| { | |
| "best_metric": 0.16089697182178497, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 0.01747742499271774, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00011651616661811826, | |
| "grad_norm": 5.975294589996338, | |
| "learning_rate": 1e-05, | |
| "loss": 1.4886, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00011651616661811826, | |
| "eval_loss": 0.7362678647041321, | |
| "eval_runtime": 1251.5529, | |
| "eval_samples_per_second": 11.55, | |
| "eval_steps_per_second": 2.888, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00023303233323623653, | |
| "grad_norm": 7.621860027313232, | |
| "learning_rate": 2e-05, | |
| "loss": 2.022, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0003495484998543548, | |
| "grad_norm": 10.003783226013184, | |
| "learning_rate": 3e-05, | |
| "loss": 2.6357, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.00046606466647247305, | |
| "grad_norm": 7.267496585845947, | |
| "learning_rate": 4e-05, | |
| "loss": 2.8787, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0005825808330905913, | |
| "grad_norm": 8.449346542358398, | |
| "learning_rate": 5e-05, | |
| "loss": 3.2433, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0006990969997087095, | |
| "grad_norm": 7.849472522735596, | |
| "learning_rate": 6e-05, | |
| "loss": 2.7579, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0008156131663268278, | |
| "grad_norm": 9.259416580200195, | |
| "learning_rate": 7e-05, | |
| "loss": 2.8856, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0009321293329449461, | |
| "grad_norm": 9.022052764892578, | |
| "learning_rate": 8e-05, | |
| "loss": 2.5201, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0010486454995630643, | |
| "grad_norm": 10.637661933898926, | |
| "learning_rate": 9e-05, | |
| "loss": 2.4325, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0011651616661811825, | |
| "grad_norm": 20.810359954833984, | |
| "learning_rate": 0.0001, | |
| "loss": 3.6425, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0012816778327993008, | |
| "grad_norm": 22.364402770996094, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 1.2994, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.001398193999417419, | |
| "grad_norm": 21.3902645111084, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 0.9994, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0015147101660355374, | |
| "grad_norm": 20.095746994018555, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 1.4923, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0016312263326536556, | |
| "grad_norm": 19.705101013183594, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 1.3813, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.001747742499271774, | |
| "grad_norm": 10.08316421508789, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 0.8004, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0018642586658898922, | |
| "grad_norm": 6.992054462432861, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 0.6966, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0019807748325080107, | |
| "grad_norm": 7.983057022094727, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 0.6435, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0020972909991261285, | |
| "grad_norm": 10.571928024291992, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 1.3488, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.002213807165744247, | |
| "grad_norm": 7.997575759887695, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 1.0088, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.002330323332362365, | |
| "grad_norm": 7.503167629241943, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 0.8738, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0024468394989804834, | |
| "grad_norm": 6.179886341094971, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 0.7921, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0025633556655986016, | |
| "grad_norm": 6.789474010467529, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 0.7108, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.00267987183221672, | |
| "grad_norm": 6.373968601226807, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 0.6128, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.002796387998834838, | |
| "grad_norm": 5.637046813964844, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 0.5566, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0029129041654529565, | |
| "grad_norm": 7.8370184898376465, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 0.5593, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0030294203320710747, | |
| "grad_norm": 5.988832950592041, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 0.559, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.003145936498689193, | |
| "grad_norm": 7.165755271911621, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 0.2816, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0032624526653073113, | |
| "grad_norm": 4.786063194274902, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 0.4154, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0033789688319254296, | |
| "grad_norm": 4.706677436828613, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.3299, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.003495484998543548, | |
| "grad_norm": 4.422995090484619, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 0.4687, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.003612001165161666, | |
| "grad_norm": 4.65325403213501, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 0.4085, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0037285173317797844, | |
| "grad_norm": 5.4333882331848145, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 0.3464, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0038450334983979027, | |
| "grad_norm": 72.83003997802734, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 0.6495, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.003961549665016021, | |
| "grad_norm": 6.982298851013184, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 0.703, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.00407806583163414, | |
| "grad_norm": 6.191880226135254, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 0.2861, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.004194581998252257, | |
| "grad_norm": 6.210261344909668, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 0.4421, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.004311098164870375, | |
| "grad_norm": 3.793665647506714, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 0.2314, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.004427614331488494, | |
| "grad_norm": 3.907691717147827, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 0.2664, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.004544130498106612, | |
| "grad_norm": 3.6872928142547607, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 0.3018, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.00466064666472473, | |
| "grad_norm": 5.131813049316406, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 0.3065, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0047771628313428485, | |
| "grad_norm": 6.162957191467285, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 0.5872, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.004893678997960967, | |
| "grad_norm": 5.168027877807617, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 0.4656, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.005010195164579085, | |
| "grad_norm": 5.235057830810547, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 0.3406, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.005126711331197203, | |
| "grad_norm": 3.2103614807128906, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 0.2492, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.0052432274978153216, | |
| "grad_norm": 4.117974281311035, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 0.4447, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.00535974366443344, | |
| "grad_norm": 4.725550174713135, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 0.3768, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.005476259831051558, | |
| "grad_norm": 3.656212091445923, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 0.3273, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.005592775997669676, | |
| "grad_norm": 3.0423667430877686, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.1449, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.005709292164287795, | |
| "grad_norm": 5.057799339294434, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 0.3125, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.005825808330905913, | |
| "grad_norm": 4.145799160003662, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 0.2972, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.005825808330905913, | |
| "eval_loss": 0.30729740858078003, | |
| "eval_runtime": 1259.5242, | |
| "eval_samples_per_second": 11.477, | |
| "eval_steps_per_second": 2.869, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.005942324497524031, | |
| "grad_norm": 7.646812438964844, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 1.4163, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.0060588406641421495, | |
| "grad_norm": 6.700872898101807, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 1.4841, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.006175356830760268, | |
| "grad_norm": 6.486911296844482, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 2.3511, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.006291872997378386, | |
| "grad_norm": 20.43660545349121, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 1.9787, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.006408389163996504, | |
| "grad_norm": 7.909752368927002, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 2.3232, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.006524905330614623, | |
| "grad_norm": 5.747597694396973, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 2.2553, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.006641421497232741, | |
| "grad_norm": 5.942625999450684, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 2.6, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.006757937663850859, | |
| "grad_norm": 6.343461513519287, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 2.6722, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.006874453830468977, | |
| "grad_norm": 7.53312873840332, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 2.3428, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.006990969997087096, | |
| "grad_norm": 6.786001682281494, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 0.8934, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.007107486163705214, | |
| "grad_norm": 6.07697057723999, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 0.5482, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.007224002330323332, | |
| "grad_norm": 7.817957401275635, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 0.8168, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.0073405184969414505, | |
| "grad_norm": 10.404017448425293, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 1.0485, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.007457034663559569, | |
| "grad_norm": 6.8413777351379395, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 0.5882, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.007573550830177687, | |
| "grad_norm": 6.327642917633057, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 0.4513, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.007690066996795805, | |
| "grad_norm": 7.955966949462891, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 0.5974, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.007806583163413924, | |
| "grad_norm": 7.306300163269043, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 1.0354, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.007923099330032043, | |
| "grad_norm": 5.990658760070801, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 0.6796, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.00803961549665016, | |
| "grad_norm": 4.4180121421813965, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 0.5267, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.00815613166326828, | |
| "grad_norm": 3.890469551086426, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 0.46, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.008272647829886397, | |
| "grad_norm": 4.6100239753723145, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 0.4944, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.008389163996504514, | |
| "grad_norm": 5.460227966308594, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 0.4118, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.008505680163122633, | |
| "grad_norm": 5.597787857055664, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 0.4636, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.00862219632974075, | |
| "grad_norm": 4.456701755523682, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 0.4148, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.00873871249635887, | |
| "grad_norm": 4.373245716094971, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 0.4957, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.008855228662976987, | |
| "grad_norm": 6.624417781829834, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 0.4018, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.008971744829595106, | |
| "grad_norm": 3.571817398071289, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 0.4453, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.009088260996213224, | |
| "grad_norm": 4.248070240020752, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 0.3584, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.009204777162831343, | |
| "grad_norm": 9.709259033203125, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 0.4822, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.00932129332944946, | |
| "grad_norm": 3.1241631507873535, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 0.2599, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.00943780949606758, | |
| "grad_norm": 4.64263916015625, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 0.4409, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.009554325662685697, | |
| "grad_norm": 4.032393455505371, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 0.3145, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.009670841829303816, | |
| "grad_norm": 5.52547550201416, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 0.64, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.009787357995921933, | |
| "grad_norm": 3.4348347187042236, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 0.222, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.009903874162540053, | |
| "grad_norm": 2.8319878578186035, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 0.3056, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.01002039032915817, | |
| "grad_norm": 1.929783582687378, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.1418, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.01013690649577629, | |
| "grad_norm": 5.322389602661133, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 0.3162, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.010253422662394407, | |
| "grad_norm": 6.367090225219727, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 0.2716, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.010369938829012526, | |
| "grad_norm": 4.7843337059021, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 0.2803, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.010486454995630643, | |
| "grad_norm": 3.9384782314300537, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 0.2986, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.010602971162248762, | |
| "grad_norm": 3.9873108863830566, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 0.1867, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.01071948732886688, | |
| "grad_norm": 5.899569511413574, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 0.4297, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.010836003495484999, | |
| "grad_norm": 3.702864646911621, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 0.3268, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.010952519662103116, | |
| "grad_norm": 3.3945624828338623, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 0.2245, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.011069035828721235, | |
| "grad_norm": 3.3172385692596436, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 0.2557, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.011185551995339353, | |
| "grad_norm": 3.4038662910461426, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 0.2956, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.011302068161957472, | |
| "grad_norm": 3.525500535964966, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 0.1912, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.01141858432857559, | |
| "grad_norm": 4.306077480316162, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 0.3404, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.011535100495193708, | |
| "grad_norm": 3.4483957290649414, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 0.1822, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.011651616661811826, | |
| "grad_norm": 3.101605176925659, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 0.1502, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.011651616661811826, | |
| "eval_loss": 0.2293005883693695, | |
| "eval_runtime": 1260.8588, | |
| "eval_samples_per_second": 11.464, | |
| "eval_steps_per_second": 2.866, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.011768132828429945, | |
| "grad_norm": 4.722489833831787, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 0.84, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.011884648995048062, | |
| "grad_norm": 3.4606871604919434, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 1.0449, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.012001165161666182, | |
| "grad_norm": 4.10048770904541, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 1.6692, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.012117681328284299, | |
| "grad_norm": 3.9006295204162598, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 1.6042, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.012234197494902418, | |
| "grad_norm": 4.057922840118408, | |
| "learning_rate": 5e-05, | |
| "loss": 1.9697, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.012350713661520536, | |
| "grad_norm": 4.708436012268066, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 2.2564, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.012467229828138655, | |
| "grad_norm": 4.315474510192871, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 2.0766, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.012583745994756772, | |
| "grad_norm": 5.208872318267822, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 1.971, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.012700262161374891, | |
| "grad_norm": 5.911382675170898, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 2.4559, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.012816778327993009, | |
| "grad_norm": 9.173274993896484, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 1.7305, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.012933294494611128, | |
| "grad_norm": 9.052083969116211, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 0.7558, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.013049810661229245, | |
| "grad_norm": 3.7256786823272705, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 0.391, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.013166326827847364, | |
| "grad_norm": 6.508945465087891, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 0.8569, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.013282842994465482, | |
| "grad_norm": 5.268301486968994, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 0.602, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.013399359161083601, | |
| "grad_norm": 4.693594932556152, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 0.4195, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.013515875327701718, | |
| "grad_norm": 6.453338623046875, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 0.7095, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.013632391494319837, | |
| "grad_norm": 6.161351680755615, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 0.8745, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.013748907660937955, | |
| "grad_norm": 8.09244155883789, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 0.7898, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.013865423827556074, | |
| "grad_norm": 5.4810662269592285, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 0.5553, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.013981939994174191, | |
| "grad_norm": 6.248891353607178, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 0.5408, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.01409845616079231, | |
| "grad_norm": 3.8639261722564697, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 0.2722, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.014214972327410428, | |
| "grad_norm": 4.835658550262451, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 0.3573, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.014331488494028547, | |
| "grad_norm": 3.8692128658294678, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 0.35, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.014448004660646665, | |
| "grad_norm": 3.284202814102173, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.3006, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.014564520827264784, | |
| "grad_norm": 4.325440883636475, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 0.2835, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.014681036993882901, | |
| "grad_norm": 3.831855297088623, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 0.2502, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.01479755316050102, | |
| "grad_norm": 3.6768546104431152, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 0.3475, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.014914069327119138, | |
| "grad_norm": 5.103757858276367, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 0.4149, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.015030585493737257, | |
| "grad_norm": 3.294356346130371, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 0.3167, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.015147101660355374, | |
| "grad_norm": 3.4960432052612305, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 0.4214, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.015263617826973493, | |
| "grad_norm": 3.2754383087158203, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 0.3199, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.01538013399359161, | |
| "grad_norm": 3.661444664001465, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 0.3551, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.01549665016020973, | |
| "grad_norm": 2.610337972640991, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 0.194, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.015613166326827847, | |
| "grad_norm": 2.5162415504455566, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 0.2553, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.015729682493445966, | |
| "grad_norm": 2.5253489017486572, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 0.1866, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.015846198660064086, | |
| "grad_norm": 2.481363296508789, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 0.1762, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.0159627148266822, | |
| "grad_norm": 2.67958402633667, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 0.1806, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.01607923099330032, | |
| "grad_norm": 5.1212477684021, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 0.5425, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.01619574715991844, | |
| "grad_norm": 4.429457664489746, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 0.4215, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.01631226332653656, | |
| "grad_norm": 2.466294050216675, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 0.2199, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.016428779493154674, | |
| "grad_norm": 4.927537441253662, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 0.3468, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.016545295659772793, | |
| "grad_norm": 3.9179611206054688, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 0.276, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.016661811826390913, | |
| "grad_norm": 2.29433536529541, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 0.1986, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.01677832799300903, | |
| "grad_norm": 3.070161819458008, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 0.3549, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.016894844159627147, | |
| "grad_norm": 2.7507412433624268, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 0.2875, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.017011360326245267, | |
| "grad_norm": 5.55330753326416, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 0.5538, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.017127876492863386, | |
| "grad_norm": 1.976783037185669, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 0.0727, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.0172443926594815, | |
| "grad_norm": 3.006279945373535, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 0.1586, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.01736090882609962, | |
| "grad_norm": 2.96524977684021, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 0.1961, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.01747742499271774, | |
| "grad_norm": 3.3227920532226562, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 0.2915, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.01747742499271774, | |
| "eval_loss": 0.16089697182178497, | |
| "eval_runtime": 1262.2602, | |
| "eval_samples_per_second": 11.452, | |
| "eval_steps_per_second": 2.863, | |
| "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": 2.56148643608789e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |