| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.4140012070006036, | |
| "eval_steps": 500, | |
| "global_step": 1500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 53.90464782714844, | |
| "learning_rate": 9.946323134728933e-06, | |
| "loss": 5.2914, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 53.05311965942383, | |
| "learning_rate": 9.892646269457864e-06, | |
| "loss": 4.9539, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 64.5956802368164, | |
| "learning_rate": 9.838969404186796e-06, | |
| "loss": 4.7374, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 54.43781280517578, | |
| "learning_rate": 9.785292538915728e-06, | |
| "loss": 4.5988, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 48.784854888916016, | |
| "learning_rate": 9.73161567364466e-06, | |
| "loss": 4.4899, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 56.800575256347656, | |
| "learning_rate": 9.677938808373591e-06, | |
| "loss": 4.4391, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 51.58082962036133, | |
| "learning_rate": 9.624261943102525e-06, | |
| "loss": 4.3347, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 50.625308990478516, | |
| "learning_rate": 9.570585077831455e-06, | |
| "loss": 4.3145, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 46.56022262573242, | |
| "learning_rate": 9.516908212560388e-06, | |
| "loss": 4.1742, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 51.267852783203125, | |
| "learning_rate": 9.463231347289318e-06, | |
| "loss": 4.1308, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 52.627098083496094, | |
| "learning_rate": 9.40955448201825e-06, | |
| "loss": 4.1251, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 51.49211120605469, | |
| "learning_rate": 9.355877616747183e-06, | |
| "loss": 4.073, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 50.88691329956055, | |
| "learning_rate": 9.302200751476115e-06, | |
| "loss": 3.9998, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 45.862796783447266, | |
| "learning_rate": 9.248523886205046e-06, | |
| "loss": 3.8666, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 51.27406692504883, | |
| "learning_rate": 9.194847020933978e-06, | |
| "loss": 3.9477, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 47.50687026977539, | |
| "learning_rate": 9.14117015566291e-06, | |
| "loss": 3.7899, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 48.58837127685547, | |
| "learning_rate": 9.087493290391842e-06, | |
| "loss": 3.7295, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 48.3990592956543, | |
| "learning_rate": 9.033816425120775e-06, | |
| "loss": 3.7023, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 46.30027770996094, | |
| "learning_rate": 8.980139559849705e-06, | |
| "loss": 3.6479, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 47.15484619140625, | |
| "learning_rate": 8.926462694578637e-06, | |
| "loss": 3.6161, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 49.22386932373047, | |
| "learning_rate": 8.87278582930757e-06, | |
| "loss": 3.5503, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 45.485557556152344, | |
| "learning_rate": 8.819108964036502e-06, | |
| "loss": 3.4459, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 50.84394454956055, | |
| "learning_rate": 8.765432098765432e-06, | |
| "loss": 3.4918, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 43.17815399169922, | |
| "learning_rate": 8.711755233494365e-06, | |
| "loss": 3.3527, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 41.90092849731445, | |
| "learning_rate": 8.658078368223295e-06, | |
| "loss": 3.347, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 43.23625564575195, | |
| "learning_rate": 8.60440150295223e-06, | |
| "loss": 3.3433, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 44.05680847167969, | |
| "learning_rate": 8.55072463768116e-06, | |
| "loss": 3.2831, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 46.9146842956543, | |
| "learning_rate": 8.497047772410092e-06, | |
| "loss": 3.2932, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 48.542701721191406, | |
| "learning_rate": 8.443370907139024e-06, | |
| "loss": 3.2108, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 39.1197509765625, | |
| "learning_rate": 8.389694041867955e-06, | |
| "loss": 3.1704, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 45.300697326660156, | |
| "learning_rate": 8.336017176596887e-06, | |
| "loss": 3.2239, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 47.439823150634766, | |
| "learning_rate": 8.28234031132582e-06, | |
| "loss": 3.1709, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 42.67869567871094, | |
| "learning_rate": 8.228663446054752e-06, | |
| "loss": 3.0708, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 46.0591926574707, | |
| "learning_rate": 8.174986580783682e-06, | |
| "loss": 3.0959, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 42.061805725097656, | |
| "learning_rate": 8.121309715512614e-06, | |
| "loss": 3.0442, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 44.487159729003906, | |
| "learning_rate": 8.067632850241547e-06, | |
| "loss": 2.8881, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 44.4419059753418, | |
| "learning_rate": 8.013955984970479e-06, | |
| "loss": 2.9994, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 40.00767135620117, | |
| "learning_rate": 7.96027911969941e-06, | |
| "loss": 2.8931, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 44.14014434814453, | |
| "learning_rate": 7.906602254428342e-06, | |
| "loss": 2.9288, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 42.418888092041016, | |
| "learning_rate": 7.852925389157274e-06, | |
| "loss": 2.7776, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 49.98628234863281, | |
| "learning_rate": 7.799248523886206e-06, | |
| "loss": 2.8408, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 39.240352630615234, | |
| "learning_rate": 7.745571658615137e-06, | |
| "loss": 2.7995, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 43.185569763183594, | |
| "learning_rate": 7.691894793344069e-06, | |
| "loss": 2.7486, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 37.41790008544922, | |
| "learning_rate": 7.638217928073001e-06, | |
| "loss": 2.7459, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 38.520973205566406, | |
| "learning_rate": 7.584541062801934e-06, | |
| "loss": 2.6785, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 42.01523971557617, | |
| "learning_rate": 7.530864197530865e-06, | |
| "loss": 2.7745, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 40.55296325683594, | |
| "learning_rate": 7.477187332259796e-06, | |
| "loss": 2.7046, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 38.31270217895508, | |
| "learning_rate": 7.423510466988728e-06, | |
| "loss": 2.6057, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 37.31733322143555, | |
| "learning_rate": 7.369833601717661e-06, | |
| "loss": 2.5229, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 37.55893325805664, | |
| "learning_rate": 7.316156736446592e-06, | |
| "loss": 2.5846, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 40.64212417602539, | |
| "learning_rate": 7.262479871175524e-06, | |
| "loss": 2.4992, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 41.34495544433594, | |
| "learning_rate": 7.208803005904456e-06, | |
| "loss": 2.5575, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 38.53882598876953, | |
| "learning_rate": 7.155126140633387e-06, | |
| "loss": 2.5468, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 39.23023986816406, | |
| "learning_rate": 7.10144927536232e-06, | |
| "loss": 2.5009, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 35.68672180175781, | |
| "learning_rate": 7.047772410091251e-06, | |
| "loss": 2.2929, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 38.0786247253418, | |
| "learning_rate": 6.994095544820183e-06, | |
| "loss": 2.3178, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 42.21394348144531, | |
| "learning_rate": 6.940418679549115e-06, | |
| "loss": 2.4645, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 37.197696685791016, | |
| "learning_rate": 6.886741814278046e-06, | |
| "loss": 2.4797, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 40.341514587402344, | |
| "learning_rate": 6.833064949006979e-06, | |
| "loss": 2.3048, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 36.51962661743164, | |
| "learning_rate": 6.779388083735911e-06, | |
| "loss": 2.4652, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 35.32337951660156, | |
| "learning_rate": 6.725711218464842e-06, | |
| "loss": 2.2979, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 40.75404739379883, | |
| "learning_rate": 6.672034353193773e-06, | |
| "loss": 2.354, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 30.498510360717773, | |
| "learning_rate": 6.6183574879227065e-06, | |
| "loss": 1.8592, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 31.13385772705078, | |
| "learning_rate": 6.564680622651638e-06, | |
| "loss": 1.7776, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 34.9401969909668, | |
| "learning_rate": 6.511003757380569e-06, | |
| "loss": 1.9329, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 32.90480041503906, | |
| "learning_rate": 6.457326892109501e-06, | |
| "loss": 1.8312, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 32.94902420043945, | |
| "learning_rate": 6.403650026838433e-06, | |
| "loss": 1.8556, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 32.02881622314453, | |
| "learning_rate": 6.349973161567365e-06, | |
| "loss": 1.8142, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 32.752323150634766, | |
| "learning_rate": 6.296296296296297e-06, | |
| "loss": 1.7429, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 31.938289642333984, | |
| "learning_rate": 6.242619431025229e-06, | |
| "loss": 1.821, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 32.64255142211914, | |
| "learning_rate": 6.18894256575416e-06, | |
| "loss": 1.7492, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 30.172483444213867, | |
| "learning_rate": 6.135265700483092e-06, | |
| "loss": 1.7661, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 32.1895637512207, | |
| "learning_rate": 6.081588835212025e-06, | |
| "loss": 1.6979, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 32.555870056152344, | |
| "learning_rate": 6.027911969940956e-06, | |
| "loss": 1.778, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 31.702539443969727, | |
| "learning_rate": 5.974235104669888e-06, | |
| "loss": 1.7263, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 32.07310104370117, | |
| "learning_rate": 5.920558239398819e-06, | |
| "loss": 1.5491, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 31.130224227905273, | |
| "learning_rate": 5.866881374127752e-06, | |
| "loss": 1.6249, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 39.838436126708984, | |
| "learning_rate": 5.8132045088566835e-06, | |
| "loss": 1.6721, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 33.75567626953125, | |
| "learning_rate": 5.759527643585615e-06, | |
| "loss": 1.6779, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 31.251935958862305, | |
| "learning_rate": 5.705850778314546e-06, | |
| "loss": 1.6401, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 31.644649505615234, | |
| "learning_rate": 5.652173913043479e-06, | |
| "loss": 1.6647, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 30.14424705505371, | |
| "learning_rate": 5.598497047772411e-06, | |
| "loss": 1.6887, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 29.70695686340332, | |
| "learning_rate": 5.544820182501342e-06, | |
| "loss": 1.5387, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 31.330068588256836, | |
| "learning_rate": 5.4911433172302745e-06, | |
| "loss": 1.6414, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 32.08658981323242, | |
| "learning_rate": 5.437466451959206e-06, | |
| "loss": 1.6158, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 33.42084503173828, | |
| "learning_rate": 5.383789586688137e-06, | |
| "loss": 1.6728, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 32.10792922973633, | |
| "learning_rate": 5.3301127214170704e-06, | |
| "loss": 1.5365, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 34.231239318847656, | |
| "learning_rate": 5.276435856146002e-06, | |
| "loss": 1.584, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 32.19587326049805, | |
| "learning_rate": 5.222758990874933e-06, | |
| "loss": 1.6233, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 30.36279296875, | |
| "learning_rate": 5.169082125603865e-06, | |
| "loss": 1.5246, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 33.34714889526367, | |
| "learning_rate": 5.115405260332798e-06, | |
| "loss": 1.5514, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 32.581424713134766, | |
| "learning_rate": 5.061728395061729e-06, | |
| "loss": 1.495, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 33.158203125, | |
| "learning_rate": 5.0080515297906606e-06, | |
| "loss": 1.5546, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 29.796606063842773, | |
| "learning_rate": 4.954374664519592e-06, | |
| "loss": 1.5141, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 31.936180114746094, | |
| "learning_rate": 4.900697799248524e-06, | |
| "loss": 1.5156, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 30.770095825195312, | |
| "learning_rate": 4.847020933977456e-06, | |
| "loss": 1.5262, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 32.497520446777344, | |
| "learning_rate": 4.793344068706388e-06, | |
| "loss": 1.5303, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 31.067218780517578, | |
| "learning_rate": 4.739667203435319e-06, | |
| "loss": 1.4994, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 27.720073699951172, | |
| "learning_rate": 4.6859903381642516e-06, | |
| "loss": 1.4268, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 30.310941696166992, | |
| "learning_rate": 4.632313472893184e-06, | |
| "loss": 1.4636, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 33.62602996826172, | |
| "learning_rate": 4.578636607622115e-06, | |
| "loss": 1.4783, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 28.9564266204834, | |
| "learning_rate": 4.5249597423510475e-06, | |
| "loss": 1.359, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 29.886262893676758, | |
| "learning_rate": 4.471282877079979e-06, | |
| "loss": 1.4405, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 26.291038513183594, | |
| "learning_rate": 4.417606011808911e-06, | |
| "loss": 1.3914, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 30.628904342651367, | |
| "learning_rate": 4.3639291465378425e-06, | |
| "loss": 1.4335, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 27.96939468383789, | |
| "learning_rate": 4.310252281266775e-06, | |
| "loss": 1.3577, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 29.119224548339844, | |
| "learning_rate": 4.256575415995706e-06, | |
| "loss": 1.3808, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 30.36097526550293, | |
| "learning_rate": 4.202898550724638e-06, | |
| "loss": 1.3545, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 30.843242645263672, | |
| "learning_rate": 4.14922168545357e-06, | |
| "loss": 1.3751, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 29.29217529296875, | |
| "learning_rate": 4.095544820182501e-06, | |
| "loss": 1.3649, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 30.685625076293945, | |
| "learning_rate": 4.0418679549114335e-06, | |
| "loss": 1.4354, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 26.101669311523438, | |
| "learning_rate": 3.988191089640365e-06, | |
| "loss": 1.3355, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 29.12729835510254, | |
| "learning_rate": 3.934514224369297e-06, | |
| "loss": 1.3568, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 27.82271957397461, | |
| "learning_rate": 3.880837359098229e-06, | |
| "loss": 1.3702, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 26.432327270507812, | |
| "learning_rate": 3.827160493827161e-06, | |
| "loss": 1.3231, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 30.632972717285156, | |
| "learning_rate": 3.7734836285560927e-06, | |
| "loss": 1.3283, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 27.142309188842773, | |
| "learning_rate": 3.7198067632850245e-06, | |
| "loss": 1.3159, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 27.63045310974121, | |
| "learning_rate": 3.6661298980139563e-06, | |
| "loss": 1.3777, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 30.256242752075195, | |
| "learning_rate": 3.612453032742888e-06, | |
| "loss": 1.2845, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 28.592174530029297, | |
| "learning_rate": 3.5587761674718204e-06, | |
| "loss": 1.3163, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 29.088247299194336, | |
| "learning_rate": 3.505099302200752e-06, | |
| "loss": 1.3145, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 27.801074981689453, | |
| "learning_rate": 3.4514224369296832e-06, | |
| "loss": 1.3675, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 28.81484603881836, | |
| "learning_rate": 3.3977455716586155e-06, | |
| "loss": 1.2854, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 28.966217041015625, | |
| "learning_rate": 3.3440687063875473e-06, | |
| "loss": 1.3431, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 23.021453857421875, | |
| "learning_rate": 3.290391841116479e-06, | |
| "loss": 1.0516, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 25.622419357299805, | |
| "learning_rate": 3.236714975845411e-06, | |
| "loss": 0.9667, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 26.45795249938965, | |
| "learning_rate": 3.1830381105743428e-06, | |
| "loss": 0.9341, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 26.28618812561035, | |
| "learning_rate": 3.1293612453032746e-06, | |
| "loss": 0.9024, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 24.80799102783203, | |
| "learning_rate": 3.075684380032206e-06, | |
| "loss": 0.8784, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "grad_norm": 25.70199966430664, | |
| "learning_rate": 3.0220075147611383e-06, | |
| "loss": 0.9307, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 24.88735580444336, | |
| "learning_rate": 2.9683306494900697e-06, | |
| "loss": 0.9037, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 26.51141929626465, | |
| "learning_rate": 2.914653784219002e-06, | |
| "loss": 0.9884, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 25.662946701049805, | |
| "learning_rate": 2.8609769189479338e-06, | |
| "loss": 0.8631, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 22.733741760253906, | |
| "learning_rate": 2.8073000536768656e-06, | |
| "loss": 0.9273, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 24.159793853759766, | |
| "learning_rate": 2.7536231884057974e-06, | |
| "loss": 0.9262, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 23.92421531677246, | |
| "learning_rate": 2.699946323134729e-06, | |
| "loss": 0.9048, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 28.564496994018555, | |
| "learning_rate": 2.646269457863661e-06, | |
| "loss": 0.9879, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 25.430883407592773, | |
| "learning_rate": 2.5925925925925925e-06, | |
| "loss": 0.892, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 23.307687759399414, | |
| "learning_rate": 2.5389157273215247e-06, | |
| "loss": 0.894, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 25.83247184753418, | |
| "learning_rate": 2.4852388620504566e-06, | |
| "loss": 0.8817, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 25.72507095336914, | |
| "learning_rate": 2.4315619967793884e-06, | |
| "loss": 0.9155, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 26.67945098876953, | |
| "learning_rate": 2.3778851315083202e-06, | |
| "loss": 0.9325, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 25.82522964477539, | |
| "learning_rate": 2.324208266237252e-06, | |
| "loss": 0.9055, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 22.66315269470215, | |
| "learning_rate": 2.270531400966184e-06, | |
| "loss": 0.8631, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 25.832313537597656, | |
| "learning_rate": 2.2168545356951157e-06, | |
| "loss": 0.9539, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 23.865262985229492, | |
| "learning_rate": 2.163177670424047e-06, | |
| "loss": 0.8461, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 23.32217025756836, | |
| "learning_rate": 2.109500805152979e-06, | |
| "loss": 0.8595, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 24.299062728881836, | |
| "learning_rate": 2.0558239398819112e-06, | |
| "loss": 0.899, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 25.582359313964844, | |
| "learning_rate": 2.002147074610843e-06, | |
| "loss": 0.898, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 25.416006088256836, | |
| "learning_rate": 1.948470209339775e-06, | |
| "loss": 0.898, | |
| "step": 1500 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1863, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 5730542923874304.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |