| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.2738853503184713, | |
| "eval_steps": 10000000, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.012738853503184714, | |
| "grad_norm": 14.034008421036804, | |
| "learning_rate": 6.369426751592357e-09, | |
| "loss": 1.9212, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.025477707006369428, | |
| "grad_norm": 14.324525585475893, | |
| "learning_rate": 1.2738853503184714e-08, | |
| "loss": 1.9573, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03821656050955414, | |
| "grad_norm": 13.359417294747553, | |
| "learning_rate": 1.910828025477707e-08, | |
| "loss": 1.9572, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.050955414012738856, | |
| "grad_norm": 14.330701922979292, | |
| "learning_rate": 2.5477707006369427e-08, | |
| "loss": 1.9322, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06369426751592357, | |
| "grad_norm": 14.315516059019966, | |
| "learning_rate": 3.184713375796178e-08, | |
| "loss": 1.924, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07643312101910828, | |
| "grad_norm": 13.502552068344668, | |
| "learning_rate": 3.821656050955414e-08, | |
| "loss": 1.9413, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.08917197452229299, | |
| "grad_norm": 13.518442083664025, | |
| "learning_rate": 4.458598726114649e-08, | |
| "loss": 1.9446, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.10191082802547771, | |
| "grad_norm": 12.383858046483768, | |
| "learning_rate": 5.0955414012738854e-08, | |
| "loss": 1.9417, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.11464968152866242, | |
| "grad_norm": 11.723789916894649, | |
| "learning_rate": 5.732484076433121e-08, | |
| "loss": 1.8923, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.12738853503184713, | |
| "grad_norm": 11.079881895874683, | |
| "learning_rate": 6.369426751592356e-08, | |
| "loss": 1.8899, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.14012738853503184, | |
| "grad_norm": 11.331614677061625, | |
| "learning_rate": 7.006369426751591e-08, | |
| "loss": 1.8947, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15286624203821655, | |
| "grad_norm": 8.897895249557337, | |
| "learning_rate": 7.643312101910828e-08, | |
| "loss": 1.8655, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.16560509554140126, | |
| "grad_norm": 7.851143763222098, | |
| "learning_rate": 8.280254777070063e-08, | |
| "loss": 1.8388, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.17834394904458598, | |
| "grad_norm": 7.651003274597473, | |
| "learning_rate": 8.917197452229298e-08, | |
| "loss": 1.8185, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1910828025477707, | |
| "grad_norm": 6.852701313431956, | |
| "learning_rate": 9.554140127388536e-08, | |
| "loss": 1.7775, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.20382165605095542, | |
| "grad_norm": 4.332528021529095, | |
| "learning_rate": 1.0191082802547771e-07, | |
| "loss": 1.7623, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.21656050955414013, | |
| "grad_norm": 3.5221886562887184, | |
| "learning_rate": 1.0828025477707006e-07, | |
| "loss": 1.6837, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.22929936305732485, | |
| "grad_norm": 3.149818516156245, | |
| "learning_rate": 1.1464968152866242e-07, | |
| "loss": 1.6979, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.24203821656050956, | |
| "grad_norm": 2.9107656285941355, | |
| "learning_rate": 1.2101910828025477e-07, | |
| "loss": 1.6961, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.25477707006369427, | |
| "grad_norm": 2.795449899540731, | |
| "learning_rate": 1.2738853503184713e-07, | |
| "loss": 1.6375, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.267515923566879, | |
| "grad_norm": 2.28687122905737, | |
| "learning_rate": 1.3375796178343948e-07, | |
| "loss": 1.6353, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2802547770700637, | |
| "grad_norm": 2.1744397918507112, | |
| "learning_rate": 1.4012738853503183e-07, | |
| "loss": 1.6551, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2929936305732484, | |
| "grad_norm": 1.9913861609957166, | |
| "learning_rate": 1.464968152866242e-07, | |
| "loss": 1.6084, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3057324840764331, | |
| "grad_norm": 3.6644916122308606, | |
| "learning_rate": 1.5286624203821656e-07, | |
| "loss": 1.6159, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3184713375796178, | |
| "grad_norm": 1.9386884261467865, | |
| "learning_rate": 1.592356687898089e-07, | |
| "loss": 1.5966, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.33121019108280253, | |
| "grad_norm": 1.801727064360074, | |
| "learning_rate": 1.6560509554140126e-07, | |
| "loss": 1.5382, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.34394904458598724, | |
| "grad_norm": 1.791207448353307, | |
| "learning_rate": 1.719745222929936e-07, | |
| "loss": 1.5677, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.35668789808917195, | |
| "grad_norm": 1.785800105876019, | |
| "learning_rate": 1.7834394904458596e-07, | |
| "loss": 1.5679, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.36942675159235666, | |
| "grad_norm": 1.7105114239580628, | |
| "learning_rate": 1.847133757961783e-07, | |
| "loss": 1.5676, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3821656050955414, | |
| "grad_norm": 1.635688328958137, | |
| "learning_rate": 1.9108280254777072e-07, | |
| "loss": 1.5159, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.39490445859872614, | |
| "grad_norm": 1.672905022352275, | |
| "learning_rate": 1.9745222929936307e-07, | |
| "loss": 1.5458, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.40764331210191085, | |
| "grad_norm": 1.667559991729344, | |
| "learning_rate": 2.0382165605095542e-07, | |
| "loss": 1.5553, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.42038216560509556, | |
| "grad_norm": 1.8961844126982597, | |
| "learning_rate": 2.1019108280254777e-07, | |
| "loss": 1.5726, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.43312101910828027, | |
| "grad_norm": 1.704794668355483, | |
| "learning_rate": 2.1656050955414012e-07, | |
| "loss": 1.5338, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.445859872611465, | |
| "grad_norm": 1.7655206338682627, | |
| "learning_rate": 2.2292993630573247e-07, | |
| "loss": 1.5407, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.4585987261146497, | |
| "grad_norm": 1.6563900983762812, | |
| "learning_rate": 2.2929936305732485e-07, | |
| "loss": 1.5357, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.4713375796178344, | |
| "grad_norm": 1.8360088416494706, | |
| "learning_rate": 2.356687898089172e-07, | |
| "loss": 1.5579, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.4840764331210191, | |
| "grad_norm": 1.6539352097565592, | |
| "learning_rate": 2.4203821656050955e-07, | |
| "loss": 1.5138, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.4968152866242038, | |
| "grad_norm": 1.6386927867862713, | |
| "learning_rate": 2.484076433121019e-07, | |
| "loss": 1.5311, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5095541401273885, | |
| "grad_norm": 1.6316402031128345, | |
| "learning_rate": 2.5477707006369425e-07, | |
| "loss": 1.4975, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5222929936305732, | |
| "grad_norm": 1.6279765149394207, | |
| "learning_rate": 2.611464968152866e-07, | |
| "loss": 1.5508, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.535031847133758, | |
| "grad_norm": 1.7198212849602736, | |
| "learning_rate": 2.6751592356687895e-07, | |
| "loss": 1.5293, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5477707006369427, | |
| "grad_norm": 1.6536550947528905, | |
| "learning_rate": 2.738853503184713e-07, | |
| "loss": 1.5422, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.5605095541401274, | |
| "grad_norm": 1.7104890897146838, | |
| "learning_rate": 2.8025477707006366e-07, | |
| "loss": 1.505, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.5732484076433121, | |
| "grad_norm": 1.7028427858115733, | |
| "learning_rate": 2.86624203821656e-07, | |
| "loss": 1.5273, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.5859872611464968, | |
| "grad_norm": 1.606699689644091, | |
| "learning_rate": 2.929936305732484e-07, | |
| "loss": 1.5315, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.5987261146496815, | |
| "grad_norm": 1.7074522582102618, | |
| "learning_rate": 2.9936305732484076e-07, | |
| "loss": 1.5321, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.6114649681528662, | |
| "grad_norm": 2.1566787662429756, | |
| "learning_rate": 3.057324840764331e-07, | |
| "loss": 1.5356, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.6242038216560509, | |
| "grad_norm": 1.6361324291726362, | |
| "learning_rate": 3.1210191082802546e-07, | |
| "loss": 1.5333, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.6369426751592356, | |
| "grad_norm": 1.5544549161506893, | |
| "learning_rate": 3.184713375796178e-07, | |
| "loss": 1.4812, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6496815286624203, | |
| "grad_norm": 1.8652030639545352, | |
| "learning_rate": 3.2484076433121017e-07, | |
| "loss": 1.5465, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.6624203821656051, | |
| "grad_norm": 1.713752710097288, | |
| "learning_rate": 3.312101910828025e-07, | |
| "loss": 1.5054, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.6751592356687898, | |
| "grad_norm": 1.6986436389098123, | |
| "learning_rate": 3.3757961783439487e-07, | |
| "loss": 1.5356, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.6878980891719745, | |
| "grad_norm": 1.6494260445407405, | |
| "learning_rate": 3.439490445859872e-07, | |
| "loss": 1.4968, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.7006369426751592, | |
| "grad_norm": 1.6145130069998217, | |
| "learning_rate": 3.5031847133757957e-07, | |
| "loss": 1.5371, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.7133757961783439, | |
| "grad_norm": 1.5117351196478945, | |
| "learning_rate": 3.566878980891719e-07, | |
| "loss": 1.5064, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.7261146496815286, | |
| "grad_norm": 1.5389329444189115, | |
| "learning_rate": 3.6305732484076427e-07, | |
| "loss": 1.4885, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.7388535031847133, | |
| "grad_norm": 1.7080520598619022, | |
| "learning_rate": 3.694267515923566e-07, | |
| "loss": 1.491, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.7515923566878981, | |
| "grad_norm": 1.4506261889056746, | |
| "learning_rate": 3.757961783439491e-07, | |
| "loss": 1.4792, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.7643312101910829, | |
| "grad_norm": 1.5912892418189286, | |
| "learning_rate": 3.8216560509554143e-07, | |
| "loss": 1.5181, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.7770700636942676, | |
| "grad_norm": 1.7347562868273487, | |
| "learning_rate": 3.885350318471338e-07, | |
| "loss": 1.5001, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.7898089171974523, | |
| "grad_norm": 1.7247795867566735, | |
| "learning_rate": 3.9490445859872613e-07, | |
| "loss": 1.5643, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.802547770700637, | |
| "grad_norm": 1.8030660386471178, | |
| "learning_rate": 4.012738853503185e-07, | |
| "loss": 1.4885, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.8152866242038217, | |
| "grad_norm": 1.847905567407947, | |
| "learning_rate": 4.0764331210191083e-07, | |
| "loss": 1.5169, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.8280254777070064, | |
| "grad_norm": 1.7953671081333586, | |
| "learning_rate": 4.140127388535032e-07, | |
| "loss": 1.5251, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.8407643312101911, | |
| "grad_norm": 1.588856325601936, | |
| "learning_rate": 4.2038216560509554e-07, | |
| "loss": 1.4935, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.8535031847133758, | |
| "grad_norm": 1.6474093001494565, | |
| "learning_rate": 4.267515923566879e-07, | |
| "loss": 1.5197, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.8662420382165605, | |
| "grad_norm": 1.8040366493948004, | |
| "learning_rate": 4.3312101910828024e-07, | |
| "loss": 1.532, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.8789808917197452, | |
| "grad_norm": 1.649448453977979, | |
| "learning_rate": 4.394904458598726e-07, | |
| "loss": 1.5326, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.89171974522293, | |
| "grad_norm": 1.7529453909608548, | |
| "learning_rate": 4.4585987261146494e-07, | |
| "loss": 1.5016, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.9044585987261147, | |
| "grad_norm": 1.6610601648124064, | |
| "learning_rate": 4.522292993630573e-07, | |
| "loss": 1.5081, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.9171974522292994, | |
| "grad_norm": 1.4933572459758815, | |
| "learning_rate": 4.585987261146497e-07, | |
| "loss": 1.508, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.9299363057324841, | |
| "grad_norm": 1.5974193238698267, | |
| "learning_rate": 4.6496815286624205e-07, | |
| "loss": 1.5202, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.9426751592356688, | |
| "grad_norm": 1.619176263750235, | |
| "learning_rate": 4.713375796178344e-07, | |
| "loss": 1.4972, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.9554140127388535, | |
| "grad_norm": 1.5369589866278408, | |
| "learning_rate": 4.777070063694267e-07, | |
| "loss": 1.4748, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.9681528662420382, | |
| "grad_norm": 1.7257692046855166, | |
| "learning_rate": 4.840764331210191e-07, | |
| "loss": 1.5267, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.9808917197452229, | |
| "grad_norm": 2.1407422935686338, | |
| "learning_rate": 4.904458598726115e-07, | |
| "loss": 1.5097, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.9936305732484076, | |
| "grad_norm": 1.9406879702100648, | |
| "learning_rate": 4.968152866242038e-07, | |
| "loss": 1.511, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.0063694267515924, | |
| "grad_norm": 1.439554415506634, | |
| "learning_rate": 5.031847133757962e-07, | |
| "loss": 1.5122, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.019108280254777, | |
| "grad_norm": 1.5106128675527724, | |
| "learning_rate": 5.095541401273885e-07, | |
| "loss": 1.4956, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.0318471337579618, | |
| "grad_norm": 1.6554308659382122, | |
| "learning_rate": 5.159235668789809e-07, | |
| "loss": 1.5018, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.0445859872611465, | |
| "grad_norm": 2.026680377715203, | |
| "learning_rate": 5.222929936305732e-07, | |
| "loss": 1.4776, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.0573248407643312, | |
| "grad_norm": 1.5139010788192246, | |
| "learning_rate": 5.286624203821656e-07, | |
| "loss": 1.4545, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.070063694267516, | |
| "grad_norm": 1.4273185604405294, | |
| "learning_rate": 5.350318471337579e-07, | |
| "loss": 1.4834, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.0828025477707006, | |
| "grad_norm": 1.4905172698067897, | |
| "learning_rate": 5.414012738853503e-07, | |
| "loss": 1.5482, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.0955414012738853, | |
| "grad_norm": 1.507962232188269, | |
| "learning_rate": 5.477707006369426e-07, | |
| "loss": 1.4802, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.10828025477707, | |
| "grad_norm": 1.738422445567781, | |
| "learning_rate": 5.54140127388535e-07, | |
| "loss": 1.5378, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.1210191082802548, | |
| "grad_norm": 1.4572453197036355, | |
| "learning_rate": 5.605095541401273e-07, | |
| "loss": 1.5151, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.1337579617834395, | |
| "grad_norm": 2.003605940138242, | |
| "learning_rate": 5.668789808917197e-07, | |
| "loss": 1.4804, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.1464968152866242, | |
| "grad_norm": 1.6149829006958483, | |
| "learning_rate": 5.73248407643312e-07, | |
| "loss": 1.5131, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.1592356687898089, | |
| "grad_norm": 1.6728742331262298, | |
| "learning_rate": 5.796178343949044e-07, | |
| "loss": 1.5034, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.1719745222929936, | |
| "grad_norm": 1.5431486794721163, | |
| "learning_rate": 5.859872611464968e-07, | |
| "loss": 1.5124, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.1847133757961783, | |
| "grad_norm": 1.6476023906777157, | |
| "learning_rate": 5.923566878980892e-07, | |
| "loss": 1.4764, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.197452229299363, | |
| "grad_norm": 1.6894401103415613, | |
| "learning_rate": 5.987261146496815e-07, | |
| "loss": 1.5029, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.2101910828025477, | |
| "grad_norm": 1.6691648931963563, | |
| "learning_rate": 6.050955414012739e-07, | |
| "loss": 1.4969, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.2229299363057324, | |
| "grad_norm": 2.051456115160016, | |
| "learning_rate": 6.114649681528662e-07, | |
| "loss": 1.4744, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.2356687898089171, | |
| "grad_norm": 2.1062990461224245, | |
| "learning_rate": 6.178343949044586e-07, | |
| "loss": 1.4882, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.2484076433121019, | |
| "grad_norm": 1.6527758494430145, | |
| "learning_rate": 6.242038216560509e-07, | |
| "loss": 1.4664, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.2611464968152866, | |
| "grad_norm": 1.4778661306564593, | |
| "learning_rate": 6.305732484076433e-07, | |
| "loss": 1.5409, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.2738853503184713, | |
| "grad_norm": 1.635617614245788, | |
| "learning_rate": 6.369426751592356e-07, | |
| "loss": 1.4478, | |
| "step": 1000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 15700, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 152037753618432.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |