| { | |
| "best_metric": 1.1310502290725708, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 1.5584415584415585, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01038961038961039, | |
| "grad_norm": 31.83489990234375, | |
| "learning_rate": 1e-05, | |
| "loss": 8.1047, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01038961038961039, | |
| "eval_loss": 3.256531238555908, | |
| "eval_runtime": 17.3266, | |
| "eval_samples_per_second": 9.35, | |
| "eval_steps_per_second": 2.366, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02077922077922078, | |
| "grad_norm": 28.17755126953125, | |
| "learning_rate": 2e-05, | |
| "loss": 7.7535, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03116883116883117, | |
| "grad_norm": 29.767200469970703, | |
| "learning_rate": 3e-05, | |
| "loss": 7.084, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.04155844155844156, | |
| "grad_norm": 24.76503562927246, | |
| "learning_rate": 4e-05, | |
| "loss": 6.9438, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05194805194805195, | |
| "grad_norm": 18.37542152404785, | |
| "learning_rate": 5e-05, | |
| "loss": 6.2551, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06233766233766234, | |
| "grad_norm": 17.238405227661133, | |
| "learning_rate": 6e-05, | |
| "loss": 5.5819, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07272727272727272, | |
| "grad_norm": 13.965852737426758, | |
| "learning_rate": 7e-05, | |
| "loss": 5.6229, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.08311688311688312, | |
| "grad_norm": 12.185733795166016, | |
| "learning_rate": 8e-05, | |
| "loss": 4.335, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.09350649350649351, | |
| "grad_norm": 13.277860641479492, | |
| "learning_rate": 9e-05, | |
| "loss": 4.0514, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1038961038961039, | |
| "grad_norm": 13.240141868591309, | |
| "learning_rate": 0.0001, | |
| "loss": 8.1257, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11428571428571428, | |
| "grad_norm": 14.634415626525879, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 9.6236, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.12467532467532468, | |
| "grad_norm": 11.495798110961914, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 9.9399, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.13506493506493505, | |
| "grad_norm": 12.6912260055542, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 10.0119, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.14545454545454545, | |
| "grad_norm": 13.443987846374512, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 8.9694, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.15584415584415584, | |
| "grad_norm": 12.248916625976562, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 9.4825, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.16623376623376623, | |
| "grad_norm": 16.383142471313477, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 8.9163, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.17662337662337663, | |
| "grad_norm": 14.51679801940918, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 8.3302, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.18701298701298702, | |
| "grad_norm": 16.01449966430664, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 7.9958, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1974025974025974, | |
| "grad_norm": 12.227169036865234, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 7.5001, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2077922077922078, | |
| "grad_norm": 14.246201515197754, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 7.2687, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.21818181818181817, | |
| "grad_norm": 16.64252471923828, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 8.6548, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.22857142857142856, | |
| "grad_norm": 16.917905807495117, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 8.8993, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.23896103896103896, | |
| "grad_norm": 19.88192367553711, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 9.2532, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.24935064935064935, | |
| "grad_norm": 20.945693969726562, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 9.8207, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2597402597402597, | |
| "grad_norm": 51.722171783447266, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 9.3168, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2701298701298701, | |
| "grad_norm": 30.76021957397461, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 7.2464, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2805194805194805, | |
| "grad_norm": 18.43760871887207, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 4.7023, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2909090909090909, | |
| "grad_norm": 9.666842460632324, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 3.2694, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3012987012987013, | |
| "grad_norm": 9.109925270080566, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 2.8659, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.3116883116883117, | |
| "grad_norm": 9.080495834350586, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 3.6459, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3220779220779221, | |
| "grad_norm": 7.356724739074707, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 2.7477, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.33246753246753247, | |
| "grad_norm": 8.920196533203125, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 3.2656, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.34285714285714286, | |
| "grad_norm": 7.383318901062012, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 3.8959, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.35324675324675325, | |
| "grad_norm": 7.815951347351074, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 4.317, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 8.140786170959473, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 8.4663, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.37402597402597404, | |
| "grad_norm": 8.832559585571289, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 8.1094, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.38441558441558443, | |
| "grad_norm": 7.1496806144714355, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 7.9823, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3948051948051948, | |
| "grad_norm": 8.56055736541748, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 7.507, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4051948051948052, | |
| "grad_norm": 10.39842700958252, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 8.0287, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.4155844155844156, | |
| "grad_norm": 10.334918022155762, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 6.7764, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.42597402597402595, | |
| "grad_norm": 10.809833526611328, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 7.6163, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.43636363636363634, | |
| "grad_norm": 11.95631217956543, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 7.7449, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.44675324675324674, | |
| "grad_norm": 12.381674766540527, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 6.9886, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.45714285714285713, | |
| "grad_norm": 12.293375015258789, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 7.1263, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.4675324675324675, | |
| "grad_norm": 11.562889099121094, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 7.0536, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4779220779220779, | |
| "grad_norm": 12.639488220214844, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 7.5769, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4883116883116883, | |
| "grad_norm": 13.091582298278809, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 8.1122, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.4987012987012987, | |
| "grad_norm": 15.97758674621582, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 8.064, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.509090909090909, | |
| "grad_norm": 17.52106475830078, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 4.9188, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "grad_norm": 13.562044143676758, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 4.7467, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "eval_loss": 1.5564213991165161, | |
| "eval_runtime": 17.6774, | |
| "eval_samples_per_second": 9.164, | |
| "eval_steps_per_second": 2.319, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5298701298701298, | |
| "grad_norm": 9.15640926361084, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 2.5358, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5402597402597402, | |
| "grad_norm": 7.217840194702148, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 2.8447, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5506493506493506, | |
| "grad_norm": 7.621275424957275, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 4.0846, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.561038961038961, | |
| "grad_norm": 6.927586078643799, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 2.7463, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 7.786722660064697, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 3.0062, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5818181818181818, | |
| "grad_norm": 7.90624475479126, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 3.631, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5922077922077922, | |
| "grad_norm": 22.610877990722656, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 4.3947, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6025974025974026, | |
| "grad_norm": 8.461299896240234, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 8.2739, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.612987012987013, | |
| "grad_norm": 8.185771942138672, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 8.7998, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6233766233766234, | |
| "grad_norm": 7.691197395324707, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 8.1865, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6337662337662338, | |
| "grad_norm": 8.699455261230469, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 8.2093, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.6441558441558441, | |
| "grad_norm": 8.309798240661621, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 7.5779, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6545454545454545, | |
| "grad_norm": 8.643186569213867, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 6.7183, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6649350649350649, | |
| "grad_norm": 8.525787353515625, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 5.9121, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6753246753246753, | |
| "grad_norm": 9.759805679321289, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 6.8276, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6857142857142857, | |
| "grad_norm": 10.294455528259277, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 6.1688, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6961038961038961, | |
| "grad_norm": 10.0902681350708, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 7.0199, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7064935064935065, | |
| "grad_norm": 14.20248031616211, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 6.6244, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.7168831168831169, | |
| "grad_norm": 12.053751945495605, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 7.9556, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 10.587654113769531, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 6.6079, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7376623376623377, | |
| "grad_norm": 13.080458641052246, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 6.8979, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.7480519480519481, | |
| "grad_norm": 20.203798294067383, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 7.6002, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7584415584415585, | |
| "grad_norm": 10.804618835449219, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 4.3275, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7688311688311689, | |
| "grad_norm": 12.095099449157715, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 3.1685, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "grad_norm": 5.91151762008667, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 2.4299, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7896103896103897, | |
| "grad_norm": 6.148896217346191, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 2.697, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 6.840137481689453, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 2.8022, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8103896103896104, | |
| "grad_norm": 7.766355991363525, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 2.9416, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.8207792207792208, | |
| "grad_norm": 5.920592784881592, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 2.6363, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8311688311688312, | |
| "grad_norm": 6.974241733551025, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 3.4222, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8415584415584415, | |
| "grad_norm": 6.821755886077881, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 2.7414, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.8519480519480519, | |
| "grad_norm": 9.252930641174316, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 6.7547, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.8623376623376623, | |
| "grad_norm": 7.803735733032227, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 8.0916, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.8727272727272727, | |
| "grad_norm": 6.56402587890625, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 7.0716, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.8831168831168831, | |
| "grad_norm": 7.134830951690674, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 7.6986, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8935064935064935, | |
| "grad_norm": 7.818478584289551, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 7.4413, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.9038961038961039, | |
| "grad_norm": 7.542208671569824, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 6.4206, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.9142857142857143, | |
| "grad_norm": 7.44448709487915, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 6.1373, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.9246753246753247, | |
| "grad_norm": 10.097084999084473, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 6.8287, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.935064935064935, | |
| "grad_norm": 12.01990795135498, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 6.7226, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.9454545454545454, | |
| "grad_norm": 9.571053504943848, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 5.2302, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.9558441558441558, | |
| "grad_norm": 11.317047119140625, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 6.8719, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.9662337662337662, | |
| "grad_norm": 10.105382919311523, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 6.5456, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9766233766233766, | |
| "grad_norm": 10.427618980407715, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 6.4078, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.987012987012987, | |
| "grad_norm": 10.727428436279297, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 6.5427, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9974025974025974, | |
| "grad_norm": 14.104158401489258, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 7.616, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0077922077922077, | |
| "grad_norm": 8.729491233825684, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 3.2676, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.018181818181818, | |
| "grad_norm": 14.363706588745117, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 3.5505, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.0285714285714285, | |
| "grad_norm": 9.730070114135742, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 2.8847, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.0389610389610389, | |
| "grad_norm": 6.3668060302734375, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 2.1662, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0389610389610389, | |
| "eval_loss": 1.4060604572296143, | |
| "eval_runtime": 17.6741, | |
| "eval_samples_per_second": 9.166, | |
| "eval_steps_per_second": 2.32, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0493506493506493, | |
| "grad_norm": 4.832397937774658, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 1.9677, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.0597402597402596, | |
| "grad_norm": 5.833955764770508, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 2.3954, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.07012987012987, | |
| "grad_norm": 6.367222785949707, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 2.1797, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.0805194805194804, | |
| "grad_norm": 6.8857598304748535, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 2.8208, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 6.170316696166992, | |
| "learning_rate": 5e-05, | |
| "loss": 2.2855, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1012987012987012, | |
| "grad_norm": 7.0842366218566895, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 4.4597, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.1116883116883116, | |
| "grad_norm": 8.464518547058105, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 7.409, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.122077922077922, | |
| "grad_norm": 8.17165756225586, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 6.0053, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.1324675324675324, | |
| "grad_norm": 9.075725555419922, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 6.5667, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 10.42436695098877, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 6.5322, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.1532467532467532, | |
| "grad_norm": 11.784306526184082, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 5.6909, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.1636363636363636, | |
| "grad_norm": 7.83477258682251, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 4.9676, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.174025974025974, | |
| "grad_norm": 7.5017218589782715, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 5.0569, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.1844155844155844, | |
| "grad_norm": 7.871354579925537, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 5.4064, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.1948051948051948, | |
| "grad_norm": 8.291726112365723, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 5.2221, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2051948051948052, | |
| "grad_norm": 9.132806777954102, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 4.8778, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.2155844155844155, | |
| "grad_norm": 9.86052131652832, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 4.9762, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.225974025974026, | |
| "grad_norm": 9.123804092407227, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 4.8187, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.2363636363636363, | |
| "grad_norm": 10.833556175231934, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 6.0691, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.2467532467532467, | |
| "grad_norm": 14.201516151428223, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 6.4682, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.2571428571428571, | |
| "grad_norm": 10.149537086486816, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 4.3814, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.2675324675324675, | |
| "grad_norm": 9.169793128967285, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 3.09, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.277922077922078, | |
| "grad_norm": 8.979920387268066, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 3.0481, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.2883116883116883, | |
| "grad_norm": 9.17720890045166, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 2.7146, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.2987012987012987, | |
| "grad_norm": 6.523139476776123, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 3.1006, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.309090909090909, | |
| "grad_norm": 5.682675361633301, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 2.1121, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.3194805194805195, | |
| "grad_norm": 5.774822235107422, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 2.5708, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.3298701298701299, | |
| "grad_norm": 4.785064220428467, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 2.2269, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.3402597402597403, | |
| "grad_norm": 5.579965114593506, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 2.6994, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.3506493506493507, | |
| "grad_norm": 7.379159450531006, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 4.8612, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.361038961038961, | |
| "grad_norm": 10.66281509399414, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 7.0442, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.3714285714285714, | |
| "grad_norm": 8.379961967468262, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 6.1418, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.3818181818181818, | |
| "grad_norm": 9.883442878723145, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 5.6191, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.3922077922077922, | |
| "grad_norm": 8.915534973144531, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 4.9791, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.4025974025974026, | |
| "grad_norm": 8.834295272827148, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 5.1164, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.412987012987013, | |
| "grad_norm": 7.788753986358643, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 4.8313, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.4233766233766234, | |
| "grad_norm": 8.902883529663086, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 5.0584, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.4337662337662338, | |
| "grad_norm": 8.027772903442383, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 4.7429, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.4441558441558442, | |
| "grad_norm": 8.277462005615234, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 4.2385, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 9.170880317687988, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 4.6918, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.464935064935065, | |
| "grad_norm": 7.7855000495910645, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 4.2009, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.4753246753246754, | |
| "grad_norm": 10.265680313110352, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 4.6864, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.4857142857142858, | |
| "grad_norm": 11.02685260772705, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 5.5487, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.4961038961038962, | |
| "grad_norm": 14.291165351867676, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 4.987, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.5064935064935066, | |
| "grad_norm": 9.032857894897461, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 3.8984, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.516883116883117, | |
| "grad_norm": 5.082447052001953, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 2.7804, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.5272727272727273, | |
| "grad_norm": 4.521763324737549, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 2.0659, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.5376623376623377, | |
| "grad_norm": 5.00535774230957, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 2.27, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.5480519480519481, | |
| "grad_norm": 5.87699031829834, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 2.4806, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.5584415584415585, | |
| "grad_norm": 5.989110469818115, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 2.7028, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.5584415584415585, | |
| "eval_loss": 1.1310502290725708, | |
| "eval_runtime": 17.6624, | |
| "eval_samples_per_second": 9.172, | |
| "eval_steps_per_second": 2.321, | |
| "step": 150 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "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": 3.4784639604228096e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |