| { | |
| "best_metric": 9.881793975830078, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 0.051897502432695426, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00025948751216347713, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1e-05, | |
| "loss": 11.3752, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00025948751216347713, | |
| "eval_loss": 11.260618209838867, | |
| "eval_runtime": 325.6045, | |
| "eval_samples_per_second": 19.935, | |
| "eval_steps_per_second": 4.985, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0005189750243269543, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2e-05, | |
| "loss": 11.5097, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0007784625364904313, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3e-05, | |
| "loss": 10.3879, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0010379500486539085, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4e-05, | |
| "loss": 11.3483, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0012974375608173856, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5e-05, | |
| "loss": 10.8091, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0015569250729808627, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6e-05, | |
| "loss": 12.1155, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.00181641258514434, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7e-05, | |
| "loss": 11.6059, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.002075900097307817, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8e-05, | |
| "loss": 11.3364, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.002335387609471294, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9e-05, | |
| "loss": 10.8013, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0025948751216347712, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0001, | |
| "loss": 11.0769, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0028543626337982483, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 11.611, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0031138501459617254, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 11.6854, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.003373337658125203, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 11.1487, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.00363282517028868, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 10.6093, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.003892312682452157, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 10.8271, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.004151800194615634, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 10.9406, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.004411287706779112, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 11.2148, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.004670775218942588, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 10.5878, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.004930262731106066, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 10.9919, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0051897502432695424, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 11.7233, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.00544923775543302, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 11.4232, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.005708725267596497, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 11.6265, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.005968212779759974, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 12.8659, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.006227700291923451, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 11.8172, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.006487187804086928, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 10.742, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.006746675316250406, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 12.3694, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0070061628284138824, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 10.7448, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.00726565034057736, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 13.0443, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.007525137852740837, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 11.2227, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.007784625364904314, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 9.9599, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.00804411287706779, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 11.6364, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.008303600389231268, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 12.3924, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.008563087901394746, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 12.0712, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.008822575413558223, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 11.2427, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.009082062925721699, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 12.1492, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.009341550437885177, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 12.1757, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.009601037950048654, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 12.1666, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.009860525462212132, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 11.1704, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.010120012974375607, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 11.0543, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.010379500486539085, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 10.9642, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.010638987998702562, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 12.1804, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.01089847551086604, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 9.7104, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.011157963023029517, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 11.2367, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.011417450535192993, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 10.9081, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.01167693804735647, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 10.5013, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.011936425559519948, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 11.1808, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.012195913071683426, | |
| "grad_norm": 5.879723641683313e+18, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 10.4664, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.012455400583846902, | |
| "grad_norm": 5.379835232780288e+16, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 9.1474, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.012714888096010379, | |
| "grad_norm": 2.907693958903628e+18, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 10.4443, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.012974375608173857, | |
| "grad_norm": 2.4829839531751506e+18, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 9.4781, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.012974375608173857, | |
| "eval_loss": 11.092522621154785, | |
| "eval_runtime": 327.7834, | |
| "eval_samples_per_second": 19.803, | |
| "eval_steps_per_second": 4.951, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.013233863120337334, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 11.2168, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.013493350632500812, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 11.127, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.013752838144664287, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 10.7368, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.014012325656827765, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 10.7977, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.014271813168991242, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 10.5346, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.01453130068115472, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 10.682, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.014790788193318197, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 10.6385, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.015050275705481673, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 10.4535, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.01530976321764515, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 10.0309, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.015569250729808628, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 9.9491, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.015828738241972104, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 10.6585, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.01608822575413558, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 10.3761, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.01634771326629906, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 10.1725, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.016607200778462537, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 11.1301, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.016866688290626014, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 10.7923, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.01712617580278949, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 11.3163, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.01738566331495297, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 10.7584, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.017645150827116447, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 10.4221, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.01790463833927992, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 9.8293, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.018164125851443398, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 10.2411, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.018423613363606876, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 10.12, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.018683100875770353, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 10.143, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.01894258838793383, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 10.9414, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.019202075900097308, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 10.6745, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.019461563412260786, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 9.6208, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.019721050924424263, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 9.8623, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.01998053843658774, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 11.0101, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.020240025948751215, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 10.4709, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.020499513460914692, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 11.478, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.02075900097307817, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 9.6984, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.021018488485241647, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 10.6487, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.021277975997405125, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 8.0202, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.021537463509568602, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 9.9757, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.02179695102173208, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 11.4616, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.022056438533895557, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 10.4769, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.022315926046059035, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 9.9165, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.02257541355822251, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 9.4074, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.022834901070385986, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 9.9358, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.023094388582549464, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 11.0703, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.02335387609471294, | |
| "grad_norm": 1.0942056045626786e+19, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 10.2091, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.02361336360687642, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 10.4913, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.023872851119039896, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 8.8221, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.024132338631203374, | |
| "grad_norm": 7.292670987605115e+17, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 9.524, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.02439182614336685, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 9.7004, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.02465131365553033, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 10.6663, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.024910801167693803, | |
| "grad_norm": 1.776532648403927e+17, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 9.1904, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.02517028867985728, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 9.1723, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.025429776192020758, | |
| "grad_norm": 1.522239488917504e+17, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 8.8549, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.025689263704184236, | |
| "grad_norm": 7.908311256216371e+16, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 8.4775, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.025948751216347713, | |
| "grad_norm": 1.1524541434455654e+18, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 8.183, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.025948751216347713, | |
| "eval_loss": 9.895584106445312, | |
| "eval_runtime": 327.4927, | |
| "eval_samples_per_second": 19.82, | |
| "eval_steps_per_second": 4.956, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.02620823872851119, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 9.4452, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.026467726240674668, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 8.8336, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.026727213752838146, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 10.8389, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.026986701265001623, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 10.0369, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.027246188777165097, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5e-05, | |
| "loss": 8.6726, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.027505676289328575, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 9.5929, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.027765163801492052, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 10.0437, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.02802465131365553, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 9.5796, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.028284138825819007, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 9.8764, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.028543626337982485, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 9.21, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.028803113850145962, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 10.0999, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.02906260136230944, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 9.3228, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.029322088874472917, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 8.7136, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.029581576386636395, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 10.5489, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.02984106389879987, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 10.1966, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.030100551410963346, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 9.3291, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.030360038923126824, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 9.6534, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.0306195264352903, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 9.8111, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.03087901394745378, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 8.7442, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.031138501459617256, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 10.389, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.031397988971780734, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 10.7633, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.03165747648394421, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 10.4918, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.03191696399610769, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 9.7887, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.03217645150827116, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 9.4412, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.032435939020434644, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 9.8076, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.03269542653259812, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 10.4094, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.0329549140447616, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 10.1478, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.03321440155692507, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 10.0829, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.03347388906908855, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 9.9107, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.03373337658125203, | |
| "grad_norm": 5.768289237719843e+18, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 9.3743, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.0339928640934155, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 9.6898, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.03425235160557898, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 9.5515, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.03451183911774246, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 9.4834, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.03477132662990594, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 9.498, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.03503081414206941, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 10.0648, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.03529030165423289, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 9.4624, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.03554978916639637, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 9.6034, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.03580927667855984, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 9.1138, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.03606876419072332, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 10.1755, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.036328251702886796, | |
| "grad_norm": 7.506911790349943e+18, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 10.3848, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.03658773921505028, | |
| "grad_norm": 7.040298296975294e+18, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 9.15, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.03684722672721375, | |
| "grad_norm": 1.1955447325645603e+19, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 9.2077, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.03710671423937723, | |
| "grad_norm": 1.6851995916108825e+19, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 9.7299, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.037366201751540706, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 8.4475, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.03762568926370419, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 9.5374, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.03788517677586766, | |
| "grad_norm": 1.2794209770894524e+19, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 9.1347, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.038144664288031135, | |
| "grad_norm": 1.3648688807149568e+17, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 8.5315, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.038404151800194616, | |
| "grad_norm": 9.202144727929979e+17, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 8.2456, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.03866363931235809, | |
| "grad_norm": 2.6924175536443884e+18, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 8.5527, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.03892312682452157, | |
| "grad_norm": 1.018408686583808e+16, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 8.0344, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.03892312682452157, | |
| "eval_loss": 9.881793975830078, | |
| "eval_runtime": 328.0349, | |
| "eval_samples_per_second": 19.788, | |
| "eval_steps_per_second": 4.948, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.039182614336685045, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 9.5692, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.039442101848848526, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 11.052, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.039701589361012, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 10.9694, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.03996107687317548, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 9.7354, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.040220564385338956, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 10.3032, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.04048005189750243, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 9.4296, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.04073953940966591, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 10.3052, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.040999026921829385, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 9.9128, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.041258514433992866, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 9.917, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.04151800194615634, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 11.2278, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.04177748945831982, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 10.6006, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.042036976970483295, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 10.3592, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.042296464482646776, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 9.7825, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.04255595199481025, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 9.8606, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.042815439506973724, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 8.9248, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.043074927019137205, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 10.2305, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.04333441453130068, | |
| "grad_norm": Infinity, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 9.6681, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.04359390204346416, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 9.4768, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.043853389555627634, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 9.4843, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.044112877067791115, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 10.4947, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.04437236457995459, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 11.0011, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.04463185209211807, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 9.4437, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.044891339604281544, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 10.1913, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.04515082711644502, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 10.5443, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.0454103146286085, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 9.9686, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.04566980214077197, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 9.7866, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.045929289652935454, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 10.0205, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.04618877716509893, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 9.7417, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.04644826467726241, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 9.3322, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.04670775218942588, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 8.5783, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.046967239701589364, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 10.4689, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.04722672721375284, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 9.973, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.04748621472591631, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 9.5476, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.04774570223807979, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 9.8646, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.04800518975024327, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 10.186, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.04826467726240675, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 9.271, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.04852416477457022, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 10.0606, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.0487836522867337, | |
| "grad_norm": Infinity, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 10.024, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.04904313979889718, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 9.8846, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.04930262731106066, | |
| "grad_norm": Infinity, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 9.9528, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.04956211482322413, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 10.4708, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.049821602335387606, | |
| "grad_norm": Infinity, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 9.2269, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.05008108984755109, | |
| "grad_norm": 1.0026961929764864e+18, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 9.39, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.05034057735971456, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 9.5845, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.05060006487187804, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 9.365, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.050859552384041516, | |
| "grad_norm": 4.629500305820615e+18, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 8.9018, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.051119039896205, | |
| "grad_norm": 1.2355594241609564e+18, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 9.1315, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.05137852740836847, | |
| "grad_norm": 1.3955587619430597e+19, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 8.0062, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.05163801492053195, | |
| "grad_norm": 1.5268942713936937e+18, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 8.8241, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.051897502432695426, | |
| "grad_norm": 2.4915881814183117e+17, | |
| "learning_rate": 0.0, | |
| "loss": 8.1527, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.051897502432695426, | |
| "eval_loss": 9.936193466186523, | |
| "eval_runtime": 327.7703, | |
| "eval_samples_per_second": 19.804, | |
| "eval_steps_per_second": 4.952, | |
| "step": 200 | |
| } | |
| ], | |
| "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": 1 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.5409973586139546e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |