, | |
| "loss": 0.7579, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.28329703211784363, | |
| "learning_rate": 3.535e-05, | |
| "loss": 0.7399, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.2471865713596344, | |
| "learning_rate": 3.585e-05, | |
| "loss": 0.765, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.314908504486084, | |
| "learning_rate": 3.635e-05, | |
| "loss": 0.7529, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.21466995775699615, | |
| "learning_rate": 3.685e-05, | |
| "loss": 0.7365, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.2545580565929413, | |
| "learning_rate": 3.735e-05, | |
| "loss": 0.7458, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.2651059329509735, | |
| "learning_rate": 3.7850000000000005e-05, | |
| "loss": 0.7578, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.2408379763364792, | |
| "learning_rate": 3.8350000000000004e-05, | |
| "loss": 0.7283, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.2575875520706177, | |
| "learning_rate": 3.885e-05, | |
| "loss": 0.7367, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.28182050585746765, | |
| "learning_rate": 3.935e-05, | |
| "loss": 0.7784, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.20579397678375244, | |
| "learning_rate": 3.9850000000000006e-05, | |
| "loss": 0.7533, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.1906532198190689, | |
| "learning_rate": 4.0350000000000005e-05, | |
| "loss": 0.7544, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.18311142921447754, | |
| "learning_rate": 4.085e-05, | |
| "loss": 0.75, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.1910839080810547, | |
| "learning_rate": 4.135e-05, | |
| "loss": 0.7538, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.2178005427122116, | |
| "learning_rate": 4.185e-05, | |
| "loss": 0.7456, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.27642378211021423, | |
| "learning_rate": 4.235e-05, | |
| "loss": 0.744, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.26685312390327454, | |
| "learning_rate": 4.285e-05, | |
| "loss": 0.7611, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.2021845579147339, | |
| "learning_rate": 4.335e-05, | |
| "loss": 0.758, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.20516808331012726, | |
| "learning_rate": 4.385e-05, | |
| "loss": 0.7404, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.20619668066501617, | |
| "learning_rate": 4.435e-05, | |
| "loss": 0.7379, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.22439967095851898, | |
| "learning_rate": 4.4850000000000006e-05, | |
| "loss": 0.7495, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.2192295491695404, | |
| "learning_rate": 4.5350000000000005e-05, | |
| "loss": 0.7377, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.26503369212150574, | |
| "learning_rate": 4.585e-05, | |
| "loss": 0.7481, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.2249353975057602, | |
| "learning_rate": 4.635e-05, | |
| "loss": 0.742, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.21413160860538483, | |
| "learning_rate": 4.685000000000001e-05, | |
| "loss": 0.7485, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.20971478521823883, | |
| "learning_rate": 4.735e-05, | |
| "loss": 0.7339, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.20295798778533936, | |
| "learning_rate": 4.785e-05, | |
| "loss": 0.722, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.1957596242427826, | |
| "learning_rate": 4.835e-05, | |
| "loss": 0.7538, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.18014560639858246, | |
| "learning_rate": 4.885e-05, | |
| "loss": 0.7251, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.19334861636161804, | |
| "learning_rate": 4.935e-05, | |
| "loss": 0.7444, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.1753440499305725, | |
| "learning_rate": 4.9850000000000006e-05, | |
| "loss": 0.7376, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "eval_loss": 0.73988938331604, | |
| "eval_runtime": 42.9535, | |
| "eval_samples_per_second": 46.562, | |
| "eval_steps_per_second": 0.372, | |
| "step": 1000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 10345, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 2.5288751418619986e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |