| { | |
| "best_metric": 2.1530466079711914, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-100", | |
| "epoch": 2.995670995670996, | |
| "eval_steps": 50, | |
| "global_step": 173, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.017316017316017316, | |
| "grad_norm": 2.8331797122955322, | |
| "learning_rate": 1e-05, | |
| "loss": 2.1187, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.017316017316017316, | |
| "eval_loss": 3.7099266052246094, | |
| "eval_runtime": 4.4282, | |
| "eval_samples_per_second": 22.131, | |
| "eval_steps_per_second": 5.646, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03463203463203463, | |
| "grad_norm": 5.01226806640625, | |
| "learning_rate": 2e-05, | |
| "loss": 2.7977, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05194805194805195, | |
| "grad_norm": 4.797194480895996, | |
| "learning_rate": 3e-05, | |
| "loss": 2.7871, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.06926406926406926, | |
| "grad_norm": 4.131883144378662, | |
| "learning_rate": 4e-05, | |
| "loss": 2.7443, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.08658008658008658, | |
| "grad_norm": 3.087463855743408, | |
| "learning_rate": 5e-05, | |
| "loss": 2.8541, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.1038961038961039, | |
| "grad_norm": 2.660694122314453, | |
| "learning_rate": 6e-05, | |
| "loss": 2.9244, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 2.8298134803771973, | |
| "learning_rate": 7e-05, | |
| "loss": 2.6589, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.13852813852813853, | |
| "grad_norm": 3.4778504371643066, | |
| "learning_rate": 8e-05, | |
| "loss": 2.8337, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.15584415584415584, | |
| "grad_norm": 3.807507038116455, | |
| "learning_rate": 9e-05, | |
| "loss": 2.892, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.17316017316017315, | |
| "grad_norm": 3.6214892864227295, | |
| "learning_rate": 0.0001, | |
| "loss": 2.8247, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.19047619047619047, | |
| "grad_norm": 3.1061809062957764, | |
| "learning_rate": 9.999071352056675e-05, | |
| "loss": 2.5751, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.2077922077922078, | |
| "grad_norm": 4.615087509155273, | |
| "learning_rate": 9.9962857531815e-05, | |
| "loss": 2.7903, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.22510822510822512, | |
| "grad_norm": 4.513574600219727, | |
| "learning_rate": 9.99164423811074e-05, | |
| "loss": 2.8595, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 5.717080593109131, | |
| "learning_rate": 9.985148530977767e-05, | |
| "loss": 3.224, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.2597402597402597, | |
| "grad_norm": 1.9179377555847168, | |
| "learning_rate": 9.976801044672608e-05, | |
| "loss": 1.8395, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.27705627705627706, | |
| "grad_norm": 1.8630417585372925, | |
| "learning_rate": 9.966604879945659e-05, | |
| "loss": 2.4277, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2943722943722944, | |
| "grad_norm": 1.6356393098831177, | |
| "learning_rate": 9.954563824255878e-05, | |
| "loss": 2.3636, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3116883116883117, | |
| "grad_norm": 1.710647463798523, | |
| "learning_rate": 9.940682350363912e-05, | |
| "loss": 2.2757, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.329004329004329, | |
| "grad_norm": 1.6765035390853882, | |
| "learning_rate": 9.924965614670629e-05, | |
| "loss": 2.1883, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.3463203463203463, | |
| "grad_norm": 1.8588552474975586, | |
| "learning_rate": 9.907419455301741e-05, | |
| "loss": 2.0356, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 2.132901191711426, | |
| "learning_rate": 9.888050389939172e-05, | |
| "loss": 2.3527, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.38095238095238093, | |
| "grad_norm": 2.075359582901001, | |
| "learning_rate": 9.866865613400008e-05, | |
| "loss": 2.3003, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.39826839826839827, | |
| "grad_norm": 2.2094523906707764, | |
| "learning_rate": 9.843872994963911e-05, | |
| "loss": 2.3735, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.4155844155844156, | |
| "grad_norm": 2.4340925216674805, | |
| "learning_rate": 9.819081075450014e-05, | |
| "loss": 2.5142, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.4329004329004329, | |
| "grad_norm": 2.3378214836120605, | |
| "learning_rate": 9.792499064044342e-05, | |
| "loss": 2.5139, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.45021645021645024, | |
| "grad_norm": 2.4948999881744385, | |
| "learning_rate": 9.764136834878986e-05, | |
| "loss": 2.5467, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.4675324675324675, | |
| "grad_norm": 2.667940139770508, | |
| "learning_rate": 9.734004923364257e-05, | |
| "loss": 2.4101, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 5.119446277618408, | |
| "learning_rate": 9.702114522275216e-05, | |
| "loss": 3.1252, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5021645021645021, | |
| "grad_norm": 1.3318713903427124, | |
| "learning_rate": 9.66847747759402e-05, | |
| "loss": 1.6226, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "grad_norm": 2.2476823329925537, | |
| "learning_rate": 9.63310628410961e-05, | |
| "loss": 2.2634, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5367965367965368, | |
| "grad_norm": 1.6670700311660767, | |
| "learning_rate": 9.596014080776423e-05, | |
| "loss": 2.0752, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5541125541125541, | |
| "grad_norm": 1.5613207817077637, | |
| "learning_rate": 9.557214645833792e-05, | |
| "loss": 2.0647, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 1.5586175918579102, | |
| "learning_rate": 9.516722391687902e-05, | |
| "loss": 2.0191, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5887445887445888, | |
| "grad_norm": 1.7481515407562256, | |
| "learning_rate": 9.474552359558166e-05, | |
| "loss": 2.0408, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 1.908803939819336, | |
| "learning_rate": 9.43072021389003e-05, | |
| "loss": 2.251, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6233766233766234, | |
| "grad_norm": 1.9676694869995117, | |
| "learning_rate": 9.38524223653626e-05, | |
| "loss": 2.1518, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6406926406926406, | |
| "grad_norm": 2.245049238204956, | |
| "learning_rate": 9.338135320708911e-05, | |
| "loss": 2.3752, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.658008658008658, | |
| "grad_norm": 2.3313148021698, | |
| "learning_rate": 9.289416964704185e-05, | |
| "loss": 2.6293, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6753246753246753, | |
| "grad_norm": 2.3561019897460938, | |
| "learning_rate": 9.239105265402525e-05, | |
| "loss": 2.3578, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.6926406926406926, | |
| "grad_norm": 2.4642536640167236, | |
| "learning_rate": 9.187218911546362e-05, | |
| "loss": 2.3347, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.70995670995671, | |
| "grad_norm": 2.577402353286743, | |
| "learning_rate": 9.133777176798013e-05, | |
| "loss": 2.3815, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 4.283364772796631, | |
| "learning_rate": 9.078799912580304e-05, | |
| "loss": 2.3154, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.7445887445887446, | |
| "grad_norm": 1.0042027235031128, | |
| "learning_rate": 9.022307540702576e-05, | |
| "loss": 1.7462, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7619047619047619, | |
| "grad_norm": 1.7617307901382446, | |
| "learning_rate": 8.964321045774807e-05, | |
| "loss": 2.0569, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "grad_norm": 1.7388392686843872, | |
| "learning_rate": 8.904861967412703e-05, | |
| "loss": 2.0223, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.7965367965367965, | |
| "grad_norm": 1.6279188394546509, | |
| "learning_rate": 8.843952392236594e-05, | |
| "loss": 1.8667, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.8138528138528138, | |
| "grad_norm": 1.613118290901184, | |
| "learning_rate": 8.781614945667169e-05, | |
| "loss": 2.1793, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.8311688311688312, | |
| "grad_norm": 1.5092192888259888, | |
| "learning_rate": 8.717872783521047e-05, | |
| "loss": 2.0122, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.8484848484848485, | |
| "grad_norm": 1.6618455648422241, | |
| "learning_rate": 8.65274958340934e-05, | |
| "loss": 1.8772, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8658008658008658, | |
| "grad_norm": 1.820896863937378, | |
| "learning_rate": 8.586269535942385e-05, | |
| "loss": 2.1534, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8658008658008658, | |
| "eval_loss": 2.2589545249938965, | |
| "eval_runtime": 4.4124, | |
| "eval_samples_per_second": 22.21, | |
| "eval_steps_per_second": 5.666, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8831168831168831, | |
| "grad_norm": 2.0972702503204346, | |
| "learning_rate": 8.518457335743926e-05, | |
| "loss": 2.3293, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.9004329004329005, | |
| "grad_norm": 2.350532293319702, | |
| "learning_rate": 8.449338172278059e-05, | |
| "loss": 2.4826, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.9177489177489178, | |
| "grad_norm": 2.3178534507751465, | |
| "learning_rate": 8.378937720492384e-05, | |
| "loss": 2.2085, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.935064935064935, | |
| "grad_norm": 2.572354793548584, | |
| "learning_rate": 8.307282131280804e-05, | |
| "loss": 2.2593, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.9523809523809523, | |
| "grad_norm": 2.8676466941833496, | |
| "learning_rate": 8.23439802176954e-05, | |
| "loss": 2.401, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 5.183324813842773, | |
| "learning_rate": 8.160312465429952e-05, | |
| "loss": 2.8153, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.987012987012987, | |
| "grad_norm": 1.302344560623169, | |
| "learning_rate": 8.085052982021847e-05, | |
| "loss": 1.8029, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.0043290043290043, | |
| "grad_norm": 3.5733864307403564, | |
| "learning_rate": 8.008647527371023e-05, | |
| "loss": 2.9339, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.0216450216450217, | |
| "grad_norm": 0.99031662940979, | |
| "learning_rate": 7.931124482984802e-05, | |
| "loss": 1.6368, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.0389610389610389, | |
| "grad_norm": 1.045397162437439, | |
| "learning_rate": 7.85251264550948e-05, | |
| "loss": 1.7296, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0562770562770563, | |
| "grad_norm": 1.1642181873321533, | |
| "learning_rate": 7.772841216033533e-05, | |
| "loss": 1.7076, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.0735930735930737, | |
| "grad_norm": 1.2963627576828003, | |
| "learning_rate": 7.692139789240611e-05, | |
| "loss": 1.6386, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 1.37860107421875, | |
| "learning_rate": 7.610438342416319e-05, | |
| "loss": 1.5183, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.1082251082251082, | |
| "grad_norm": 1.6060508489608765, | |
| "learning_rate": 7.527767224312883e-05, | |
| "loss": 1.8744, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.1255411255411256, | |
| "grad_norm": 1.3911950588226318, | |
| "learning_rate": 7.44415714387582e-05, | |
| "loss": 1.4808, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 1.5222082138061523, | |
| "learning_rate": 7.359639158836828e-05, | |
| "loss": 1.7056, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.1601731601731602, | |
| "grad_norm": 1.7811347246170044, | |
| "learning_rate": 7.274244664177097e-05, | |
| "loss": 1.8924, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.1774891774891776, | |
| "grad_norm": 1.972655177116394, | |
| "learning_rate": 7.188005380465364e-05, | |
| "loss": 1.7403, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.1948051948051948, | |
| "grad_norm": 2.413468837738037, | |
| "learning_rate": 7.10095334207501e-05, | |
| "loss": 1.7472, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.2121212121212122, | |
| "grad_norm": 2.587477207183838, | |
| "learning_rate": 7.013120885284598e-05, | |
| "loss": 1.9488, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2294372294372296, | |
| "grad_norm": 2.9659059047698975, | |
| "learning_rate": 6.924540636266272e-05, | |
| "loss": 1.7928, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.2467532467532467, | |
| "grad_norm": 3.635132074356079, | |
| "learning_rate": 6.835245498966461e-05, | |
| "loss": 1.8512, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.2640692640692641, | |
| "grad_norm": 1.1850864887237549, | |
| "learning_rate": 6.745268642883404e-05, | |
| "loss": 1.3889, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.2813852813852815, | |
| "grad_norm": 1.5134196281433105, | |
| "learning_rate": 6.654643490746042e-05, | |
| "loss": 1.8089, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.2987012987012987, | |
| "grad_norm": 1.5268129110336304, | |
| "learning_rate": 6.563403706098833e-05, | |
| "loss": 1.913, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.316017316017316, | |
| "grad_norm": 1.3524274826049805, | |
| "learning_rate": 6.471583180797121e-05, | |
| "loss": 1.317, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 1.6053553819656372, | |
| "learning_rate": 6.379216022417696e-05, | |
| "loss": 1.7508, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.3506493506493507, | |
| "grad_norm": 1.6736403703689575, | |
| "learning_rate": 6.286336541589224e-05, | |
| "loss": 1.6712, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.3679653679653678, | |
| "grad_norm": 1.6220240592956543, | |
| "learning_rate": 6.192979239247243e-05, | |
| "loss": 1.5654, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.3852813852813852, | |
| "grad_norm": 1.7063477039337158, | |
| "learning_rate": 6.0991787938184784e-05, | |
| "loss": 1.7755, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.4025974025974026, | |
| "grad_norm": 1.7277164459228516, | |
| "learning_rate": 6.004970048339226e-05, | |
| "loss": 1.5972, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.4199134199134198, | |
| "grad_norm": 1.855843186378479, | |
| "learning_rate": 5.910387997512573e-05, | |
| "loss": 1.7611, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.4372294372294372, | |
| "grad_norm": 2.0694801807403564, | |
| "learning_rate": 5.8154677747093134e-05, | |
| "loss": 1.6634, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 2.3597748279571533, | |
| "learning_rate": 5.7202446389173223e-05, | |
| "loss": 1.9045, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.4718614718614718, | |
| "grad_norm": 2.8176522254943848, | |
| "learning_rate": 5.624753961644281e-05, | |
| "loss": 1.6573, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.4891774891774892, | |
| "grad_norm": 2.932007312774658, | |
| "learning_rate": 5.5290312137786146e-05, | |
| "loss": 1.5173, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.5064935064935066, | |
| "grad_norm": 0.9135814905166626, | |
| "learning_rate": 5.433111952413495e-05, | |
| "loss": 1.4504, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.5238095238095237, | |
| "grad_norm": 1.2319753170013428, | |
| "learning_rate": 5.33703180763884e-05, | |
| "loss": 1.8403, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.5411255411255411, | |
| "grad_norm": 1.2729883193969727, | |
| "learning_rate": 5.240826469306187e-05, | |
| "loss": 1.6524, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.5584415584415585, | |
| "grad_norm": 1.4365655183792114, | |
| "learning_rate": 5.144531673771363e-05, | |
| "loss": 1.4742, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5757575757575757, | |
| "grad_norm": 1.4098918437957764, | |
| "learning_rate": 5.048183190619904e-05, | |
| "loss": 1.5275, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.593073593073593, | |
| "grad_norm": 1.5895005464553833, | |
| "learning_rate": 4.951816809380097e-05, | |
| "loss": 1.5147, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.6103896103896105, | |
| "grad_norm": 1.71897292137146, | |
| "learning_rate": 4.855468326228638e-05, | |
| "loss": 1.5553, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.6277056277056277, | |
| "grad_norm": 1.8746683597564697, | |
| "learning_rate": 4.759173530693814e-05, | |
| "loss": 1.6687, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.645021645021645, | |
| "grad_norm": 2.1632306575775146, | |
| "learning_rate": 4.6629681923611603e-05, | |
| "loss": 1.7037, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.6623376623376624, | |
| "grad_norm": 2.1969408988952637, | |
| "learning_rate": 4.566888047586507e-05, | |
| "loss": 1.8775, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.6796536796536796, | |
| "grad_norm": 2.26798152923584, | |
| "learning_rate": 4.4709687862213866e-05, | |
| "loss": 1.6915, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.696969696969697, | |
| "grad_norm": 2.2888317108154297, | |
| "learning_rate": 4.3752460383557195e-05, | |
| "loss": 1.7374, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 3.2511236667633057, | |
| "learning_rate": 4.27975536108268e-05, | |
| "loss": 1.8612, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.7316017316017316, | |
| "grad_norm": 3.068394899368286, | |
| "learning_rate": 4.1845322252906864e-05, | |
| "loss": 1.4067, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.7316017316017316, | |
| "eval_loss": 2.1530466079711914, | |
| "eval_runtime": 4.4162, | |
| "eval_samples_per_second": 22.191, | |
| "eval_steps_per_second": 5.661, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.7489177489177488, | |
| "grad_norm": 1.058007001876831, | |
| "learning_rate": 4.0896120024874286e-05, | |
| "loss": 1.5795, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.7662337662337664, | |
| "grad_norm": 1.2379379272460938, | |
| "learning_rate": 3.9950299516607766e-05, | |
| "loss": 1.4978, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.7835497835497836, | |
| "grad_norm": 1.4069570302963257, | |
| "learning_rate": 3.900821206181521e-05, | |
| "loss": 1.5601, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.8008658008658007, | |
| "grad_norm": 1.4146642684936523, | |
| "learning_rate": 3.8070207607527584e-05, | |
| "loss": 1.5123, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 1.520447015762329, | |
| "learning_rate": 3.713663458410779e-05, | |
| "loss": 1.4908, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.8354978354978355, | |
| "grad_norm": 1.6304378509521484, | |
| "learning_rate": 3.620783977582305e-05, | |
| "loss": 1.3594, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.8528138528138527, | |
| "grad_norm": 1.668250560760498, | |
| "learning_rate": 3.528416819202881e-05, | |
| "loss": 1.4907, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.87012987012987, | |
| "grad_norm": 1.842556357383728, | |
| "learning_rate": 3.43659629390117e-05, | |
| "loss": 1.7364, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.8874458874458875, | |
| "grad_norm": 2.1994338035583496, | |
| "learning_rate": 3.345356509253959e-05, | |
| "loss": 1.8325, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.9047619047619047, | |
| "grad_norm": 2.015636920928955, | |
| "learning_rate": 3.254731357116597e-05, | |
| "loss": 1.6642, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.922077922077922, | |
| "grad_norm": 2.2322309017181396, | |
| "learning_rate": 3.16475450103354e-05, | |
| "loss": 1.5802, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.9393939393939394, | |
| "grad_norm": 2.6270971298217773, | |
| "learning_rate": 3.0754593637337276e-05, | |
| "loss": 1.6923, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.9567099567099566, | |
| "grad_norm": 3.3639109134674072, | |
| "learning_rate": 2.986879114715403e-05, | |
| "loss": 1.7439, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.974025974025974, | |
| "grad_norm": 3.1718485355377197, | |
| "learning_rate": 2.899046657924992e-05, | |
| "loss": 1.4849, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.9913419913419914, | |
| "grad_norm": 2.31093168258667, | |
| "learning_rate": 2.8119946195346375e-05, | |
| "loss": 1.7909, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.0086580086580086, | |
| "grad_norm": 2.096163034439087, | |
| "learning_rate": 2.7257553358229034e-05, | |
| "loss": 1.5833, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.0259740259740258, | |
| "grad_norm": 0.9350630044937134, | |
| "learning_rate": 2.6403608411631742e-05, | |
| "loss": 1.2758, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.0432900432900434, | |
| "grad_norm": 1.2270903587341309, | |
| "learning_rate": 2.555842856124182e-05, | |
| "loss": 1.5394, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.0606060606060606, | |
| "grad_norm": 1.1900259256362915, | |
| "learning_rate": 2.472232775687119e-05, | |
| "loss": 1.1756, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.0779220779220777, | |
| "grad_norm": 1.293131709098816, | |
| "learning_rate": 2.389561657583681e-05, | |
| "loss": 1.2139, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.0952380952380953, | |
| "grad_norm": 1.4784541130065918, | |
| "learning_rate": 2.30786021075939e-05, | |
| "loss": 1.4056, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.1125541125541125, | |
| "grad_norm": 1.422511339187622, | |
| "learning_rate": 2.2271587839664672e-05, | |
| "loss": 1.0889, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.1298701298701297, | |
| "grad_norm": 1.5573115348815918, | |
| "learning_rate": 2.1474873544905205e-05, | |
| "loss": 1.2393, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.1471861471861473, | |
| "grad_norm": 1.5458406209945679, | |
| "learning_rate": 2.0688755170151996e-05, | |
| "loss": 1.2199, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.1645021645021645, | |
| "grad_norm": 1.9785674810409546, | |
| "learning_rate": 1.9913524726289784e-05, | |
| "loss": 1.3959, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 2.0440776348114014, | |
| "learning_rate": 1.914947017978153e-05, | |
| "loss": 1.3751, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.1991341991341993, | |
| "grad_norm": 2.1104283332824707, | |
| "learning_rate": 1.8396875345700497e-05, | |
| "loss": 1.1956, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.2164502164502164, | |
| "grad_norm": 2.581237554550171, | |
| "learning_rate": 1.76560197823046e-05, | |
| "loss": 1.3668, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.2337662337662336, | |
| "grad_norm": 3.1223180294036865, | |
| "learning_rate": 1.692717868719195e-05, | |
| "loss": 1.136, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.2510822510822512, | |
| "grad_norm": 2.574193239212036, | |
| "learning_rate": 1.621062279507617e-05, | |
| "loss": 1.0829, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.2683982683982684, | |
| "grad_norm": 1.1069941520690918, | |
| "learning_rate": 1.550661827721941e-05, | |
| "loss": 1.1654, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "grad_norm": 1.4014630317687988, | |
| "learning_rate": 1.4815426642560754e-05, | |
| "loss": 1.3528, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.303030303030303, | |
| "grad_norm": 1.6205848455429077, | |
| "learning_rate": 1.413730464057616e-05, | |
| "loss": 1.4042, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.3203463203463204, | |
| "grad_norm": 1.6912100315093994, | |
| "learning_rate": 1.3472504165906613e-05, | |
| "loss": 1.1654, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.3376623376623376, | |
| "grad_norm": 1.7683279514312744, | |
| "learning_rate": 1.2821272164789544e-05, | |
| "loss": 1.2481, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.354978354978355, | |
| "grad_norm": 1.8324084281921387, | |
| "learning_rate": 1.2183850543328312e-05, | |
| "loss": 1.1558, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.3722943722943723, | |
| "grad_norm": 1.9852972030639648, | |
| "learning_rate": 1.156047607763407e-05, | |
| "loss": 1.2857, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.3896103896103895, | |
| "grad_norm": 2.126631259918213, | |
| "learning_rate": 1.0951380325872979e-05, | |
| "loss": 1.2143, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.4069264069264067, | |
| "grad_norm": 2.3943769931793213, | |
| "learning_rate": 1.0356789542251938e-05, | |
| "loss": 1.2955, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.4242424242424243, | |
| "grad_norm": 2.30513596534729, | |
| "learning_rate": 9.776924592974256e-06, | |
| "loss": 1.1925, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.4415584415584415, | |
| "grad_norm": 2.5262770652770996, | |
| "learning_rate": 9.212000874196953e-06, | |
| "loss": 1.2257, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.458874458874459, | |
| "grad_norm": 3.0429446697235107, | |
| "learning_rate": 8.662228232019876e-06, | |
| "loss": 1.4461, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.4761904761904763, | |
| "grad_norm": 3.428879737854004, | |
| "learning_rate": 8.127810884536403e-06, | |
| "loss": 1.0103, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.4935064935064934, | |
| "grad_norm": 2.0732643604278564, | |
| "learning_rate": 7.60894734597476e-06, | |
| "loss": 1.0342, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.5108225108225106, | |
| "grad_norm": 1.3209905624389648, | |
| "learning_rate": 7.105830352958142e-06, | |
| "loss": 1.4006, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.5281385281385282, | |
| "grad_norm": 1.4162795543670654, | |
| "learning_rate": 6.618646792910893e-06, | |
| "loss": 1.2731, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 1.7020889520645142, | |
| "learning_rate": 6.147577634637414e-06, | |
| "loss": 1.3335, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.562770562770563, | |
| "grad_norm": 1.5427604913711548, | |
| "learning_rate": 5.692797861099719e-06, | |
| "loss": 1.0563, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.58008658008658, | |
| "grad_norm": 1.9081095457077026, | |
| "learning_rate": 5.25447640441834e-06, | |
| "loss": 1.2153, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.5974025974025974, | |
| "grad_norm": 1.9644572734832764, | |
| "learning_rate": 4.832776083120982e-06, | |
| "loss": 1.3009, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.5974025974025974, | |
| "eval_loss": 2.225019931793213, | |
| "eval_runtime": 4.4108, | |
| "eval_samples_per_second": 22.218, | |
| "eval_steps_per_second": 5.668, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.6147186147186146, | |
| "grad_norm": 1.814753532409668, | |
| "learning_rate": 4.427853541662091e-06, | |
| "loss": 1.1572, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.632034632034632, | |
| "grad_norm": 2.255131959915161, | |
| "learning_rate": 4.039859192235779e-06, | |
| "loss": 1.3291, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.6493506493506493, | |
| "grad_norm": 2.160221576690674, | |
| "learning_rate": 3.668937158903901e-06, | |
| "loss": 1.3002, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 2.4960460662841797, | |
| "learning_rate": 3.315225224059809e-06, | |
| "loss": 1.387, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.683982683982684, | |
| "grad_norm": 2.546381711959839, | |
| "learning_rate": 2.9788547772478416e-06, | |
| "loss": 1.3375, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.7012987012987013, | |
| "grad_norm": 2.7530722618103027, | |
| "learning_rate": 2.6599507663574384e-06, | |
| "loss": 1.2888, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.7186147186147185, | |
| "grad_norm": 3.3948094844818115, | |
| "learning_rate": 2.3586316512101416e-06, | |
| "loss": 1.0494, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.7359307359307357, | |
| "grad_norm": 2.2093355655670166, | |
| "learning_rate": 2.0750093595565733e-06, | |
| "loss": 0.9195, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.7532467532467533, | |
| "grad_norm": 1.2557899951934814, | |
| "learning_rate": 1.8091892454998594e-06, | |
| "loss": 1.3048, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.7705627705627704, | |
| "grad_norm": 1.4298012256622314, | |
| "learning_rate": 1.5612700503608968e-06, | |
| "loss": 1.2899, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.787878787878788, | |
| "grad_norm": 1.55137300491333, | |
| "learning_rate": 1.33134386599994e-06, | |
| "loss": 1.3592, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.8051948051948052, | |
| "grad_norm": 1.5704561471939087, | |
| "learning_rate": 1.1194961006082972e-06, | |
| "loss": 1.0256, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.8225108225108224, | |
| "grad_norm": 1.9115115404129028, | |
| "learning_rate": 9.258054469825972e-07, | |
| "loss": 1.2241, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.8398268398268396, | |
| "grad_norm": 1.8863279819488525, | |
| "learning_rate": 7.503438532937168e-07, | |
| "loss": 1.2463, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 2.0536916255950928, | |
| "learning_rate": 5.931764963608866e-07, | |
| "loss": 1.2806, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.8744588744588744, | |
| "grad_norm": 2.066479206085205, | |
| "learning_rate": 4.543617574412184e-07, | |
| "loss": 1.2572, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.891774891774892, | |
| "grad_norm": 2.4157185554504395, | |
| "learning_rate": 3.339512005434309e-07, | |
| "loss": 1.4229, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 2.4809465408325195, | |
| "learning_rate": 2.319895532739369e-07, | |
| "loss": 1.3262, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.9264069264069263, | |
| "grad_norm": 2.556966781616211, | |
| "learning_rate": 1.4851469022234e-07, | |
| "loss": 1.2532, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.9437229437229435, | |
| "grad_norm": 2.8384437561035156, | |
| "learning_rate": 8.355761889260461e-08, | |
| "loss": 1.2835, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.961038961038961, | |
| "grad_norm": 3.8424901962280273, | |
| "learning_rate": 3.7142468185014104e-08, | |
| "loss": 1.2521, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.9783549783549783, | |
| "grad_norm": 2.5043461322784424, | |
| "learning_rate": 9.286479433257e-09, | |
| "loss": 1.1914, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.995670995670996, | |
| "grad_norm": 3.945042371749878, | |
| "learning_rate": 0.0, | |
| "loss": 1.3781, | |
| "step": 173 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 173, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 1 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.155617850143539e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |