| { | |
| "best_metric": 0.2598925232887268, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 0.035557662676306744, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00023705108450871164, | |
| "grad_norm": 45.60728073120117, | |
| "learning_rate": 1e-05, | |
| "loss": 7.2208, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00023705108450871164, | |
| "eval_loss": 2.17527437210083, | |
| "eval_runtime": 512.9389, | |
| "eval_samples_per_second": 13.852, | |
| "eval_steps_per_second": 3.464, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0004741021690174233, | |
| "grad_norm": 39.479434967041016, | |
| "learning_rate": 2e-05, | |
| "loss": 5.9475, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0007111532535261348, | |
| "grad_norm": 43.14913558959961, | |
| "learning_rate": 3e-05, | |
| "loss": 6.5429, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0009482043380348466, | |
| "grad_norm": 38.2624626159668, | |
| "learning_rate": 4e-05, | |
| "loss": 4.6751, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0011852554225435582, | |
| "grad_norm": 37.59519577026367, | |
| "learning_rate": 5e-05, | |
| "loss": 3.7288, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0014223065070522697, | |
| "grad_norm": 24.643327713012695, | |
| "learning_rate": 6e-05, | |
| "loss": 2.9582, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0016593575915609814, | |
| "grad_norm": 20.228885650634766, | |
| "learning_rate": 7e-05, | |
| "loss": 2.3467, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.001896408676069693, | |
| "grad_norm": 15.846456527709961, | |
| "learning_rate": 8e-05, | |
| "loss": 1.4973, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.002133459760578405, | |
| "grad_norm": 14.652710914611816, | |
| "learning_rate": 9e-05, | |
| "loss": 1.5795, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0023705108450871163, | |
| "grad_norm": 22.30735969543457, | |
| "learning_rate": 0.0001, | |
| "loss": 1.6046, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.002607561929595828, | |
| "grad_norm": 20.14059066772461, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 2.1066, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0028446130141045393, | |
| "grad_norm": 14.741179466247559, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 1.6849, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0030816640986132513, | |
| "grad_norm": 15.680244445800781, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 1.6815, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0033187151831219628, | |
| "grad_norm": 11.952381134033203, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 1.6821, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0035557662676306743, | |
| "grad_norm": 7.957185745239258, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 0.8576, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.003792817352139386, | |
| "grad_norm": 8.777189254760742, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 1.4227, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.004029868436648098, | |
| "grad_norm": 15.89772891998291, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 1.3271, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.00426691952115681, | |
| "grad_norm": 13.817143440246582, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 1.7801, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.004503970605665521, | |
| "grad_norm": 15.342754364013672, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 1.3992, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.004741021690174233, | |
| "grad_norm": 9.978917121887207, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 1.1139, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.004978072774682945, | |
| "grad_norm": 10.110090255737305, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 1.3036, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.005215123859191656, | |
| "grad_norm": 12.066587448120117, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 1.3331, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.005452174943700368, | |
| "grad_norm": 15.906118392944336, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 1.6102, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.005689226028209079, | |
| "grad_norm": 12.463726997375488, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 1.4063, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.005926277112717791, | |
| "grad_norm": 10.469868659973145, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 1.4652, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0061633281972265025, | |
| "grad_norm": 8.34838581085205, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 1.0534, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.006400379281735214, | |
| "grad_norm": 7.536215782165527, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 1.2063, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0066374303662439255, | |
| "grad_norm": 7.370880603790283, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 0.9058, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0068744814507526375, | |
| "grad_norm": 9.6860933303833, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 1.1714, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0071115325352613485, | |
| "grad_norm": 13.223746299743652, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 1.2399, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0073485836197700605, | |
| "grad_norm": 10.730929374694824, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 1.1303, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.007585634704278772, | |
| "grad_norm": 6.884200096130371, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 0.9594, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.007822685788787484, | |
| "grad_norm": 11.763479232788086, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 1.6842, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.008059736873296195, | |
| "grad_norm": 10.257864952087402, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 1.4945, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.008296787957804906, | |
| "grad_norm": 13.270899772644043, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 1.2097, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.00853383904231362, | |
| "grad_norm": 11.5889253616333, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 1.3319, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.00877089012682233, | |
| "grad_norm": 11.069531440734863, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 1.564, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.009007941211331041, | |
| "grad_norm": 9.586097717285156, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 1.5775, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.009244992295839754, | |
| "grad_norm": 9.339739799499512, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 1.2764, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.009482043380348465, | |
| "grad_norm": 8.734746932983398, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 1.3781, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.009719094464857176, | |
| "grad_norm": 11.1151704788208, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 1.3118, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.00995614554936589, | |
| "grad_norm": 10.246061325073242, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 1.5707, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.0101931966338746, | |
| "grad_norm": 12.3263521194458, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 1.723, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.010430247718383311, | |
| "grad_norm": 8.47308349609375, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 1.1802, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.010667298802892022, | |
| "grad_norm": 8.325446128845215, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 1.1147, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.010904349887400735, | |
| "grad_norm": 12.091687202453613, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 1.0594, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.011141400971909446, | |
| "grad_norm": 7.746819019317627, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 0.797, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.011378452056418157, | |
| "grad_norm": 9.007611274719238, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.9117, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.01161550314092687, | |
| "grad_norm": 12.215007781982422, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 1.7131, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.011852554225435581, | |
| "grad_norm": 14.961793899536133, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 1.4138, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.011852554225435581, | |
| "eval_loss": 0.2936895191669464, | |
| "eval_runtime": 518.4074, | |
| "eval_samples_per_second": 13.705, | |
| "eval_steps_per_second": 3.428, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.012089605309944292, | |
| "grad_norm": 9.533975601196289, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 1.2758, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.012326656394453005, | |
| "grad_norm": 10.313169479370117, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 1.0634, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.012563707478961716, | |
| "grad_norm": 10.298772811889648, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 1.292, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.012800758563470427, | |
| "grad_norm": 7.554491996765137, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 1.1744, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.01303780964797914, | |
| "grad_norm": 11.438995361328125, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 1.545, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.013274860732487851, | |
| "grad_norm": 9.73892879486084, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 1.3023, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.013511911816996562, | |
| "grad_norm": 7.812572956085205, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 1.3008, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.013748962901505275, | |
| "grad_norm": 9.590088844299316, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 1.5249, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.013986013986013986, | |
| "grad_norm": 9.948052406311035, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 1.3549, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.014223065070522697, | |
| "grad_norm": 8.376433372497559, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 1.2803, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.01446011615503141, | |
| "grad_norm": 5.502856254577637, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 0.9784, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.014697167239540121, | |
| "grad_norm": 9.21668815612793, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 1.4134, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.014934218324048832, | |
| "grad_norm": 6.221367835998535, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 0.979, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.015171269408557545, | |
| "grad_norm": 9.170941352844238, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 1.1757, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.015408320493066256, | |
| "grad_norm": 7.874046802520752, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 1.0453, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.01564537157757497, | |
| "grad_norm": 9.328206062316895, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 1.1295, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.01588242266208368, | |
| "grad_norm": 10.148048400878906, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 0.8948, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.01611947374659239, | |
| "grad_norm": 10.749184608459473, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 1.6974, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.016356524831101102, | |
| "grad_norm": 11.889942169189453, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 1.4894, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.016593575915609813, | |
| "grad_norm": 11.449923515319824, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 1.3378, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.016830627000118524, | |
| "grad_norm": 7.078919410705566, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 0.9582, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.01706767808462724, | |
| "grad_norm": 9.08558177947998, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 0.7153, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.01730472916913595, | |
| "grad_norm": 8.467891693115234, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 1.3214, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.01754178025364466, | |
| "grad_norm": 10.016956329345703, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 1.3502, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.017778831338153372, | |
| "grad_norm": 8.021065711975098, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 1.1391, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.018015882422662083, | |
| "grad_norm": 10.226483345031738, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 1.2666, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.018252933507170794, | |
| "grad_norm": 7.5994672775268555, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 0.929, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.01848998459167951, | |
| "grad_norm": 11.9320650100708, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 1.8044, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.01872703567618822, | |
| "grad_norm": 9.507772445678711, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 0.9235, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.01896408676069693, | |
| "grad_norm": 8.77246379852295, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 1.0449, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.01920113784520564, | |
| "grad_norm": 11.77663803100586, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 1.0512, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.019438188929714353, | |
| "grad_norm": 10.001542091369629, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 1.2113, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.019675240014223064, | |
| "grad_norm": 9.057281494140625, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 1.1445, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.01991229109873178, | |
| "grad_norm": 18.364543914794922, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 1.159, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.02014934218324049, | |
| "grad_norm": 9.89426326751709, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 0.9635, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.0203863932677492, | |
| "grad_norm": 12.017746925354004, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 1.0893, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.02062344435225791, | |
| "grad_norm": 11.03870964050293, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 0.8324, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.020860495436766623, | |
| "grad_norm": 10.81103515625, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 1.472, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.021097546521275334, | |
| "grad_norm": 27.316665649414062, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 1.0544, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.021334597605784045, | |
| "grad_norm": 20.078157424926758, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 0.8053, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.02157164869029276, | |
| "grad_norm": 9.962789535522461, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 1.0016, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.02180869977480147, | |
| "grad_norm": 17.85474967956543, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 1.117, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.02204575085931018, | |
| "grad_norm": 15.2191162109375, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 1.0432, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.022282801943818892, | |
| "grad_norm": 28.267955780029297, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 0.8571, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.022519853028327604, | |
| "grad_norm": 6.8763108253479, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 0.6518, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.022756904112836315, | |
| "grad_norm": 8.670818328857422, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 0.9891, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.02299395519734503, | |
| "grad_norm": 9.819561958312988, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 1.4, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.02323100628185374, | |
| "grad_norm": 7.537906169891357, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 1.0282, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.02346805736636245, | |
| "grad_norm": 17.41543960571289, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 1.2158, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.023705108450871162, | |
| "grad_norm": 9.822668075561523, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 1.0042, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.023705108450871162, | |
| "eval_loss": 0.28296175599098206, | |
| "eval_runtime": 516.2865, | |
| "eval_samples_per_second": 13.762, | |
| "eval_steps_per_second": 3.442, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.023942159535379873, | |
| "grad_norm": 8.359138488769531, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 1.1768, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.024179210619888584, | |
| "grad_norm": 8.67495346069336, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 1.3943, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.0244162617043973, | |
| "grad_norm": 9.854082107543945, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 1.4637, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.02465331278890601, | |
| "grad_norm": 6.110072612762451, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 0.7395, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.02489036387341472, | |
| "grad_norm": 6.735311508178711, | |
| "learning_rate": 5e-05, | |
| "loss": 0.905, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.025127414957923432, | |
| "grad_norm": 8.923399925231934, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 0.8893, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.025364466042432143, | |
| "grad_norm": 8.221948623657227, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 1.6142, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.025601517126940854, | |
| "grad_norm": 8.158872604370117, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 1.5885, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.02583856821144957, | |
| "grad_norm": 7.332838535308838, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 1.0663, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.02607561929595828, | |
| "grad_norm": 7.769575119018555, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 1.2486, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.02631267038046699, | |
| "grad_norm": 7.445107460021973, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 1.4, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.026549721464975702, | |
| "grad_norm": 8.416275978088379, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 1.2037, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.026786772549484413, | |
| "grad_norm": 10.549809455871582, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 1.3073, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.027023823633993124, | |
| "grad_norm": 6.388961315155029, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 0.7881, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.02726087471850184, | |
| "grad_norm": 8.752462387084961, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 1.1797, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.02749792580301055, | |
| "grad_norm": 8.613337516784668, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 1.3528, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.02773497688751926, | |
| "grad_norm": 9.686309814453125, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 1.5578, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.027972027972027972, | |
| "grad_norm": 6.918097496032715, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 1.4049, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.028209079056536683, | |
| "grad_norm": 8.275176048278809, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 0.9878, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.028446130141045394, | |
| "grad_norm": 6.292196273803711, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 1.0339, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.028683181225554105, | |
| "grad_norm": 7.8501505851745605, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 1.453, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.02892023231006282, | |
| "grad_norm": 5.446768760681152, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 0.7386, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.02915728339457153, | |
| "grad_norm": 6.707988739013672, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 0.8834, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.029394334479080242, | |
| "grad_norm": 7.342574596405029, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.9934, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.029631385563588953, | |
| "grad_norm": 7.558194160461426, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 0.8803, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.029868436648097664, | |
| "grad_norm": 6.624544620513916, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 0.8641, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.030105487732606375, | |
| "grad_norm": 8.050422668457031, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 1.3738, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.03034253881711509, | |
| "grad_norm": 8.919082641601562, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 1.2314, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.0305795899016238, | |
| "grad_norm": 33.773231506347656, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 0.6851, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.030816640986132512, | |
| "grad_norm": 10.311910629272461, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 1.3308, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.031053692070641223, | |
| "grad_norm": 8.833625793457031, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 1.1228, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.03129074315514994, | |
| "grad_norm": 7.480016708374023, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 0.944, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.031527794239658645, | |
| "grad_norm": 7.066028118133545, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 0.6924, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.03176484532416736, | |
| "grad_norm": 4.683218479156494, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 0.6277, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.03200189640867607, | |
| "grad_norm": 7.2049078941345215, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 1.0737, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.03223894749318478, | |
| "grad_norm": 9.224268913269043, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 1.0738, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.032475998577693496, | |
| "grad_norm": 10.263715744018555, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 0.9315, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.032713049662202204, | |
| "grad_norm": 9.596421241760254, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 0.9987, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.03295010074671092, | |
| "grad_norm": 7.585577011108398, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 1.1157, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.033187151831219626, | |
| "grad_norm": 6.539881706237793, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 0.9617, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.03342420291572834, | |
| "grad_norm": 9.355979919433594, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 1.1429, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.03366125400023705, | |
| "grad_norm": 9.769394874572754, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 1.2177, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.03389830508474576, | |
| "grad_norm": 10.125142097473145, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 0.7486, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.03413535616925448, | |
| "grad_norm": 8.567256927490234, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 0.8809, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.034372407253763185, | |
| "grad_norm": 9.61689567565918, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 0.8924, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.0346094583382719, | |
| "grad_norm": 8.733377456665039, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 0.7403, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.03484650942278061, | |
| "grad_norm": 9.422379493713379, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 0.7349, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.03508356050728932, | |
| "grad_norm": 7.522639751434326, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 0.978, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.035320611591798036, | |
| "grad_norm": 9.866047859191895, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 1.1348, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.035557662676306744, | |
| "grad_norm": 9.662226676940918, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 1.0662, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.035557662676306744, | |
| "eval_loss": 0.2598925232887268, | |
| "eval_runtime": 516.1483, | |
| "eval_samples_per_second": 13.765, | |
| "eval_steps_per_second": 3.443, | |
| "step": 150 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.150074508503941e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |