| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 2.9954430379746837, |
| "eval_steps": 500, |
| "global_step": 1479, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.020253164556962026, |
| "grad_norm": 2.8689478961832267, |
| "learning_rate": 5e-06, |
| "loss": 0.7568, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.04050632911392405, |
| "grad_norm": 3.001868856913536, |
| "learning_rate": 5e-06, |
| "loss": 0.6505, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.060759493670886074, |
| "grad_norm": 1.7446226923969068, |
| "learning_rate": 5e-06, |
| "loss": 0.6289, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.0810126582278481, |
| "grad_norm": 1.6182227627431862, |
| "learning_rate": 5e-06, |
| "loss": 0.6135, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.10126582278481013, |
| "grad_norm": 1.4561815879205204, |
| "learning_rate": 5e-06, |
| "loss": 0.6032, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.12151898734177215, |
| "grad_norm": 1.868314705510585, |
| "learning_rate": 5e-06, |
| "loss": 0.598, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.14177215189873418, |
| "grad_norm": 2.221839850792739, |
| "learning_rate": 5e-06, |
| "loss": 0.5943, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.1620253164556962, |
| "grad_norm": 1.957422608645371, |
| "learning_rate": 5e-06, |
| "loss": 0.5882, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.18227848101265823, |
| "grad_norm": 1.8894540634342296, |
| "learning_rate": 5e-06, |
| "loss": 0.5862, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.20253164556962025, |
| "grad_norm": 1.536518597653581, |
| "learning_rate": 5e-06, |
| "loss": 0.5887, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.22278481012658227, |
| "grad_norm": 1.5918014919176409, |
| "learning_rate": 5e-06, |
| "loss": 0.581, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.2430379746835443, |
| "grad_norm": 1.6229766772085499, |
| "learning_rate": 5e-06, |
| "loss": 0.5791, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.26329113924050634, |
| "grad_norm": 1.8008938637594714, |
| "learning_rate": 5e-06, |
| "loss": 0.5791, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.28354430379746837, |
| "grad_norm": 1.3032513968415904, |
| "learning_rate": 5e-06, |
| "loss": 0.569, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.3037974683544304, |
| "grad_norm": 1.336374473392279, |
| "learning_rate": 5e-06, |
| "loss": 0.5789, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.3240506329113924, |
| "grad_norm": 1.2036867936611222, |
| "learning_rate": 5e-06, |
| "loss": 0.5778, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.34430379746835443, |
| "grad_norm": 1.402710637277177, |
| "learning_rate": 5e-06, |
| "loss": 0.5782, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.36455696202531646, |
| "grad_norm": 1.8260946142647767, |
| "learning_rate": 5e-06, |
| "loss": 0.5737, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.3848101265822785, |
| "grad_norm": 2.1303478863380407, |
| "learning_rate": 5e-06, |
| "loss": 0.5731, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.4050632911392405, |
| "grad_norm": 2.307038000540032, |
| "learning_rate": 5e-06, |
| "loss": 0.567, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.4253164556962025, |
| "grad_norm": 1.894495451172693, |
| "learning_rate": 5e-06, |
| "loss": 0.5621, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.44556962025316454, |
| "grad_norm": 1.6444980357120855, |
| "learning_rate": 5e-06, |
| "loss": 0.5671, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.46582278481012657, |
| "grad_norm": 1.9198629798093232, |
| "learning_rate": 5e-06, |
| "loss": 0.5658, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.4860759493670886, |
| "grad_norm": 2.2873751282619046, |
| "learning_rate": 5e-06, |
| "loss": 0.5619, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.5063291139240507, |
| "grad_norm": 1.4353410284363248, |
| "learning_rate": 5e-06, |
| "loss": 0.5658, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.5265822784810127, |
| "grad_norm": 2.0051814044166414, |
| "learning_rate": 5e-06, |
| "loss": 0.5519, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.5468354430379747, |
| "grad_norm": 1.4757564591331935, |
| "learning_rate": 5e-06, |
| "loss": 0.5617, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.5670886075949367, |
| "grad_norm": 1.6282246239795468, |
| "learning_rate": 5e-06, |
| "loss": 0.5574, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.5873417721518988, |
| "grad_norm": 1.2305572479943359, |
| "learning_rate": 5e-06, |
| "loss": 0.5585, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.6075949367088608, |
| "grad_norm": 1.2706982136132412, |
| "learning_rate": 5e-06, |
| "loss": 0.5585, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.6278481012658228, |
| "grad_norm": 1.2232981728516181, |
| "learning_rate": 5e-06, |
| "loss": 0.5643, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.6481012658227848, |
| "grad_norm": 1.1516553737442161, |
| "learning_rate": 5e-06, |
| "loss": 0.5608, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.6683544303797468, |
| "grad_norm": 1.349232938651194, |
| "learning_rate": 5e-06, |
| "loss": 0.5605, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.6886075949367089, |
| "grad_norm": 1.4739907121269211, |
| "learning_rate": 5e-06, |
| "loss": 0.5599, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.7088607594936709, |
| "grad_norm": 1.2169660233676007, |
| "learning_rate": 5e-06, |
| "loss": 0.5609, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.7291139240506329, |
| "grad_norm": 1.2065677522582892, |
| "learning_rate": 5e-06, |
| "loss": 0.5628, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.7493670886075949, |
| "grad_norm": 1.3074719959050782, |
| "learning_rate": 5e-06, |
| "loss": 0.5553, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.769620253164557, |
| "grad_norm": 1.3044439725094406, |
| "learning_rate": 5e-06, |
| "loss": 0.5519, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.789873417721519, |
| "grad_norm": 1.5949500566625288, |
| "learning_rate": 5e-06, |
| "loss": 0.5547, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.810126582278481, |
| "grad_norm": 1.112807791877861, |
| "learning_rate": 5e-06, |
| "loss": 0.5549, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.830379746835443, |
| "grad_norm": 1.212766430255019, |
| "learning_rate": 5e-06, |
| "loss": 0.5497, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.850632911392405, |
| "grad_norm": 1.1632232875371673, |
| "learning_rate": 5e-06, |
| "loss": 0.5578, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.8708860759493671, |
| "grad_norm": 1.1326313099880627, |
| "learning_rate": 5e-06, |
| "loss": 0.551, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.8911392405063291, |
| "grad_norm": 1.3963870137646481, |
| "learning_rate": 5e-06, |
| "loss": 0.5565, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.9113924050632911, |
| "grad_norm": 1.1003111250626216, |
| "learning_rate": 5e-06, |
| "loss": 0.5475, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.9316455696202531, |
| "grad_norm": 1.068292273731364, |
| "learning_rate": 5e-06, |
| "loss": 0.5517, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.9518987341772152, |
| "grad_norm": 1.137113934432503, |
| "learning_rate": 5e-06, |
| "loss": 0.5527, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.9721518987341772, |
| "grad_norm": 1.614059874947369, |
| "learning_rate": 5e-06, |
| "loss": 0.5498, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.9924050632911392, |
| "grad_norm": 1.6034425576801679, |
| "learning_rate": 5e-06, |
| "loss": 0.5465, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.9984810126582279, |
| "eval_loss": 0.06862536817789078, |
| "eval_runtime": 508.7992, |
| "eval_samples_per_second": 26.148, |
| "eval_steps_per_second": 0.409, |
| "step": 493 |
| }, |
| { |
| "epoch": 1.0126582278481013, |
| "grad_norm": 1.7336127404292327, |
| "learning_rate": 5e-06, |
| "loss": 0.5067, |
| "step": 500 |
| }, |
| { |
| "epoch": 1.0329113924050632, |
| "grad_norm": 1.4762279605615538, |
| "learning_rate": 5e-06, |
| "loss": 0.4726, |
| "step": 510 |
| }, |
| { |
| "epoch": 1.0531645569620254, |
| "grad_norm": 1.8368447972911353, |
| "learning_rate": 5e-06, |
| "loss": 0.4689, |
| "step": 520 |
| }, |
| { |
| "epoch": 1.0734177215189873, |
| "grad_norm": 2.013270791256026, |
| "learning_rate": 5e-06, |
| "loss": 0.472, |
| "step": 530 |
| }, |
| { |
| "epoch": 1.0936708860759494, |
| "grad_norm": 1.4632352066803245, |
| "learning_rate": 5e-06, |
| "loss": 0.4688, |
| "step": 540 |
| }, |
| { |
| "epoch": 1.1139240506329113, |
| "grad_norm": 1.669648908257983, |
| "learning_rate": 5e-06, |
| "loss": 0.4717, |
| "step": 550 |
| }, |
| { |
| "epoch": 1.1341772151898735, |
| "grad_norm": 1.4524541423429347, |
| "learning_rate": 5e-06, |
| "loss": 0.4689, |
| "step": 560 |
| }, |
| { |
| "epoch": 1.1544303797468354, |
| "grad_norm": 1.8669027245651724, |
| "learning_rate": 5e-06, |
| "loss": 0.4749, |
| "step": 570 |
| }, |
| { |
| "epoch": 1.1746835443037975, |
| "grad_norm": 1.7850707694175234, |
| "learning_rate": 5e-06, |
| "loss": 0.4732, |
| "step": 580 |
| }, |
| { |
| "epoch": 1.1949367088607594, |
| "grad_norm": 1.883521820325262, |
| "learning_rate": 5e-06, |
| "loss": 0.4672, |
| "step": 590 |
| }, |
| { |
| "epoch": 1.2151898734177216, |
| "grad_norm": 1.4258811553718265, |
| "learning_rate": 5e-06, |
| "loss": 0.4745, |
| "step": 600 |
| }, |
| { |
| "epoch": 1.2354430379746835, |
| "grad_norm": 1.644969173589512, |
| "learning_rate": 5e-06, |
| "loss": 0.471, |
| "step": 610 |
| }, |
| { |
| "epoch": 1.2556962025316456, |
| "grad_norm": 1.426328376259139, |
| "learning_rate": 5e-06, |
| "loss": 0.4763, |
| "step": 620 |
| }, |
| { |
| "epoch": 1.2759493670886077, |
| "grad_norm": 1.3727142665042786, |
| "learning_rate": 5e-06, |
| "loss": 0.4758, |
| "step": 630 |
| }, |
| { |
| "epoch": 1.2962025316455696, |
| "grad_norm": 1.354317940438644, |
| "learning_rate": 5e-06, |
| "loss": 0.4752, |
| "step": 640 |
| }, |
| { |
| "epoch": 1.3164556962025316, |
| "grad_norm": 1.4162232950056415, |
| "learning_rate": 5e-06, |
| "loss": 0.4797, |
| "step": 650 |
| }, |
| { |
| "epoch": 1.3367088607594937, |
| "grad_norm": 1.3995914361407802, |
| "learning_rate": 5e-06, |
| "loss": 0.4681, |
| "step": 660 |
| }, |
| { |
| "epoch": 1.3569620253164558, |
| "grad_norm": 1.3749681142826822, |
| "learning_rate": 5e-06, |
| "loss": 0.4816, |
| "step": 670 |
| }, |
| { |
| "epoch": 1.3772151898734177, |
| "grad_norm": 1.2628569327422936, |
| "learning_rate": 5e-06, |
| "loss": 0.4785, |
| "step": 680 |
| }, |
| { |
| "epoch": 1.3974683544303796, |
| "grad_norm": 1.1219572126848418, |
| "learning_rate": 5e-06, |
| "loss": 0.4777, |
| "step": 690 |
| }, |
| { |
| "epoch": 1.4177215189873418, |
| "grad_norm": 1.1286403040641895, |
| "learning_rate": 5e-06, |
| "loss": 0.4757, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.437974683544304, |
| "grad_norm": 1.2151331373022498, |
| "learning_rate": 5e-06, |
| "loss": 0.4809, |
| "step": 710 |
| }, |
| { |
| "epoch": 1.4582278481012658, |
| "grad_norm": 1.3695718863731767, |
| "learning_rate": 5e-06, |
| "loss": 0.4809, |
| "step": 720 |
| }, |
| { |
| "epoch": 1.4784810126582277, |
| "grad_norm": 1.311112240498471, |
| "learning_rate": 5e-06, |
| "loss": 0.4832, |
| "step": 730 |
| }, |
| { |
| "epoch": 1.4987341772151899, |
| "grad_norm": 1.2465757257045427, |
| "learning_rate": 5e-06, |
| "loss": 0.4759, |
| "step": 740 |
| }, |
| { |
| "epoch": 1.518987341772152, |
| "grad_norm": 1.194065701898256, |
| "learning_rate": 5e-06, |
| "loss": 0.4784, |
| "step": 750 |
| }, |
| { |
| "epoch": 1.539240506329114, |
| "grad_norm": 1.23627515464373, |
| "learning_rate": 5e-06, |
| "loss": 0.4795, |
| "step": 760 |
| }, |
| { |
| "epoch": 1.5594936708860758, |
| "grad_norm": 1.2846596778216188, |
| "learning_rate": 5e-06, |
| "loss": 0.4785, |
| "step": 770 |
| }, |
| { |
| "epoch": 1.579746835443038, |
| "grad_norm": 1.139161411477993, |
| "learning_rate": 5e-06, |
| "loss": 0.4789, |
| "step": 780 |
| }, |
| { |
| "epoch": 1.6, |
| "grad_norm": 1.3387138721449912, |
| "learning_rate": 5e-06, |
| "loss": 0.4811, |
| "step": 790 |
| }, |
| { |
| "epoch": 1.620253164556962, |
| "grad_norm": 1.1004200883858706, |
| "learning_rate": 5e-06, |
| "loss": 0.4771, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.640506329113924, |
| "grad_norm": 1.1473680211097377, |
| "learning_rate": 5e-06, |
| "loss": 0.4697, |
| "step": 810 |
| }, |
| { |
| "epoch": 1.660759493670886, |
| "grad_norm": 1.2889331947392633, |
| "learning_rate": 5e-06, |
| "loss": 0.4795, |
| "step": 820 |
| }, |
| { |
| "epoch": 1.6810126582278482, |
| "grad_norm": 1.2344589623177893, |
| "learning_rate": 5e-06, |
| "loss": 0.4747, |
| "step": 830 |
| }, |
| { |
| "epoch": 1.70126582278481, |
| "grad_norm": 1.1678583016837603, |
| "learning_rate": 5e-06, |
| "loss": 0.4765, |
| "step": 840 |
| }, |
| { |
| "epoch": 1.721518987341772, |
| "grad_norm": 1.2673691237610725, |
| "learning_rate": 5e-06, |
| "loss": 0.4787, |
| "step": 850 |
| }, |
| { |
| "epoch": 1.7417721518987341, |
| "grad_norm": 1.3406707202706432, |
| "learning_rate": 5e-06, |
| "loss": 0.4784, |
| "step": 860 |
| }, |
| { |
| "epoch": 1.7620253164556963, |
| "grad_norm": 1.1790031287381286, |
| "learning_rate": 5e-06, |
| "loss": 0.4816, |
| "step": 870 |
| }, |
| { |
| "epoch": 1.7822784810126582, |
| "grad_norm": 1.2136035177998101, |
| "learning_rate": 5e-06, |
| "loss": 0.4824, |
| "step": 880 |
| }, |
| { |
| "epoch": 1.80253164556962, |
| "grad_norm": 1.1400186893660278, |
| "learning_rate": 5e-06, |
| "loss": 0.4786, |
| "step": 890 |
| }, |
| { |
| "epoch": 1.8227848101265822, |
| "grad_norm": 1.241551295379632, |
| "learning_rate": 5e-06, |
| "loss": 0.4782, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.8430379746835444, |
| "grad_norm": 1.164147007484972, |
| "learning_rate": 5e-06, |
| "loss": 0.4805, |
| "step": 910 |
| }, |
| { |
| "epoch": 1.8632911392405065, |
| "grad_norm": 1.2304573608166283, |
| "learning_rate": 5e-06, |
| "loss": 0.4795, |
| "step": 920 |
| }, |
| { |
| "epoch": 1.8835443037974684, |
| "grad_norm": 1.085909714785085, |
| "learning_rate": 5e-06, |
| "loss": 0.4763, |
| "step": 930 |
| }, |
| { |
| "epoch": 1.9037974683544303, |
| "grad_norm": 1.2049197342008424, |
| "learning_rate": 5e-06, |
| "loss": 0.4801, |
| "step": 940 |
| }, |
| { |
| "epoch": 1.9240506329113924, |
| "grad_norm": 1.4182445064455464, |
| "learning_rate": 5e-06, |
| "loss": 0.4807, |
| "step": 950 |
| }, |
| { |
| "epoch": 1.9443037974683546, |
| "grad_norm": 1.3989582624297905, |
| "learning_rate": 5e-06, |
| "loss": 0.4846, |
| "step": 960 |
| }, |
| { |
| "epoch": 1.9645569620253165, |
| "grad_norm": 1.2747713735640067, |
| "learning_rate": 5e-06, |
| "loss": 0.4827, |
| "step": 970 |
| }, |
| { |
| "epoch": 1.9848101265822784, |
| "grad_norm": 1.1281460860893593, |
| "learning_rate": 5e-06, |
| "loss": 0.4833, |
| "step": 980 |
| }, |
| { |
| "epoch": 1.998987341772152, |
| "eval_loss": 0.06866126507520676, |
| "eval_runtime": 508.4487, |
| "eval_samples_per_second": 26.166, |
| "eval_steps_per_second": 0.409, |
| "step": 987 |
| }, |
| { |
| "epoch": 2.0050632911392405, |
| "grad_norm": 2.038487562355176, |
| "learning_rate": 5e-06, |
| "loss": 0.4557, |
| "step": 990 |
| }, |
| { |
| "epoch": 2.0253164556962027, |
| "grad_norm": 1.7768517961980532, |
| "learning_rate": 5e-06, |
| "loss": 0.3926, |
| "step": 1000 |
| }, |
| { |
| "epoch": 2.0455696202531644, |
| "grad_norm": 1.533703900767279, |
| "learning_rate": 5e-06, |
| "loss": 0.3891, |
| "step": 1010 |
| }, |
| { |
| "epoch": 2.0658227848101265, |
| "grad_norm": 1.4568677312204936, |
| "learning_rate": 5e-06, |
| "loss": 0.3852, |
| "step": 1020 |
| }, |
| { |
| "epoch": 2.0860759493670886, |
| "grad_norm": 1.4076328993412954, |
| "learning_rate": 5e-06, |
| "loss": 0.3857, |
| "step": 1030 |
| }, |
| { |
| "epoch": 2.1063291139240508, |
| "grad_norm": 1.2865521789174106, |
| "learning_rate": 5e-06, |
| "loss": 0.3835, |
| "step": 1040 |
| }, |
| { |
| "epoch": 2.1265822784810124, |
| "grad_norm": 1.3441219402172357, |
| "learning_rate": 5e-06, |
| "loss": 0.3936, |
| "step": 1050 |
| }, |
| { |
| "epoch": 2.1468354430379746, |
| "grad_norm": 1.4310276803158983, |
| "learning_rate": 5e-06, |
| "loss": 0.3881, |
| "step": 1060 |
| }, |
| { |
| "epoch": 2.1670886075949367, |
| "grad_norm": 1.3973941947351598, |
| "learning_rate": 5e-06, |
| "loss": 0.3919, |
| "step": 1070 |
| }, |
| { |
| "epoch": 2.187341772151899, |
| "grad_norm": 1.3437412522316032, |
| "learning_rate": 5e-06, |
| "loss": 0.3896, |
| "step": 1080 |
| }, |
| { |
| "epoch": 2.207594936708861, |
| "grad_norm": 1.460671268065879, |
| "learning_rate": 5e-06, |
| "loss": 0.3877, |
| "step": 1090 |
| }, |
| { |
| "epoch": 2.2278481012658227, |
| "grad_norm": 1.423951076769063, |
| "learning_rate": 5e-06, |
| "loss": 0.3969, |
| "step": 1100 |
| }, |
| { |
| "epoch": 2.248101265822785, |
| "grad_norm": 1.2900417026263609, |
| "learning_rate": 5e-06, |
| "loss": 0.3928, |
| "step": 1110 |
| }, |
| { |
| "epoch": 2.268354430379747, |
| "grad_norm": 1.45820727705065, |
| "learning_rate": 5e-06, |
| "loss": 0.3956, |
| "step": 1120 |
| }, |
| { |
| "epoch": 2.2886075949367086, |
| "grad_norm": 1.3731405883122467, |
| "learning_rate": 5e-06, |
| "loss": 0.3963, |
| "step": 1130 |
| }, |
| { |
| "epoch": 2.3088607594936708, |
| "grad_norm": 1.3707160494580402, |
| "learning_rate": 5e-06, |
| "loss": 0.3922, |
| "step": 1140 |
| }, |
| { |
| "epoch": 2.329113924050633, |
| "grad_norm": 1.5877019868138311, |
| "learning_rate": 5e-06, |
| "loss": 0.3987, |
| "step": 1150 |
| }, |
| { |
| "epoch": 2.349367088607595, |
| "grad_norm": 1.8180386160312771, |
| "learning_rate": 5e-06, |
| "loss": 0.3989, |
| "step": 1160 |
| }, |
| { |
| "epoch": 2.369620253164557, |
| "grad_norm": 1.5896397183121131, |
| "learning_rate": 5e-06, |
| "loss": 0.3992, |
| "step": 1170 |
| }, |
| { |
| "epoch": 2.389873417721519, |
| "grad_norm": 1.513891217454247, |
| "learning_rate": 5e-06, |
| "loss": 0.3989, |
| "step": 1180 |
| }, |
| { |
| "epoch": 2.410126582278481, |
| "grad_norm": 1.306055682802281, |
| "learning_rate": 5e-06, |
| "loss": 0.3979, |
| "step": 1190 |
| }, |
| { |
| "epoch": 2.430379746835443, |
| "grad_norm": 1.413731403999378, |
| "learning_rate": 5e-06, |
| "loss": 0.3969, |
| "step": 1200 |
| }, |
| { |
| "epoch": 2.4506329113924052, |
| "grad_norm": 1.301408947711459, |
| "learning_rate": 5e-06, |
| "loss": 0.3964, |
| "step": 1210 |
| }, |
| { |
| "epoch": 2.470886075949367, |
| "grad_norm": 1.2647095937599742, |
| "learning_rate": 5e-06, |
| "loss": 0.4, |
| "step": 1220 |
| }, |
| { |
| "epoch": 2.491139240506329, |
| "grad_norm": 1.3506237496972513, |
| "learning_rate": 5e-06, |
| "loss": 0.4001, |
| "step": 1230 |
| }, |
| { |
| "epoch": 2.511392405063291, |
| "grad_norm": 1.454117052700542, |
| "learning_rate": 5e-06, |
| "loss": 0.4039, |
| "step": 1240 |
| }, |
| { |
| "epoch": 2.5316455696202533, |
| "grad_norm": 1.4375989405158807, |
| "learning_rate": 5e-06, |
| "loss": 0.3977, |
| "step": 1250 |
| }, |
| { |
| "epoch": 2.5518987341772155, |
| "grad_norm": 1.649455092325561, |
| "learning_rate": 5e-06, |
| "loss": 0.3936, |
| "step": 1260 |
| }, |
| { |
| "epoch": 2.572151898734177, |
| "grad_norm": 1.5707031422488011, |
| "learning_rate": 5e-06, |
| "loss": 0.4, |
| "step": 1270 |
| }, |
| { |
| "epoch": 2.5924050632911393, |
| "grad_norm": 1.3358180734037475, |
| "learning_rate": 5e-06, |
| "loss": 0.4013, |
| "step": 1280 |
| }, |
| { |
| "epoch": 2.6126582278481014, |
| "grad_norm": 1.2969805529597713, |
| "learning_rate": 5e-06, |
| "loss": 0.4064, |
| "step": 1290 |
| }, |
| { |
| "epoch": 2.632911392405063, |
| "grad_norm": 1.4958463505521509, |
| "learning_rate": 5e-06, |
| "loss": 0.4009, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.6531645569620252, |
| "grad_norm": 1.4049951287522473, |
| "learning_rate": 5e-06, |
| "loss": 0.4008, |
| "step": 1310 |
| }, |
| { |
| "epoch": 2.6734177215189874, |
| "grad_norm": 1.423575070005059, |
| "learning_rate": 5e-06, |
| "loss": 0.4038, |
| "step": 1320 |
| }, |
| { |
| "epoch": 2.6936708860759495, |
| "grad_norm": 1.3543397934898849, |
| "learning_rate": 5e-06, |
| "loss": 0.4058, |
| "step": 1330 |
| }, |
| { |
| "epoch": 2.7139240506329116, |
| "grad_norm": 1.412173031525196, |
| "learning_rate": 5e-06, |
| "loss": 0.4041, |
| "step": 1340 |
| }, |
| { |
| "epoch": 2.7341772151898733, |
| "grad_norm": 1.2640152971127463, |
| "learning_rate": 5e-06, |
| "loss": 0.4035, |
| "step": 1350 |
| }, |
| { |
| "epoch": 2.7544303797468355, |
| "grad_norm": 1.355988535378844, |
| "learning_rate": 5e-06, |
| "loss": 0.3998, |
| "step": 1360 |
| }, |
| { |
| "epoch": 2.7746835443037976, |
| "grad_norm": 1.4297573153619936, |
| "learning_rate": 5e-06, |
| "loss": 0.4053, |
| "step": 1370 |
| }, |
| { |
| "epoch": 2.7949367088607593, |
| "grad_norm": 1.356862002651741, |
| "learning_rate": 5e-06, |
| "loss": 0.4035, |
| "step": 1380 |
| }, |
| { |
| "epoch": 2.8151898734177214, |
| "grad_norm": 1.4946852230976382, |
| "learning_rate": 5e-06, |
| "loss": 0.4061, |
| "step": 1390 |
| }, |
| { |
| "epoch": 2.8354430379746836, |
| "grad_norm": 1.305693925680122, |
| "learning_rate": 5e-06, |
| "loss": 0.3996, |
| "step": 1400 |
| }, |
| { |
| "epoch": 2.8556962025316457, |
| "grad_norm": 1.2829753261584584, |
| "learning_rate": 5e-06, |
| "loss": 0.4079, |
| "step": 1410 |
| }, |
| { |
| "epoch": 2.875949367088608, |
| "grad_norm": 1.2500077779574505, |
| "learning_rate": 5e-06, |
| "loss": 0.4039, |
| "step": 1420 |
| }, |
| { |
| "epoch": 2.8962025316455695, |
| "grad_norm": 1.31399827384334, |
| "learning_rate": 5e-06, |
| "loss": 0.4047, |
| "step": 1430 |
| }, |
| { |
| "epoch": 2.9164556962025316, |
| "grad_norm": 1.261569107275823, |
| "learning_rate": 5e-06, |
| "loss": 0.4069, |
| "step": 1440 |
| }, |
| { |
| "epoch": 2.9367088607594938, |
| "grad_norm": 1.4114779852554502, |
| "learning_rate": 5e-06, |
| "loss": 0.4158, |
| "step": 1450 |
| }, |
| { |
| "epoch": 2.9569620253164555, |
| "grad_norm": 1.4216851039571943, |
| "learning_rate": 5e-06, |
| "loss": 0.4035, |
| "step": 1460 |
| }, |
| { |
| "epoch": 2.9772151898734176, |
| "grad_norm": 1.3036666558612016, |
| "learning_rate": 5e-06, |
| "loss": 0.4076, |
| "step": 1470 |
| }, |
| { |
| "epoch": 2.9954430379746837, |
| "eval_loss": 0.07229307293891907, |
| "eval_runtime": 506.6356, |
| "eval_samples_per_second": 26.26, |
| "eval_steps_per_second": 0.411, |
| "step": 1479 |
| }, |
| { |
| "epoch": 2.9954430379746837, |
| "step": 1479, |
| "total_flos": 2477170706350080.0, |
| "train_loss": 0.4832688944818524, |
| "train_runtime": 84767.9876, |
| "train_samples_per_second": 8.945, |
| "train_steps_per_second": 0.017 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 1479, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 2477170706350080.0, |
| "train_batch_size": 8, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
