diff --git "a/epoch_10/checkpoint-3300/trainer_state.json" "b/epoch_10/checkpoint-3300/trainer_state.json" new file mode 100644--- /dev/null +++ "b/epoch_10/checkpoint-3300/trainer_state.json" @@ -0,0 +1,23134 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.75, + "eval_steps": 500, + "global_step": 3300, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00022727272727272727, + "grad_norm": 3.1564557868690186, + "learning_rate": 1.25e-06, + "loss": 0.1583, + "step": 1 + }, + { + "epoch": 0.00045454545454545455, + "grad_norm": 3.0871206953440655, + "learning_rate": 1.249999840689502e-06, + "loss": 0.0746, + "step": 2 + }, + { + "epoch": 0.0006818181818181819, + "grad_norm": 2.8524761685234723, + "learning_rate": 1.2499993627580887e-06, + "loss": 0.1094, + "step": 3 + }, + { + "epoch": 0.0009090909090909091, + "grad_norm": 2.2737654241081366, + "learning_rate": 1.2499985662060041e-06, + "loss": 0.1014, + "step": 4 + }, + { + "epoch": 0.0011363636363636363, + "grad_norm": 2.0708585630946468, + "learning_rate": 1.249997451033654e-06, + "loss": 0.112, + "step": 5 + }, + { + "epoch": 0.0013636363636363637, + "grad_norm": 1.8787087949048937, + "learning_rate": 1.249996017241607e-06, + "loss": 0.0937, + "step": 6 + }, + { + "epoch": 0.001590909090909091, + "grad_norm": 1.8669892625691389, + "learning_rate": 1.249994264830594e-06, + "loss": 0.0735, + "step": 7 + }, + { + "epoch": 0.0018181818181818182, + "grad_norm": 2.1763334417027425, + "learning_rate": 1.2499921938015086e-06, + "loss": 0.0785, + "step": 8 + }, + { + "epoch": 0.0020454545454545456, + "grad_norm": 2.7707875750281823, + "learning_rate": 1.2499898041554066e-06, + "loss": 0.1213, + "step": 9 + }, + { + "epoch": 0.0022727272727272726, + "grad_norm": 2.898646282028457, + "learning_rate": 1.2499870958935056e-06, + "loss": 0.1414, + "step": 10 + }, + { + "epoch": 0.0025, + "grad_norm": 2.760213572746958, + "learning_rate": 1.2499840690171872e-06, + "loss": 0.1505, + "step": 11 + }, + { + "epoch": 0.0027272727272727275, + "grad_norm": 4.526664405862104, + "learning_rate": 1.2499807235279937e-06, + "loss": 0.0927, + "step": 12 + }, + { + "epoch": 0.0029545454545454545, + "grad_norm": 1.1602592455565957, + "learning_rate": 1.249977059427631e-06, + "loss": 0.0988, + "step": 13 + }, + { + "epoch": 0.003181818181818182, + "grad_norm": 2.448196546544741, + "learning_rate": 1.2499730767179668e-06, + "loss": 0.1542, + "step": 14 + }, + { + "epoch": 0.003409090909090909, + "grad_norm": 1.2786979356593762, + "learning_rate": 1.2499687754010318e-06, + "loss": 0.0712, + "step": 15 + }, + { + "epoch": 0.0036363636363636364, + "grad_norm": 1.863521368884654, + "learning_rate": 1.2499641554790185e-06, + "loss": 0.0972, + "step": 16 + }, + { + "epoch": 0.003863636363636364, + "grad_norm": 3.4338545021748463, + "learning_rate": 1.2499592169542823e-06, + "loss": 0.1425, + "step": 17 + }, + { + "epoch": 0.004090909090909091, + "grad_norm": 2.431787184289732, + "learning_rate": 1.2499539598293406e-06, + "loss": 0.0727, + "step": 18 + }, + { + "epoch": 0.004318181818181818, + "grad_norm": 4.09922410750763, + "learning_rate": 1.2499483841068736e-06, + "loss": 0.1478, + "step": 19 + }, + { + "epoch": 0.004545454545454545, + "grad_norm": 1.9180043241114382, + "learning_rate": 1.2499424897897237e-06, + "loss": 0.0626, + "step": 20 + }, + { + "epoch": 0.004772727272727273, + "grad_norm": 3.1401027564219706, + "learning_rate": 1.2499362768808958e-06, + "loss": 0.0821, + "step": 21 + }, + { + "epoch": 0.005, + "grad_norm": 2.5736052104454825, + "learning_rate": 1.2499297453835574e-06, + "loss": 0.2294, + "step": 22 + }, + { + "epoch": 0.005227272727272727, + "grad_norm": 2.1908031698645307, + "learning_rate": 1.2499228953010379e-06, + "loss": 0.068, + "step": 23 + }, + { + "epoch": 0.005454545454545455, + "grad_norm": 2.7986141410716545, + "learning_rate": 1.2499157266368298e-06, + "loss": 0.0694, + "step": 24 + }, + { + "epoch": 0.005681818181818182, + "grad_norm": 2.216881506903992, + "learning_rate": 1.2499082393945871e-06, + "loss": 0.12, + "step": 25 + }, + { + "epoch": 0.005909090909090909, + "grad_norm": 4.466210564333436, + "learning_rate": 1.2499004335781272e-06, + "loss": 0.1879, + "step": 26 + }, + { + "epoch": 0.006136363636363636, + "grad_norm": 2.7418052007687295, + "learning_rate": 1.2498923091914293e-06, + "loss": 0.1108, + "step": 27 + }, + { + "epoch": 0.006363636363636364, + "grad_norm": 1.3142670132507652, + "learning_rate": 1.249883866238635e-06, + "loss": 0.0504, + "step": 28 + }, + { + "epoch": 0.006590909090909091, + "grad_norm": 1.9151950776224913, + "learning_rate": 1.2498751047240488e-06, + "loss": 0.1039, + "step": 29 + }, + { + "epoch": 0.006818181818181818, + "grad_norm": 2.891969345154028, + "learning_rate": 1.2498660246521371e-06, + "loss": 0.1498, + "step": 30 + }, + { + "epoch": 0.007045454545454546, + "grad_norm": 3.9171329520161087, + "learning_rate": 1.2498566260275289e-06, + "loss": 0.2158, + "step": 31 + }, + { + "epoch": 0.007272727272727273, + "grad_norm": 2.1431658757611176, + "learning_rate": 1.2498469088550156e-06, + "loss": 0.1436, + "step": 32 + }, + { + "epoch": 0.0075, + "grad_norm": 2.6610469921166717, + "learning_rate": 1.2498368731395507e-06, + "loss": 0.1393, + "step": 33 + }, + { + "epoch": 0.007727272727272728, + "grad_norm": 2.377264363902789, + "learning_rate": 1.2498265188862505e-06, + "loss": 0.1755, + "step": 34 + }, + { + "epoch": 0.007954545454545454, + "grad_norm": 2.982370727132437, + "learning_rate": 1.2498158461003935e-06, + "loss": 0.1799, + "step": 35 + }, + { + "epoch": 0.008181818181818182, + "grad_norm": 2.0834076454595216, + "learning_rate": 1.2498048547874208e-06, + "loss": 0.0629, + "step": 36 + }, + { + "epoch": 0.00840909090909091, + "grad_norm": 1.6460145378113114, + "learning_rate": 1.2497935449529355e-06, + "loss": 0.1084, + "step": 37 + }, + { + "epoch": 0.008636363636363636, + "grad_norm": 2.822555458451542, + "learning_rate": 1.2497819166027035e-06, + "loss": 0.0875, + "step": 38 + }, + { + "epoch": 0.008863636363636363, + "grad_norm": 2.302417169304832, + "learning_rate": 1.2497699697426523e-06, + "loss": 0.083, + "step": 39 + }, + { + "epoch": 0.00909090909090909, + "grad_norm": 1.9534618045201375, + "learning_rate": 1.2497577043788732e-06, + "loss": 0.0677, + "step": 40 + }, + { + "epoch": 0.009318181818181817, + "grad_norm": 2.137698662950649, + "learning_rate": 1.2497451205176183e-06, + "loss": 0.1114, + "step": 41 + }, + { + "epoch": 0.009545454545454546, + "grad_norm": 1.7768261933514558, + "learning_rate": 1.2497322181653032e-06, + "loss": 0.097, + "step": 42 + }, + { + "epoch": 0.009772727272727273, + "grad_norm": 1.5397860594922692, + "learning_rate": 1.249718997328505e-06, + "loss": 0.0934, + "step": 43 + }, + { + "epoch": 0.01, + "grad_norm": 3.2731744122160866, + "learning_rate": 1.2497054580139642e-06, + "loss": 0.1237, + "step": 44 + }, + { + "epoch": 0.010227272727272727, + "grad_norm": 3.974334828169075, + "learning_rate": 1.2496916002285823e-06, + "loss": 0.1476, + "step": 45 + }, + { + "epoch": 0.010454545454545454, + "grad_norm": 3.088552043173043, + "learning_rate": 1.2496774239794246e-06, + "loss": 0.1153, + "step": 46 + }, + { + "epoch": 0.010681818181818181, + "grad_norm": 2.5693338482707704, + "learning_rate": 1.2496629292737176e-06, + "loss": 0.1505, + "step": 47 + }, + { + "epoch": 0.01090909090909091, + "grad_norm": 2.914658976439222, + "learning_rate": 1.249648116118851e-06, + "loss": 0.0825, + "step": 48 + }, + { + "epoch": 0.011136363636363637, + "grad_norm": 1.5723669201020165, + "learning_rate": 1.2496329845223759e-06, + "loss": 0.1345, + "step": 49 + }, + { + "epoch": 0.011363636363636364, + "grad_norm": 2.049916859226786, + "learning_rate": 1.2496175344920069e-06, + "loss": 0.0693, + "step": 50 + }, + { + "epoch": 0.011590909090909091, + "grad_norm": 2.086372008443818, + "learning_rate": 1.24960176603562e-06, + "loss": 0.0614, + "step": 51 + }, + { + "epoch": 0.011818181818181818, + "grad_norm": 2.4927038659392986, + "learning_rate": 1.2495856791612538e-06, + "loss": 0.1084, + "step": 52 + }, + { + "epoch": 0.012045454545454545, + "grad_norm": 1.8086524849206982, + "learning_rate": 1.2495692738771095e-06, + "loss": 0.121, + "step": 53 + }, + { + "epoch": 0.012272727272727272, + "grad_norm": 1.3817882879062044, + "learning_rate": 1.2495525501915503e-06, + "loss": 0.0597, + "step": 54 + }, + { + "epoch": 0.0125, + "grad_norm": 2.896657660201753, + "learning_rate": 1.2495355081131017e-06, + "loss": 0.1074, + "step": 55 + }, + { + "epoch": 0.012727272727272728, + "grad_norm": 1.6624835583942623, + "learning_rate": 1.249518147650452e-06, + "loss": 0.0797, + "step": 56 + }, + { + "epoch": 0.012954545454545455, + "grad_norm": 2.48719066036232, + "learning_rate": 1.249500468812451e-06, + "loss": 0.1102, + "step": 57 + }, + { + "epoch": 0.013181818181818182, + "grad_norm": 2.6981918257828945, + "learning_rate": 1.2494824716081117e-06, + "loss": 0.1138, + "step": 58 + }, + { + "epoch": 0.013409090909090909, + "grad_norm": 2.037769809610619, + "learning_rate": 1.2494641560466087e-06, + "loss": 0.0855, + "step": 59 + }, + { + "epoch": 0.013636363636363636, + "grad_norm": 4.19876298075569, + "learning_rate": 1.249445522137279e-06, + "loss": 0.1618, + "step": 60 + }, + { + "epoch": 0.013863636363636364, + "grad_norm": 2.985233501757055, + "learning_rate": 1.2494265698896224e-06, + "loss": 0.1315, + "step": 61 + }, + { + "epoch": 0.014090909090909091, + "grad_norm": 3.943719228894905, + "learning_rate": 1.2494072993133005e-06, + "loss": 0.101, + "step": 62 + }, + { + "epoch": 0.014318181818181818, + "grad_norm": 1.5166367196464041, + "learning_rate": 1.2493877104181373e-06, + "loss": 0.0699, + "step": 63 + }, + { + "epoch": 0.014545454545454545, + "grad_norm": 2.785524997719726, + "learning_rate": 1.249367803214119e-06, + "loss": 0.0704, + "step": 64 + }, + { + "epoch": 0.014772727272727272, + "grad_norm": 3.7434422644571996, + "learning_rate": 1.2493475777113945e-06, + "loss": 0.0795, + "step": 65 + }, + { + "epoch": 0.015, + "grad_norm": 2.651551208185626, + "learning_rate": 1.2493270339202742e-06, + "loss": 0.1855, + "step": 66 + }, + { + "epoch": 0.015227272727272726, + "grad_norm": 2.131594310137375, + "learning_rate": 1.2493061718512314e-06, + "loss": 0.0596, + "step": 67 + }, + { + "epoch": 0.015454545454545455, + "grad_norm": 3.5637959978757188, + "learning_rate": 1.2492849915149013e-06, + "loss": 0.1314, + "step": 68 + }, + { + "epoch": 0.015681818181818182, + "grad_norm": 3.067312708438428, + "learning_rate": 1.2492634929220817e-06, + "loss": 0.0977, + "step": 69 + }, + { + "epoch": 0.015909090909090907, + "grad_norm": 4.795853215029535, + "learning_rate": 1.2492416760837326e-06, + "loss": 0.1065, + "step": 70 + }, + { + "epoch": 0.016136363636363636, + "grad_norm": 2.7085485675098893, + "learning_rate": 1.2492195410109757e-06, + "loss": 0.202, + "step": 71 + }, + { + "epoch": 0.016363636363636365, + "grad_norm": 2.220924336670223, + "learning_rate": 1.2491970877150955e-06, + "loss": 0.1259, + "step": 72 + }, + { + "epoch": 0.01659090909090909, + "grad_norm": 4.382863674488229, + "learning_rate": 1.2491743162075384e-06, + "loss": 0.0927, + "step": 73 + }, + { + "epoch": 0.01681818181818182, + "grad_norm": 3.0024931990622723, + "learning_rate": 1.2491512264999135e-06, + "loss": 0.1347, + "step": 74 + }, + { + "epoch": 0.017045454545454544, + "grad_norm": 2.2776832807746774, + "learning_rate": 1.2491278186039916e-06, + "loss": 0.1098, + "step": 75 + }, + { + "epoch": 0.017272727272727273, + "grad_norm": 2.2894350646089867, + "learning_rate": 1.2491040925317057e-06, + "loss": 0.0898, + "step": 76 + }, + { + "epoch": 0.0175, + "grad_norm": 5.139266732236653, + "learning_rate": 1.2490800482951515e-06, + "loss": 0.2217, + "step": 77 + }, + { + "epoch": 0.017727272727272727, + "grad_norm": 2.6591556970939196, + "learning_rate": 1.2490556859065865e-06, + "loss": 0.1037, + "step": 78 + }, + { + "epoch": 0.017954545454545456, + "grad_norm": 0.6899563031613882, + "learning_rate": 1.2490310053784301e-06, + "loss": 0.0394, + "step": 79 + }, + { + "epoch": 0.01818181818181818, + "grad_norm": 1.750014384494401, + "learning_rate": 1.249006006723265e-06, + "loss": 0.0443, + "step": 80 + }, + { + "epoch": 0.01840909090909091, + "grad_norm": 3.6255161610135915, + "learning_rate": 1.2489806899538349e-06, + "loss": 0.1413, + "step": 81 + }, + { + "epoch": 0.018636363636363635, + "grad_norm": 1.6647213381152102, + "learning_rate": 1.2489550550830462e-06, + "loss": 0.0789, + "step": 82 + }, + { + "epoch": 0.018863636363636364, + "grad_norm": 2.511141928773147, + "learning_rate": 1.2489291021239674e-06, + "loss": 0.1646, + "step": 83 + }, + { + "epoch": 0.019090909090909092, + "grad_norm": 4.057508003278786, + "learning_rate": 1.2489028310898293e-06, + "loss": 0.109, + "step": 84 + }, + { + "epoch": 0.019318181818181818, + "grad_norm": 2.3478641229707944, + "learning_rate": 1.2488762419940244e-06, + "loss": 0.1639, + "step": 85 + }, + { + "epoch": 0.019545454545454546, + "grad_norm": 3.227542896767507, + "learning_rate": 1.248849334850108e-06, + "loss": 0.1302, + "step": 86 + }, + { + "epoch": 0.01977272727272727, + "grad_norm": 1.4309512769932444, + "learning_rate": 1.2488221096717967e-06, + "loss": 0.0749, + "step": 87 + }, + { + "epoch": 0.02, + "grad_norm": 2.6539459562469814, + "learning_rate": 1.2487945664729703e-06, + "loss": 0.1174, + "step": 88 + }, + { + "epoch": 0.020227272727272726, + "grad_norm": 2.9633208868780625, + "learning_rate": 1.2487667052676699e-06, + "loss": 0.1693, + "step": 89 + }, + { + "epoch": 0.020454545454545454, + "grad_norm": 2.5940064856818648, + "learning_rate": 1.2487385260700987e-06, + "loss": 0.1559, + "step": 90 + }, + { + "epoch": 0.020681818181818183, + "grad_norm": 1.4986977005682232, + "learning_rate": 1.2487100288946228e-06, + "loss": 0.0894, + "step": 91 + }, + { + "epoch": 0.02090909090909091, + "grad_norm": 2.770064610473978, + "learning_rate": 1.2486812137557693e-06, + "loss": 0.0787, + "step": 92 + }, + { + "epoch": 0.021136363636363637, + "grad_norm": 3.5155481343826374, + "learning_rate": 1.2486520806682283e-06, + "loss": 0.1111, + "step": 93 + }, + { + "epoch": 0.021363636363636362, + "grad_norm": 3.2397105678231286, + "learning_rate": 1.248622629646852e-06, + "loss": 0.1673, + "step": 94 + }, + { + "epoch": 0.02159090909090909, + "grad_norm": 4.485071390736067, + "learning_rate": 1.2485928607066537e-06, + "loss": 0.1313, + "step": 95 + }, + { + "epoch": 0.02181818181818182, + "grad_norm": 2.868784239959084, + "learning_rate": 1.2485627738628097e-06, + "loss": 0.1766, + "step": 96 + }, + { + "epoch": 0.022045454545454545, + "grad_norm": 1.9572548550585382, + "learning_rate": 1.248532369130658e-06, + "loss": 0.0701, + "step": 97 + }, + { + "epoch": 0.022272727272727274, + "grad_norm": 3.562549040586559, + "learning_rate": 1.2485016465256987e-06, + "loss": 0.0929, + "step": 98 + }, + { + "epoch": 0.0225, + "grad_norm": 3.0769713343074203, + "learning_rate": 1.2484706060635945e-06, + "loss": 0.0713, + "step": 99 + }, + { + "epoch": 0.022727272727272728, + "grad_norm": 4.307852803631376, + "learning_rate": 1.248439247760169e-06, + "loss": 0.1084, + "step": 100 + }, + { + "epoch": 0.022954545454545453, + "grad_norm": 4.869491313841007, + "learning_rate": 1.2484075716314085e-06, + "loss": 0.0797, + "step": 101 + }, + { + "epoch": 0.023181818181818182, + "grad_norm": 3.1659915462648813, + "learning_rate": 1.2483755776934616e-06, + "loss": 0.216, + "step": 102 + }, + { + "epoch": 0.02340909090909091, + "grad_norm": 2.250688009576782, + "learning_rate": 1.2483432659626384e-06, + "loss": 0.0816, + "step": 103 + }, + { + "epoch": 0.023636363636363636, + "grad_norm": 2.112637705852539, + "learning_rate": 1.2483106364554115e-06, + "loss": 0.1052, + "step": 104 + }, + { + "epoch": 0.023863636363636365, + "grad_norm": 3.466753301527032, + "learning_rate": 1.248277689188415e-06, + "loss": 0.0873, + "step": 105 + }, + { + "epoch": 0.02409090909090909, + "grad_norm": 2.6080421255934123, + "learning_rate": 1.248244424178445e-06, + "loss": 0.1938, + "step": 106 + }, + { + "epoch": 0.02431818181818182, + "grad_norm": 5.17559784844226, + "learning_rate": 1.2482108414424602e-06, + "loss": 0.1054, + "step": 107 + }, + { + "epoch": 0.024545454545454544, + "grad_norm": 3.3553491682740755, + "learning_rate": 1.2481769409975805e-06, + "loss": 0.1261, + "step": 108 + }, + { + "epoch": 0.024772727272727273, + "grad_norm": 2.967140573926061, + "learning_rate": 1.2481427228610881e-06, + "loss": 0.0776, + "step": 109 + }, + { + "epoch": 0.025, + "grad_norm": 2.259269672745337, + "learning_rate": 1.2481081870504278e-06, + "loss": 0.0998, + "step": 110 + }, + { + "epoch": 0.025227272727272727, + "grad_norm": 2.77211471027501, + "learning_rate": 1.2480733335832052e-06, + "loss": 0.0603, + "step": 111 + }, + { + "epoch": 0.025454545454545455, + "grad_norm": 2.3290232368520987, + "learning_rate": 1.2480381624771882e-06, + "loss": 0.0599, + "step": 112 + }, + { + "epoch": 0.02568181818181818, + "grad_norm": 3.309853760155977, + "learning_rate": 1.2480026737503073e-06, + "loss": 0.155, + "step": 113 + }, + { + "epoch": 0.02590909090909091, + "grad_norm": 4.067381132717093, + "learning_rate": 1.2479668674206543e-06, + "loss": 0.1333, + "step": 114 + }, + { + "epoch": 0.026136363636363635, + "grad_norm": 2.332432589859219, + "learning_rate": 1.2479307435064827e-06, + "loss": 0.0771, + "step": 115 + }, + { + "epoch": 0.026363636363636363, + "grad_norm": 1.4462895712083488, + "learning_rate": 1.2478943020262087e-06, + "loss": 0.0914, + "step": 116 + }, + { + "epoch": 0.026590909090909092, + "grad_norm": 1.853218151783297, + "learning_rate": 1.2478575429984097e-06, + "loss": 0.0991, + "step": 117 + }, + { + "epoch": 0.026818181818181817, + "grad_norm": 2.1321542365873096, + "learning_rate": 1.2478204664418254e-06, + "loss": 0.0828, + "step": 118 + }, + { + "epoch": 0.027045454545454546, + "grad_norm": 5.7750606868415195, + "learning_rate": 1.2477830723753567e-06, + "loss": 0.1457, + "step": 119 + }, + { + "epoch": 0.02727272727272727, + "grad_norm": 2.5984710080875724, + "learning_rate": 1.2477453608180673e-06, + "loss": 0.0709, + "step": 120 + }, + { + "epoch": 0.0275, + "grad_norm": 2.5476777232607075, + "learning_rate": 1.2477073317891822e-06, + "loss": 0.1472, + "step": 121 + }, + { + "epoch": 0.02772727272727273, + "grad_norm": 3.6137631809940656, + "learning_rate": 1.2476689853080883e-06, + "loss": 0.0813, + "step": 122 + }, + { + "epoch": 0.027954545454545454, + "grad_norm": 2.3243197327455807, + "learning_rate": 1.2476303213943346e-06, + "loss": 0.0851, + "step": 123 + }, + { + "epoch": 0.028181818181818183, + "grad_norm": 3.0761678542607904, + "learning_rate": 1.2475913400676314e-06, + "loss": 0.0948, + "step": 124 + }, + { + "epoch": 0.028409090909090908, + "grad_norm": 6.155367368201849, + "learning_rate": 1.2475520413478516e-06, + "loss": 0.2232, + "step": 125 + }, + { + "epoch": 0.028636363636363637, + "grad_norm": 3.0633068628188846, + "learning_rate": 1.247512425255029e-06, + "loss": 0.0692, + "step": 126 + }, + { + "epoch": 0.028863636363636362, + "grad_norm": 4.061351369979323, + "learning_rate": 1.2474724918093594e-06, + "loss": 0.1308, + "step": 127 + }, + { + "epoch": 0.02909090909090909, + "grad_norm": 2.338316158740006, + "learning_rate": 1.2474322410312012e-06, + "loss": 0.0802, + "step": 128 + }, + { + "epoch": 0.02931818181818182, + "grad_norm": 3.5437216396921865, + "learning_rate": 1.247391672941074e-06, + "loss": 0.0787, + "step": 129 + }, + { + "epoch": 0.029545454545454545, + "grad_norm": 3.0196728035741787, + "learning_rate": 1.2473507875596586e-06, + "loss": 0.1619, + "step": 130 + }, + { + "epoch": 0.029772727272727274, + "grad_norm": 2.9764561040359188, + "learning_rate": 1.2473095849077984e-06, + "loss": 0.1352, + "step": 131 + }, + { + "epoch": 0.03, + "grad_norm": 2.871503655115075, + "learning_rate": 1.2472680650064984e-06, + "loss": 0.137, + "step": 132 + }, + { + "epoch": 0.030227272727272728, + "grad_norm": 3.449733673285195, + "learning_rate": 1.247226227876925e-06, + "loss": 0.1507, + "step": 133 + }, + { + "epoch": 0.030454545454545453, + "grad_norm": 1.4229134277451103, + "learning_rate": 1.2471840735404066e-06, + "loss": 0.0537, + "step": 134 + }, + { + "epoch": 0.03068181818181818, + "grad_norm": 5.21547359860219, + "learning_rate": 1.2471416020184332e-06, + "loss": 0.1219, + "step": 135 + }, + { + "epoch": 0.03090909090909091, + "grad_norm": 3.9463925409166736, + "learning_rate": 1.2470988133326564e-06, + "loss": 0.1006, + "step": 136 + }, + { + "epoch": 0.031136363636363636, + "grad_norm": 2.028557481884059, + "learning_rate": 1.2470557075048897e-06, + "loss": 0.0806, + "step": 137 + }, + { + "epoch": 0.031363636363636364, + "grad_norm": 1.3511272458735586, + "learning_rate": 1.2470122845571081e-06, + "loss": 0.0381, + "step": 138 + }, + { + "epoch": 0.03159090909090909, + "grad_norm": 2.240995192644135, + "learning_rate": 1.2469685445114486e-06, + "loss": 0.0778, + "step": 139 + }, + { + "epoch": 0.031818181818181815, + "grad_norm": 2.2989914292484133, + "learning_rate": 1.2469244873902089e-06, + "loss": 0.1052, + "step": 140 + }, + { + "epoch": 0.032045454545454544, + "grad_norm": 1.9668113409612322, + "learning_rate": 1.2468801132158499e-06, + "loss": 0.0794, + "step": 141 + }, + { + "epoch": 0.03227272727272727, + "grad_norm": 3.339777447340142, + "learning_rate": 1.2468354220109926e-06, + "loss": 0.1062, + "step": 142 + }, + { + "epoch": 0.0325, + "grad_norm": 1.6068963394967977, + "learning_rate": 1.2467904137984208e-06, + "loss": 0.0719, + "step": 143 + }, + { + "epoch": 0.03272727272727273, + "grad_norm": 3.8214218982394095, + "learning_rate": 1.246745088601079e-06, + "loss": 0.081, + "step": 144 + }, + { + "epoch": 0.03295454545454545, + "grad_norm": 4.306724243856774, + "learning_rate": 1.246699446442074e-06, + "loss": 0.2016, + "step": 145 + }, + { + "epoch": 0.03318181818181818, + "grad_norm": 3.3488764528414436, + "learning_rate": 1.2466534873446738e-06, + "loss": 0.1674, + "step": 146 + }, + { + "epoch": 0.03340909090909091, + "grad_norm": 3.340918188093561, + "learning_rate": 1.246607211332308e-06, + "loss": 0.1423, + "step": 147 + }, + { + "epoch": 0.03363636363636364, + "grad_norm": 4.298208459135629, + "learning_rate": 1.2465606184285679e-06, + "loss": 0.1084, + "step": 148 + }, + { + "epoch": 0.03386363636363637, + "grad_norm": 4.83332401437178, + "learning_rate": 1.2465137086572057e-06, + "loss": 0.1378, + "step": 149 + }, + { + "epoch": 0.03409090909090909, + "grad_norm": 2.0100987856536676, + "learning_rate": 1.2464664820421365e-06, + "loss": 0.0992, + "step": 150 + }, + { + "epoch": 0.03431818181818182, + "grad_norm": 2.2440363044043385, + "learning_rate": 1.246418938607436e-06, + "loss": 0.1303, + "step": 151 + }, + { + "epoch": 0.034545454545454546, + "grad_norm": 3.025573774710823, + "learning_rate": 1.246371078377341e-06, + "loss": 0.0974, + "step": 152 + }, + { + "epoch": 0.034772727272727275, + "grad_norm": 1.937901805124536, + "learning_rate": 1.246322901376251e-06, + "loss": 0.0973, + "step": 153 + }, + { + "epoch": 0.035, + "grad_norm": 3.1041211804266986, + "learning_rate": 1.2462744076287257e-06, + "loss": 0.0931, + "step": 154 + }, + { + "epoch": 0.035227272727272725, + "grad_norm": 3.607430386694668, + "learning_rate": 1.2462255971594874e-06, + "loss": 0.0905, + "step": 155 + }, + { + "epoch": 0.035454545454545454, + "grad_norm": 2.8474368927478846, + "learning_rate": 1.2461764699934192e-06, + "loss": 0.0778, + "step": 156 + }, + { + "epoch": 0.03568181818181818, + "grad_norm": 4.692528793624785, + "learning_rate": 1.2461270261555657e-06, + "loss": 0.1346, + "step": 157 + }, + { + "epoch": 0.03590909090909091, + "grad_norm": 5.1932548605192705, + "learning_rate": 1.246077265671133e-06, + "loss": 0.1588, + "step": 158 + }, + { + "epoch": 0.03613636363636363, + "grad_norm": 2.235652523117108, + "learning_rate": 1.2460271885654891e-06, + "loss": 0.0814, + "step": 159 + }, + { + "epoch": 0.03636363636363636, + "grad_norm": 1.4366983962818436, + "learning_rate": 1.2459767948641626e-06, + "loss": 0.0497, + "step": 160 + }, + { + "epoch": 0.03659090909090909, + "grad_norm": 3.3244431186602115, + "learning_rate": 1.2459260845928438e-06, + "loss": 0.1346, + "step": 161 + }, + { + "epoch": 0.03681818181818182, + "grad_norm": 4.421323767971664, + "learning_rate": 1.245875057777385e-06, + "loss": 0.1625, + "step": 162 + }, + { + "epoch": 0.03704545454545455, + "grad_norm": 2.202743342981436, + "learning_rate": 1.245823714443799e-06, + "loss": 0.0641, + "step": 163 + }, + { + "epoch": 0.03727272727272727, + "grad_norm": 3.1490163814852314, + "learning_rate": 1.24577205461826e-06, + "loss": 0.0915, + "step": 164 + }, + { + "epoch": 0.0375, + "grad_norm": 3.90617230224661, + "learning_rate": 1.2457200783271044e-06, + "loss": 0.1211, + "step": 165 + }, + { + "epoch": 0.03772727272727273, + "grad_norm": 3.5673136863497583, + "learning_rate": 1.245667785596829e-06, + "loss": 0.1931, + "step": 166 + }, + { + "epoch": 0.037954545454545456, + "grad_norm": 3.662037324018389, + "learning_rate": 1.2456151764540924e-06, + "loss": 0.1059, + "step": 167 + }, + { + "epoch": 0.038181818181818185, + "grad_norm": 5.2427862043265465, + "learning_rate": 1.2455622509257147e-06, + "loss": 0.1561, + "step": 168 + }, + { + "epoch": 0.03840909090909091, + "grad_norm": 3.813217911826591, + "learning_rate": 1.2455090090386765e-06, + "loss": 0.147, + "step": 169 + }, + { + "epoch": 0.038636363636363635, + "grad_norm": 5.979161844721714, + "learning_rate": 1.2454554508201205e-06, + "loss": 0.081, + "step": 170 + }, + { + "epoch": 0.038863636363636364, + "grad_norm": 3.297904441228195, + "learning_rate": 1.2454015762973505e-06, + "loss": 0.127, + "step": 171 + }, + { + "epoch": 0.03909090909090909, + "grad_norm": 2.4111997938022394, + "learning_rate": 1.2453473854978307e-06, + "loss": 0.0901, + "step": 172 + }, + { + "epoch": 0.03931818181818182, + "grad_norm": 2.9178197728538184, + "learning_rate": 1.2452928784491877e-06, + "loss": 0.1601, + "step": 173 + }, + { + "epoch": 0.03954545454545454, + "grad_norm": 2.2838805088879695, + "learning_rate": 1.245238055179209e-06, + "loss": 0.1106, + "step": 174 + }, + { + "epoch": 0.03977272727272727, + "grad_norm": 4.225610243550824, + "learning_rate": 1.245182915715843e-06, + "loss": 0.1188, + "step": 175 + }, + { + "epoch": 0.04, + "grad_norm": 6.132241816785104, + "learning_rate": 1.2451274600871991e-06, + "loss": 0.1397, + "step": 176 + }, + { + "epoch": 0.04022727272727273, + "grad_norm": 2.2994994620518385, + "learning_rate": 1.245071688321549e-06, + "loss": 0.0712, + "step": 177 + }, + { + "epoch": 0.04045454545454545, + "grad_norm": 4.457853679200538, + "learning_rate": 1.2450156004473238e-06, + "loss": 0.221, + "step": 178 + }, + { + "epoch": 0.04068181818181818, + "grad_norm": 1.4608180509638122, + "learning_rate": 1.2449591964931173e-06, + "loss": 0.1186, + "step": 179 + }, + { + "epoch": 0.04090909090909091, + "grad_norm": 3.488167641269867, + "learning_rate": 1.2449024764876841e-06, + "loss": 0.1287, + "step": 180 + }, + { + "epoch": 0.04113636363636364, + "grad_norm": 2.8836309634419797, + "learning_rate": 1.2448454404599393e-06, + "loss": 0.19, + "step": 181 + }, + { + "epoch": 0.041363636363636366, + "grad_norm": 3.007058803650201, + "learning_rate": 1.2447880884389597e-06, + "loss": 0.0736, + "step": 182 + }, + { + "epoch": 0.04159090909090909, + "grad_norm": 4.660901649803379, + "learning_rate": 1.2447304204539827e-06, + "loss": 0.0864, + "step": 183 + }, + { + "epoch": 0.04181818181818182, + "grad_norm": 3.359526252715459, + "learning_rate": 1.2446724365344076e-06, + "loss": 0.0803, + "step": 184 + }, + { + "epoch": 0.042045454545454546, + "grad_norm": 4.972634500324508, + "learning_rate": 1.2446141367097936e-06, + "loss": 0.1098, + "step": 185 + }, + { + "epoch": 0.042272727272727274, + "grad_norm": 3.6931508786255116, + "learning_rate": 1.244555521009862e-06, + "loss": 0.124, + "step": 186 + }, + { + "epoch": 0.0425, + "grad_norm": 4.126558171996079, + "learning_rate": 1.2444965894644946e-06, + "loss": 0.12, + "step": 187 + }, + { + "epoch": 0.042727272727272725, + "grad_norm": 4.717973383855569, + "learning_rate": 1.2444373421037345e-06, + "loss": 0.177, + "step": 188 + }, + { + "epoch": 0.042954545454545454, + "grad_norm": 2.691838829669658, + "learning_rate": 1.2443777789577852e-06, + "loss": 0.0881, + "step": 189 + }, + { + "epoch": 0.04318181818181818, + "grad_norm": 3.081374121810987, + "learning_rate": 1.244317900057012e-06, + "loss": 0.0772, + "step": 190 + }, + { + "epoch": 0.04340909090909091, + "grad_norm": 3.0459745454991687, + "learning_rate": 1.2442577054319405e-06, + "loss": 0.1122, + "step": 191 + }, + { + "epoch": 0.04363636363636364, + "grad_norm": 4.226915945079344, + "learning_rate": 1.2441971951132578e-06, + "loss": 0.1152, + "step": 192 + }, + { + "epoch": 0.04386363636363636, + "grad_norm": 3.016846350804669, + "learning_rate": 1.2441363691318114e-06, + "loss": 0.0972, + "step": 193 + }, + { + "epoch": 0.04409090909090909, + "grad_norm": 2.922147572450178, + "learning_rate": 1.2440752275186102e-06, + "loss": 0.0676, + "step": 194 + }, + { + "epoch": 0.04431818181818182, + "grad_norm": 2.9933027146166435, + "learning_rate": 1.244013770304824e-06, + "loss": 0.1054, + "step": 195 + }, + { + "epoch": 0.04454545454545455, + "grad_norm": 1.6272047352465335, + "learning_rate": 1.2439519975217828e-06, + "loss": 0.0358, + "step": 196 + }, + { + "epoch": 0.04477272727272727, + "grad_norm": 4.001149751281069, + "learning_rate": 1.2438899092009783e-06, + "loss": 0.1005, + "step": 197 + }, + { + "epoch": 0.045, + "grad_norm": 2.442023342894882, + "learning_rate": 1.2438275053740624e-06, + "loss": 0.0875, + "step": 198 + }, + { + "epoch": 0.04522727272727273, + "grad_norm": 2.3726057341901057, + "learning_rate": 1.2437647860728487e-06, + "loss": 0.0798, + "step": 199 + }, + { + "epoch": 0.045454545454545456, + "grad_norm": 3.7627291366778053, + "learning_rate": 1.2437017513293107e-06, + "loss": 0.1381, + "step": 200 + }, + { + "epoch": 0.045681818181818185, + "grad_norm": 4.404542465289423, + "learning_rate": 1.243638401175583e-06, + "loss": 0.0983, + "step": 201 + }, + { + "epoch": 0.045909090909090906, + "grad_norm": 2.5741320063902444, + "learning_rate": 1.2435747356439614e-06, + "loss": 0.1584, + "step": 202 + }, + { + "epoch": 0.046136363636363635, + "grad_norm": 1.8778235613476721, + "learning_rate": 1.2435107547669022e-06, + "loss": 0.0628, + "step": 203 + }, + { + "epoch": 0.046363636363636364, + "grad_norm": 3.8848862850347725, + "learning_rate": 1.2434464585770226e-06, + "loss": 0.0752, + "step": 204 + }, + { + "epoch": 0.04659090909090909, + "grad_norm": 2.4163364429527836, + "learning_rate": 1.2433818471070998e-06, + "loss": 0.1294, + "step": 205 + }, + { + "epoch": 0.04681818181818182, + "grad_norm": 4.851886053457437, + "learning_rate": 1.2433169203900726e-06, + "loss": 0.2216, + "step": 206 + }, + { + "epoch": 0.04704545454545454, + "grad_norm": 4.043858256143806, + "learning_rate": 1.2432516784590405e-06, + "loss": 0.1007, + "step": 207 + }, + { + "epoch": 0.04727272727272727, + "grad_norm": 4.598823628105842, + "learning_rate": 1.243186121347263e-06, + "loss": 0.1299, + "step": 208 + }, + { + "epoch": 0.0475, + "grad_norm": 1.3564985485137444, + "learning_rate": 1.243120249088161e-06, + "loss": 0.0348, + "step": 209 + }, + { + "epoch": 0.04772727272727273, + "grad_norm": 2.7173202718214085, + "learning_rate": 1.2430540617153156e-06, + "loss": 0.0777, + "step": 210 + }, + { + "epoch": 0.04795454545454545, + "grad_norm": 1.9123604147804552, + "learning_rate": 1.2429875592624685e-06, + "loss": 0.1068, + "step": 211 + }, + { + "epoch": 0.04818181818181818, + "grad_norm": 2.984878058237683, + "learning_rate": 1.2429207417635226e-06, + "loss": 0.117, + "step": 212 + }, + { + "epoch": 0.04840909090909091, + "grad_norm": 3.747706279792091, + "learning_rate": 1.242853609252541e-06, + "loss": 0.1706, + "step": 213 + }, + { + "epoch": 0.04863636363636364, + "grad_norm": 3.9985415806012465, + "learning_rate": 1.2427861617637472e-06, + "loss": 0.1684, + "step": 214 + }, + { + "epoch": 0.048863636363636366, + "grad_norm": 2.6300406906354277, + "learning_rate": 1.2427183993315256e-06, + "loss": 0.0607, + "step": 215 + }, + { + "epoch": 0.04909090909090909, + "grad_norm": 2.217014265436301, + "learning_rate": 1.2426503219904213e-06, + "loss": 0.0899, + "step": 216 + }, + { + "epoch": 0.04931818181818182, + "grad_norm": 3.165836427930639, + "learning_rate": 1.242581929775139e-06, + "loss": 0.1, + "step": 217 + }, + { + "epoch": 0.049545454545454545, + "grad_norm": 3.9852536006190165, + "learning_rate": 1.2425132227205456e-06, + "loss": 0.1146, + "step": 218 + }, + { + "epoch": 0.049772727272727274, + "grad_norm": 2.1481742805579542, + "learning_rate": 1.2424442008616667e-06, + "loss": 0.0677, + "step": 219 + }, + { + "epoch": 0.05, + "grad_norm": 2.1785322536157983, + "learning_rate": 1.2423748642336894e-06, + "loss": 0.0552, + "step": 220 + }, + { + "epoch": 0.050227272727272725, + "grad_norm": 2.722599219763063, + "learning_rate": 1.2423052128719611e-06, + "loss": 0.0801, + "step": 221 + }, + { + "epoch": 0.05045454545454545, + "grad_norm": 2.801639397882677, + "learning_rate": 1.24223524681199e-06, + "loss": 0.1094, + "step": 222 + }, + { + "epoch": 0.05068181818181818, + "grad_norm": 3.1931318042189387, + "learning_rate": 1.2421649660894438e-06, + "loss": 0.1316, + "step": 223 + }, + { + "epoch": 0.05090909090909091, + "grad_norm": 3.560889234046392, + "learning_rate": 1.2420943707401514e-06, + "loss": 0.0868, + "step": 224 + }, + { + "epoch": 0.05113636363636364, + "grad_norm": 2.871586263263952, + "learning_rate": 1.2420234608001017e-06, + "loss": 0.178, + "step": 225 + }, + { + "epoch": 0.05136363636363636, + "grad_norm": 3.3558917259297374, + "learning_rate": 1.2419522363054446e-06, + "loss": 0.1124, + "step": 226 + }, + { + "epoch": 0.05159090909090909, + "grad_norm": 2.653370112966987, + "learning_rate": 1.2418806972924893e-06, + "loss": 0.0639, + "step": 227 + }, + { + "epoch": 0.05181818181818182, + "grad_norm": 2.318706061305113, + "learning_rate": 1.2418088437977063e-06, + "loss": 0.1034, + "step": 228 + }, + { + "epoch": 0.05204545454545455, + "grad_norm": 1.603067220826378, + "learning_rate": 1.241736675857726e-06, + "loss": 0.0531, + "step": 229 + }, + { + "epoch": 0.05227272727272727, + "grad_norm": 3.5221313850695357, + "learning_rate": 1.241664193509339e-06, + "loss": 0.0705, + "step": 230 + }, + { + "epoch": 0.0525, + "grad_norm": 2.9115542611733316, + "learning_rate": 1.2415913967894966e-06, + "loss": 0.1676, + "step": 231 + }, + { + "epoch": 0.05272727272727273, + "grad_norm": 5.2387329766984285, + "learning_rate": 1.2415182857353098e-06, + "loss": 0.1835, + "step": 232 + }, + { + "epoch": 0.052954545454545456, + "grad_norm": 2.6925585427755108, + "learning_rate": 1.2414448603840504e-06, + "loss": 0.1428, + "step": 233 + }, + { + "epoch": 0.053181818181818184, + "grad_norm": 2.4112938787700346, + "learning_rate": 1.24137112077315e-06, + "loss": 0.1466, + "step": 234 + }, + { + "epoch": 0.053409090909090906, + "grad_norm": 3.612503049076161, + "learning_rate": 1.2412970669402005e-06, + "loss": 0.2306, + "step": 235 + }, + { + "epoch": 0.053636363636363635, + "grad_norm": 2.9353124286458474, + "learning_rate": 1.2412226989229542e-06, + "loss": 0.1645, + "step": 236 + }, + { + "epoch": 0.053863636363636364, + "grad_norm": 3.0972610435351724, + "learning_rate": 1.2411480167593237e-06, + "loss": 0.0899, + "step": 237 + }, + { + "epoch": 0.05409090909090909, + "grad_norm": 3.024935159985577, + "learning_rate": 1.241073020487381e-06, + "loss": 0.1422, + "step": 238 + }, + { + "epoch": 0.05431818181818182, + "grad_norm": 1.6842156495364666, + "learning_rate": 1.2409977101453591e-06, + "loss": 0.0629, + "step": 239 + }, + { + "epoch": 0.05454545454545454, + "grad_norm": 2.997434405589183, + "learning_rate": 1.2409220857716506e-06, + "loss": 0.1086, + "step": 240 + }, + { + "epoch": 0.05477272727272727, + "grad_norm": 4.95026820635043, + "learning_rate": 1.2408461474048083e-06, + "loss": 0.1407, + "step": 241 + }, + { + "epoch": 0.055, + "grad_norm": 5.36764802412918, + "learning_rate": 1.240769895083545e-06, + "loss": 0.1281, + "step": 242 + }, + { + "epoch": 0.05522727272727273, + "grad_norm": 4.49029573820107, + "learning_rate": 1.2406933288467337e-06, + "loss": 0.253, + "step": 243 + }, + { + "epoch": 0.05545454545454546, + "grad_norm": 2.0848040362493068, + "learning_rate": 1.2406164487334077e-06, + "loss": 0.1165, + "step": 244 + }, + { + "epoch": 0.05568181818181818, + "grad_norm": 3.847975248165923, + "learning_rate": 1.2405392547827594e-06, + "loss": 0.1036, + "step": 245 + }, + { + "epoch": 0.05590909090909091, + "grad_norm": 2.980948774517248, + "learning_rate": 1.2404617470341423e-06, + "loss": 0.0895, + "step": 246 + }, + { + "epoch": 0.05613636363636364, + "grad_norm": 4.224887043568875, + "learning_rate": 1.2403839255270693e-06, + "loss": 0.1178, + "step": 247 + }, + { + "epoch": 0.056363636363636366, + "grad_norm": 3.5498600106406712, + "learning_rate": 1.2403057903012128e-06, + "loss": 0.1154, + "step": 248 + }, + { + "epoch": 0.05659090909090909, + "grad_norm": 3.4621989317160615, + "learning_rate": 1.240227341396406e-06, + "loss": 0.0909, + "step": 249 + }, + { + "epoch": 0.056818181818181816, + "grad_norm": 1.3590722330326725, + "learning_rate": 1.2401485788526418e-06, + "loss": 0.0845, + "step": 250 + }, + { + "epoch": 0.057045454545454545, + "grad_norm": 2.9462631300946054, + "learning_rate": 1.2400695027100725e-06, + "loss": 0.1211, + "step": 251 + }, + { + "epoch": 0.057272727272727274, + "grad_norm": 3.238051336111907, + "learning_rate": 1.2399901130090112e-06, + "loss": 0.1434, + "step": 252 + }, + { + "epoch": 0.0575, + "grad_norm": 3.604496531024588, + "learning_rate": 1.2399104097899295e-06, + "loss": 0.1815, + "step": 253 + }, + { + "epoch": 0.057727272727272724, + "grad_norm": 1.9942030163646114, + "learning_rate": 1.2398303930934601e-06, + "loss": 0.0736, + "step": 254 + }, + { + "epoch": 0.05795454545454545, + "grad_norm": 3.6088574724314495, + "learning_rate": 1.2397500629603948e-06, + "loss": 0.1003, + "step": 255 + }, + { + "epoch": 0.05818181818181818, + "grad_norm": 2.805828529161223, + "learning_rate": 1.2396694194316851e-06, + "loss": 0.077, + "step": 256 + }, + { + "epoch": 0.05840909090909091, + "grad_norm": 3.43534343443617, + "learning_rate": 1.2395884625484433e-06, + "loss": 0.2032, + "step": 257 + }, + { + "epoch": 0.05863636363636364, + "grad_norm": 4.178732290307962, + "learning_rate": 1.2395071923519403e-06, + "loss": 0.144, + "step": 258 + }, + { + "epoch": 0.05886363636363636, + "grad_norm": 2.831807463605748, + "learning_rate": 1.2394256088836069e-06, + "loss": 0.0978, + "step": 259 + }, + { + "epoch": 0.05909090909090909, + "grad_norm": 2.3657900224583917, + "learning_rate": 1.2393437121850342e-06, + "loss": 0.1267, + "step": 260 + }, + { + "epoch": 0.05931818181818182, + "grad_norm": 3.5073361352341523, + "learning_rate": 1.2392615022979723e-06, + "loss": 0.1194, + "step": 261 + }, + { + "epoch": 0.05954545454545455, + "grad_norm": 3.5313427432809266, + "learning_rate": 1.2391789792643317e-06, + "loss": 0.1054, + "step": 262 + }, + { + "epoch": 0.059772727272727276, + "grad_norm": 2.382859154728883, + "learning_rate": 1.2390961431261814e-06, + "loss": 0.1299, + "step": 263 + }, + { + "epoch": 0.06, + "grad_norm": 2.1468709471911644, + "learning_rate": 1.2390129939257515e-06, + "loss": 0.1156, + "step": 264 + }, + { + "epoch": 0.060227272727272727, + "grad_norm": 3.3016022546150334, + "learning_rate": 1.2389295317054306e-06, + "loss": 0.1323, + "step": 265 + }, + { + "epoch": 0.060454545454545455, + "grad_norm": 4.731490369570349, + "learning_rate": 1.238845756507767e-06, + "loss": 0.1367, + "step": 266 + }, + { + "epoch": 0.060681818181818184, + "grad_norm": 3.4209347943679775, + "learning_rate": 1.2387616683754691e-06, + "loss": 0.1027, + "step": 267 + }, + { + "epoch": 0.060909090909090906, + "grad_norm": 2.350963720331163, + "learning_rate": 1.2386772673514044e-06, + "loss": 0.0575, + "step": 268 + }, + { + "epoch": 0.061136363636363635, + "grad_norm": 2.9880295035280993, + "learning_rate": 1.2385925534786e-06, + "loss": 0.1085, + "step": 269 + }, + { + "epoch": 0.06136363636363636, + "grad_norm": 2.56285122384255, + "learning_rate": 1.2385075268002423e-06, + "loss": 0.0793, + "step": 270 + }, + { + "epoch": 0.06159090909090909, + "grad_norm": 2.3937810091687886, + "learning_rate": 1.2384221873596775e-06, + "loss": 0.0735, + "step": 271 + }, + { + "epoch": 0.06181818181818182, + "grad_norm": 2.5001507655210014, + "learning_rate": 1.2383365352004111e-06, + "loss": 0.091, + "step": 272 + }, + { + "epoch": 0.06204545454545454, + "grad_norm": 3.700184736568469, + "learning_rate": 1.238250570366108e-06, + "loss": 0.1512, + "step": 273 + }, + { + "epoch": 0.06227272727272727, + "grad_norm": 3.6453160558956768, + "learning_rate": 1.2381642929005927e-06, + "loss": 0.1285, + "step": 274 + }, + { + "epoch": 0.0625, + "grad_norm": 5.266382511826849, + "learning_rate": 1.238077702847849e-06, + "loss": 0.1588, + "step": 275 + }, + { + "epoch": 0.06272727272727273, + "grad_norm": 3.545403410529095, + "learning_rate": 1.2379908002520191e-06, + "loss": 0.1662, + "step": 276 + }, + { + "epoch": 0.06295454545454546, + "grad_norm": 1.6938809103905639, + "learning_rate": 1.2379035851574063e-06, + "loss": 0.0613, + "step": 277 + }, + { + "epoch": 0.06318181818181819, + "grad_norm": 2.575887005944469, + "learning_rate": 1.237816057608472e-06, + "loss": 0.1428, + "step": 278 + }, + { + "epoch": 0.06340909090909091, + "grad_norm": 4.516399201106766, + "learning_rate": 1.2377282176498371e-06, + "loss": 0.0772, + "step": 279 + }, + { + "epoch": 0.06363636363636363, + "grad_norm": 3.424446945691615, + "learning_rate": 1.2376400653262817e-06, + "loss": 0.1653, + "step": 280 + }, + { + "epoch": 0.06386363636363636, + "grad_norm": 2.3338340326091975, + "learning_rate": 1.237551600682746e-06, + "loss": 0.0652, + "step": 281 + }, + { + "epoch": 0.06409090909090909, + "grad_norm": 3.674262565265678, + "learning_rate": 1.237462823764328e-06, + "loss": 0.1507, + "step": 282 + }, + { + "epoch": 0.06431818181818182, + "grad_norm": 2.8097098717075837, + "learning_rate": 1.2373737346162857e-06, + "loss": 0.0783, + "step": 283 + }, + { + "epoch": 0.06454545454545454, + "grad_norm": 2.1134022943660886, + "learning_rate": 1.2372843332840364e-06, + "loss": 0.0508, + "step": 284 + }, + { + "epoch": 0.06477272727272727, + "grad_norm": 2.941982472133499, + "learning_rate": 1.2371946198131563e-06, + "loss": 0.0746, + "step": 285 + }, + { + "epoch": 0.065, + "grad_norm": 2.6655175306751206, + "learning_rate": 1.2371045942493804e-06, + "loss": 0.1341, + "step": 286 + }, + { + "epoch": 0.06522727272727273, + "grad_norm": 2.889228330006131, + "learning_rate": 1.2370142566386038e-06, + "loss": 0.1226, + "step": 287 + }, + { + "epoch": 0.06545454545454546, + "grad_norm": 2.475509412428948, + "learning_rate": 1.2369236070268795e-06, + "loss": 0.0771, + "step": 288 + }, + { + "epoch": 0.06568181818181819, + "grad_norm": 2.8734234433410064, + "learning_rate": 1.2368326454604201e-06, + "loss": 0.1377, + "step": 289 + }, + { + "epoch": 0.0659090909090909, + "grad_norm": 2.96918137109757, + "learning_rate": 1.2367413719855976e-06, + "loss": 0.0622, + "step": 290 + }, + { + "epoch": 0.06613636363636363, + "grad_norm": 2.0985807506742797, + "learning_rate": 1.2366497866489423e-06, + "loss": 0.0607, + "step": 291 + }, + { + "epoch": 0.06636363636363636, + "grad_norm": 1.6736951160467164, + "learning_rate": 1.236557889497144e-06, + "loss": 0.0637, + "step": 292 + }, + { + "epoch": 0.06659090909090909, + "grad_norm": 2.91353804544837, + "learning_rate": 1.236465680577051e-06, + "loss": 0.0953, + "step": 293 + }, + { + "epoch": 0.06681818181818182, + "grad_norm": 1.9139717897460742, + "learning_rate": 1.2363731599356712e-06, + "loss": 0.1109, + "step": 294 + }, + { + "epoch": 0.06704545454545455, + "grad_norm": 3.3861661956749303, + "learning_rate": 1.2362803276201709e-06, + "loss": 0.2097, + "step": 295 + }, + { + "epoch": 0.06727272727272728, + "grad_norm": 3.49955831825288, + "learning_rate": 1.2361871836778755e-06, + "loss": 0.0737, + "step": 296 + }, + { + "epoch": 0.0675, + "grad_norm": 4.109601016905457, + "learning_rate": 1.236093728156269e-06, + "loss": 0.1312, + "step": 297 + }, + { + "epoch": 0.06772727272727273, + "grad_norm": 3.9447522756754574, + "learning_rate": 1.2359999611029944e-06, + "loss": 0.1285, + "step": 298 + }, + { + "epoch": 0.06795454545454545, + "grad_norm": 3.5777088957585335, + "learning_rate": 1.2359058825658534e-06, + "loss": 0.094, + "step": 299 + }, + { + "epoch": 0.06818181818181818, + "grad_norm": 3.24614699904347, + "learning_rate": 1.2358114925928073e-06, + "loss": 0.1062, + "step": 300 + }, + { + "epoch": 0.0684090909090909, + "grad_norm": 4.2885475809761, + "learning_rate": 1.2357167912319747e-06, + "loss": 0.1294, + "step": 301 + }, + { + "epoch": 0.06863636363636363, + "grad_norm": 5.024114239014274, + "learning_rate": 1.2356217785316344e-06, + "loss": 0.0944, + "step": 302 + }, + { + "epoch": 0.06886363636363636, + "grad_norm": 2.7271177890337337, + "learning_rate": 1.235526454540223e-06, + "loss": 0.1642, + "step": 303 + }, + { + "epoch": 0.06909090909090909, + "grad_norm": 3.0461067706851375, + "learning_rate": 1.2354308193063358e-06, + "loss": 0.1035, + "step": 304 + }, + { + "epoch": 0.06931818181818182, + "grad_norm": 3.897760951805898, + "learning_rate": 1.2353348728787274e-06, + "loss": 0.1534, + "step": 305 + }, + { + "epoch": 0.06954545454545455, + "grad_norm": 4.59091057607257, + "learning_rate": 1.2352386153063107e-06, + "loss": 0.0949, + "step": 306 + }, + { + "epoch": 0.06977272727272728, + "grad_norm": 4.685536648799799, + "learning_rate": 1.2351420466381566e-06, + "loss": 0.1191, + "step": 307 + }, + { + "epoch": 0.07, + "grad_norm": 4.1690535040432195, + "learning_rate": 1.235045166923496e-06, + "loss": 0.136, + "step": 308 + }, + { + "epoch": 0.07022727272727272, + "grad_norm": 1.9966520119641828, + "learning_rate": 1.2349479762117171e-06, + "loss": 0.0882, + "step": 309 + }, + { + "epoch": 0.07045454545454545, + "grad_norm": 3.1351608130766118, + "learning_rate": 1.2348504745523673e-06, + "loss": 0.1607, + "step": 310 + }, + { + "epoch": 0.07068181818181818, + "grad_norm": 3.8684992555244704, + "learning_rate": 1.2347526619951523e-06, + "loss": 0.1974, + "step": 311 + }, + { + "epoch": 0.07090909090909091, + "grad_norm": 3.4173493046863705, + "learning_rate": 1.2346545385899358e-06, + "loss": 0.1091, + "step": 312 + }, + { + "epoch": 0.07113636363636364, + "grad_norm": 3.9107300294820084, + "learning_rate": 1.2345561043867413e-06, + "loss": 0.1624, + "step": 313 + }, + { + "epoch": 0.07136363636363637, + "grad_norm": 4.65440679481081, + "learning_rate": 1.2344573594357493e-06, + "loss": 0.1092, + "step": 314 + }, + { + "epoch": 0.0715909090909091, + "grad_norm": 2.318004893532808, + "learning_rate": 1.2343583037872998e-06, + "loss": 0.0524, + "step": 315 + }, + { + "epoch": 0.07181818181818182, + "grad_norm": 2.468838769041821, + "learning_rate": 1.2342589374918905e-06, + "loss": 0.1709, + "step": 316 + }, + { + "epoch": 0.07204545454545455, + "grad_norm": 2.4083330429805017, + "learning_rate": 1.2341592606001777e-06, + "loss": 0.1566, + "step": 317 + }, + { + "epoch": 0.07227272727272727, + "grad_norm": 2.5980442872540404, + "learning_rate": 1.2340592731629758e-06, + "loss": 0.0865, + "step": 318 + }, + { + "epoch": 0.0725, + "grad_norm": 3.2649125347809145, + "learning_rate": 1.2339589752312581e-06, + "loss": 0.1097, + "step": 319 + }, + { + "epoch": 0.07272727272727272, + "grad_norm": 3.0138364809855376, + "learning_rate": 1.233858366856156e-06, + "loss": 0.1277, + "step": 320 + }, + { + "epoch": 0.07295454545454545, + "grad_norm": 2.9719665774949133, + "learning_rate": 1.2337574480889585e-06, + "loss": 0.1352, + "step": 321 + }, + { + "epoch": 0.07318181818181818, + "grad_norm": 2.6159876608694366, + "learning_rate": 1.2336562189811138e-06, + "loss": 0.077, + "step": 322 + }, + { + "epoch": 0.07340909090909091, + "grad_norm": 2.9798690182317817, + "learning_rate": 1.2335546795842276e-06, + "loss": 0.0994, + "step": 323 + }, + { + "epoch": 0.07363636363636364, + "grad_norm": 4.36952254070313, + "learning_rate": 1.233452829950064e-06, + "loss": 0.1089, + "step": 324 + }, + { + "epoch": 0.07386363636363637, + "grad_norm": 3.983941617998435, + "learning_rate": 1.2333506701305453e-06, + "loss": 0.1086, + "step": 325 + }, + { + "epoch": 0.0740909090909091, + "grad_norm": 2.942877892081033, + "learning_rate": 1.2332482001777522e-06, + "loss": 0.1187, + "step": 326 + }, + { + "epoch": 0.07431818181818182, + "grad_norm": 4.829523635851263, + "learning_rate": 1.233145420143923e-06, + "loss": 0.1388, + "step": 327 + }, + { + "epoch": 0.07454545454545454, + "grad_norm": 3.4238852452941098, + "learning_rate": 1.2330423300814542e-06, + "loss": 0.1621, + "step": 328 + }, + { + "epoch": 0.07477272727272727, + "grad_norm": 2.4085626464128738, + "learning_rate": 1.2329389300429008e-06, + "loss": 0.0884, + "step": 329 + }, + { + "epoch": 0.075, + "grad_norm": 1.5646102261458528, + "learning_rate": 1.232835220080975e-06, + "loss": 0.0394, + "step": 330 + }, + { + "epoch": 0.07522727272727273, + "grad_norm": 3.2692833754571473, + "learning_rate": 1.232731200248548e-06, + "loss": 0.1507, + "step": 331 + }, + { + "epoch": 0.07545454545454545, + "grad_norm": 5.056602519711096, + "learning_rate": 1.232626870598648e-06, + "loss": 0.1145, + "step": 332 + }, + { + "epoch": 0.07568181818181818, + "grad_norm": 3.9084820724279523, + "learning_rate": 1.2325222311844617e-06, + "loss": 0.2015, + "step": 333 + }, + { + "epoch": 0.07590909090909091, + "grad_norm": 3.410622960920858, + "learning_rate": 1.2324172820593339e-06, + "loss": 0.0986, + "step": 334 + }, + { + "epoch": 0.07613636363636364, + "grad_norm": 3.9624702950417316, + "learning_rate": 1.2323120232767667e-06, + "loss": 0.1273, + "step": 335 + }, + { + "epoch": 0.07636363636363637, + "grad_norm": 2.955000733026103, + "learning_rate": 1.2322064548904202e-06, + "loss": 0.0875, + "step": 336 + }, + { + "epoch": 0.07659090909090908, + "grad_norm": 4.903064091889978, + "learning_rate": 1.232100576954113e-06, + "loss": 0.1367, + "step": 337 + }, + { + "epoch": 0.07681818181818181, + "grad_norm": 1.916111414455401, + "learning_rate": 1.2319943895218205e-06, + "loss": 0.0972, + "step": 338 + }, + { + "epoch": 0.07704545454545454, + "grad_norm": 2.084553789495333, + "learning_rate": 1.2318878926476765e-06, + "loss": 0.1488, + "step": 339 + }, + { + "epoch": 0.07727272727272727, + "grad_norm": 3.0390232417733265, + "learning_rate": 1.2317810863859728e-06, + "loss": 0.1023, + "step": 340 + }, + { + "epoch": 0.0775, + "grad_norm": 2.566586044086669, + "learning_rate": 1.231673970791158e-06, + "loss": 0.0825, + "step": 341 + }, + { + "epoch": 0.07772727272727273, + "grad_norm": 1.4143761259409724, + "learning_rate": 1.231566545917839e-06, + "loss": 0.0678, + "step": 342 + }, + { + "epoch": 0.07795454545454546, + "grad_norm": 3.695680197725065, + "learning_rate": 1.2314588118207808e-06, + "loss": 0.1733, + "step": 343 + }, + { + "epoch": 0.07818181818181819, + "grad_norm": 3.584500933397159, + "learning_rate": 1.2313507685549054e-06, + "loss": 0.0783, + "step": 344 + }, + { + "epoch": 0.07840909090909091, + "grad_norm": 2.699699383437907, + "learning_rate": 1.231242416175292e-06, + "loss": 0.0769, + "step": 345 + }, + { + "epoch": 0.07863636363636364, + "grad_norm": 8.433938005021442, + "learning_rate": 1.2311337547371785e-06, + "loss": 0.3417, + "step": 346 + }, + { + "epoch": 0.07886363636363636, + "grad_norm": 3.210046372468604, + "learning_rate": 1.2310247842959597e-06, + "loss": 0.1642, + "step": 347 + }, + { + "epoch": 0.07909090909090909, + "grad_norm": 2.3937539000754535, + "learning_rate": 1.230915504907188e-06, + "loss": 0.0772, + "step": 348 + }, + { + "epoch": 0.07931818181818182, + "grad_norm": 4.382301194492111, + "learning_rate": 1.2308059166265734e-06, + "loss": 0.151, + "step": 349 + }, + { + "epoch": 0.07954545454545454, + "grad_norm": 2.1183386290201933, + "learning_rate": 1.2306960195099833e-06, + "loss": 0.0647, + "step": 350 + }, + { + "epoch": 0.07977272727272727, + "grad_norm": 2.914854238929018, + "learning_rate": 1.2305858136134422e-06, + "loss": 0.1442, + "step": 351 + }, + { + "epoch": 0.08, + "grad_norm": 2.880009628674947, + "learning_rate": 1.2304752989931327e-06, + "loss": 0.0989, + "step": 352 + }, + { + "epoch": 0.08022727272727273, + "grad_norm": 1.7499362501646205, + "learning_rate": 1.2303644757053945e-06, + "loss": 0.0739, + "step": 353 + }, + { + "epoch": 0.08045454545454546, + "grad_norm": 2.3710050840314563, + "learning_rate": 1.2302533438067247e-06, + "loss": 0.1101, + "step": 354 + }, + { + "epoch": 0.08068181818181819, + "grad_norm": 2.775229157421502, + "learning_rate": 1.230141903353777e-06, + "loss": 0.0975, + "step": 355 + }, + { + "epoch": 0.0809090909090909, + "grad_norm": 2.8777692899564413, + "learning_rate": 1.2300301544033636e-06, + "loss": 0.0919, + "step": 356 + }, + { + "epoch": 0.08113636363636363, + "grad_norm": 2.8727841883067575, + "learning_rate": 1.2299180970124533e-06, + "loss": 0.1527, + "step": 357 + }, + { + "epoch": 0.08136363636363636, + "grad_norm": 1.422509800090171, + "learning_rate": 1.2298057312381723e-06, + "loss": 0.0579, + "step": 358 + }, + { + "epoch": 0.08159090909090909, + "grad_norm": 2.973743441610179, + "learning_rate": 1.2296930571378035e-06, + "loss": 0.0834, + "step": 359 + }, + { + "epoch": 0.08181818181818182, + "grad_norm": 4.312132603239237, + "learning_rate": 1.229580074768788e-06, + "loss": 0.1906, + "step": 360 + }, + { + "epoch": 0.08204545454545455, + "grad_norm": 3.5531200393827804, + "learning_rate": 1.2294667841887234e-06, + "loss": 0.1326, + "step": 361 + }, + { + "epoch": 0.08227272727272728, + "grad_norm": 2.8912038290653834, + "learning_rate": 1.2293531854553642e-06, + "loss": 0.0602, + "step": 362 + }, + { + "epoch": 0.0825, + "grad_norm": 2.0460550512427873, + "learning_rate": 1.2292392786266225e-06, + "loss": 0.0972, + "step": 363 + }, + { + "epoch": 0.08272727272727273, + "grad_norm": 3.0491624100309953, + "learning_rate": 1.2291250637605672e-06, + "loss": 0.068, + "step": 364 + }, + { + "epoch": 0.08295454545454546, + "grad_norm": 2.645099862357967, + "learning_rate": 1.2290105409154244e-06, + "loss": 0.0582, + "step": 365 + }, + { + "epoch": 0.08318181818181818, + "grad_norm": 3.23588490313887, + "learning_rate": 1.2288957101495772e-06, + "loss": 0.0726, + "step": 366 + }, + { + "epoch": 0.0834090909090909, + "grad_norm": 2.718702366667663, + "learning_rate": 1.2287805715215651e-06, + "loss": 0.087, + "step": 367 + }, + { + "epoch": 0.08363636363636363, + "grad_norm": 2.5177340514280035, + "learning_rate": 1.2286651250900858e-06, + "loss": 0.0692, + "step": 368 + }, + { + "epoch": 0.08386363636363636, + "grad_norm": 3.565939235709026, + "learning_rate": 1.2285493709139925e-06, + "loss": 0.1349, + "step": 369 + }, + { + "epoch": 0.08409090909090909, + "grad_norm": 2.5115314513567752, + "learning_rate": 1.2284333090522962e-06, + "loss": 0.0561, + "step": 370 + }, + { + "epoch": 0.08431818181818182, + "grad_norm": 3.5773653486237906, + "learning_rate": 1.2283169395641647e-06, + "loss": 0.0834, + "step": 371 + }, + { + "epoch": 0.08454545454545455, + "grad_norm": 2.34166758968944, + "learning_rate": 1.228200262508922e-06, + "loss": 0.108, + "step": 372 + }, + { + "epoch": 0.08477272727272728, + "grad_norm": 4.6462223685616255, + "learning_rate": 1.2280832779460494e-06, + "loss": 0.1493, + "step": 373 + }, + { + "epoch": 0.085, + "grad_norm": 1.831905947789295, + "learning_rate": 1.2279659859351853e-06, + "loss": 0.0753, + "step": 374 + }, + { + "epoch": 0.08522727272727272, + "grad_norm": 4.2849863677899025, + "learning_rate": 1.2278483865361239e-06, + "loss": 0.0992, + "step": 375 + }, + { + "epoch": 0.08545454545454545, + "grad_norm": 1.4117935293833792, + "learning_rate": 1.227730479808817e-06, + "loss": 0.0341, + "step": 376 + }, + { + "epoch": 0.08568181818181818, + "grad_norm": 5.713125552144446, + "learning_rate": 1.2276122658133723e-06, + "loss": 0.1282, + "step": 377 + }, + { + "epoch": 0.08590909090909091, + "grad_norm": 3.697554767267387, + "learning_rate": 1.2274937446100548e-06, + "loss": 0.0993, + "step": 378 + }, + { + "epoch": 0.08613636363636364, + "grad_norm": 3.524641522952936, + "learning_rate": 1.227374916259286e-06, + "loss": 0.0936, + "step": 379 + }, + { + "epoch": 0.08636363636363636, + "grad_norm": 2.335828502970838, + "learning_rate": 1.2272557808216433e-06, + "loss": 0.0648, + "step": 380 + }, + { + "epoch": 0.0865909090909091, + "grad_norm": 3.445574949871101, + "learning_rate": 1.2271363383578619e-06, + "loss": 0.0715, + "step": 381 + }, + { + "epoch": 0.08681818181818182, + "grad_norm": 2.181025688884454, + "learning_rate": 1.2270165889288325e-06, + "loss": 0.112, + "step": 382 + }, + { + "epoch": 0.08704545454545455, + "grad_norm": 2.5603501910266417, + "learning_rate": 1.2268965325956022e-06, + "loss": 0.0623, + "step": 383 + }, + { + "epoch": 0.08727272727272728, + "grad_norm": 3.306333508413848, + "learning_rate": 1.2267761694193752e-06, + "loss": 0.0814, + "step": 384 + }, + { + "epoch": 0.0875, + "grad_norm": 4.433098920527244, + "learning_rate": 1.2266554994615121e-06, + "loss": 0.1043, + "step": 385 + }, + { + "epoch": 0.08772727272727272, + "grad_norm": 4.258435056932738, + "learning_rate": 1.2265345227835295e-06, + "loss": 0.1707, + "step": 386 + }, + { + "epoch": 0.08795454545454545, + "grad_norm": 4.216978904374584, + "learning_rate": 1.2264132394471007e-06, + "loss": 0.1638, + "step": 387 + }, + { + "epoch": 0.08818181818181818, + "grad_norm": 2.7287499547986336, + "learning_rate": 1.226291649514055e-06, + "loss": 0.0535, + "step": 388 + }, + { + "epoch": 0.08840909090909091, + "grad_norm": 3.983021252773258, + "learning_rate": 1.226169753046378e-06, + "loss": 0.1665, + "step": 389 + }, + { + "epoch": 0.08863636363636364, + "grad_norm": 1.5800728762904996, + "learning_rate": 1.2260475501062121e-06, + "loss": 0.0609, + "step": 390 + }, + { + "epoch": 0.08886363636363637, + "grad_norm": 3.992922742817948, + "learning_rate": 1.2259250407558553e-06, + "loss": 0.1, + "step": 391 + }, + { + "epoch": 0.0890909090909091, + "grad_norm": 2.779682545458947, + "learning_rate": 1.2258022250577622e-06, + "loss": 0.0966, + "step": 392 + }, + { + "epoch": 0.08931818181818182, + "grad_norm": 3.655542737877026, + "learning_rate": 1.2256791030745434e-06, + "loss": 0.1374, + "step": 393 + }, + { + "epoch": 0.08954545454545454, + "grad_norm": 2.500478617406216, + "learning_rate": 1.225555674868966e-06, + "loss": 0.0886, + "step": 394 + }, + { + "epoch": 0.08977272727272727, + "grad_norm": 5.143537623057685, + "learning_rate": 1.2254319405039524e-06, + "loss": 0.1367, + "step": 395 + }, + { + "epoch": 0.09, + "grad_norm": 3.3327322494465035, + "learning_rate": 1.2253079000425818e-06, + "loss": 0.0913, + "step": 396 + }, + { + "epoch": 0.09022727272727273, + "grad_norm": 1.7869807265219433, + "learning_rate": 1.2251835535480895e-06, + "loss": 0.0977, + "step": 397 + }, + { + "epoch": 0.09045454545454545, + "grad_norm": 3.8096113850412237, + "learning_rate": 1.2250589010838662e-06, + "loss": 0.1512, + "step": 398 + }, + { + "epoch": 0.09068181818181818, + "grad_norm": 4.216213052356721, + "learning_rate": 1.224933942713459e-06, + "loss": 0.1808, + "step": 399 + }, + { + "epoch": 0.09090909090909091, + "grad_norm": 3.592401972291036, + "learning_rate": 1.2248086785005709e-06, + "loss": 0.0993, + "step": 400 + }, + { + "epoch": 0.09113636363636364, + "grad_norm": 5.7914383694762765, + "learning_rate": 1.2246831085090611e-06, + "loss": 0.1745, + "step": 401 + }, + { + "epoch": 0.09136363636363637, + "grad_norm": 2.117479261808125, + "learning_rate": 1.2245572328029438e-06, + "loss": 0.0784, + "step": 402 + }, + { + "epoch": 0.0915909090909091, + "grad_norm": 3.2248905523062956, + "learning_rate": 1.22443105144639e-06, + "loss": 0.1426, + "step": 403 + }, + { + "epoch": 0.09181818181818181, + "grad_norm": 2.643077722507744, + "learning_rate": 1.224304564503726e-06, + "loss": 0.1381, + "step": 404 + }, + { + "epoch": 0.09204545454545454, + "grad_norm": 4.5379695065109615, + "learning_rate": 1.224177772039434e-06, + "loss": 0.1243, + "step": 405 + }, + { + "epoch": 0.09227272727272727, + "grad_norm": 2.9591360798904627, + "learning_rate": 1.2240506741181523e-06, + "loss": 0.1227, + "step": 406 + }, + { + "epoch": 0.0925, + "grad_norm": 3.971875431298172, + "learning_rate": 1.2239232708046745e-06, + "loss": 0.0832, + "step": 407 + }, + { + "epoch": 0.09272727272727273, + "grad_norm": 3.072371783830302, + "learning_rate": 1.2237955621639496e-06, + "loss": 0.1162, + "step": 408 + }, + { + "epoch": 0.09295454545454546, + "grad_norm": 1.794641744951481, + "learning_rate": 1.223667548261083e-06, + "loss": 0.0587, + "step": 409 + }, + { + "epoch": 0.09318181818181819, + "grad_norm": 3.191749699166827, + "learning_rate": 1.2235392291613353e-06, + "loss": 0.0851, + "step": 410 + }, + { + "epoch": 0.09340909090909091, + "grad_norm": 3.662085038328866, + "learning_rate": 1.2234106049301228e-06, + "loss": 0.1025, + "step": 411 + }, + { + "epoch": 0.09363636363636364, + "grad_norm": 3.494582410758984, + "learning_rate": 1.2232816756330173e-06, + "loss": 0.1086, + "step": 412 + }, + { + "epoch": 0.09386363636363636, + "grad_norm": 3.346192757277286, + "learning_rate": 1.223152441335746e-06, + "loss": 0.1224, + "step": 413 + }, + { + "epoch": 0.09409090909090909, + "grad_norm": 3.3167149607638122, + "learning_rate": 1.223022902104192e-06, + "loss": 0.1104, + "step": 414 + }, + { + "epoch": 0.09431818181818181, + "grad_norm": 2.2080794870277267, + "learning_rate": 1.2228930580043931e-06, + "loss": 0.0511, + "step": 415 + }, + { + "epoch": 0.09454545454545454, + "grad_norm": 3.5852155567021957, + "learning_rate": 1.2227629091025437e-06, + "loss": 0.1165, + "step": 416 + }, + { + "epoch": 0.09477272727272727, + "grad_norm": 4.080273349233287, + "learning_rate": 1.2226324554649921e-06, + "loss": 0.1031, + "step": 417 + }, + { + "epoch": 0.095, + "grad_norm": 1.8438060849960656, + "learning_rate": 1.222501697158243e-06, + "loss": 0.0751, + "step": 418 + }, + { + "epoch": 0.09522727272727273, + "grad_norm": 3.3510017123652305, + "learning_rate": 1.2223706342489565e-06, + "loss": 0.0985, + "step": 419 + }, + { + "epoch": 0.09545454545454546, + "grad_norm": 3.030731737092712, + "learning_rate": 1.222239266803947e-06, + "loss": 0.1089, + "step": 420 + }, + { + "epoch": 0.09568181818181819, + "grad_norm": 2.5989480599697226, + "learning_rate": 1.2221075948901856e-06, + "loss": 0.1052, + "step": 421 + }, + { + "epoch": 0.0959090909090909, + "grad_norm": 2.54014851347867, + "learning_rate": 1.221975618574797e-06, + "loss": 0.086, + "step": 422 + }, + { + "epoch": 0.09613636363636363, + "grad_norm": 3.2992221492601317, + "learning_rate": 1.2218433379250623e-06, + "loss": 0.0795, + "step": 423 + }, + { + "epoch": 0.09636363636363636, + "grad_norm": 3.176443675461042, + "learning_rate": 1.2217107530084174e-06, + "loss": 0.1231, + "step": 424 + }, + { + "epoch": 0.09659090909090909, + "grad_norm": 3.0570377085904883, + "learning_rate": 1.2215778638924527e-06, + "loss": 0.0865, + "step": 425 + }, + { + "epoch": 0.09681818181818182, + "grad_norm": 4.106706978977622, + "learning_rate": 1.221444670644915e-06, + "loss": 0.0911, + "step": 426 + }, + { + "epoch": 0.09704545454545455, + "grad_norm": 4.170021502107259, + "learning_rate": 1.221311173333705e-06, + "loss": 0.0917, + "step": 427 + }, + { + "epoch": 0.09727272727272727, + "grad_norm": 3.381683750510494, + "learning_rate": 1.2211773720268784e-06, + "loss": 0.0921, + "step": 428 + }, + { + "epoch": 0.0975, + "grad_norm": 3.0435870331860913, + "learning_rate": 1.2210432667926467e-06, + "loss": 0.1012, + "step": 429 + }, + { + "epoch": 0.09772727272727273, + "grad_norm": 2.818167489252515, + "learning_rate": 1.2209088576993757e-06, + "loss": 0.1523, + "step": 430 + }, + { + "epoch": 0.09795454545454546, + "grad_norm": 4.359715563785674, + "learning_rate": 1.2207741448155867e-06, + "loss": 0.0967, + "step": 431 + }, + { + "epoch": 0.09818181818181818, + "grad_norm": 2.7773735957677137, + "learning_rate": 1.220639128209955e-06, + "loss": 0.1025, + "step": 432 + }, + { + "epoch": 0.0984090909090909, + "grad_norm": 2.272772407483095, + "learning_rate": 1.2205038079513113e-06, + "loss": 0.0854, + "step": 433 + }, + { + "epoch": 0.09863636363636363, + "grad_norm": 4.702441375803285, + "learning_rate": 1.2203681841086409e-06, + "loss": 0.1115, + "step": 434 + }, + { + "epoch": 0.09886363636363636, + "grad_norm": 2.382424771327282, + "learning_rate": 1.2202322567510843e-06, + "loss": 0.0789, + "step": 435 + }, + { + "epoch": 0.09909090909090909, + "grad_norm": 4.101847192253393, + "learning_rate": 1.2200960259479362e-06, + "loss": 0.1171, + "step": 436 + }, + { + "epoch": 0.09931818181818182, + "grad_norm": 2.7790127243978513, + "learning_rate": 1.219959491768646e-06, + "loss": 0.1539, + "step": 437 + }, + { + "epoch": 0.09954545454545455, + "grad_norm": 4.106290395070255, + "learning_rate": 1.2198226542828183e-06, + "loss": 0.0917, + "step": 438 + }, + { + "epoch": 0.09977272727272728, + "grad_norm": 1.6201437073059783, + "learning_rate": 1.219685513560212e-06, + "loss": 0.0533, + "step": 439 + }, + { + "epoch": 0.1, + "grad_norm": 1.5706619304924578, + "learning_rate": 1.2195480696707401e-06, + "loss": 0.0434, + "step": 440 + }, + { + "epoch": 0.10022727272727272, + "grad_norm": 3.2837696987403535, + "learning_rate": 1.2194103226844711e-06, + "loss": 0.1895, + "step": 441 + }, + { + "epoch": 0.10045454545454545, + "grad_norm": 3.0487629356689077, + "learning_rate": 1.2192722726716272e-06, + "loss": 0.1791, + "step": 442 + }, + { + "epoch": 0.10068181818181818, + "grad_norm": 2.864945221356336, + "learning_rate": 1.2191339197025857e-06, + "loss": 0.1006, + "step": 443 + }, + { + "epoch": 0.1009090909090909, + "grad_norm": 3.328999333526785, + "learning_rate": 1.2189952638478778e-06, + "loss": 0.12, + "step": 444 + }, + { + "epoch": 0.10113636363636364, + "grad_norm": 6.824535899317012, + "learning_rate": 1.2188563051781894e-06, + "loss": 0.107, + "step": 445 + }, + { + "epoch": 0.10136363636363636, + "grad_norm": 2.970207101777574, + "learning_rate": 1.2187170437643608e-06, + "loss": 0.1015, + "step": 446 + }, + { + "epoch": 0.10159090909090909, + "grad_norm": 1.8730743795381233, + "learning_rate": 1.2185774796773864e-06, + "loss": 0.0755, + "step": 447 + }, + { + "epoch": 0.10181818181818182, + "grad_norm": 1.7354586509157446, + "learning_rate": 1.2184376129884156e-06, + "loss": 0.0496, + "step": 448 + }, + { + "epoch": 0.10204545454545455, + "grad_norm": 4.080468082238516, + "learning_rate": 1.2182974437687512e-06, + "loss": 0.0928, + "step": 449 + }, + { + "epoch": 0.10227272727272728, + "grad_norm": 1.8482335715828038, + "learning_rate": 1.2181569720898503e-06, + "loss": 0.0657, + "step": 450 + }, + { + "epoch": 0.1025, + "grad_norm": 3.124794345318771, + "learning_rate": 1.2180161980233249e-06, + "loss": 0.1249, + "step": 451 + }, + { + "epoch": 0.10272727272727272, + "grad_norm": 4.467542714725325, + "learning_rate": 1.2178751216409404e-06, + "loss": 0.1201, + "step": 452 + }, + { + "epoch": 0.10295454545454545, + "grad_norm": 3.735204267539462, + "learning_rate": 1.217733743014617e-06, + "loss": 0.1153, + "step": 453 + }, + { + "epoch": 0.10318181818181818, + "grad_norm": 3.8454326050978116, + "learning_rate": 1.2175920622164284e-06, + "loss": 0.1061, + "step": 454 + }, + { + "epoch": 0.10340909090909091, + "grad_norm": 3.2273986480705594, + "learning_rate": 1.2174500793186024e-06, + "loss": 0.1641, + "step": 455 + }, + { + "epoch": 0.10363636363636364, + "grad_norm": 2.072272718882929, + "learning_rate": 1.2173077943935212e-06, + "loss": 0.0762, + "step": 456 + }, + { + "epoch": 0.10386363636363637, + "grad_norm": 4.204228694774329, + "learning_rate": 1.2171652075137209e-06, + "loss": 0.0929, + "step": 457 + }, + { + "epoch": 0.1040909090909091, + "grad_norm": 2.4050603296632715, + "learning_rate": 1.2170223187518908e-06, + "loss": 0.0891, + "step": 458 + }, + { + "epoch": 0.10431818181818182, + "grad_norm": 3.2518785787165534, + "learning_rate": 1.216879128180875e-06, + "loss": 0.0776, + "step": 459 + }, + { + "epoch": 0.10454545454545454, + "grad_norm": 1.6948526180067895, + "learning_rate": 1.2167356358736714e-06, + "loss": 0.0481, + "step": 460 + }, + { + "epoch": 0.10477272727272727, + "grad_norm": 4.104910408043853, + "learning_rate": 1.2165918419034312e-06, + "loss": 0.1031, + "step": 461 + }, + { + "epoch": 0.105, + "grad_norm": 3.2745461768223003, + "learning_rate": 1.2164477463434599e-06, + "loss": 0.1118, + "step": 462 + }, + { + "epoch": 0.10522727272727272, + "grad_norm": 3.58775286474145, + "learning_rate": 1.216303349267216e-06, + "loss": 0.0918, + "step": 463 + }, + { + "epoch": 0.10545454545454545, + "grad_norm": 2.0315975482504354, + "learning_rate": 1.2161586507483126e-06, + "loss": 0.073, + "step": 464 + }, + { + "epoch": 0.10568181818181818, + "grad_norm": 5.10951651637718, + "learning_rate": 1.2160136508605156e-06, + "loss": 0.1138, + "step": 465 + }, + { + "epoch": 0.10590909090909091, + "grad_norm": 2.199555010883721, + "learning_rate": 1.2158683496777457e-06, + "loss": 0.0759, + "step": 466 + }, + { + "epoch": 0.10613636363636364, + "grad_norm": 7.390425381374746, + "learning_rate": 1.215722747274076e-06, + "loss": 0.1778, + "step": 467 + }, + { + "epoch": 0.10636363636363637, + "grad_norm": 3.3969297444237925, + "learning_rate": 1.2155768437237342e-06, + "loss": 0.1422, + "step": 468 + }, + { + "epoch": 0.1065909090909091, + "grad_norm": 3.707282959931986, + "learning_rate": 1.2154306391011003e-06, + "loss": 0.1551, + "step": 469 + }, + { + "epoch": 0.10681818181818181, + "grad_norm": 1.942479482754022, + "learning_rate": 1.215284133480709e-06, + "loss": 0.0634, + "step": 470 + }, + { + "epoch": 0.10704545454545454, + "grad_norm": 3.677402904477567, + "learning_rate": 1.2151373269372476e-06, + "loss": 0.1869, + "step": 471 + }, + { + "epoch": 0.10727272727272727, + "grad_norm": 3.044979112756595, + "learning_rate": 1.2149902195455574e-06, + "loss": 0.1213, + "step": 472 + }, + { + "epoch": 0.1075, + "grad_norm": 4.9193341182656605, + "learning_rate": 1.2148428113806326e-06, + "loss": 0.2508, + "step": 473 + }, + { + "epoch": 0.10772727272727273, + "grad_norm": 2.7779693073379335, + "learning_rate": 1.214695102517621e-06, + "loss": 0.124, + "step": 474 + }, + { + "epoch": 0.10795454545454546, + "grad_norm": 3.211210019518054, + "learning_rate": 1.214547093031824e-06, + "loss": 0.1231, + "step": 475 + }, + { + "epoch": 0.10818181818181818, + "grad_norm": 2.008978878561022, + "learning_rate": 1.2143987829986953e-06, + "loss": 0.0672, + "step": 476 + }, + { + "epoch": 0.10840909090909091, + "grad_norm": 1.928961052455212, + "learning_rate": 1.2142501724938425e-06, + "loss": 0.1028, + "step": 477 + }, + { + "epoch": 0.10863636363636364, + "grad_norm": 1.864918540956773, + "learning_rate": 1.2141012615930266e-06, + "loss": 0.0553, + "step": 478 + }, + { + "epoch": 0.10886363636363636, + "grad_norm": 3.347544298479479, + "learning_rate": 1.2139520503721614e-06, + "loss": 0.1771, + "step": 479 + }, + { + "epoch": 0.10909090909090909, + "grad_norm": 2.989975130941474, + "learning_rate": 1.2138025389073133e-06, + "loss": 0.0802, + "step": 480 + }, + { + "epoch": 0.10931818181818181, + "grad_norm": 3.687482090917259, + "learning_rate": 1.2136527272747032e-06, + "loss": 0.0811, + "step": 481 + }, + { + "epoch": 0.10954545454545454, + "grad_norm": 3.249420071251407, + "learning_rate": 1.2135026155507033e-06, + "loss": 0.0963, + "step": 482 + }, + { + "epoch": 0.10977272727272727, + "grad_norm": 3.152509668833972, + "learning_rate": 1.2133522038118398e-06, + "loss": 0.0943, + "step": 483 + }, + { + "epoch": 0.11, + "grad_norm": 2.5869070668642618, + "learning_rate": 1.2132014921347917e-06, + "loss": 0.075, + "step": 484 + }, + { + "epoch": 0.11022727272727273, + "grad_norm": 3.4670725879292146, + "learning_rate": 1.213050480596391e-06, + "loss": 0.1183, + "step": 485 + }, + { + "epoch": 0.11045454545454546, + "grad_norm": 2.5519406460348337, + "learning_rate": 1.2128991692736223e-06, + "loss": 0.0613, + "step": 486 + }, + { + "epoch": 0.11068181818181819, + "grad_norm": 2.11024294228615, + "learning_rate": 1.2127475582436232e-06, + "loss": 0.0704, + "step": 487 + }, + { + "epoch": 0.11090909090909092, + "grad_norm": 3.270442483188052, + "learning_rate": 1.2125956475836837e-06, + "loss": 0.1098, + "step": 488 + }, + { + "epoch": 0.11113636363636363, + "grad_norm": 3.696356279452533, + "learning_rate": 1.2124434373712473e-06, + "loss": 0.2228, + "step": 489 + }, + { + "epoch": 0.11136363636363636, + "grad_norm": 2.524477089039239, + "learning_rate": 1.2122909276839095e-06, + "loss": 0.0835, + "step": 490 + }, + { + "epoch": 0.11159090909090909, + "grad_norm": 3.011341540909158, + "learning_rate": 1.2121381185994192e-06, + "loss": 0.0997, + "step": 491 + }, + { + "epoch": 0.11181818181818182, + "grad_norm": 1.8314402211744092, + "learning_rate": 1.211985010195677e-06, + "loss": 0.0492, + "step": 492 + }, + { + "epoch": 0.11204545454545455, + "grad_norm": 2.279695549709343, + "learning_rate": 1.2118316025507369e-06, + "loss": 0.0926, + "step": 493 + }, + { + "epoch": 0.11227272727272727, + "grad_norm": 3.623903633913838, + "learning_rate": 1.2116778957428046e-06, + "loss": 0.1712, + "step": 494 + }, + { + "epoch": 0.1125, + "grad_norm": 2.2409293046632537, + "learning_rate": 1.2115238898502395e-06, + "loss": 0.0465, + "step": 495 + }, + { + "epoch": 0.11272727272727273, + "grad_norm": 3.7349766209955106, + "learning_rate": 1.2113695849515527e-06, + "loss": 0.1544, + "step": 496 + }, + { + "epoch": 0.11295454545454546, + "grad_norm": 4.647699836801936, + "learning_rate": 1.2112149811254076e-06, + "loss": 0.1349, + "step": 497 + }, + { + "epoch": 0.11318181818181818, + "grad_norm": 3.0341743955241602, + "learning_rate": 1.2110600784506202e-06, + "loss": 0.0831, + "step": 498 + }, + { + "epoch": 0.1134090909090909, + "grad_norm": 3.881301371468742, + "learning_rate": 1.2109048770061593e-06, + "loss": 0.099, + "step": 499 + }, + { + "epoch": 0.11363636363636363, + "grad_norm": 3.0402740523486873, + "learning_rate": 1.2107493768711453e-06, + "loss": 0.0978, + "step": 500 + }, + { + "epoch": 0.11386363636363636, + "grad_norm": 3.000760767985402, + "learning_rate": 1.2105935781248513e-06, + "loss": 0.1205, + "step": 501 + }, + { + "epoch": 0.11409090909090909, + "grad_norm": 3.438632741851898, + "learning_rate": 1.2104374808467023e-06, + "loss": 0.1558, + "step": 502 + }, + { + "epoch": 0.11431818181818182, + "grad_norm": 3.282001842034137, + "learning_rate": 1.2102810851162758e-06, + "loss": 0.1085, + "step": 503 + }, + { + "epoch": 0.11454545454545455, + "grad_norm": 3.0542969344903184, + "learning_rate": 1.2101243910133013e-06, + "loss": 0.0638, + "step": 504 + }, + { + "epoch": 0.11477272727272728, + "grad_norm": 1.9077001079674953, + "learning_rate": 1.2099673986176604e-06, + "loss": 0.0846, + "step": 505 + }, + { + "epoch": 0.115, + "grad_norm": 4.637444296114528, + "learning_rate": 1.2098101080093873e-06, + "loss": 0.2038, + "step": 506 + }, + { + "epoch": 0.11522727272727273, + "grad_norm": 3.619692095699777, + "learning_rate": 1.2096525192686673e-06, + "loss": 0.2106, + "step": 507 + }, + { + "epoch": 0.11545454545454545, + "grad_norm": 4.377926370826819, + "learning_rate": 1.209494632475838e-06, + "loss": 0.0939, + "step": 508 + }, + { + "epoch": 0.11568181818181818, + "grad_norm": 2.5080931840901863, + "learning_rate": 1.2093364477113893e-06, + "loss": 0.0651, + "step": 509 + }, + { + "epoch": 0.1159090909090909, + "grad_norm": 2.4233336784930612, + "learning_rate": 1.2091779650559628e-06, + "loss": 0.072, + "step": 510 + }, + { + "epoch": 0.11613636363636363, + "grad_norm": 2.0275482779191174, + "learning_rate": 1.209019184590352e-06, + "loss": 0.0737, + "step": 511 + }, + { + "epoch": 0.11636363636363636, + "grad_norm": 2.310551052174877, + "learning_rate": 1.2088601063955018e-06, + "loss": 0.0587, + "step": 512 + }, + { + "epoch": 0.11659090909090909, + "grad_norm": 3.9605777687594643, + "learning_rate": 1.20870073055251e-06, + "loss": 0.11, + "step": 513 + }, + { + "epoch": 0.11681818181818182, + "grad_norm": 4.629579208573231, + "learning_rate": 1.2085410571426244e-06, + "loss": 0.1959, + "step": 514 + }, + { + "epoch": 0.11704545454545455, + "grad_norm": 1.2644581499421943, + "learning_rate": 1.208381086247246e-06, + "loss": 0.061, + "step": 515 + }, + { + "epoch": 0.11727272727272728, + "grad_norm": 1.6192671709158455, + "learning_rate": 1.2082208179479272e-06, + "loss": 0.072, + "step": 516 + }, + { + "epoch": 0.1175, + "grad_norm": 4.144386976162768, + "learning_rate": 1.2080602523263715e-06, + "loss": 0.1269, + "step": 517 + }, + { + "epoch": 0.11772727272727272, + "grad_norm": 5.825934210672073, + "learning_rate": 1.207899389464434e-06, + "loss": 0.1571, + "step": 518 + }, + { + "epoch": 0.11795454545454545, + "grad_norm": 2.804910898128096, + "learning_rate": 1.2077382294441218e-06, + "loss": 0.086, + "step": 519 + }, + { + "epoch": 0.11818181818181818, + "grad_norm": 2.7133158930138985, + "learning_rate": 1.2075767723475932e-06, + "loss": 0.0642, + "step": 520 + }, + { + "epoch": 0.11840909090909091, + "grad_norm": 5.286400216578927, + "learning_rate": 1.2074150182571579e-06, + "loss": 0.1888, + "step": 521 + }, + { + "epoch": 0.11863636363636364, + "grad_norm": 3.397145415157825, + "learning_rate": 1.2072529672552771e-06, + "loss": 0.1186, + "step": 522 + }, + { + "epoch": 0.11886363636363637, + "grad_norm": 3.2642858351017385, + "learning_rate": 1.2070906194245634e-06, + "loss": 0.0885, + "step": 523 + }, + { + "epoch": 0.1190909090909091, + "grad_norm": 2.3563086292879487, + "learning_rate": 1.2069279748477812e-06, + "loss": 0.0654, + "step": 524 + }, + { + "epoch": 0.11931818181818182, + "grad_norm": 1.852507555484346, + "learning_rate": 1.206765033607845e-06, + "loss": 0.0789, + "step": 525 + }, + { + "epoch": 0.11954545454545455, + "grad_norm": 4.856134913872863, + "learning_rate": 1.2066017957878212e-06, + "loss": 0.1834, + "step": 526 + }, + { + "epoch": 0.11977272727272727, + "grad_norm": 2.6640492451084556, + "learning_rate": 1.2064382614709276e-06, + "loss": 0.1437, + "step": 527 + }, + { + "epoch": 0.12, + "grad_norm": 3.2207800318387507, + "learning_rate": 1.206274430740533e-06, + "loss": 0.188, + "step": 528 + }, + { + "epoch": 0.12022727272727272, + "grad_norm": 4.11518851820918, + "learning_rate": 1.2061103036801573e-06, + "loss": 0.1482, + "step": 529 + }, + { + "epoch": 0.12045454545454545, + "grad_norm": 5.313187827562059, + "learning_rate": 1.2059458803734712e-06, + "loss": 0.1351, + "step": 530 + }, + { + "epoch": 0.12068181818181818, + "grad_norm": 2.0868957453810273, + "learning_rate": 1.2057811609042968e-06, + "loss": 0.0712, + "step": 531 + }, + { + "epoch": 0.12090909090909091, + "grad_norm": 3.750446712425499, + "learning_rate": 1.205616145356607e-06, + "loss": 0.1348, + "step": 532 + }, + { + "epoch": 0.12113636363636364, + "grad_norm": 2.227505661725015, + "learning_rate": 1.2054508338145257e-06, + "loss": 0.0936, + "step": 533 + }, + { + "epoch": 0.12136363636363637, + "grad_norm": 4.120704580636114, + "learning_rate": 1.2052852263623274e-06, + "loss": 0.1092, + "step": 534 + }, + { + "epoch": 0.1215909090909091, + "grad_norm": 2.7432990999032763, + "learning_rate": 1.2051193230844382e-06, + "loss": 0.0729, + "step": 535 + }, + { + "epoch": 0.12181818181818181, + "grad_norm": 3.7818779836224437, + "learning_rate": 1.2049531240654343e-06, + "loss": 0.1108, + "step": 536 + }, + { + "epoch": 0.12204545454545454, + "grad_norm": 3.1366345037936227, + "learning_rate": 1.2047866293900428e-06, + "loss": 0.1622, + "step": 537 + }, + { + "epoch": 0.12227272727272727, + "grad_norm": 2.875327093691517, + "learning_rate": 1.2046198391431415e-06, + "loss": 0.1321, + "step": 538 + }, + { + "epoch": 0.1225, + "grad_norm": 2.648122703453979, + "learning_rate": 1.2044527534097595e-06, + "loss": 0.097, + "step": 539 + }, + { + "epoch": 0.12272727272727273, + "grad_norm": 1.6393062873629956, + "learning_rate": 1.2042853722750756e-06, + "loss": 0.0454, + "step": 540 + }, + { + "epoch": 0.12295454545454546, + "grad_norm": 5.359359873367382, + "learning_rate": 1.2041176958244197e-06, + "loss": 0.1038, + "step": 541 + }, + { + "epoch": 0.12318181818181818, + "grad_norm": 3.5133283828272623, + "learning_rate": 1.2039497241432724e-06, + "loss": 0.0739, + "step": 542 + }, + { + "epoch": 0.12340909090909091, + "grad_norm": 1.8604827276275182, + "learning_rate": 1.2037814573172642e-06, + "loss": 0.077, + "step": 543 + }, + { + "epoch": 0.12363636363636364, + "grad_norm": 4.4318356593910035, + "learning_rate": 1.2036128954321768e-06, + "loss": 0.1304, + "step": 544 + }, + { + "epoch": 0.12386363636363637, + "grad_norm": 3.922792866259573, + "learning_rate": 1.2034440385739418e-06, + "loss": 0.0993, + "step": 545 + }, + { + "epoch": 0.12409090909090909, + "grad_norm": 3.9512881199876073, + "learning_rate": 1.2032748868286415e-06, + "loss": 0.1589, + "step": 546 + }, + { + "epoch": 0.12431818181818181, + "grad_norm": 3.462313519254494, + "learning_rate": 1.2031054402825082e-06, + "loss": 0.0734, + "step": 547 + }, + { + "epoch": 0.12454545454545454, + "grad_norm": 3.3057861908192656, + "learning_rate": 1.2029356990219248e-06, + "loss": 0.1348, + "step": 548 + }, + { + "epoch": 0.12477272727272727, + "grad_norm": 3.223977288932668, + "learning_rate": 1.2027656631334242e-06, + "loss": 0.1887, + "step": 549 + }, + { + "epoch": 0.125, + "grad_norm": 4.9277717305548006, + "learning_rate": 1.2025953327036897e-06, + "loss": 0.1629, + "step": 550 + }, + { + "epoch": 0.12522727272727271, + "grad_norm": 2.9568147489411616, + "learning_rate": 1.2024247078195542e-06, + "loss": 0.0951, + "step": 551 + }, + { + "epoch": 0.12545454545454546, + "grad_norm": 2.7716605979755182, + "learning_rate": 1.202253788568002e-06, + "loss": 0.0956, + "step": 552 + }, + { + "epoch": 0.12568181818181817, + "grad_norm": 2.564268273251238, + "learning_rate": 1.202082575036166e-06, + "loss": 0.1367, + "step": 553 + }, + { + "epoch": 0.12590909090909091, + "grad_norm": 2.8526862414378322, + "learning_rate": 1.2019110673113302e-06, + "loss": 0.1075, + "step": 554 + }, + { + "epoch": 0.12613636363636363, + "grad_norm": 5.5229813826793315, + "learning_rate": 1.2017392654809278e-06, + "loss": 0.1054, + "step": 555 + }, + { + "epoch": 0.12636363636363637, + "grad_norm": 2.7238182500044794, + "learning_rate": 1.2015671696325423e-06, + "loss": 0.1323, + "step": 556 + }, + { + "epoch": 0.1265909090909091, + "grad_norm": 4.175398220182507, + "learning_rate": 1.2013947798539073e-06, + "loss": 0.0757, + "step": 557 + }, + { + "epoch": 0.12681818181818183, + "grad_norm": 2.394274914481463, + "learning_rate": 1.2012220962329058e-06, + "loss": 0.0519, + "step": 558 + }, + { + "epoch": 0.12704545454545454, + "grad_norm": 3.1676135108802463, + "learning_rate": 1.201049118857571e-06, + "loss": 0.1133, + "step": 559 + }, + { + "epoch": 0.12727272727272726, + "grad_norm": 5.704490571588163, + "learning_rate": 1.2008758478160853e-06, + "loss": 0.1623, + "step": 560 + }, + { + "epoch": 0.1275, + "grad_norm": 2.2408515449382875, + "learning_rate": 1.2007022831967813e-06, + "loss": 0.0823, + "step": 561 + }, + { + "epoch": 0.12772727272727272, + "grad_norm": 2.860111215399375, + "learning_rate": 1.2005284250881417e-06, + "loss": 0.0966, + "step": 562 + }, + { + "epoch": 0.12795454545454546, + "grad_norm": 3.6587552112745354, + "learning_rate": 1.2003542735787973e-06, + "loss": 0.1507, + "step": 563 + }, + { + "epoch": 0.12818181818181817, + "grad_norm": 1.8287971343258551, + "learning_rate": 1.20017982875753e-06, + "loss": 0.0526, + "step": 564 + }, + { + "epoch": 0.12840909090909092, + "grad_norm": 2.0598344947462572, + "learning_rate": 1.2000050907132705e-06, + "loss": 0.057, + "step": 565 + }, + { + "epoch": 0.12863636363636363, + "grad_norm": 4.6119483078979595, + "learning_rate": 1.1998300595350993e-06, + "loss": 0.1943, + "step": 566 + }, + { + "epoch": 0.12886363636363637, + "grad_norm": 3.17573639775011, + "learning_rate": 1.1996547353122461e-06, + "loss": 0.0951, + "step": 567 + }, + { + "epoch": 0.1290909090909091, + "grad_norm": 4.240549393179977, + "learning_rate": 1.1994791181340897e-06, + "loss": 0.1353, + "step": 568 + }, + { + "epoch": 0.1293181818181818, + "grad_norm": 3.913561720838828, + "learning_rate": 1.1993032080901593e-06, + "loss": 0.1204, + "step": 569 + }, + { + "epoch": 0.12954545454545455, + "grad_norm": 3.5940294239948747, + "learning_rate": 1.1991270052701323e-06, + "loss": 0.1364, + "step": 570 + }, + { + "epoch": 0.12977272727272726, + "grad_norm": 2.4515799882553027, + "learning_rate": 1.1989505097638357e-06, + "loss": 0.1679, + "step": 571 + }, + { + "epoch": 0.13, + "grad_norm": 1.7492542711277368, + "learning_rate": 1.198773721661246e-06, + "loss": 0.0804, + "step": 572 + }, + { + "epoch": 0.13022727272727272, + "grad_norm": 3.2990053875732093, + "learning_rate": 1.1985966410524883e-06, + "loss": 0.0866, + "step": 573 + }, + { + "epoch": 0.13045454545454546, + "grad_norm": 2.3457141093478784, + "learning_rate": 1.1984192680278376e-06, + "loss": 0.0516, + "step": 574 + }, + { + "epoch": 0.13068181818181818, + "grad_norm": 2.134601565498266, + "learning_rate": 1.1982416026777172e-06, + "loss": 0.0771, + "step": 575 + }, + { + "epoch": 0.13090909090909092, + "grad_norm": 2.1076945213159535, + "learning_rate": 1.1980636450926999e-06, + "loss": 0.0694, + "step": 576 + }, + { + "epoch": 0.13113636363636363, + "grad_norm": 1.4383376329301214, + "learning_rate": 1.1978853953635074e-06, + "loss": 0.0515, + "step": 577 + }, + { + "epoch": 0.13136363636363638, + "grad_norm": 3.060462093745668, + "learning_rate": 1.1977068535810101e-06, + "loss": 0.1077, + "step": 578 + }, + { + "epoch": 0.1315909090909091, + "grad_norm": 3.5199149141450157, + "learning_rate": 1.1975280198362276e-06, + "loss": 0.1471, + "step": 579 + }, + { + "epoch": 0.1318181818181818, + "grad_norm": 3.283303033777514, + "learning_rate": 1.1973488942203282e-06, + "loss": 0.1003, + "step": 580 + }, + { + "epoch": 0.13204545454545455, + "grad_norm": 2.7274884821164562, + "learning_rate": 1.197169476824629e-06, + "loss": 0.0991, + "step": 581 + }, + { + "epoch": 0.13227272727272726, + "grad_norm": 2.8445324401658874, + "learning_rate": 1.1969897677405956e-06, + "loss": 0.1115, + "step": 582 + }, + { + "epoch": 0.1325, + "grad_norm": 5.023911656304322, + "learning_rate": 1.1968097670598428e-06, + "loss": 0.2189, + "step": 583 + }, + { + "epoch": 0.13272727272727272, + "grad_norm": 2.6658727836557055, + "learning_rate": 1.1966294748741336e-06, + "loss": 0.0574, + "step": 584 + }, + { + "epoch": 0.13295454545454546, + "grad_norm": 2.155192949943212, + "learning_rate": 1.19644889127538e-06, + "loss": 0.0662, + "step": 585 + }, + { + "epoch": 0.13318181818181818, + "grad_norm": 5.185895348668454, + "learning_rate": 1.1962680163556424e-06, + "loss": 0.1174, + "step": 586 + }, + { + "epoch": 0.13340909090909092, + "grad_norm": 2.2768955235479686, + "learning_rate": 1.1960868502071294e-06, + "loss": 0.1488, + "step": 587 + }, + { + "epoch": 0.13363636363636364, + "grad_norm": 2.15947631768917, + "learning_rate": 1.1959053929221984e-06, + "loss": 0.1091, + "step": 588 + }, + { + "epoch": 0.13386363636363635, + "grad_norm": 3.5100559294299347, + "learning_rate": 1.1957236445933553e-06, + "loss": 0.1116, + "step": 589 + }, + { + "epoch": 0.1340909090909091, + "grad_norm": 2.946513848269364, + "learning_rate": 1.1955416053132542e-06, + "loss": 0.0614, + "step": 590 + }, + { + "epoch": 0.1343181818181818, + "grad_norm": 3.340713899254648, + "learning_rate": 1.1953592751746976e-06, + "loss": 0.1067, + "step": 591 + }, + { + "epoch": 0.13454545454545455, + "grad_norm": 2.5778727430589714, + "learning_rate": 1.1951766542706362e-06, + "loss": 0.0679, + "step": 592 + }, + { + "epoch": 0.13477272727272727, + "grad_norm": 3.2224245336451722, + "learning_rate": 1.1949937426941689e-06, + "loss": 0.1085, + "step": 593 + }, + { + "epoch": 0.135, + "grad_norm": 3.359498676188438, + "learning_rate": 1.1948105405385428e-06, + "loss": 0.1725, + "step": 594 + }, + { + "epoch": 0.13522727272727272, + "grad_norm": 3.0433127389455987, + "learning_rate": 1.1946270478971533e-06, + "loss": 0.126, + "step": 595 + }, + { + "epoch": 0.13545454545454547, + "grad_norm": 2.7071052078673907, + "learning_rate": 1.194443264863544e-06, + "loss": 0.1076, + "step": 596 + }, + { + "epoch": 0.13568181818181818, + "grad_norm": 2.3551668039606124, + "learning_rate": 1.1942591915314058e-06, + "loss": 0.1311, + "step": 597 + }, + { + "epoch": 0.1359090909090909, + "grad_norm": 3.666501070316695, + "learning_rate": 1.1940748279945784e-06, + "loss": 0.1022, + "step": 598 + }, + { + "epoch": 0.13613636363636364, + "grad_norm": 2.2003349882930308, + "learning_rate": 1.1938901743470494e-06, + "loss": 0.1038, + "step": 599 + }, + { + "epoch": 0.13636363636363635, + "grad_norm": 3.128777788636594, + "learning_rate": 1.1937052306829534e-06, + "loss": 0.0807, + "step": 600 + }, + { + "epoch": 0.1365909090909091, + "grad_norm": 3.578582761882501, + "learning_rate": 1.1935199970965741e-06, + "loss": 0.095, + "step": 601 + }, + { + "epoch": 0.1368181818181818, + "grad_norm": 3.3499443547141303, + "learning_rate": 1.193334473682342e-06, + "loss": 0.1319, + "step": 602 + }, + { + "epoch": 0.13704545454545455, + "grad_norm": 2.3485015487514684, + "learning_rate": 1.193148660534836e-06, + "loss": 0.1161, + "step": 603 + }, + { + "epoch": 0.13727272727272727, + "grad_norm": 2.506771300239319, + "learning_rate": 1.1929625577487825e-06, + "loss": 0.1036, + "step": 604 + }, + { + "epoch": 0.1375, + "grad_norm": 4.180978541036742, + "learning_rate": 1.1927761654190552e-06, + "loss": 0.1093, + "step": 605 + }, + { + "epoch": 0.13772727272727273, + "grad_norm": 2.274271918110917, + "learning_rate": 1.192589483640676e-06, + "loss": 0.0488, + "step": 606 + }, + { + "epoch": 0.13795454545454544, + "grad_norm": 3.9989728601369743, + "learning_rate": 1.1924025125088138e-06, + "loss": 0.1202, + "step": 607 + }, + { + "epoch": 0.13818181818181818, + "grad_norm": 5.604231304748778, + "learning_rate": 1.1922152521187854e-06, + "loss": 0.2126, + "step": 608 + }, + { + "epoch": 0.1384090909090909, + "grad_norm": 2.134884091781087, + "learning_rate": 1.1920277025660553e-06, + "loss": 0.0631, + "step": 609 + }, + { + "epoch": 0.13863636363636364, + "grad_norm": 4.173570592411882, + "learning_rate": 1.1918398639462345e-06, + "loss": 0.1042, + "step": 610 + }, + { + "epoch": 0.13886363636363636, + "grad_norm": 6.864550241306282, + "learning_rate": 1.1916517363550821e-06, + "loss": 0.2171, + "step": 611 + }, + { + "epoch": 0.1390909090909091, + "grad_norm": 2.631229130251518, + "learning_rate": 1.1914633198885047e-06, + "loss": 0.1257, + "step": 612 + }, + { + "epoch": 0.1393181818181818, + "grad_norm": 0.8426286857023338, + "learning_rate": 1.1912746146425555e-06, + "loss": 0.0251, + "step": 613 + }, + { + "epoch": 0.13954545454545456, + "grad_norm": 3.798074645257706, + "learning_rate": 1.1910856207134352e-06, + "loss": 0.1613, + "step": 614 + }, + { + "epoch": 0.13977272727272727, + "grad_norm": 3.4664147461018913, + "learning_rate": 1.1908963381974916e-06, + "loss": 0.0716, + "step": 615 + }, + { + "epoch": 0.14, + "grad_norm": 7.371412477513619, + "learning_rate": 1.19070676719122e-06, + "loss": 0.3521, + "step": 616 + }, + { + "epoch": 0.14022727272727273, + "grad_norm": 3.157479047877215, + "learning_rate": 1.1905169077912623e-06, + "loss": 0.1253, + "step": 617 + }, + { + "epoch": 0.14045454545454544, + "grad_norm": 2.685317401054625, + "learning_rate": 1.1903267600944077e-06, + "loss": 0.0734, + "step": 618 + }, + { + "epoch": 0.14068181818181819, + "grad_norm": 2.1415402542548554, + "learning_rate": 1.1901363241975921e-06, + "loss": 0.0548, + "step": 619 + }, + { + "epoch": 0.1409090909090909, + "grad_norm": 2.9848043799184185, + "learning_rate": 1.1899456001978987e-06, + "loss": 0.1125, + "step": 620 + }, + { + "epoch": 0.14113636363636364, + "grad_norm": 4.506518783865914, + "learning_rate": 1.1897545881925573e-06, + "loss": 0.1421, + "step": 621 + }, + { + "epoch": 0.14136363636363636, + "grad_norm": 2.6891958121755914, + "learning_rate": 1.1895632882789447e-06, + "loss": 0.0955, + "step": 622 + }, + { + "epoch": 0.1415909090909091, + "grad_norm": 2.465824632315862, + "learning_rate": 1.1893717005545843e-06, + "loss": 0.1053, + "step": 623 + }, + { + "epoch": 0.14181818181818182, + "grad_norm": 3.0201452818122916, + "learning_rate": 1.189179825117146e-06, + "loss": 0.1213, + "step": 624 + }, + { + "epoch": 0.14204545454545456, + "grad_norm": 3.329101304932354, + "learning_rate": 1.1889876620644472e-06, + "loss": 0.0792, + "step": 625 + }, + { + "epoch": 0.14227272727272727, + "grad_norm": 3.046606101528462, + "learning_rate": 1.1887952114944509e-06, + "loss": 0.0698, + "step": 626 + }, + { + "epoch": 0.1425, + "grad_norm": 1.5949836606987318, + "learning_rate": 1.1886024735052676e-06, + "loss": 0.0763, + "step": 627 + }, + { + "epoch": 0.14272727272727273, + "grad_norm": 2.314090195158352, + "learning_rate": 1.1884094481951535e-06, + "loss": 0.11, + "step": 628 + }, + { + "epoch": 0.14295454545454545, + "grad_norm": 4.028948054738105, + "learning_rate": 1.1882161356625122e-06, + "loss": 0.1311, + "step": 629 + }, + { + "epoch": 0.1431818181818182, + "grad_norm": 1.7515725063837653, + "learning_rate": 1.1880225360058925e-06, + "loss": 0.0919, + "step": 630 + }, + { + "epoch": 0.1434090909090909, + "grad_norm": 2.9033989421868096, + "learning_rate": 1.1878286493239907e-06, + "loss": 0.133, + "step": 631 + }, + { + "epoch": 0.14363636363636365, + "grad_norm": 2.2494301771344642, + "learning_rate": 1.187634475715649e-06, + "loss": 0.0791, + "step": 632 + }, + { + "epoch": 0.14386363636363636, + "grad_norm": 1.3701099057464936, + "learning_rate": 1.1874400152798557e-06, + "loss": 0.0481, + "step": 633 + }, + { + "epoch": 0.1440909090909091, + "grad_norm": 3.6710250802809234, + "learning_rate": 1.1872452681157453e-06, + "loss": 0.085, + "step": 634 + }, + { + "epoch": 0.14431818181818182, + "grad_norm": 2.443957657160077, + "learning_rate": 1.1870502343225992e-06, + "loss": 0.0647, + "step": 635 + }, + { + "epoch": 0.14454545454545453, + "grad_norm": 2.0793326121841846, + "learning_rate": 1.186854913999844e-06, + "loss": 0.1012, + "step": 636 + }, + { + "epoch": 0.14477272727272728, + "grad_norm": 2.892727939632014, + "learning_rate": 1.1866593072470527e-06, + "loss": 0.0996, + "step": 637 + }, + { + "epoch": 0.145, + "grad_norm": 1.913055121072542, + "learning_rate": 1.1864634141639448e-06, + "loss": 0.0531, + "step": 638 + }, + { + "epoch": 0.14522727272727273, + "grad_norm": 3.5432153609693495, + "learning_rate": 1.1862672348503848e-06, + "loss": 0.1218, + "step": 639 + }, + { + "epoch": 0.14545454545454545, + "grad_norm": 2.2036943399739917, + "learning_rate": 1.186070769406384e-06, + "loss": 0.0739, + "step": 640 + }, + { + "epoch": 0.1456818181818182, + "grad_norm": 3.5143590647841636, + "learning_rate": 1.185874017932099e-06, + "loss": 0.0911, + "step": 641 + }, + { + "epoch": 0.1459090909090909, + "grad_norm": 2.493140405671084, + "learning_rate": 1.1856769805278327e-06, + "loss": 0.0835, + "step": 642 + }, + { + "epoch": 0.14613636363636365, + "grad_norm": 3.073337249659016, + "learning_rate": 1.1854796572940332e-06, + "loss": 0.166, + "step": 643 + }, + { + "epoch": 0.14636363636363636, + "grad_norm": 1.6993044436545248, + "learning_rate": 1.1852820483312951e-06, + "loss": 0.0817, + "step": 644 + }, + { + "epoch": 0.14659090909090908, + "grad_norm": 3.4739177837922393, + "learning_rate": 1.1850841537403577e-06, + "loss": 0.0717, + "step": 645 + }, + { + "epoch": 0.14681818181818182, + "grad_norm": 3.4398766150994833, + "learning_rate": 1.1848859736221062e-06, + "loss": 0.1449, + "step": 646 + }, + { + "epoch": 0.14704545454545453, + "grad_norm": 5.82184982437933, + "learning_rate": 1.1846875080775724e-06, + "loss": 0.1468, + "step": 647 + }, + { + "epoch": 0.14727272727272728, + "grad_norm": 3.335638666408031, + "learning_rate": 1.1844887572079322e-06, + "loss": 0.1316, + "step": 648 + }, + { + "epoch": 0.1475, + "grad_norm": 3.101326236389216, + "learning_rate": 1.1842897211145075e-06, + "loss": 0.0948, + "step": 649 + }, + { + "epoch": 0.14772727272727273, + "grad_norm": 2.9810958233139973, + "learning_rate": 1.1840903998987657e-06, + "loss": 0.0792, + "step": 650 + }, + { + "epoch": 0.14795454545454545, + "grad_norm": 2.063537416474323, + "learning_rate": 1.1838907936623196e-06, + "loss": 0.0733, + "step": 651 + }, + { + "epoch": 0.1481818181818182, + "grad_norm": 2.559089255573767, + "learning_rate": 1.183690902506927e-06, + "loss": 0.0749, + "step": 652 + }, + { + "epoch": 0.1484090909090909, + "grad_norm": 3.319331738740318, + "learning_rate": 1.1834907265344913e-06, + "loss": 0.1844, + "step": 653 + }, + { + "epoch": 0.14863636363636365, + "grad_norm": 2.2452512483980773, + "learning_rate": 1.1832902658470608e-06, + "loss": 0.1165, + "step": 654 + }, + { + "epoch": 0.14886363636363636, + "grad_norm": 3.961645280409906, + "learning_rate": 1.1830895205468293e-06, + "loss": 0.1088, + "step": 655 + }, + { + "epoch": 0.14909090909090908, + "grad_norm": 2.2411016758925055, + "learning_rate": 1.182888490736135e-06, + "loss": 0.082, + "step": 656 + }, + { + "epoch": 0.14931818181818182, + "grad_norm": 3.982938567395181, + "learning_rate": 1.1826871765174622e-06, + "loss": 0.1279, + "step": 657 + }, + { + "epoch": 0.14954545454545454, + "grad_norm": 3.8472629805431753, + "learning_rate": 1.1824855779934392e-06, + "loss": 0.0806, + "step": 658 + }, + { + "epoch": 0.14977272727272728, + "grad_norm": 3.119210512568414, + "learning_rate": 1.1822836952668397e-06, + "loss": 0.1166, + "step": 659 + }, + { + "epoch": 0.15, + "grad_norm": 3.444563836716263, + "learning_rate": 1.182081528440582e-06, + "loss": 0.1556, + "step": 660 + }, + { + "epoch": 0.15022727272727274, + "grad_norm": 2.332361190943201, + "learning_rate": 1.18187907761773e-06, + "loss": 0.0801, + "step": 661 + }, + { + "epoch": 0.15045454545454545, + "grad_norm": 4.378773787277541, + "learning_rate": 1.1816763429014917e-06, + "loss": 0.181, + "step": 662 + }, + { + "epoch": 0.1506818181818182, + "grad_norm": 3.0462084426413756, + "learning_rate": 1.1814733243952193e-06, + "loss": 0.1252, + "step": 663 + }, + { + "epoch": 0.1509090909090909, + "grad_norm": 2.9809447577195844, + "learning_rate": 1.1812700222024111e-06, + "loss": 0.1239, + "step": 664 + }, + { + "epoch": 0.15113636363636362, + "grad_norm": 3.8979422274589455, + "learning_rate": 1.1810664364267092e-06, + "loss": 0.0959, + "step": 665 + }, + { + "epoch": 0.15136363636363637, + "grad_norm": 2.603933804827277, + "learning_rate": 1.1808625671718999e-06, + "loss": 0.0711, + "step": 666 + }, + { + "epoch": 0.15159090909090908, + "grad_norm": 4.264269902864219, + "learning_rate": 1.1806584145419144e-06, + "loss": 0.1744, + "step": 667 + }, + { + "epoch": 0.15181818181818182, + "grad_norm": 3.624752658022269, + "learning_rate": 1.1804539786408292e-06, + "loss": 0.1256, + "step": 668 + }, + { + "epoch": 0.15204545454545454, + "grad_norm": 4.091230439217564, + "learning_rate": 1.1802492595728634e-06, + "loss": 0.1198, + "step": 669 + }, + { + "epoch": 0.15227272727272728, + "grad_norm": 2.9300008130310222, + "learning_rate": 1.1800442574423823e-06, + "loss": 0.0887, + "step": 670 + }, + { + "epoch": 0.1525, + "grad_norm": 3.523441706434944, + "learning_rate": 1.1798389723538942e-06, + "loss": 0.0771, + "step": 671 + }, + { + "epoch": 0.15272727272727274, + "grad_norm": 3.692998888266014, + "learning_rate": 1.1796334044120522e-06, + "loss": 0.1534, + "step": 672 + }, + { + "epoch": 0.15295454545454545, + "grad_norm": 3.5410929810366825, + "learning_rate": 1.1794275537216534e-06, + "loss": 0.1291, + "step": 673 + }, + { + "epoch": 0.15318181818181817, + "grad_norm": 2.3053550202142374, + "learning_rate": 1.1792214203876396e-06, + "loss": 0.0807, + "step": 674 + }, + { + "epoch": 0.1534090909090909, + "grad_norm": 2.4383667101146838, + "learning_rate": 1.1790150045150958e-06, + "loss": 0.1094, + "step": 675 + }, + { + "epoch": 0.15363636363636363, + "grad_norm": 3.1943774587578697, + "learning_rate": 1.1788083062092518e-06, + "loss": 0.0697, + "step": 676 + }, + { + "epoch": 0.15386363636363637, + "grad_norm": 2.3029150188751153, + "learning_rate": 1.1786013255754808e-06, + "loss": 0.1161, + "step": 677 + }, + { + "epoch": 0.15409090909090908, + "grad_norm": 3.8304151073044976, + "learning_rate": 1.1783940627193002e-06, + "loss": 0.1358, + "step": 678 + }, + { + "epoch": 0.15431818181818183, + "grad_norm": 2.537599788611614, + "learning_rate": 1.1781865177463717e-06, + "loss": 0.0936, + "step": 679 + }, + { + "epoch": 0.15454545454545454, + "grad_norm": 1.9155458625625725, + "learning_rate": 1.1779786907625e-06, + "loss": 0.0958, + "step": 680 + }, + { + "epoch": 0.15477272727272728, + "grad_norm": 2.002463538608908, + "learning_rate": 1.177770581873634e-06, + "loss": 0.0656, + "step": 681 + }, + { + "epoch": 0.155, + "grad_norm": 3.6218868543424576, + "learning_rate": 1.1775621911858665e-06, + "loss": 0.1501, + "step": 682 + }, + { + "epoch": 0.1552272727272727, + "grad_norm": 3.593769936603755, + "learning_rate": 1.1773535188054336e-06, + "loss": 0.1113, + "step": 683 + }, + { + "epoch": 0.15545454545454546, + "grad_norm": 4.567215505192269, + "learning_rate": 1.177144564838715e-06, + "loss": 0.1377, + "step": 684 + }, + { + "epoch": 0.15568181818181817, + "grad_norm": 3.9430899061276734, + "learning_rate": 1.1769353293922341e-06, + "loss": 0.1441, + "step": 685 + }, + { + "epoch": 0.1559090909090909, + "grad_norm": 3.5287441960723633, + "learning_rate": 1.1767258125726584e-06, + "loss": 0.0789, + "step": 686 + }, + { + "epoch": 0.15613636363636363, + "grad_norm": 4.719876001633977, + "learning_rate": 1.1765160144867972e-06, + "loss": 0.1325, + "step": 687 + }, + { + "epoch": 0.15636363636363637, + "grad_norm": 6.246920774452161, + "learning_rate": 1.176305935241605e-06, + "loss": 0.135, + "step": 688 + }, + { + "epoch": 0.1565909090909091, + "grad_norm": 3.92283294243239, + "learning_rate": 1.1760955749441786e-06, + "loss": 0.0894, + "step": 689 + }, + { + "epoch": 0.15681818181818183, + "grad_norm": 3.6624543698448955, + "learning_rate": 1.1758849337017587e-06, + "loss": 0.0825, + "step": 690 + }, + { + "epoch": 0.15704545454545454, + "grad_norm": 2.392177011821537, + "learning_rate": 1.175674011621728e-06, + "loss": 0.0841, + "step": 691 + }, + { + "epoch": 0.1572727272727273, + "grad_norm": 4.237324683508599, + "learning_rate": 1.1754628088116138e-06, + "loss": 0.0982, + "step": 692 + }, + { + "epoch": 0.1575, + "grad_norm": 4.2210592048022475, + "learning_rate": 1.1752513253790861e-06, + "loss": 0.1303, + "step": 693 + }, + { + "epoch": 0.15772727272727272, + "grad_norm": 2.725390442178886, + "learning_rate": 1.1750395614319576e-06, + "loss": 0.071, + "step": 694 + }, + { + "epoch": 0.15795454545454546, + "grad_norm": 5.200799852051973, + "learning_rate": 1.174827517078184e-06, + "loss": 0.1553, + "step": 695 + }, + { + "epoch": 0.15818181818181817, + "grad_norm": 6.03732484526907, + "learning_rate": 1.1746151924258644e-06, + "loss": 0.1253, + "step": 696 + }, + { + "epoch": 0.15840909090909092, + "grad_norm": 4.421033049553361, + "learning_rate": 1.1744025875832405e-06, + "loss": 0.1416, + "step": 697 + }, + { + "epoch": 0.15863636363636363, + "grad_norm": 2.0786119052806176, + "learning_rate": 1.174189702658697e-06, + "loss": 0.0647, + "step": 698 + }, + { + "epoch": 0.15886363636363637, + "grad_norm": 3.3198446731825917, + "learning_rate": 1.173976537760761e-06, + "loss": 0.1627, + "step": 699 + }, + { + "epoch": 0.1590909090909091, + "grad_norm": 4.0021155173324345, + "learning_rate": 1.1737630929981026e-06, + "loss": 0.125, + "step": 700 + }, + { + "epoch": 0.15931818181818183, + "grad_norm": 4.420975500308553, + "learning_rate": 1.1735493684795348e-06, + "loss": 0.1224, + "step": 701 + }, + { + "epoch": 0.15954545454545455, + "grad_norm": 2.944405277794975, + "learning_rate": 1.173335364314013e-06, + "loss": 0.1023, + "step": 702 + }, + { + "epoch": 0.15977272727272726, + "grad_norm": 3.4196656248503765, + "learning_rate": 1.173121080610635e-06, + "loss": 0.1157, + "step": 703 + }, + { + "epoch": 0.16, + "grad_norm": 2.1264439028178, + "learning_rate": 1.1729065174786414e-06, + "loss": 0.0518, + "step": 704 + }, + { + "epoch": 0.16022727272727272, + "grad_norm": 2.199405159906099, + "learning_rate": 1.1726916750274148e-06, + "loss": 0.1096, + "step": 705 + }, + { + "epoch": 0.16045454545454546, + "grad_norm": 2.645214422172338, + "learning_rate": 1.1724765533664808e-06, + "loss": 0.0507, + "step": 706 + }, + { + "epoch": 0.16068181818181818, + "grad_norm": 3.7610341760027284, + "learning_rate": 1.1722611526055073e-06, + "loss": 0.0866, + "step": 707 + }, + { + "epoch": 0.16090909090909092, + "grad_norm": 2.3219500542040032, + "learning_rate": 1.1720454728543034e-06, + "loss": 0.0604, + "step": 708 + }, + { + "epoch": 0.16113636363636363, + "grad_norm": 4.755721287440461, + "learning_rate": 1.171829514222822e-06, + "loss": 0.2037, + "step": 709 + }, + { + "epoch": 0.16136363636363638, + "grad_norm": 3.7612911000368308, + "learning_rate": 1.1716132768211572e-06, + "loss": 0.1342, + "step": 710 + }, + { + "epoch": 0.1615909090909091, + "grad_norm": 8.293683997911904, + "learning_rate": 1.1713967607595455e-06, + "loss": 0.1672, + "step": 711 + }, + { + "epoch": 0.1618181818181818, + "grad_norm": 3.964012710726141, + "learning_rate": 1.1711799661483653e-06, + "loss": 0.0778, + "step": 712 + }, + { + "epoch": 0.16204545454545455, + "grad_norm": 2.2404179991424575, + "learning_rate": 1.170962893098137e-06, + "loss": 0.0565, + "step": 713 + }, + { + "epoch": 0.16227272727272726, + "grad_norm": 1.9328320314401497, + "learning_rate": 1.1707455417195231e-06, + "loss": 0.077, + "step": 714 + }, + { + "epoch": 0.1625, + "grad_norm": 2.9441362131463116, + "learning_rate": 1.170527912123328e-06, + "loss": 0.0765, + "step": 715 + }, + { + "epoch": 0.16272727272727272, + "grad_norm": 4.329109442397501, + "learning_rate": 1.1703100044204984e-06, + "loss": 0.0953, + "step": 716 + }, + { + "epoch": 0.16295454545454546, + "grad_norm": 3.5713673629683904, + "learning_rate": 1.1700918187221214e-06, + "loss": 0.1576, + "step": 717 + }, + { + "epoch": 0.16318181818181818, + "grad_norm": 2.246960488592579, + "learning_rate": 1.169873355139427e-06, + "loss": 0.0694, + "step": 718 + }, + { + "epoch": 0.16340909090909092, + "grad_norm": 2.020881450674326, + "learning_rate": 1.1696546137837865e-06, + "loss": 0.0946, + "step": 719 + }, + { + "epoch": 0.16363636363636364, + "grad_norm": 2.740348718285887, + "learning_rate": 1.169435594766713e-06, + "loss": 0.0817, + "step": 720 + }, + { + "epoch": 0.16386363636363635, + "grad_norm": 1.7314284256627515, + "learning_rate": 1.1692162981998608e-06, + "loss": 0.1009, + "step": 721 + }, + { + "epoch": 0.1640909090909091, + "grad_norm": 3.3266677847846995, + "learning_rate": 1.1689967241950263e-06, + "loss": 0.089, + "step": 722 + }, + { + "epoch": 0.1643181818181818, + "grad_norm": 1.755824745041743, + "learning_rate": 1.168776872864146e-06, + "loss": 0.0709, + "step": 723 + }, + { + "epoch": 0.16454545454545455, + "grad_norm": 4.576832156304644, + "learning_rate": 1.1685567443192996e-06, + "loss": 0.1322, + "step": 724 + }, + { + "epoch": 0.16477272727272727, + "grad_norm": 3.3537810168372526, + "learning_rate": 1.168336338672707e-06, + "loss": 0.137, + "step": 725 + }, + { + "epoch": 0.165, + "grad_norm": 3.83015338910802, + "learning_rate": 1.1681156560367296e-06, + "loss": 0.1443, + "step": 726 + }, + { + "epoch": 0.16522727272727272, + "grad_norm": 3.8733617575663173, + "learning_rate": 1.1678946965238697e-06, + "loss": 0.209, + "step": 727 + }, + { + "epoch": 0.16545454545454547, + "grad_norm": 1.6989197516038992, + "learning_rate": 1.1676734602467713e-06, + "loss": 0.0471, + "step": 728 + }, + { + "epoch": 0.16568181818181818, + "grad_norm": 5.241436581266166, + "learning_rate": 1.1674519473182192e-06, + "loss": 0.0842, + "step": 729 + }, + { + "epoch": 0.16590909090909092, + "grad_norm": 2.878713840863321, + "learning_rate": 1.1672301578511392e-06, + "loss": 0.1135, + "step": 730 + }, + { + "epoch": 0.16613636363636364, + "grad_norm": 1.928845528223969, + "learning_rate": 1.167008091958598e-06, + "loss": 0.0552, + "step": 731 + }, + { + "epoch": 0.16636363636363635, + "grad_norm": 2.4174286141337933, + "learning_rate": 1.1667857497538037e-06, + "loss": 0.0556, + "step": 732 + }, + { + "epoch": 0.1665909090909091, + "grad_norm": 3.2329134838761493, + "learning_rate": 1.166563131350105e-06, + "loss": 0.1427, + "step": 733 + }, + { + "epoch": 0.1668181818181818, + "grad_norm": 2.813126509147356, + "learning_rate": 1.166340236860991e-06, + "loss": 0.1662, + "step": 734 + }, + { + "epoch": 0.16704545454545455, + "grad_norm": 2.9201085192791494, + "learning_rate": 1.166117066400092e-06, + "loss": 0.125, + "step": 735 + }, + { + "epoch": 0.16727272727272727, + "grad_norm": 3.191862604606099, + "learning_rate": 1.1658936200811789e-06, + "loss": 0.1119, + "step": 736 + }, + { + "epoch": 0.1675, + "grad_norm": 2.699796433803432, + "learning_rate": 1.1656698980181633e-06, + "loss": 0.1217, + "step": 737 + }, + { + "epoch": 0.16772727272727272, + "grad_norm": 3.538530142721879, + "learning_rate": 1.1654459003250971e-06, + "loss": 0.0691, + "step": 738 + }, + { + "epoch": 0.16795454545454547, + "grad_norm": 2.6070942677517923, + "learning_rate": 1.1652216271161728e-06, + "loss": 0.0565, + "step": 739 + }, + { + "epoch": 0.16818181818181818, + "grad_norm": 3.887246723964787, + "learning_rate": 1.1649970785057238e-06, + "loss": 0.1807, + "step": 740 + }, + { + "epoch": 0.1684090909090909, + "grad_norm": 2.759394903870801, + "learning_rate": 1.1647722546082232e-06, + "loss": 0.0616, + "step": 741 + }, + { + "epoch": 0.16863636363636364, + "grad_norm": 2.536242923816874, + "learning_rate": 1.164547155538285e-06, + "loss": 0.0958, + "step": 742 + }, + { + "epoch": 0.16886363636363635, + "grad_norm": 4.118117887483789, + "learning_rate": 1.1643217814106633e-06, + "loss": 0.0876, + "step": 743 + }, + { + "epoch": 0.1690909090909091, + "grad_norm": 4.078654365260598, + "learning_rate": 1.1640961323402522e-06, + "loss": 0.0988, + "step": 744 + }, + { + "epoch": 0.1693181818181818, + "grad_norm": 3.0828097044712037, + "learning_rate": 1.163870208442086e-06, + "loss": 0.1264, + "step": 745 + }, + { + "epoch": 0.16954545454545455, + "grad_norm": 3.062974378424102, + "learning_rate": 1.1636440098313398e-06, + "loss": 0.1078, + "step": 746 + }, + { + "epoch": 0.16977272727272727, + "grad_norm": 2.292498000650939, + "learning_rate": 1.1634175366233278e-06, + "loss": 0.1396, + "step": 747 + }, + { + "epoch": 0.17, + "grad_norm": 2.8740313269870352, + "learning_rate": 1.1631907889335046e-06, + "loss": 0.0612, + "step": 748 + }, + { + "epoch": 0.17022727272727273, + "grad_norm": 3.934309592606626, + "learning_rate": 1.162963766877465e-06, + "loss": 0.0892, + "step": 749 + }, + { + "epoch": 0.17045454545454544, + "grad_norm": 2.8254071182276745, + "learning_rate": 1.1627364705709428e-06, + "loss": 0.133, + "step": 750 + }, + { + "epoch": 0.17068181818181818, + "grad_norm": 4.93486206217756, + "learning_rate": 1.1625089001298129e-06, + "loss": 0.1305, + "step": 751 + }, + { + "epoch": 0.1709090909090909, + "grad_norm": 4.931372452787778, + "learning_rate": 1.1622810556700889e-06, + "loss": 0.2143, + "step": 752 + }, + { + "epoch": 0.17113636363636364, + "grad_norm": 3.4399850531012706, + "learning_rate": 1.1620529373079246e-06, + "loss": 0.1141, + "step": 753 + }, + { + "epoch": 0.17136363636363636, + "grad_norm": 2.695962220389638, + "learning_rate": 1.1618245451596128e-06, + "loss": 0.0696, + "step": 754 + }, + { + "epoch": 0.1715909090909091, + "grad_norm": 0.9438166658732292, + "learning_rate": 1.161595879341587e-06, + "loss": 0.0272, + "step": 755 + }, + { + "epoch": 0.17181818181818181, + "grad_norm": 3.6425295982084336, + "learning_rate": 1.1613669399704192e-06, + "loss": 0.1209, + "step": 756 + }, + { + "epoch": 0.17204545454545456, + "grad_norm": 3.0562439281759777, + "learning_rate": 1.1611377271628213e-06, + "loss": 0.1224, + "step": 757 + }, + { + "epoch": 0.17227272727272727, + "grad_norm": 2.0260641858137154, + "learning_rate": 1.1609082410356447e-06, + "loss": 0.0463, + "step": 758 + }, + { + "epoch": 0.1725, + "grad_norm": 3.881360237771756, + "learning_rate": 1.1606784817058797e-06, + "loss": 0.1443, + "step": 759 + }, + { + "epoch": 0.17272727272727273, + "grad_norm": 4.920258329077468, + "learning_rate": 1.1604484492906562e-06, + "loss": 0.1928, + "step": 760 + }, + { + "epoch": 0.17295454545454544, + "grad_norm": 4.678867199792111, + "learning_rate": 1.1602181439072432e-06, + "loss": 0.0745, + "step": 761 + }, + { + "epoch": 0.1731818181818182, + "grad_norm": 4.512860912312718, + "learning_rate": 1.1599875656730492e-06, + "loss": 0.1652, + "step": 762 + }, + { + "epoch": 0.1734090909090909, + "grad_norm": 1.6457019733022824, + "learning_rate": 1.159756714705621e-06, + "loss": 0.0806, + "step": 763 + }, + { + "epoch": 0.17363636363636364, + "grad_norm": 1.9859359375789238, + "learning_rate": 1.1595255911226456e-06, + "loss": 0.0614, + "step": 764 + }, + { + "epoch": 0.17386363636363636, + "grad_norm": 3.091855498137776, + "learning_rate": 1.1592941950419475e-06, + "loss": 0.0941, + "step": 765 + }, + { + "epoch": 0.1740909090909091, + "grad_norm": 3.19444562241392, + "learning_rate": 1.1590625265814918e-06, + "loss": 0.1728, + "step": 766 + }, + { + "epoch": 0.17431818181818182, + "grad_norm": 2.1360602458385602, + "learning_rate": 1.1588305858593811e-06, + "loss": 0.077, + "step": 767 + }, + { + "epoch": 0.17454545454545456, + "grad_norm": 4.158284403412673, + "learning_rate": 1.1585983729938575e-06, + "loss": 0.104, + "step": 768 + }, + { + "epoch": 0.17477272727272727, + "grad_norm": 3.263432773676213, + "learning_rate": 1.1583658881033013e-06, + "loss": 0.0968, + "step": 769 + }, + { + "epoch": 0.175, + "grad_norm": 4.121141338929559, + "learning_rate": 1.158133131306232e-06, + "loss": 0.0811, + "step": 770 + }, + { + "epoch": 0.17522727272727273, + "grad_norm": 3.610070838376739, + "learning_rate": 1.1579001027213078e-06, + "loss": 0.0787, + "step": 771 + }, + { + "epoch": 0.17545454545454545, + "grad_norm": 3.4342892106173424, + "learning_rate": 1.1576668024673248e-06, + "loss": 0.0825, + "step": 772 + }, + { + "epoch": 0.1756818181818182, + "grad_norm": 3.111710435324402, + "learning_rate": 1.157433230663218e-06, + "loss": 0.0752, + "step": 773 + }, + { + "epoch": 0.1759090909090909, + "grad_norm": 3.601238284194029, + "learning_rate": 1.1571993874280611e-06, + "loss": 0.1352, + "step": 774 + }, + { + "epoch": 0.17613636363636365, + "grad_norm": 2.6535049416196514, + "learning_rate": 1.1569652728810658e-06, + "loss": 0.0802, + "step": 775 + }, + { + "epoch": 0.17636363636363636, + "grad_norm": 4.135370866476488, + "learning_rate": 1.156730887141582e-06, + "loss": 0.1253, + "step": 776 + }, + { + "epoch": 0.1765909090909091, + "grad_norm": 3.3574842071078614, + "learning_rate": 1.156496230329098e-06, + "loss": 0.1816, + "step": 777 + }, + { + "epoch": 0.17681818181818182, + "grad_norm": 1.4563969441054139, + "learning_rate": 1.1562613025632406e-06, + "loss": 0.0333, + "step": 778 + }, + { + "epoch": 0.17704545454545453, + "grad_norm": 2.115189213131307, + "learning_rate": 1.1560261039637744e-06, + "loss": 0.0391, + "step": 779 + }, + { + "epoch": 0.17727272727272728, + "grad_norm": 3.151419061116544, + "learning_rate": 1.1557906346506021e-06, + "loss": 0.0794, + "step": 780 + }, + { + "epoch": 0.1775, + "grad_norm": 4.474835234569177, + "learning_rate": 1.1555548947437642e-06, + "loss": 0.1448, + "step": 781 + }, + { + "epoch": 0.17772727272727273, + "grad_norm": 3.9564645686320477, + "learning_rate": 1.1553188843634399e-06, + "loss": 0.0948, + "step": 782 + }, + { + "epoch": 0.17795454545454545, + "grad_norm": 4.166484993015325, + "learning_rate": 1.1550826036299455e-06, + "loss": 0.1863, + "step": 783 + }, + { + "epoch": 0.1781818181818182, + "grad_norm": 4.219755482501418, + "learning_rate": 1.1548460526637354e-06, + "loss": 0.1917, + "step": 784 + }, + { + "epoch": 0.1784090909090909, + "grad_norm": 5.062985347209895, + "learning_rate": 1.1546092315854017e-06, + "loss": 0.1081, + "step": 785 + }, + { + "epoch": 0.17863636363636365, + "grad_norm": 2.386537802996193, + "learning_rate": 1.1543721405156744e-06, + "loss": 0.0996, + "step": 786 + }, + { + "epoch": 0.17886363636363636, + "grad_norm": 2.1778672692961734, + "learning_rate": 1.154134779575421e-06, + "loss": 0.1112, + "step": 787 + }, + { + "epoch": 0.17909090909090908, + "grad_norm": 3.1874664823281047, + "learning_rate": 1.1538971488856465e-06, + "loss": 0.0909, + "step": 788 + }, + { + "epoch": 0.17931818181818182, + "grad_norm": 2.491214246337047, + "learning_rate": 1.153659248567494e-06, + "loss": 0.0669, + "step": 789 + }, + { + "epoch": 0.17954545454545454, + "grad_norm": 2.636713363199442, + "learning_rate": 1.1534210787422425e-06, + "loss": 0.0781, + "step": 790 + }, + { + "epoch": 0.17977272727272728, + "grad_norm": 2.1557068773481456, + "learning_rate": 1.1531826395313104e-06, + "loss": 0.0524, + "step": 791 + }, + { + "epoch": 0.18, + "grad_norm": 5.769207254944547, + "learning_rate": 1.152943931056252e-06, + "loss": 0.281, + "step": 792 + }, + { + "epoch": 0.18022727272727274, + "grad_norm": 3.3586606679709936, + "learning_rate": 1.1527049534387595e-06, + "loss": 0.1963, + "step": 793 + }, + { + "epoch": 0.18045454545454545, + "grad_norm": 2.936113305720301, + "learning_rate": 1.1524657068006622e-06, + "loss": 0.0656, + "step": 794 + }, + { + "epoch": 0.1806818181818182, + "grad_norm": 2.0405530911531016, + "learning_rate": 1.1522261912639266e-06, + "loss": 0.0585, + "step": 795 + }, + { + "epoch": 0.1809090909090909, + "grad_norm": 3.0192839207172413, + "learning_rate": 1.151986406950656e-06, + "loss": 0.0905, + "step": 796 + }, + { + "epoch": 0.18113636363636362, + "grad_norm": 3.697720053741981, + "learning_rate": 1.1517463539830908e-06, + "loss": 0.1993, + "step": 797 + }, + { + "epoch": 0.18136363636363637, + "grad_norm": 1.986111330717534, + "learning_rate": 1.1515060324836088e-06, + "loss": 0.0708, + "step": 798 + }, + { + "epoch": 0.18159090909090908, + "grad_norm": 3.951465859999876, + "learning_rate": 1.151265442574724e-06, + "loss": 0.2031, + "step": 799 + }, + { + "epoch": 0.18181818181818182, + "grad_norm": 4.61905370493032, + "learning_rate": 1.151024584379088e-06, + "loss": 0.1098, + "step": 800 + }, + { + "epoch": 0.18204545454545454, + "grad_norm": 3.867575536372285, + "learning_rate": 1.1507834580194883e-06, + "loss": 0.17, + "step": 801 + }, + { + "epoch": 0.18227272727272728, + "grad_norm": 2.288884928836873, + "learning_rate": 1.15054206361885e-06, + "loss": 0.0515, + "step": 802 + }, + { + "epoch": 0.1825, + "grad_norm": 3.7389399908616348, + "learning_rate": 1.1503004013002341e-06, + "loss": 0.0951, + "step": 803 + }, + { + "epoch": 0.18272727272727274, + "grad_norm": 5.6499972029633545, + "learning_rate": 1.1500584711868388e-06, + "loss": 0.1508, + "step": 804 + }, + { + "epoch": 0.18295454545454545, + "grad_norm": 3.797091938549479, + "learning_rate": 1.1498162734019983e-06, + "loss": 0.0982, + "step": 805 + }, + { + "epoch": 0.1831818181818182, + "grad_norm": 5.437372581085665, + "learning_rate": 1.1495738080691834e-06, + "loss": 0.134, + "step": 806 + }, + { + "epoch": 0.1834090909090909, + "grad_norm": 5.605488520067354, + "learning_rate": 1.1493310753120017e-06, + "loss": 0.1869, + "step": 807 + }, + { + "epoch": 0.18363636363636363, + "grad_norm": 3.410884346668656, + "learning_rate": 1.1490880752541967e-06, + "loss": 0.1151, + "step": 808 + }, + { + "epoch": 0.18386363636363637, + "grad_norm": 2.496107182404278, + "learning_rate": 1.148844808019648e-06, + "loss": 0.1254, + "step": 809 + }, + { + "epoch": 0.18409090909090908, + "grad_norm": 2.5010611206664253, + "learning_rate": 1.1486012737323718e-06, + "loss": 0.0632, + "step": 810 + }, + { + "epoch": 0.18431818181818183, + "grad_norm": 6.3840587248432925, + "learning_rate": 1.1483574725165207e-06, + "loss": 0.2068, + "step": 811 + }, + { + "epoch": 0.18454545454545454, + "grad_norm": 1.1693017176882667, + "learning_rate": 1.1481134044963825e-06, + "loss": 0.0453, + "step": 812 + }, + { + "epoch": 0.18477272727272728, + "grad_norm": 2.5986432528058394, + "learning_rate": 1.1478690697963814e-06, + "loss": 0.1016, + "step": 813 + }, + { + "epoch": 0.185, + "grad_norm": 2.5746770787719453, + "learning_rate": 1.1476244685410784e-06, + "loss": 0.0642, + "step": 814 + }, + { + "epoch": 0.18522727272727274, + "grad_norm": 3.39089320442834, + "learning_rate": 1.147379600855169e-06, + "loss": 0.1876, + "step": 815 + }, + { + "epoch": 0.18545454545454546, + "grad_norm": 3.1117172658023313, + "learning_rate": 1.1471344668634854e-06, + "loss": 0.1246, + "step": 816 + }, + { + "epoch": 0.18568181818181817, + "grad_norm": 3.59479608999141, + "learning_rate": 1.1468890666909954e-06, + "loss": 0.148, + "step": 817 + }, + { + "epoch": 0.1859090909090909, + "grad_norm": 4.040931884819615, + "learning_rate": 1.1466434004628023e-06, + "loss": 0.1352, + "step": 818 + }, + { + "epoch": 0.18613636363636363, + "grad_norm": 3.523166963086899, + "learning_rate": 1.1463974683041455e-06, + "loss": 0.1357, + "step": 819 + }, + { + "epoch": 0.18636363636363637, + "grad_norm": 3.1919367634166957, + "learning_rate": 1.1461512703403992e-06, + "loss": 0.1144, + "step": 820 + }, + { + "epoch": 0.18659090909090909, + "grad_norm": 2.74843977378851, + "learning_rate": 1.1459048066970736e-06, + "loss": 0.084, + "step": 821 + }, + { + "epoch": 0.18681818181818183, + "grad_norm": 2.333067256985285, + "learning_rate": 1.1456580774998146e-06, + "loss": 0.0815, + "step": 822 + }, + { + "epoch": 0.18704545454545454, + "grad_norm": 2.8285465665063607, + "learning_rate": 1.1454110828744027e-06, + "loss": 0.0968, + "step": 823 + }, + { + "epoch": 0.18727272727272729, + "grad_norm": 3.2628114799373877, + "learning_rate": 1.1451638229467547e-06, + "loss": 0.0792, + "step": 824 + }, + { + "epoch": 0.1875, + "grad_norm": 2.7169244214423585, + "learning_rate": 1.1449162978429218e-06, + "loss": 0.0661, + "step": 825 + }, + { + "epoch": 0.18772727272727271, + "grad_norm": 4.461674334515633, + "learning_rate": 1.144668507689091e-06, + "loss": 0.1074, + "step": 826 + }, + { + "epoch": 0.18795454545454546, + "grad_norm": 3.9093425705449563, + "learning_rate": 1.1444204526115837e-06, + "loss": 0.1075, + "step": 827 + }, + { + "epoch": 0.18818181818181817, + "grad_norm": 3.0801888841111, + "learning_rate": 1.144172132736857e-06, + "loss": 0.0743, + "step": 828 + }, + { + "epoch": 0.18840909090909091, + "grad_norm": 2.497736335549103, + "learning_rate": 1.1439235481915028e-06, + "loss": 0.0636, + "step": 829 + }, + { + "epoch": 0.18863636363636363, + "grad_norm": 4.036541868534562, + "learning_rate": 1.1436746991022479e-06, + "loss": 0.0751, + "step": 830 + }, + { + "epoch": 0.18886363636363637, + "grad_norm": 2.0919249932718205, + "learning_rate": 1.143425585595954e-06, + "loss": 0.0837, + "step": 831 + }, + { + "epoch": 0.1890909090909091, + "grad_norm": 3.4801565927300677, + "learning_rate": 1.1431762077996174e-06, + "loss": 0.1468, + "step": 832 + }, + { + "epoch": 0.18931818181818183, + "grad_norm": 2.084375174684, + "learning_rate": 1.1429265658403698e-06, + "loss": 0.0424, + "step": 833 + }, + { + "epoch": 0.18954545454545454, + "grad_norm": 3.5196390678506377, + "learning_rate": 1.1426766598454768e-06, + "loss": 0.0829, + "step": 834 + }, + { + "epoch": 0.18977272727272726, + "grad_norm": 2.2068595873641894, + "learning_rate": 1.1424264899423383e-06, + "loss": 0.0886, + "step": 835 + }, + { + "epoch": 0.19, + "grad_norm": 3.3459857680580765, + "learning_rate": 1.1421760562584901e-06, + "loss": 0.1178, + "step": 836 + }, + { + "epoch": 0.19022727272727272, + "grad_norm": 2.0411126728527216, + "learning_rate": 1.1419253589216012e-06, + "loss": 0.1094, + "step": 837 + }, + { + "epoch": 0.19045454545454546, + "grad_norm": 3.829790474004804, + "learning_rate": 1.1416743980594758e-06, + "loss": 0.09, + "step": 838 + }, + { + "epoch": 0.19068181818181817, + "grad_norm": 2.7388804788347527, + "learning_rate": 1.1414231738000521e-06, + "loss": 0.1271, + "step": 839 + }, + { + "epoch": 0.19090909090909092, + "grad_norm": 2.834565204110733, + "learning_rate": 1.1411716862714027e-06, + "loss": 0.0722, + "step": 840 + }, + { + "epoch": 0.19113636363636363, + "grad_norm": 4.000841536291183, + "learning_rate": 1.1409199356017339e-06, + "loss": 0.1025, + "step": 841 + }, + { + "epoch": 0.19136363636363637, + "grad_norm": 3.2525668025020287, + "learning_rate": 1.140667921919387e-06, + "loss": 0.1952, + "step": 842 + }, + { + "epoch": 0.1915909090909091, + "grad_norm": 3.3381337964962055, + "learning_rate": 1.1404156453528367e-06, + "loss": 0.1035, + "step": 843 + }, + { + "epoch": 0.1918181818181818, + "grad_norm": 4.5141519965291685, + "learning_rate": 1.1401631060306921e-06, + "loss": 0.1128, + "step": 844 + }, + { + "epoch": 0.19204545454545455, + "grad_norm": 3.0066688758504956, + "learning_rate": 1.1399103040816963e-06, + "loss": 0.1182, + "step": 845 + }, + { + "epoch": 0.19227272727272726, + "grad_norm": 4.723192498623169, + "learning_rate": 1.1396572396347257e-06, + "loss": 0.2221, + "step": 846 + }, + { + "epoch": 0.1925, + "grad_norm": 2.8797172585471578, + "learning_rate": 1.1394039128187914e-06, + "loss": 0.1145, + "step": 847 + }, + { + "epoch": 0.19272727272727272, + "grad_norm": 5.144087657615666, + "learning_rate": 1.1391503237630375e-06, + "loss": 0.141, + "step": 848 + }, + { + "epoch": 0.19295454545454546, + "grad_norm": 2.973541767107418, + "learning_rate": 1.1388964725967423e-06, + "loss": 0.085, + "step": 849 + }, + { + "epoch": 0.19318181818181818, + "grad_norm": 1.575144106160146, + "learning_rate": 1.138642359449317e-06, + "loss": 0.0433, + "step": 850 + }, + { + "epoch": 0.19340909090909092, + "grad_norm": 3.3567469540831123, + "learning_rate": 1.1383879844503073e-06, + "loss": 0.0672, + "step": 851 + }, + { + "epoch": 0.19363636363636363, + "grad_norm": 2.8085788785735906, + "learning_rate": 1.1381333477293918e-06, + "loss": 0.0958, + "step": 852 + }, + { + "epoch": 0.19386363636363638, + "grad_norm": 3.7814802768345297, + "learning_rate": 1.137878449416383e-06, + "loss": 0.1161, + "step": 853 + }, + { + "epoch": 0.1940909090909091, + "grad_norm": 2.629569362701198, + "learning_rate": 1.137623289641226e-06, + "loss": 0.1446, + "step": 854 + }, + { + "epoch": 0.1943181818181818, + "grad_norm": 2.0645333851737186, + "learning_rate": 1.1373678685339994e-06, + "loss": 0.065, + "step": 855 + }, + { + "epoch": 0.19454545454545455, + "grad_norm": 1.8126540163890332, + "learning_rate": 1.137112186224916e-06, + "loss": 0.1073, + "step": 856 + }, + { + "epoch": 0.19477272727272726, + "grad_norm": 4.304917434651155, + "learning_rate": 1.1368562428443205e-06, + "loss": 0.1071, + "step": 857 + }, + { + "epoch": 0.195, + "grad_norm": 4.8593705421564035, + "learning_rate": 1.1366000385226913e-06, + "loss": 0.1335, + "step": 858 + }, + { + "epoch": 0.19522727272727272, + "grad_norm": 2.301130421885278, + "learning_rate": 1.1363435733906398e-06, + "loss": 0.0525, + "step": 859 + }, + { + "epoch": 0.19545454545454546, + "grad_norm": 4.242517304753756, + "learning_rate": 1.13608684757891e-06, + "loss": 0.1251, + "step": 860 + }, + { + "epoch": 0.19568181818181818, + "grad_norm": 1.8661570786899095, + "learning_rate": 1.1358298612183793e-06, + "loss": 0.0556, + "step": 861 + }, + { + "epoch": 0.19590909090909092, + "grad_norm": 2.5444055328479163, + "learning_rate": 1.135572614440058e-06, + "loss": 0.1222, + "step": 862 + }, + { + "epoch": 0.19613636363636364, + "grad_norm": 3.2357632589129697, + "learning_rate": 1.1353151073750882e-06, + "loss": 0.0935, + "step": 863 + }, + { + "epoch": 0.19636363636363635, + "grad_norm": 3.9981685641430738, + "learning_rate": 1.1350573401547457e-06, + "loss": 0.1166, + "step": 864 + }, + { + "epoch": 0.1965909090909091, + "grad_norm": 4.379534172721873, + "learning_rate": 1.1347993129104386e-06, + "loss": 0.1062, + "step": 865 + }, + { + "epoch": 0.1968181818181818, + "grad_norm": 4.768650537412762, + "learning_rate": 1.1345410257737078e-06, + "loss": 0.1199, + "step": 866 + }, + { + "epoch": 0.19704545454545455, + "grad_norm": 2.853710045167137, + "learning_rate": 1.1342824788762258e-06, + "loss": 0.0893, + "step": 867 + }, + { + "epoch": 0.19727272727272727, + "grad_norm": 2.335354308975055, + "learning_rate": 1.1340236723497985e-06, + "loss": 0.0851, + "step": 868 + }, + { + "epoch": 0.1975, + "grad_norm": 2.8747515548498814, + "learning_rate": 1.133764606326364e-06, + "loss": 0.0773, + "step": 869 + }, + { + "epoch": 0.19772727272727272, + "grad_norm": 2.1784702004332033, + "learning_rate": 1.1335052809379921e-06, + "loss": 0.0661, + "step": 870 + }, + { + "epoch": 0.19795454545454547, + "grad_norm": 4.106341975791218, + "learning_rate": 1.1332456963168854e-06, + "loss": 0.1477, + "step": 871 + }, + { + "epoch": 0.19818181818181818, + "grad_norm": 4.977389286172251, + "learning_rate": 1.1329858525953785e-06, + "loss": 0.1869, + "step": 872 + }, + { + "epoch": 0.1984090909090909, + "grad_norm": 2.317151812877045, + "learning_rate": 1.132725749905938e-06, + "loss": 0.077, + "step": 873 + }, + { + "epoch": 0.19863636363636364, + "grad_norm": 2.552204007391219, + "learning_rate": 1.132465388381163e-06, + "loss": 0.1181, + "step": 874 + }, + { + "epoch": 0.19886363636363635, + "grad_norm": 3.6600451842047352, + "learning_rate": 1.1322047681537834e-06, + "loss": 0.0594, + "step": 875 + }, + { + "epoch": 0.1990909090909091, + "grad_norm": 2.5082881600578317, + "learning_rate": 1.131943889356662e-06, + "loss": 0.0994, + "step": 876 + }, + { + "epoch": 0.1993181818181818, + "grad_norm": 2.1261281732928103, + "learning_rate": 1.1316827521227935e-06, + "loss": 0.0561, + "step": 877 + }, + { + "epoch": 0.19954545454545455, + "grad_norm": 5.257878400649931, + "learning_rate": 1.1314213565853036e-06, + "loss": 0.1005, + "step": 878 + }, + { + "epoch": 0.19977272727272727, + "grad_norm": 2.576253488114408, + "learning_rate": 1.1311597028774503e-06, + "loss": 0.0891, + "step": 879 + }, + { + "epoch": 0.2, + "grad_norm": 2.8279020911423713, + "learning_rate": 1.1308977911326229e-06, + "loss": 0.0759, + "step": 880 + }, + { + "epoch": 0.20022727272727273, + "grad_norm": 4.484852640090644, + "learning_rate": 1.1306356214843423e-06, + "loss": 0.2537, + "step": 881 + }, + { + "epoch": 0.20045454545454544, + "grad_norm": 7.272905373261484, + "learning_rate": 1.1303731940662608e-06, + "loss": 0.1503, + "step": 882 + }, + { + "epoch": 0.20068181818181818, + "grad_norm": 4.774637236618464, + "learning_rate": 1.1301105090121624e-06, + "loss": 0.1941, + "step": 883 + }, + { + "epoch": 0.2009090909090909, + "grad_norm": 3.438028953018756, + "learning_rate": 1.1298475664559622e-06, + "loss": 0.1303, + "step": 884 + }, + { + "epoch": 0.20113636363636364, + "grad_norm": 2.774482146550907, + "learning_rate": 1.1295843665317067e-06, + "loss": 0.0704, + "step": 885 + }, + { + "epoch": 0.20136363636363636, + "grad_norm": 2.7265476753353646, + "learning_rate": 1.1293209093735732e-06, + "loss": 0.08, + "step": 886 + }, + { + "epoch": 0.2015909090909091, + "grad_norm": 4.43422406285228, + "learning_rate": 1.129057195115871e-06, + "loss": 0.1307, + "step": 887 + }, + { + "epoch": 0.2018181818181818, + "grad_norm": 3.6975839958235643, + "learning_rate": 1.1287932238930397e-06, + "loss": 0.1302, + "step": 888 + }, + { + "epoch": 0.20204545454545456, + "grad_norm": 3.831455980792235, + "learning_rate": 1.12852899583965e-06, + "loss": 0.1898, + "step": 889 + }, + { + "epoch": 0.20227272727272727, + "grad_norm": 2.436933066118174, + "learning_rate": 1.1282645110904036e-06, + "loss": 0.098, + "step": 890 + }, + { + "epoch": 0.2025, + "grad_norm": 1.7773442635177201, + "learning_rate": 1.1279997697801334e-06, + "loss": 0.0899, + "step": 891 + }, + { + "epoch": 0.20272727272727273, + "grad_norm": 2.1032029335871543, + "learning_rate": 1.1277347720438028e-06, + "loss": 0.145, + "step": 892 + }, + { + "epoch": 0.20295454545454544, + "grad_norm": 2.770331507409027, + "learning_rate": 1.127469518016506e-06, + "loss": 0.1044, + "step": 893 + }, + { + "epoch": 0.20318181818181819, + "grad_norm": 3.5441778156585295, + "learning_rate": 1.1272040078334675e-06, + "loss": 0.1497, + "step": 894 + }, + { + "epoch": 0.2034090909090909, + "grad_norm": 2.6893741919723695, + "learning_rate": 1.1269382416300429e-06, + "loss": 0.0777, + "step": 895 + }, + { + "epoch": 0.20363636363636364, + "grad_norm": 2.2841890720733473, + "learning_rate": 1.126672219541718e-06, + "loss": 0.118, + "step": 896 + }, + { + "epoch": 0.20386363636363636, + "grad_norm": 5.540763430217786, + "learning_rate": 1.1264059417041095e-06, + "loss": 0.1391, + "step": 897 + }, + { + "epoch": 0.2040909090909091, + "grad_norm": 2.8610742683601242, + "learning_rate": 1.1261394082529638e-06, + "loss": 0.0646, + "step": 898 + }, + { + "epoch": 0.20431818181818182, + "grad_norm": 2.872031002882643, + "learning_rate": 1.1258726193241578e-06, + "loss": 0.0948, + "step": 899 + }, + { + "epoch": 0.20454545454545456, + "grad_norm": 3.8751449143375227, + "learning_rate": 1.1256055750536992e-06, + "loss": 0.0832, + "step": 900 + }, + { + "epoch": 0.20477272727272727, + "grad_norm": 2.5799813110994547, + "learning_rate": 1.1253382755777253e-06, + "loss": 0.065, + "step": 901 + }, + { + "epoch": 0.205, + "grad_norm": 3.444138034227412, + "learning_rate": 1.1250707210325036e-06, + "loss": 0.1795, + "step": 902 + }, + { + "epoch": 0.20522727272727273, + "grad_norm": 5.460686045839307, + "learning_rate": 1.1248029115544319e-06, + "loss": 0.1481, + "step": 903 + }, + { + "epoch": 0.20545454545454545, + "grad_norm": 2.2262385807894316, + "learning_rate": 1.1245348472800372e-06, + "loss": 0.1118, + "step": 904 + }, + { + "epoch": 0.2056818181818182, + "grad_norm": 3.1727301779256627, + "learning_rate": 1.1242665283459775e-06, + "loss": 0.1029, + "step": 905 + }, + { + "epoch": 0.2059090909090909, + "grad_norm": 2.4975066772858803, + "learning_rate": 1.1239979548890398e-06, + "loss": 0.074, + "step": 906 + }, + { + "epoch": 0.20613636363636365, + "grad_norm": 4.086147455795527, + "learning_rate": 1.1237291270461413e-06, + "loss": 0.107, + "step": 907 + }, + { + "epoch": 0.20636363636363636, + "grad_norm": 1.4946254182340941, + "learning_rate": 1.1234600449543288e-06, + "loss": 0.0776, + "step": 908 + }, + { + "epoch": 0.2065909090909091, + "grad_norm": 2.3309728888609236, + "learning_rate": 1.123190708750778e-06, + "loss": 0.102, + "step": 909 + }, + { + "epoch": 0.20681818181818182, + "grad_norm": 3.2589772617261548, + "learning_rate": 1.1229211185727957e-06, + "loss": 0.0809, + "step": 910 + }, + { + "epoch": 0.20704545454545453, + "grad_norm": 2.2062518677859733, + "learning_rate": 1.1226512745578166e-06, + "loss": 0.0651, + "step": 911 + }, + { + "epoch": 0.20727272727272728, + "grad_norm": 3.5393690725859095, + "learning_rate": 1.1223811768434057e-06, + "loss": 0.1055, + "step": 912 + }, + { + "epoch": 0.2075, + "grad_norm": 2.8541449566637405, + "learning_rate": 1.122110825567257e-06, + "loss": 0.1093, + "step": 913 + }, + { + "epoch": 0.20772727272727273, + "grad_norm": 4.957816351715077, + "learning_rate": 1.1218402208671938e-06, + "loss": 0.13, + "step": 914 + }, + { + "epoch": 0.20795454545454545, + "grad_norm": 2.3856979214899052, + "learning_rate": 1.1215693628811688e-06, + "loss": 0.0851, + "step": 915 + }, + { + "epoch": 0.2081818181818182, + "grad_norm": 2.636273651043917, + "learning_rate": 1.1212982517472636e-06, + "loss": 0.095, + "step": 916 + }, + { + "epoch": 0.2084090909090909, + "grad_norm": 3.628394270454844, + "learning_rate": 1.1210268876036888e-06, + "loss": 0.0791, + "step": 917 + }, + { + "epoch": 0.20863636363636365, + "grad_norm": 3.245880589677805, + "learning_rate": 1.1207552705887841e-06, + "loss": 0.0988, + "step": 918 + }, + { + "epoch": 0.20886363636363636, + "grad_norm": 3.54027197356293, + "learning_rate": 1.1204834008410184e-06, + "loss": 0.1314, + "step": 919 + }, + { + "epoch": 0.20909090909090908, + "grad_norm": 2.5694677638108256, + "learning_rate": 1.1202112784989891e-06, + "loss": 0.0757, + "step": 920 + }, + { + "epoch": 0.20931818181818182, + "grad_norm": 5.172062888883635, + "learning_rate": 1.1199389037014221e-06, + "loss": 0.19, + "step": 921 + }, + { + "epoch": 0.20954545454545453, + "grad_norm": 4.417472036448215, + "learning_rate": 1.1196662765871725e-06, + "loss": 0.1033, + "step": 922 + }, + { + "epoch": 0.20977272727272728, + "grad_norm": 2.967313003114298, + "learning_rate": 1.119393397295224e-06, + "loss": 0.1437, + "step": 923 + }, + { + "epoch": 0.21, + "grad_norm": 2.230029214295421, + "learning_rate": 1.1191202659646883e-06, + "loss": 0.1167, + "step": 924 + }, + { + "epoch": 0.21022727272727273, + "grad_norm": 3.0089456972634214, + "learning_rate": 1.1188468827348066e-06, + "loss": 0.1701, + "step": 925 + }, + { + "epoch": 0.21045454545454545, + "grad_norm": 1.5419612850832471, + "learning_rate": 1.1185732477449475e-06, + "loss": 0.0352, + "step": 926 + }, + { + "epoch": 0.2106818181818182, + "grad_norm": 1.089912645411623, + "learning_rate": 1.1182993611346084e-06, + "loss": 0.0474, + "step": 927 + }, + { + "epoch": 0.2109090909090909, + "grad_norm": 2.376305552810638, + "learning_rate": 1.1180252230434151e-06, + "loss": 0.1099, + "step": 928 + }, + { + "epoch": 0.21113636363636365, + "grad_norm": 2.9229857750989963, + "learning_rate": 1.1177508336111215e-06, + "loss": 0.0904, + "step": 929 + }, + { + "epoch": 0.21136363636363636, + "grad_norm": 2.9386514942478197, + "learning_rate": 1.1174761929776094e-06, + "loss": 0.1036, + "step": 930 + }, + { + "epoch": 0.21159090909090908, + "grad_norm": 3.70985402340233, + "learning_rate": 1.1172013012828887e-06, + "loss": 0.155, + "step": 931 + }, + { + "epoch": 0.21181818181818182, + "grad_norm": 3.023041463909678, + "learning_rate": 1.1169261586670976e-06, + "loss": 0.0769, + "step": 932 + }, + { + "epoch": 0.21204545454545454, + "grad_norm": 2.418338668648173, + "learning_rate": 1.1166507652705021e-06, + "loss": 0.0898, + "step": 933 + }, + { + "epoch": 0.21227272727272728, + "grad_norm": 1.5932507975771357, + "learning_rate": 1.1163751212334962e-06, + "loss": 0.0754, + "step": 934 + }, + { + "epoch": 0.2125, + "grad_norm": 4.299118638141458, + "learning_rate": 1.116099226696601e-06, + "loss": 0.092, + "step": 935 + }, + { + "epoch": 0.21272727272727274, + "grad_norm": 1.3404163487083518, + "learning_rate": 1.1158230818004656e-06, + "loss": 0.0804, + "step": 936 + }, + { + "epoch": 0.21295454545454545, + "grad_norm": 2.9391351265281123, + "learning_rate": 1.1155466866858677e-06, + "loss": 0.0908, + "step": 937 + }, + { + "epoch": 0.2131818181818182, + "grad_norm": 2.655188398881594, + "learning_rate": 1.1152700414937111e-06, + "loss": 0.0986, + "step": 938 + }, + { + "epoch": 0.2134090909090909, + "grad_norm": 2.8106235341663828, + "learning_rate": 1.114993146365028e-06, + "loss": 0.1547, + "step": 939 + }, + { + "epoch": 0.21363636363636362, + "grad_norm": 3.6297463148915456, + "learning_rate": 1.1147160014409779e-06, + "loss": 0.1103, + "step": 940 + }, + { + "epoch": 0.21386363636363637, + "grad_norm": 2.8524582915545005, + "learning_rate": 1.1144386068628472e-06, + "loss": 0.0772, + "step": 941 + }, + { + "epoch": 0.21409090909090908, + "grad_norm": 3.1120254153406948, + "learning_rate": 1.1141609627720501e-06, + "loss": 0.0716, + "step": 942 + }, + { + "epoch": 0.21431818181818182, + "grad_norm": 4.166340096987895, + "learning_rate": 1.1138830693101277e-06, + "loss": 0.0905, + "step": 943 + }, + { + "epoch": 0.21454545454545454, + "grad_norm": 3.44698959353396, + "learning_rate": 1.1136049266187481e-06, + "loss": 0.1115, + "step": 944 + }, + { + "epoch": 0.21477272727272728, + "grad_norm": 5.103850539890016, + "learning_rate": 1.113326534839707e-06, + "loss": 0.1221, + "step": 945 + }, + { + "epoch": 0.215, + "grad_norm": 3.4001181793894477, + "learning_rate": 1.1130478941149268e-06, + "loss": 0.1072, + "step": 946 + }, + { + "epoch": 0.21522727272727274, + "grad_norm": 1.4398984693810495, + "learning_rate": 1.1127690045864561e-06, + "loss": 0.057, + "step": 947 + }, + { + "epoch": 0.21545454545454545, + "grad_norm": 2.139917325562843, + "learning_rate": 1.1124898663964718e-06, + "loss": 0.0567, + "step": 948 + }, + { + "epoch": 0.21568181818181817, + "grad_norm": 2.6360047341409096, + "learning_rate": 1.1122104796872763e-06, + "loss": 0.0874, + "step": 949 + }, + { + "epoch": 0.2159090909090909, + "grad_norm": 1.9418779910396977, + "learning_rate": 1.1119308446012993e-06, + "loss": 0.08, + "step": 950 + }, + { + "epoch": 0.21613636363636363, + "grad_norm": 3.0841562509889493, + "learning_rate": 1.111650961281097e-06, + "loss": 0.1253, + "step": 951 + }, + { + "epoch": 0.21636363636363637, + "grad_norm": 3.0421489372648622, + "learning_rate": 1.111370829869352e-06, + "loss": 0.0752, + "step": 952 + }, + { + "epoch": 0.21659090909090908, + "grad_norm": 2.59526236210807, + "learning_rate": 1.1110904505088738e-06, + "loss": 0.1283, + "step": 953 + }, + { + "epoch": 0.21681818181818183, + "grad_norm": 2.7191104654458416, + "learning_rate": 1.1108098233425977e-06, + "loss": 0.0683, + "step": 954 + }, + { + "epoch": 0.21704545454545454, + "grad_norm": 3.500085636965691, + "learning_rate": 1.1105289485135855e-06, + "loss": 0.1666, + "step": 955 + }, + { + "epoch": 0.21727272727272728, + "grad_norm": 2.693817982180609, + "learning_rate": 1.1102478261650258e-06, + "loss": 0.1293, + "step": 956 + }, + { + "epoch": 0.2175, + "grad_norm": 2.6763048421495483, + "learning_rate": 1.1099664564402327e-06, + "loss": 0.0543, + "step": 957 + }, + { + "epoch": 0.2177272727272727, + "grad_norm": 3.24729570899025, + "learning_rate": 1.1096848394826467e-06, + "loss": 0.0802, + "step": 958 + }, + { + "epoch": 0.21795454545454546, + "grad_norm": 2.919029786989127, + "learning_rate": 1.1094029754358343e-06, + "loss": 0.1037, + "step": 959 + }, + { + "epoch": 0.21818181818181817, + "grad_norm": 3.623639918767071, + "learning_rate": 1.1091208644434882e-06, + "loss": 0.0802, + "step": 960 + }, + { + "epoch": 0.2184090909090909, + "grad_norm": 3.968975467882094, + "learning_rate": 1.1088385066494267e-06, + "loss": 0.1528, + "step": 961 + }, + { + "epoch": 0.21863636363636363, + "grad_norm": 3.15057704771838, + "learning_rate": 1.1085559021975937e-06, + "loss": 0.1234, + "step": 962 + }, + { + "epoch": 0.21886363636363637, + "grad_norm": 2.6709843917686036, + "learning_rate": 1.1082730512320597e-06, + "loss": 0.0802, + "step": 963 + }, + { + "epoch": 0.2190909090909091, + "grad_norm": 2.8299447934939392, + "learning_rate": 1.1079899538970196e-06, + "loss": 0.175, + "step": 964 + }, + { + "epoch": 0.21931818181818183, + "grad_norm": 4.3456158484941465, + "learning_rate": 1.1077066103367954e-06, + "loss": 0.1304, + "step": 965 + }, + { + "epoch": 0.21954545454545454, + "grad_norm": 2.9640615768038687, + "learning_rate": 1.1074230206958332e-06, + "loss": 0.0978, + "step": 966 + }, + { + "epoch": 0.2197727272727273, + "grad_norm": 3.023816834500527, + "learning_rate": 1.1071391851187052e-06, + "loss": 0.1221, + "step": 967 + }, + { + "epoch": 0.22, + "grad_norm": 2.97550683203516, + "learning_rate": 1.1068551037501093e-06, + "loss": 0.2092, + "step": 968 + }, + { + "epoch": 0.22022727272727272, + "grad_norm": 2.8314882467134086, + "learning_rate": 1.1065707767348684e-06, + "loss": 0.1076, + "step": 969 + }, + { + "epoch": 0.22045454545454546, + "grad_norm": 3.6105411878655294, + "learning_rate": 1.1062862042179302e-06, + "loss": 0.0938, + "step": 970 + }, + { + "epoch": 0.22068181818181817, + "grad_norm": 3.9761655423322995, + "learning_rate": 1.1060013863443683e-06, + "loss": 0.118, + "step": 971 + }, + { + "epoch": 0.22090909090909092, + "grad_norm": 3.1122566983148263, + "learning_rate": 1.1057163232593808e-06, + "loss": 0.0769, + "step": 972 + }, + { + "epoch": 0.22113636363636363, + "grad_norm": 2.7806972715712104, + "learning_rate": 1.1054310151082913e-06, + "loss": 0.1501, + "step": 973 + }, + { + "epoch": 0.22136363636363637, + "grad_norm": 3.1084208389456762, + "learning_rate": 1.1051454620365475e-06, + "loss": 0.1068, + "step": 974 + }, + { + "epoch": 0.2215909090909091, + "grad_norm": 1.7533252783178466, + "learning_rate": 1.1048596641897233e-06, + "loss": 0.0819, + "step": 975 + }, + { + "epoch": 0.22181818181818183, + "grad_norm": 2.854155714159375, + "learning_rate": 1.104573621713516e-06, + "loss": 0.2365, + "step": 976 + }, + { + "epoch": 0.22204545454545455, + "grad_norm": 1.6819497397911294, + "learning_rate": 1.1042873347537485e-06, + "loss": 0.0362, + "step": 977 + }, + { + "epoch": 0.22227272727272726, + "grad_norm": 3.4077534670067635, + "learning_rate": 1.1040008034563682e-06, + "loss": 0.0866, + "step": 978 + }, + { + "epoch": 0.2225, + "grad_norm": 2.1573687666392316, + "learning_rate": 1.1037140279674468e-06, + "loss": 0.0698, + "step": 979 + }, + { + "epoch": 0.22272727272727272, + "grad_norm": 2.3392318655049693, + "learning_rate": 1.1034270084331803e-06, + "loss": 0.0797, + "step": 980 + }, + { + "epoch": 0.22295454545454546, + "grad_norm": 2.1296081662600734, + "learning_rate": 1.1031397449998896e-06, + "loss": 0.1405, + "step": 981 + }, + { + "epoch": 0.22318181818181818, + "grad_norm": 2.8460089165570452, + "learning_rate": 1.10285223781402e-06, + "loss": 0.0731, + "step": 982 + }, + { + "epoch": 0.22340909090909092, + "grad_norm": 2.8334340814496635, + "learning_rate": 1.1025644870221405e-06, + "loss": 0.1504, + "step": 983 + }, + { + "epoch": 0.22363636363636363, + "grad_norm": 3.446019615343611, + "learning_rate": 1.1022764927709447e-06, + "loss": 0.0605, + "step": 984 + }, + { + "epoch": 0.22386363636363638, + "grad_norm": 3.097380365284322, + "learning_rate": 1.1019882552072502e-06, + "loss": 0.118, + "step": 985 + }, + { + "epoch": 0.2240909090909091, + "grad_norm": 6.496170788687828, + "learning_rate": 1.101699774477999e-06, + "loss": 0.1331, + "step": 986 + }, + { + "epoch": 0.2243181818181818, + "grad_norm": 2.9428983013033765, + "learning_rate": 1.1014110507302563e-06, + "loss": 0.0701, + "step": 987 + }, + { + "epoch": 0.22454545454545455, + "grad_norm": 2.05414278673617, + "learning_rate": 1.1011220841112118e-06, + "loss": 0.0596, + "step": 988 + }, + { + "epoch": 0.22477272727272726, + "grad_norm": 3.0359800204017087, + "learning_rate": 1.1008328747681788e-06, + "loss": 0.0995, + "step": 989 + }, + { + "epoch": 0.225, + "grad_norm": 4.361680045062287, + "learning_rate": 1.1005434228485945e-06, + "loss": 0.1094, + "step": 990 + }, + { + "epoch": 0.22522727272727272, + "grad_norm": 3.950076969846753, + "learning_rate": 1.1002537285000196e-06, + "loss": 0.0978, + "step": 991 + }, + { + "epoch": 0.22545454545454546, + "grad_norm": 2.9762040848859677, + "learning_rate": 1.099963791870138e-06, + "loss": 0.1443, + "step": 992 + }, + { + "epoch": 0.22568181818181818, + "grad_norm": 5.024103465586588, + "learning_rate": 1.099673613106758e-06, + "loss": 0.2074, + "step": 993 + }, + { + "epoch": 0.22590909090909092, + "grad_norm": 2.1861585122545137, + "learning_rate": 1.0993831923578107e-06, + "loss": 0.0761, + "step": 994 + }, + { + "epoch": 0.22613636363636364, + "grad_norm": 2.072397310473323, + "learning_rate": 1.099092529771351e-06, + "loss": 0.1005, + "step": 995 + }, + { + "epoch": 0.22636363636363635, + "grad_norm": 3.670339750568969, + "learning_rate": 1.0988016254955565e-06, + "loss": 0.0774, + "step": 996 + }, + { + "epoch": 0.2265909090909091, + "grad_norm": 3.725794384946036, + "learning_rate": 1.0985104796787285e-06, + "loss": 0.1026, + "step": 997 + }, + { + "epoch": 0.2268181818181818, + "grad_norm": 3.6322454934485733, + "learning_rate": 1.0982190924692911e-06, + "loss": 0.107, + "step": 998 + }, + { + "epoch": 0.22704545454545455, + "grad_norm": 1.4567468274314403, + "learning_rate": 1.0979274640157919e-06, + "loss": 0.0441, + "step": 999 + }, + { + "epoch": 0.22727272727272727, + "grad_norm": 2.520970431012095, + "learning_rate": 1.097635594466901e-06, + "loss": 0.0825, + "step": 1000 + }, + { + "epoch": 0.2275, + "grad_norm": 1.8186379199767186, + "learning_rate": 1.0973434839714116e-06, + "loss": 0.1005, + "step": 1001 + }, + { + "epoch": 0.22772727272727272, + "grad_norm": 3.94367300332879, + "learning_rate": 1.0970511326782398e-06, + "loss": 0.082, + "step": 1002 + }, + { + "epoch": 0.22795454545454547, + "grad_norm": 4.379848907669469, + "learning_rate": 1.0967585407364246e-06, + "loss": 0.2218, + "step": 1003 + }, + { + "epoch": 0.22818181818181818, + "grad_norm": 1.94048255803627, + "learning_rate": 1.0964657082951274e-06, + "loss": 0.1066, + "step": 1004 + }, + { + "epoch": 0.22840909090909092, + "grad_norm": 3.7788831732647785, + "learning_rate": 1.0961726355036324e-06, + "loss": 0.0955, + "step": 1005 + }, + { + "epoch": 0.22863636363636364, + "grad_norm": 2.9575383024806285, + "learning_rate": 1.0958793225113459e-06, + "loss": 0.1377, + "step": 1006 + }, + { + "epoch": 0.22886363636363635, + "grad_norm": 2.6695086579432683, + "learning_rate": 1.0955857694677971e-06, + "loss": 0.0893, + "step": 1007 + }, + { + "epoch": 0.2290909090909091, + "grad_norm": 3.2067387816766635, + "learning_rate": 1.0952919765226378e-06, + "loss": 0.0949, + "step": 1008 + }, + { + "epoch": 0.2293181818181818, + "grad_norm": 4.054634659463853, + "learning_rate": 1.0949979438256415e-06, + "loss": 0.1983, + "step": 1009 + }, + { + "epoch": 0.22954545454545455, + "grad_norm": 3.196778340437732, + "learning_rate": 1.0947036715267039e-06, + "loss": 0.0703, + "step": 1010 + }, + { + "epoch": 0.22977272727272727, + "grad_norm": 5.520026263163719, + "learning_rate": 1.0944091597758438e-06, + "loss": 0.1547, + "step": 1011 + }, + { + "epoch": 0.23, + "grad_norm": 2.4552227156109314, + "learning_rate": 1.0941144087232008e-06, + "loss": 0.1261, + "step": 1012 + }, + { + "epoch": 0.23022727272727272, + "grad_norm": 3.163944736931236, + "learning_rate": 1.0938194185190374e-06, + "loss": 0.0638, + "step": 1013 + }, + { + "epoch": 0.23045454545454547, + "grad_norm": 2.2120503722203604, + "learning_rate": 1.0935241893137376e-06, + "loss": 0.0617, + "step": 1014 + }, + { + "epoch": 0.23068181818181818, + "grad_norm": 2.6687056550410033, + "learning_rate": 1.0932287212578075e-06, + "loss": 0.0716, + "step": 1015 + }, + { + "epoch": 0.2309090909090909, + "grad_norm": 3.9174651509494085, + "learning_rate": 1.0929330145018747e-06, + "loss": 0.0933, + "step": 1016 + }, + { + "epoch": 0.23113636363636364, + "grad_norm": 2.9392520257103243, + "learning_rate": 1.0926370691966883e-06, + "loss": 0.0936, + "step": 1017 + }, + { + "epoch": 0.23136363636363635, + "grad_norm": 3.0054439327490776, + "learning_rate": 1.0923408854931202e-06, + "loss": 0.0863, + "step": 1018 + }, + { + "epoch": 0.2315909090909091, + "grad_norm": 3.9962511615012066, + "learning_rate": 1.0920444635421622e-06, + "loss": 0.164, + "step": 1019 + }, + { + "epoch": 0.2318181818181818, + "grad_norm": 2.1155702026565932, + "learning_rate": 1.0917478034949285e-06, + "loss": 0.0583, + "step": 1020 + }, + { + "epoch": 0.23204545454545455, + "grad_norm": 2.7604927565516273, + "learning_rate": 1.0914509055026545e-06, + "loss": 0.0856, + "step": 1021 + }, + { + "epoch": 0.23227272727272727, + "grad_norm": 3.1040964018256587, + "learning_rate": 1.0911537697166967e-06, + "loss": 0.0917, + "step": 1022 + }, + { + "epoch": 0.2325, + "grad_norm": 2.4074755195341373, + "learning_rate": 1.0908563962885337e-06, + "loss": 0.1086, + "step": 1023 + }, + { + "epoch": 0.23272727272727273, + "grad_norm": 3.132420152142783, + "learning_rate": 1.0905587853697638e-06, + "loss": 0.098, + "step": 1024 + }, + { + "epoch": 0.23295454545454544, + "grad_norm": 3.8986786653423464, + "learning_rate": 1.0902609371121076e-06, + "loss": 0.1329, + "step": 1025 + }, + { + "epoch": 0.23318181818181818, + "grad_norm": 2.7913827297085305, + "learning_rate": 1.089962851667406e-06, + "loss": 0.1389, + "step": 1026 + }, + { + "epoch": 0.2334090909090909, + "grad_norm": 3.6111004875460533, + "learning_rate": 1.089664529187621e-06, + "loss": 0.0801, + "step": 1027 + }, + { + "epoch": 0.23363636363636364, + "grad_norm": 3.303790015604333, + "learning_rate": 1.0893659698248358e-06, + "loss": 0.0957, + "step": 1028 + }, + { + "epoch": 0.23386363636363636, + "grad_norm": 2.4104582587904995, + "learning_rate": 1.0890671737312538e-06, + "loss": 0.0933, + "step": 1029 + }, + { + "epoch": 0.2340909090909091, + "grad_norm": 1.2705515071854425, + "learning_rate": 1.0887681410591994e-06, + "loss": 0.0714, + "step": 1030 + }, + { + "epoch": 0.23431818181818181, + "grad_norm": 4.811434051395423, + "learning_rate": 1.0884688719611176e-06, + "loss": 0.1043, + "step": 1031 + }, + { + "epoch": 0.23454545454545456, + "grad_norm": 2.4391651938058887, + "learning_rate": 1.0881693665895737e-06, + "loss": 0.0999, + "step": 1032 + }, + { + "epoch": 0.23477272727272727, + "grad_norm": 8.686326823283993, + "learning_rate": 1.0878696250972536e-06, + "loss": 0.1909, + "step": 1033 + }, + { + "epoch": 0.235, + "grad_norm": 2.1755611203800376, + "learning_rate": 1.087569647636964e-06, + "loss": 0.071, + "step": 1034 + }, + { + "epoch": 0.23522727272727273, + "grad_norm": 2.2158883396280697, + "learning_rate": 1.0872694343616312e-06, + "loss": 0.0704, + "step": 1035 + }, + { + "epoch": 0.23545454545454544, + "grad_norm": 2.087822407387583, + "learning_rate": 1.0869689854243019e-06, + "loss": 0.0843, + "step": 1036 + }, + { + "epoch": 0.2356818181818182, + "grad_norm": 3.5078081934965653, + "learning_rate": 1.0866683009781432e-06, + "loss": 0.1349, + "step": 1037 + }, + { + "epoch": 0.2359090909090909, + "grad_norm": 2.6331042373038627, + "learning_rate": 1.0863673811764419e-06, + "loss": 0.1012, + "step": 1038 + }, + { + "epoch": 0.23613636363636364, + "grad_norm": 2.2993682874178387, + "learning_rate": 1.0860662261726054e-06, + "loss": 0.0484, + "step": 1039 + }, + { + "epoch": 0.23636363636363636, + "grad_norm": 1.0403619668073416, + "learning_rate": 1.0857648361201603e-06, + "loss": 0.0501, + "step": 1040 + }, + { + "epoch": 0.2365909090909091, + "grad_norm": 5.425217295870728, + "learning_rate": 1.085463211172753e-06, + "loss": 0.1529, + "step": 1041 + }, + { + "epoch": 0.23681818181818182, + "grad_norm": 3.9756083085913216, + "learning_rate": 1.0851613514841508e-06, + "loss": 0.13, + "step": 1042 + }, + { + "epoch": 0.23704545454545456, + "grad_norm": 3.7432192413930747, + "learning_rate": 1.0848592572082391e-06, + "loss": 0.1006, + "step": 1043 + }, + { + "epoch": 0.23727272727272727, + "grad_norm": 3.4396043030899217, + "learning_rate": 1.084556928499024e-06, + "loss": 0.0699, + "step": 1044 + }, + { + "epoch": 0.2375, + "grad_norm": 2.5718606655420952, + "learning_rate": 1.0842543655106305e-06, + "loss": 0.1114, + "step": 1045 + }, + { + "epoch": 0.23772727272727273, + "grad_norm": 3.9763559701797266, + "learning_rate": 1.0839515683973035e-06, + "loss": 0.1054, + "step": 1046 + }, + { + "epoch": 0.23795454545454545, + "grad_norm": 7.466043325210298, + "learning_rate": 1.083648537313407e-06, + "loss": 0.131, + "step": 1047 + }, + { + "epoch": 0.2381818181818182, + "grad_norm": 2.3716868946146192, + "learning_rate": 1.083345272413424e-06, + "loss": 0.0792, + "step": 1048 + }, + { + "epoch": 0.2384090909090909, + "grad_norm": 3.0319954725654426, + "learning_rate": 1.0830417738519575e-06, + "loss": 0.1305, + "step": 1049 + }, + { + "epoch": 0.23863636363636365, + "grad_norm": 2.7584754293056433, + "learning_rate": 1.0827380417837287e-06, + "loss": 0.1098, + "step": 1050 + }, + { + "epoch": 0.23886363636363636, + "grad_norm": 2.904571613623144, + "learning_rate": 1.0824340763635785e-06, + "loss": 0.0797, + "step": 1051 + }, + { + "epoch": 0.2390909090909091, + "grad_norm": 2.6772635397168103, + "learning_rate": 1.0821298777464665e-06, + "loss": 0.0839, + "step": 1052 + }, + { + "epoch": 0.23931818181818182, + "grad_norm": 2.274822075599455, + "learning_rate": 1.081825446087471e-06, + "loss": 0.0688, + "step": 1053 + }, + { + "epoch": 0.23954545454545453, + "grad_norm": 2.5235214855610875, + "learning_rate": 1.0815207815417894e-06, + "loss": 0.0703, + "step": 1054 + }, + { + "epoch": 0.23977272727272728, + "grad_norm": 2.3941577475881015, + "learning_rate": 1.081215884264738e-06, + "loss": 0.0483, + "step": 1055 + }, + { + "epoch": 0.24, + "grad_norm": 3.3673495659882553, + "learning_rate": 1.0809107544117511e-06, + "loss": 0.1346, + "step": 1056 + }, + { + "epoch": 0.24022727272727273, + "grad_norm": 4.320657147163067, + "learning_rate": 1.0806053921383823e-06, + "loss": 0.0932, + "step": 1057 + }, + { + "epoch": 0.24045454545454545, + "grad_norm": 3.9548614200224215, + "learning_rate": 1.0802997976003031e-06, + "loss": 0.0952, + "step": 1058 + }, + { + "epoch": 0.2406818181818182, + "grad_norm": 2.4954608842300954, + "learning_rate": 1.0799939709533036e-06, + "loss": 0.1008, + "step": 1059 + }, + { + "epoch": 0.2409090909090909, + "grad_norm": 3.187134518513773, + "learning_rate": 1.0796879123532924e-06, + "loss": 0.1621, + "step": 1060 + }, + { + "epoch": 0.24113636363636365, + "grad_norm": 3.05296079715538, + "learning_rate": 1.0793816219562963e-06, + "loss": 0.0966, + "step": 1061 + }, + { + "epoch": 0.24136363636363636, + "grad_norm": 2.5614302974081538, + "learning_rate": 1.0790750999184598e-06, + "loss": 0.1329, + "step": 1062 + }, + { + "epoch": 0.24159090909090908, + "grad_norm": 3.84173783278149, + "learning_rate": 1.0787683463960462e-06, + "loss": 0.083, + "step": 1063 + }, + { + "epoch": 0.24181818181818182, + "grad_norm": 2.604420650190104, + "learning_rate": 1.0784613615454365e-06, + "loss": 0.0854, + "step": 1064 + }, + { + "epoch": 0.24204545454545454, + "grad_norm": 2.203252366003654, + "learning_rate": 1.0781541455231294e-06, + "loss": 0.0982, + "step": 1065 + }, + { + "epoch": 0.24227272727272728, + "grad_norm": 4.044297924815103, + "learning_rate": 1.077846698485742e-06, + "loss": 0.1194, + "step": 1066 + }, + { + "epoch": 0.2425, + "grad_norm": 2.496559723065901, + "learning_rate": 1.0775390205900084e-06, + "loss": 0.0674, + "step": 1067 + }, + { + "epoch": 0.24272727272727274, + "grad_norm": 2.3771885624811295, + "learning_rate": 1.0772311119927808e-06, + "loss": 0.0697, + "step": 1068 + }, + { + "epoch": 0.24295454545454545, + "grad_norm": 4.567381702847766, + "learning_rate": 1.0769229728510298e-06, + "loss": 0.1504, + "step": 1069 + }, + { + "epoch": 0.2431818181818182, + "grad_norm": 2.4423259837780114, + "learning_rate": 1.0766146033218417e-06, + "loss": 0.1347, + "step": 1070 + }, + { + "epoch": 0.2434090909090909, + "grad_norm": 6.4022156829521775, + "learning_rate": 1.076306003562422e-06, + "loss": 0.2103, + "step": 1071 + }, + { + "epoch": 0.24363636363636362, + "grad_norm": 3.0414937711217997, + "learning_rate": 1.0759971737300928e-06, + "loss": 0.0978, + "step": 1072 + }, + { + "epoch": 0.24386363636363637, + "grad_norm": 2.9426147955924344, + "learning_rate": 1.0756881139822934e-06, + "loss": 0.1012, + "step": 1073 + }, + { + "epoch": 0.24409090909090908, + "grad_norm": 2.7207722131416223, + "learning_rate": 1.0753788244765805e-06, + "loss": 0.1124, + "step": 1074 + }, + { + "epoch": 0.24431818181818182, + "grad_norm": 3.59614523815748, + "learning_rate": 1.0750693053706282e-06, + "loss": 0.1236, + "step": 1075 + }, + { + "epoch": 0.24454545454545454, + "grad_norm": 2.4201005806731084, + "learning_rate": 1.0747595568222268e-06, + "loss": 0.1141, + "step": 1076 + }, + { + "epoch": 0.24477272727272728, + "grad_norm": 5.149939108345512, + "learning_rate": 1.0744495789892848e-06, + "loss": 0.1068, + "step": 1077 + }, + { + "epoch": 0.245, + "grad_norm": 2.0022508432855384, + "learning_rate": 1.0741393720298263e-06, + "loss": 0.0821, + "step": 1078 + }, + { + "epoch": 0.24522727272727274, + "grad_norm": 2.54279051615943, + "learning_rate": 1.073828936101993e-06, + "loss": 0.134, + "step": 1079 + }, + { + "epoch": 0.24545454545454545, + "grad_norm": 3.671636218279181, + "learning_rate": 1.0735182713640436e-06, + "loss": 0.1231, + "step": 1080 + }, + { + "epoch": 0.2456818181818182, + "grad_norm": 1.9929652393310888, + "learning_rate": 1.0732073779743523e-06, + "loss": 0.0725, + "step": 1081 + }, + { + "epoch": 0.2459090909090909, + "grad_norm": 2.729430627099501, + "learning_rate": 1.0728962560914108e-06, + "loss": 0.0972, + "step": 1082 + }, + { + "epoch": 0.24613636363636363, + "grad_norm": 2.7644868096031736, + "learning_rate": 1.0725849058738274e-06, + "loss": 0.0635, + "step": 1083 + }, + { + "epoch": 0.24636363636363637, + "grad_norm": 4.2097583327768575, + "learning_rate": 1.0722733274803261e-06, + "loss": 0.171, + "step": 1084 + }, + { + "epoch": 0.24659090909090908, + "grad_norm": 2.632446584616852, + "learning_rate": 1.0719615210697476e-06, + "loss": 0.0739, + "step": 1085 + }, + { + "epoch": 0.24681818181818183, + "grad_norm": 2.491685390413527, + "learning_rate": 1.0716494868010488e-06, + "loss": 0.0815, + "step": 1086 + }, + { + "epoch": 0.24704545454545454, + "grad_norm": 2.989766310368257, + "learning_rate": 1.071337224833303e-06, + "loss": 0.0899, + "step": 1087 + }, + { + "epoch": 0.24727272727272728, + "grad_norm": 3.6186010355658063, + "learning_rate": 1.0710247353256988e-06, + "loss": 0.0772, + "step": 1088 + }, + { + "epoch": 0.2475, + "grad_norm": 2.6104287680708214, + "learning_rate": 1.0707120184375422e-06, + "loss": 0.1062, + "step": 1089 + }, + { + "epoch": 0.24772727272727274, + "grad_norm": 2.897428790966074, + "learning_rate": 1.0703990743282534e-06, + "loss": 0.0714, + "step": 1090 + }, + { + "epoch": 0.24795454545454546, + "grad_norm": 2.2482087891800213, + "learning_rate": 1.07008590315737e-06, + "loss": 0.0803, + "step": 1091 + }, + { + "epoch": 0.24818181818181817, + "grad_norm": 3.6688591905200223, + "learning_rate": 1.069772505084544e-06, + "loss": 0.0992, + "step": 1092 + }, + { + "epoch": 0.2484090909090909, + "grad_norm": 1.7946380493299723, + "learning_rate": 1.0694588802695443e-06, + "loss": 0.0774, + "step": 1093 + }, + { + "epoch": 0.24863636363636363, + "grad_norm": 3.7597612489294963, + "learning_rate": 1.0691450288722545e-06, + "loss": 0.1348, + "step": 1094 + }, + { + "epoch": 0.24886363636363637, + "grad_norm": 3.0933391356841655, + "learning_rate": 1.0688309510526742e-06, + "loss": 0.0889, + "step": 1095 + }, + { + "epoch": 0.24909090909090909, + "grad_norm": 3.928779673621332, + "learning_rate": 1.0685166469709181e-06, + "loss": 0.0821, + "step": 1096 + }, + { + "epoch": 0.24931818181818183, + "grad_norm": 3.8333992154963075, + "learning_rate": 1.0682021167872166e-06, + "loss": 0.0903, + "step": 1097 + }, + { + "epoch": 0.24954545454545454, + "grad_norm": 3.8470035812397096, + "learning_rate": 1.0678873606619152e-06, + "loss": 0.1932, + "step": 1098 + }, + { + "epoch": 0.24977272727272729, + "grad_norm": 1.4609102601517443, + "learning_rate": 1.0675723787554743e-06, + "loss": 0.0558, + "step": 1099 + }, + { + "epoch": 0.25, + "grad_norm": 3.027837105277264, + "learning_rate": 1.0672571712284697e-06, + "loss": 0.0689, + "step": 1100 + }, + { + "epoch": 0.25022727272727274, + "grad_norm": 1.9022194967237485, + "learning_rate": 1.0669417382415923e-06, + "loss": 0.1028, + "step": 1101 + }, + { + "epoch": 0.25045454545454543, + "grad_norm": 4.908840511238501, + "learning_rate": 1.0666260799556477e-06, + "loss": 0.2641, + "step": 1102 + }, + { + "epoch": 0.2506818181818182, + "grad_norm": 3.1349704494987076, + "learning_rate": 1.0663101965315568e-06, + "loss": 0.1048, + "step": 1103 + }, + { + "epoch": 0.2509090909090909, + "grad_norm": 2.848665923484888, + "learning_rate": 1.0659940881303545e-06, + "loss": 0.0736, + "step": 1104 + }, + { + "epoch": 0.25113636363636366, + "grad_norm": 3.0902645626569023, + "learning_rate": 1.065677754913191e-06, + "loss": 0.139, + "step": 1105 + }, + { + "epoch": 0.25136363636363634, + "grad_norm": 2.8410783582247063, + "learning_rate": 1.0653611970413311e-06, + "loss": 0.0956, + "step": 1106 + }, + { + "epoch": 0.2515909090909091, + "grad_norm": 1.8548826281717323, + "learning_rate": 1.0650444146761537e-06, + "loss": 0.0713, + "step": 1107 + }, + { + "epoch": 0.25181818181818183, + "grad_norm": 1.9519730941403817, + "learning_rate": 1.0647274079791529e-06, + "loss": 0.1079, + "step": 1108 + }, + { + "epoch": 0.2520454545454546, + "grad_norm": 3.4321252624895906, + "learning_rate": 1.0644101771119361e-06, + "loss": 0.1774, + "step": 1109 + }, + { + "epoch": 0.25227272727272726, + "grad_norm": 3.188539603748569, + "learning_rate": 1.0640927222362258e-06, + "loss": 0.1259, + "step": 1110 + }, + { + "epoch": 0.2525, + "grad_norm": 3.1786074000230466, + "learning_rate": 1.0637750435138586e-06, + "loss": 0.1291, + "step": 1111 + }, + { + "epoch": 0.25272727272727274, + "grad_norm": 2.935927476724215, + "learning_rate": 1.063457141106785e-06, + "loss": 0.0636, + "step": 1112 + }, + { + "epoch": 0.25295454545454543, + "grad_norm": 2.3538712286001346, + "learning_rate": 1.0631390151770693e-06, + "loss": 0.0669, + "step": 1113 + }, + { + "epoch": 0.2531818181818182, + "grad_norm": 4.166216153769425, + "learning_rate": 1.0628206658868906e-06, + "loss": 0.083, + "step": 1114 + }, + { + "epoch": 0.2534090909090909, + "grad_norm": 3.335386848239713, + "learning_rate": 1.0625020933985411e-06, + "loss": 0.1271, + "step": 1115 + }, + { + "epoch": 0.25363636363636366, + "grad_norm": 3.668876936762705, + "learning_rate": 1.0621832978744267e-06, + "loss": 0.129, + "step": 1116 + }, + { + "epoch": 0.25386363636363635, + "grad_norm": 2.8707870682329717, + "learning_rate": 1.0618642794770679e-06, + "loss": 0.1099, + "step": 1117 + }, + { + "epoch": 0.2540909090909091, + "grad_norm": 3.1743270833391617, + "learning_rate": 1.0615450383690978e-06, + "loss": 0.104, + "step": 1118 + }, + { + "epoch": 0.25431818181818183, + "grad_norm": 4.411136606960197, + "learning_rate": 1.0612255747132637e-06, + "loss": 0.197, + "step": 1119 + }, + { + "epoch": 0.2545454545454545, + "grad_norm": 4.522715797500118, + "learning_rate": 1.0609058886724258e-06, + "loss": 0.1594, + "step": 1120 + }, + { + "epoch": 0.25477272727272726, + "grad_norm": 2.619063527701687, + "learning_rate": 1.0605859804095587e-06, + "loss": 0.1288, + "step": 1121 + }, + { + "epoch": 0.255, + "grad_norm": 2.7346833131662303, + "learning_rate": 1.0602658500877486e-06, + "loss": 0.0712, + "step": 1122 + }, + { + "epoch": 0.25522727272727275, + "grad_norm": 3.454681131557408, + "learning_rate": 1.0599454978701965e-06, + "loss": 0.1312, + "step": 1123 + }, + { + "epoch": 0.25545454545454543, + "grad_norm": 4.836201003082417, + "learning_rate": 1.0596249239202158e-06, + "loss": 0.1443, + "step": 1124 + }, + { + "epoch": 0.2556818181818182, + "grad_norm": 3.3069584940246193, + "learning_rate": 1.0593041284012333e-06, + "loss": 0.1432, + "step": 1125 + }, + { + "epoch": 0.2559090909090909, + "grad_norm": 2.6247501899999484, + "learning_rate": 1.058983111476788e-06, + "loss": 0.1013, + "step": 1126 + }, + { + "epoch": 0.25613636363636366, + "grad_norm": 3.134728400241263, + "learning_rate": 1.0586618733105327e-06, + "loss": 0.1152, + "step": 1127 + }, + { + "epoch": 0.25636363636363635, + "grad_norm": 2.471571568700953, + "learning_rate": 1.0583404140662322e-06, + "loss": 0.087, + "step": 1128 + }, + { + "epoch": 0.2565909090909091, + "grad_norm": 3.580658889213365, + "learning_rate": 1.0580187339077647e-06, + "loss": 0.132, + "step": 1129 + }, + { + "epoch": 0.25681818181818183, + "grad_norm": 5.581104788463544, + "learning_rate": 1.0576968329991205e-06, + "loss": 0.2383, + "step": 1130 + }, + { + "epoch": 0.2570454545454545, + "grad_norm": 2.7258317898938054, + "learning_rate": 1.0573747115044025e-06, + "loss": 0.0934, + "step": 1131 + }, + { + "epoch": 0.25727272727272726, + "grad_norm": 3.337406205935018, + "learning_rate": 1.0570523695878264e-06, + "loss": 0.0909, + "step": 1132 + }, + { + "epoch": 0.2575, + "grad_norm": 3.2913923317895635, + "learning_rate": 1.05672980741372e-06, + "loss": 0.0982, + "step": 1133 + }, + { + "epoch": 0.25772727272727275, + "grad_norm": 3.86032130876738, + "learning_rate": 1.0564070251465233e-06, + "loss": 0.0968, + "step": 1134 + }, + { + "epoch": 0.25795454545454544, + "grad_norm": 3.610851924548778, + "learning_rate": 1.0560840229507888e-06, + "loss": 0.1274, + "step": 1135 + }, + { + "epoch": 0.2581818181818182, + "grad_norm": 2.619014645191683, + "learning_rate": 1.055760800991181e-06, + "loss": 0.0982, + "step": 1136 + }, + { + "epoch": 0.2584090909090909, + "grad_norm": 2.34710234066561, + "learning_rate": 1.0554373594324762e-06, + "loss": 0.1039, + "step": 1137 + }, + { + "epoch": 0.2586363636363636, + "grad_norm": 2.5846300302168443, + "learning_rate": 1.0551136984395627e-06, + "loss": 0.0616, + "step": 1138 + }, + { + "epoch": 0.25886363636363635, + "grad_norm": 3.224107555341992, + "learning_rate": 1.0547898181774413e-06, + "loss": 0.1302, + "step": 1139 + }, + { + "epoch": 0.2590909090909091, + "grad_norm": 2.8589358044845046, + "learning_rate": 1.0544657188112238e-06, + "loss": 0.1141, + "step": 1140 + }, + { + "epoch": 0.25931818181818184, + "grad_norm": 2.05674128732612, + "learning_rate": 1.0541414005061334e-06, + "loss": 0.0745, + "step": 1141 + }, + { + "epoch": 0.2595454545454545, + "grad_norm": 2.9269850969986213, + "learning_rate": 1.0538168634275067e-06, + "loss": 0.0779, + "step": 1142 + }, + { + "epoch": 0.25977272727272727, + "grad_norm": 2.3597357066384617, + "learning_rate": 1.0534921077407898e-06, + "loss": 0.0564, + "step": 1143 + }, + { + "epoch": 0.26, + "grad_norm": 4.955020639379597, + "learning_rate": 1.0531671336115409e-06, + "loss": 0.1245, + "step": 1144 + }, + { + "epoch": 0.26022727272727275, + "grad_norm": 4.611165413983898, + "learning_rate": 1.0528419412054306e-06, + "loss": 0.1802, + "step": 1145 + }, + { + "epoch": 0.26045454545454544, + "grad_norm": 5.571400061249043, + "learning_rate": 1.052516530688239e-06, + "loss": 0.2667, + "step": 1146 + }, + { + "epoch": 0.2606818181818182, + "grad_norm": 2.584603848382816, + "learning_rate": 1.0521909022258589e-06, + "loss": 0.079, + "step": 1147 + }, + { + "epoch": 0.2609090909090909, + "grad_norm": 2.292024862093378, + "learning_rate": 1.0518650559842934e-06, + "loss": 0.1041, + "step": 1148 + }, + { + "epoch": 0.2611363636363636, + "grad_norm": 2.9635701988967225, + "learning_rate": 1.0515389921296568e-06, + "loss": 0.1298, + "step": 1149 + }, + { + "epoch": 0.26136363636363635, + "grad_norm": 2.6060951733849, + "learning_rate": 1.0512127108281741e-06, + "loss": 0.0976, + "step": 1150 + }, + { + "epoch": 0.2615909090909091, + "grad_norm": 3.336142510912546, + "learning_rate": 1.0508862122461822e-06, + "loss": 0.1088, + "step": 1151 + }, + { + "epoch": 0.26181818181818184, + "grad_norm": 2.515431097786272, + "learning_rate": 1.050559496550127e-06, + "loss": 0.0764, + "step": 1152 + }, + { + "epoch": 0.2620454545454545, + "grad_norm": 1.9117201500769616, + "learning_rate": 1.0502325639065669e-06, + "loss": 0.0804, + "step": 1153 + }, + { + "epoch": 0.26227272727272727, + "grad_norm": 2.5360828225316574, + "learning_rate": 1.0499054144821695e-06, + "loss": 0.0996, + "step": 1154 + }, + { + "epoch": 0.2625, + "grad_norm": 2.9624683555496976, + "learning_rate": 1.0495780484437138e-06, + "loss": 0.1037, + "step": 1155 + }, + { + "epoch": 0.26272727272727275, + "grad_norm": 4.521898912545863, + "learning_rate": 1.0492504659580888e-06, + "loss": 0.1655, + "step": 1156 + }, + { + "epoch": 0.26295454545454544, + "grad_norm": 2.879417400715769, + "learning_rate": 1.0489226671922938e-06, + "loss": 0.0801, + "step": 1157 + }, + { + "epoch": 0.2631818181818182, + "grad_norm": 4.203946358705877, + "learning_rate": 1.048594652313439e-06, + "loss": 0.1803, + "step": 1158 + }, + { + "epoch": 0.2634090909090909, + "grad_norm": 3.461286047268551, + "learning_rate": 1.0482664214887436e-06, + "loss": 0.0864, + "step": 1159 + }, + { + "epoch": 0.2636363636363636, + "grad_norm": 2.5585120257094975, + "learning_rate": 1.047937974885538e-06, + "loss": 0.0764, + "step": 1160 + }, + { + "epoch": 0.26386363636363636, + "grad_norm": 3.033971509978034, + "learning_rate": 1.0476093126712622e-06, + "loss": 0.1625, + "step": 1161 + }, + { + "epoch": 0.2640909090909091, + "grad_norm": 1.9278718632250313, + "learning_rate": 1.0472804350134661e-06, + "loss": 0.1097, + "step": 1162 + }, + { + "epoch": 0.26431818181818184, + "grad_norm": 5.90775615003635, + "learning_rate": 1.0469513420798091e-06, + "loss": 0.2175, + "step": 1163 + }, + { + "epoch": 0.26454545454545453, + "grad_norm": 3.2554359040625456, + "learning_rate": 1.046622034038061e-06, + "loss": 0.0774, + "step": 1164 + }, + { + "epoch": 0.26477272727272727, + "grad_norm": 3.2164960084544676, + "learning_rate": 1.046292511056101e-06, + "loss": 0.1111, + "step": 1165 + }, + { + "epoch": 0.265, + "grad_norm": 4.760424798582011, + "learning_rate": 1.0459627733019172e-06, + "loss": 0.113, + "step": 1166 + }, + { + "epoch": 0.2652272727272727, + "grad_norm": 3.2307296962561405, + "learning_rate": 1.0456328209436085e-06, + "loss": 0.0782, + "step": 1167 + }, + { + "epoch": 0.26545454545454544, + "grad_norm": 2.2219191421489355, + "learning_rate": 1.0453026541493822e-06, + "loss": 0.0549, + "step": 1168 + }, + { + "epoch": 0.2656818181818182, + "grad_norm": 4.873811896656266, + "learning_rate": 1.044972273087555e-06, + "loss": 0.0896, + "step": 1169 + }, + { + "epoch": 0.26590909090909093, + "grad_norm": 2.763460087657202, + "learning_rate": 1.0446416779265535e-06, + "loss": 0.1191, + "step": 1170 + }, + { + "epoch": 0.2661363636363636, + "grad_norm": 3.3099578175490545, + "learning_rate": 1.0443108688349127e-06, + "loss": 0.1825, + "step": 1171 + }, + { + "epoch": 0.26636363636363636, + "grad_norm": 2.3047812827934546, + "learning_rate": 1.0439798459812772e-06, + "loss": 0.1085, + "step": 1172 + }, + { + "epoch": 0.2665909090909091, + "grad_norm": 3.289833569514505, + "learning_rate": 1.0436486095344e-06, + "loss": 0.1242, + "step": 1173 + }, + { + "epoch": 0.26681818181818184, + "grad_norm": 4.144472887365878, + "learning_rate": 1.0433171596631433e-06, + "loss": 0.1365, + "step": 1174 + }, + { + "epoch": 0.26704545454545453, + "grad_norm": 3.4068857746250654, + "learning_rate": 1.0429854965364786e-06, + "loss": 0.1298, + "step": 1175 + }, + { + "epoch": 0.2672727272727273, + "grad_norm": 2.784157877662139, + "learning_rate": 1.0426536203234851e-06, + "loss": 0.0672, + "step": 1176 + }, + { + "epoch": 0.2675, + "grad_norm": 3.2392138523889, + "learning_rate": 1.0423215311933516e-06, + "loss": 0.1149, + "step": 1177 + }, + { + "epoch": 0.2677272727272727, + "grad_norm": 3.7422104486465004, + "learning_rate": 1.0419892293153749e-06, + "loss": 0.1899, + "step": 1178 + }, + { + "epoch": 0.26795454545454545, + "grad_norm": 1.97757911042594, + "learning_rate": 1.04165671485896e-06, + "loss": 0.0729, + "step": 1179 + }, + { + "epoch": 0.2681818181818182, + "grad_norm": 4.8616702314773645, + "learning_rate": 1.0413239879936213e-06, + "loss": 0.1787, + "step": 1180 + }, + { + "epoch": 0.26840909090909093, + "grad_norm": 3.76160174265459, + "learning_rate": 1.0409910488889801e-06, + "loss": 0.2417, + "step": 1181 + }, + { + "epoch": 0.2686363636363636, + "grad_norm": 4.1081000200378766, + "learning_rate": 1.0406578977147672e-06, + "loss": 0.2098, + "step": 1182 + }, + { + "epoch": 0.26886363636363636, + "grad_norm": 3.3076135633560235, + "learning_rate": 1.0403245346408205e-06, + "loss": 0.1025, + "step": 1183 + }, + { + "epoch": 0.2690909090909091, + "grad_norm": 2.7507527516201296, + "learning_rate": 1.0399909598370866e-06, + "loss": 0.1149, + "step": 1184 + }, + { + "epoch": 0.26931818181818185, + "grad_norm": 3.511550704062235, + "learning_rate": 1.03965717347362e-06, + "loss": 0.1057, + "step": 1185 + }, + { + "epoch": 0.26954545454545453, + "grad_norm": 2.7465272868884973, + "learning_rate": 1.0393231757205824e-06, + "loss": 0.0869, + "step": 1186 + }, + { + "epoch": 0.2697727272727273, + "grad_norm": 3.593568654804558, + "learning_rate": 1.038988966748244e-06, + "loss": 0.2613, + "step": 1187 + }, + { + "epoch": 0.27, + "grad_norm": 3.9327152940673433, + "learning_rate": 1.0386545467269823e-06, + "loss": 0.1204, + "step": 1188 + }, + { + "epoch": 0.2702272727272727, + "grad_norm": 3.830959036284215, + "learning_rate": 1.0383199158272825e-06, + "loss": 0.1664, + "step": 1189 + }, + { + "epoch": 0.27045454545454545, + "grad_norm": 2.6234453071156194, + "learning_rate": 1.0379850742197374e-06, + "loss": 0.1107, + "step": 1190 + }, + { + "epoch": 0.2706818181818182, + "grad_norm": 2.4484267058914218, + "learning_rate": 1.037650022075047e-06, + "loss": 0.0457, + "step": 1191 + }, + { + "epoch": 0.27090909090909093, + "grad_norm": 3.6027517542579948, + "learning_rate": 1.0373147595640183e-06, + "loss": 0.1508, + "step": 1192 + }, + { + "epoch": 0.2711363636363636, + "grad_norm": 3.66557808115639, + "learning_rate": 1.0369792868575669e-06, + "loss": 0.1395, + "step": 1193 + }, + { + "epoch": 0.27136363636363636, + "grad_norm": 2.3421386966709763, + "learning_rate": 1.0366436041267142e-06, + "loss": 0.1068, + "step": 1194 + }, + { + "epoch": 0.2715909090909091, + "grad_norm": 2.364825103517622, + "learning_rate": 1.0363077115425888e-06, + "loss": 0.0658, + "step": 1195 + }, + { + "epoch": 0.2718181818181818, + "grad_norm": 2.827596504521179, + "learning_rate": 1.0359716092764268e-06, + "loss": 0.0788, + "step": 1196 + }, + { + "epoch": 0.27204545454545453, + "grad_norm": 2.712101827179471, + "learning_rate": 1.0356352974995713e-06, + "loss": 0.1092, + "step": 1197 + }, + { + "epoch": 0.2722727272727273, + "grad_norm": 3.4211025016975696, + "learning_rate": 1.0352987763834716e-06, + "loss": 0.0755, + "step": 1198 + }, + { + "epoch": 0.2725, + "grad_norm": 2.181716518453859, + "learning_rate": 1.034962046099684e-06, + "loss": 0.0967, + "step": 1199 + }, + { + "epoch": 0.2727272727272727, + "grad_norm": 1.5972840781696545, + "learning_rate": 1.0346251068198712e-06, + "loss": 0.0725, + "step": 1200 + }, + { + "epoch": 0.27295454545454545, + "grad_norm": 6.611898249169295, + "learning_rate": 1.0342879587158033e-06, + "loss": 0.1439, + "step": 1201 + }, + { + "epoch": 0.2731818181818182, + "grad_norm": 3.404224294800257, + "learning_rate": 1.0339506019593557e-06, + "loss": 0.1756, + "step": 1202 + }, + { + "epoch": 0.27340909090909093, + "grad_norm": 3.36849337311357, + "learning_rate": 1.033613036722511e-06, + "loss": 0.0858, + "step": 1203 + }, + { + "epoch": 0.2736363636363636, + "grad_norm": 2.920389669762444, + "learning_rate": 1.0332752631773578e-06, + "loss": 0.0982, + "step": 1204 + }, + { + "epoch": 0.27386363636363636, + "grad_norm": 1.9893910067578717, + "learning_rate": 1.0329372814960906e-06, + "loss": 0.0714, + "step": 1205 + }, + { + "epoch": 0.2740909090909091, + "grad_norm": 3.665121033604715, + "learning_rate": 1.0325990918510107e-06, + "loss": 0.1745, + "step": 1206 + }, + { + "epoch": 0.2743181818181818, + "grad_norm": 5.0133746273641036, + "learning_rate": 1.0322606944145248e-06, + "loss": 0.1642, + "step": 1207 + }, + { + "epoch": 0.27454545454545454, + "grad_norm": 2.942319464478918, + "learning_rate": 1.0319220893591456e-06, + "loss": 0.1582, + "step": 1208 + }, + { + "epoch": 0.2747727272727273, + "grad_norm": 3.9355611864654816, + "learning_rate": 1.031583276857492e-06, + "loss": 0.0966, + "step": 1209 + }, + { + "epoch": 0.275, + "grad_norm": 5.497230724166222, + "learning_rate": 1.0312442570822883e-06, + "loss": 0.1574, + "step": 1210 + }, + { + "epoch": 0.2752272727272727, + "grad_norm": 2.6423198890427617, + "learning_rate": 1.0309050302063647e-06, + "loss": 0.0893, + "step": 1211 + }, + { + "epoch": 0.27545454545454545, + "grad_norm": 2.919024950992986, + "learning_rate": 1.0305655964026573e-06, + "loss": 0.0677, + "step": 1212 + }, + { + "epoch": 0.2756818181818182, + "grad_norm": 3.245035209906528, + "learning_rate": 1.0302259558442065e-06, + "loss": 0.1301, + "step": 1213 + }, + { + "epoch": 0.2759090909090909, + "grad_norm": 3.088015277136867, + "learning_rate": 1.0298861087041593e-06, + "loss": 0.0619, + "step": 1214 + }, + { + "epoch": 0.2761363636363636, + "grad_norm": 3.0667942128613546, + "learning_rate": 1.0295460551557676e-06, + "loss": 0.1149, + "step": 1215 + }, + { + "epoch": 0.27636363636363637, + "grad_norm": 3.7705792579395885, + "learning_rate": 1.0292057953723883e-06, + "loss": 0.1448, + "step": 1216 + }, + { + "epoch": 0.2765909090909091, + "grad_norm": 2.296762305512623, + "learning_rate": 1.028865329527484e-06, + "loss": 0.0646, + "step": 1217 + }, + { + "epoch": 0.2768181818181818, + "grad_norm": 3.412860152081858, + "learning_rate": 1.0285246577946217e-06, + "loss": 0.1156, + "step": 1218 + }, + { + "epoch": 0.27704545454545454, + "grad_norm": 3.3551204170488687, + "learning_rate": 1.0281837803474739e-06, + "loss": 0.0913, + "step": 1219 + }, + { + "epoch": 0.2772727272727273, + "grad_norm": 2.6300648277918186, + "learning_rate": 1.0278426973598176e-06, + "loss": 0.0893, + "step": 1220 + }, + { + "epoch": 0.2775, + "grad_norm": 4.267340697987342, + "learning_rate": 1.0275014090055346e-06, + "loss": 0.1531, + "step": 1221 + }, + { + "epoch": 0.2777272727272727, + "grad_norm": 4.175658743526924, + "learning_rate": 1.0271599154586119e-06, + "loss": 0.1299, + "step": 1222 + }, + { + "epoch": 0.27795454545454545, + "grad_norm": 4.104916768940491, + "learning_rate": 1.0268182168931404e-06, + "loss": 0.2129, + "step": 1223 + }, + { + "epoch": 0.2781818181818182, + "grad_norm": 2.6913963955231575, + "learning_rate": 1.026476313483316e-06, + "loss": 0.174, + "step": 1224 + }, + { + "epoch": 0.2784090909090909, + "grad_norm": 3.1268199613817353, + "learning_rate": 1.0261342054034387e-06, + "loss": 0.0852, + "step": 1225 + }, + { + "epoch": 0.2786363636363636, + "grad_norm": 2.739388675630088, + "learning_rate": 1.025791892827913e-06, + "loss": 0.1104, + "step": 1226 + }, + { + "epoch": 0.27886363636363637, + "grad_norm": 2.578031850672837, + "learning_rate": 1.0254493759312482e-06, + "loss": 0.0791, + "step": 1227 + }, + { + "epoch": 0.2790909090909091, + "grad_norm": 3.1850947685526054, + "learning_rate": 1.0251066548880564e-06, + "loss": 0.1036, + "step": 1228 + }, + { + "epoch": 0.2793181818181818, + "grad_norm": 3.7844384233526682, + "learning_rate": 1.024763729873055e-06, + "loss": 0.1385, + "step": 1229 + }, + { + "epoch": 0.27954545454545454, + "grad_norm": 2.4412008962790432, + "learning_rate": 1.0244206010610653e-06, + "loss": 0.0531, + "step": 1230 + }, + { + "epoch": 0.2797727272727273, + "grad_norm": 3.6686135357686847, + "learning_rate": 1.0240772686270115e-06, + "loss": 0.1861, + "step": 1231 + }, + { + "epoch": 0.28, + "grad_norm": 3.0273675973376704, + "learning_rate": 1.0237337327459226e-06, + "loss": 0.1072, + "step": 1232 + }, + { + "epoch": 0.2802272727272727, + "grad_norm": 2.317028823128039, + "learning_rate": 1.0233899935929311e-06, + "loss": 0.0855, + "step": 1233 + }, + { + "epoch": 0.28045454545454546, + "grad_norm": 1.7256650936989038, + "learning_rate": 1.023046051343273e-06, + "loss": 0.0716, + "step": 1234 + }, + { + "epoch": 0.2806818181818182, + "grad_norm": 3.985469120765848, + "learning_rate": 1.0227019061722877e-06, + "loss": 0.1288, + "step": 1235 + }, + { + "epoch": 0.2809090909090909, + "grad_norm": 4.199360367786426, + "learning_rate": 1.022357558255418e-06, + "loss": 0.107, + "step": 1236 + }, + { + "epoch": 0.28113636363636363, + "grad_norm": 2.893282212714047, + "learning_rate": 1.022013007768211e-06, + "loss": 0.0709, + "step": 1237 + }, + { + "epoch": 0.28136363636363637, + "grad_norm": 3.974929013667309, + "learning_rate": 1.0216682548863155e-06, + "loss": 0.1839, + "step": 1238 + }, + { + "epoch": 0.2815909090909091, + "grad_norm": 3.3902074415185344, + "learning_rate": 1.0213232997854848e-06, + "loss": 0.1541, + "step": 1239 + }, + { + "epoch": 0.2818181818181818, + "grad_norm": 2.4337599173330955, + "learning_rate": 1.0209781426415747e-06, + "loss": 0.0899, + "step": 1240 + }, + { + "epoch": 0.28204545454545454, + "grad_norm": 2.9406645429318843, + "learning_rate": 1.020632783630544e-06, + "loss": 0.0979, + "step": 1241 + }, + { + "epoch": 0.2822727272727273, + "grad_norm": 5.490640358581111, + "learning_rate": 1.0202872229284546e-06, + "loss": 0.0868, + "step": 1242 + }, + { + "epoch": 0.2825, + "grad_norm": 3.0526553585957004, + "learning_rate": 1.0199414607114713e-06, + "loss": 0.1049, + "step": 1243 + }, + { + "epoch": 0.2827272727272727, + "grad_norm": 4.057542765796797, + "learning_rate": 1.0195954971558613e-06, + "loss": 0.1082, + "step": 1244 + }, + { + "epoch": 0.28295454545454546, + "grad_norm": 2.9727632213082407, + "learning_rate": 1.0192493324379943e-06, + "loss": 0.0979, + "step": 1245 + }, + { + "epoch": 0.2831818181818182, + "grad_norm": 3.3042578929448805, + "learning_rate": 1.0189029667343433e-06, + "loss": 0.0898, + "step": 1246 + }, + { + "epoch": 0.2834090909090909, + "grad_norm": 3.4637977960305477, + "learning_rate": 1.0185564002214831e-06, + "loss": 0.0964, + "step": 1247 + }, + { + "epoch": 0.28363636363636363, + "grad_norm": 2.5565841635288167, + "learning_rate": 1.0182096330760912e-06, + "loss": 0.1344, + "step": 1248 + }, + { + "epoch": 0.2838636363636364, + "grad_norm": 3.0353620572123416, + "learning_rate": 1.0178626654749473e-06, + "loss": 0.1397, + "step": 1249 + }, + { + "epoch": 0.2840909090909091, + "grad_norm": 2.207763006413901, + "learning_rate": 1.017515497594933e-06, + "loss": 0.0791, + "step": 1250 + }, + { + "epoch": 0.2843181818181818, + "grad_norm": 2.019429635350712, + "learning_rate": 1.0171681296130326e-06, + "loss": 0.0756, + "step": 1251 + }, + { + "epoch": 0.28454545454545455, + "grad_norm": 3.101997722382022, + "learning_rate": 1.0168205617063318e-06, + "loss": 0.0829, + "step": 1252 + }, + { + "epoch": 0.2847727272727273, + "grad_norm": 2.539656556897068, + "learning_rate": 1.0164727940520185e-06, + "loss": 0.0659, + "step": 1253 + }, + { + "epoch": 0.285, + "grad_norm": 4.227345066050936, + "learning_rate": 1.016124826827383e-06, + "loss": 0.213, + "step": 1254 + }, + { + "epoch": 0.2852272727272727, + "grad_norm": 2.5530645566250425, + "learning_rate": 1.015776660209816e-06, + "loss": 0.1004, + "step": 1255 + }, + { + "epoch": 0.28545454545454546, + "grad_norm": 4.449344380751562, + "learning_rate": 1.0154282943768106e-06, + "loss": 0.1975, + "step": 1256 + }, + { + "epoch": 0.2856818181818182, + "grad_norm": 2.2821194210751914, + "learning_rate": 1.0150797295059623e-06, + "loss": 0.0774, + "step": 1257 + }, + { + "epoch": 0.2859090909090909, + "grad_norm": 4.043121990235564, + "learning_rate": 1.0147309657749663e-06, + "loss": 0.0918, + "step": 1258 + }, + { + "epoch": 0.28613636363636363, + "grad_norm": 4.041844848020356, + "learning_rate": 1.0143820033616206e-06, + "loss": 0.1983, + "step": 1259 + }, + { + "epoch": 0.2863636363636364, + "grad_norm": 3.5655863841244124, + "learning_rate": 1.0140328424438242e-06, + "loss": 0.1476, + "step": 1260 + }, + { + "epoch": 0.2865909090909091, + "grad_norm": 3.324615622925245, + "learning_rate": 1.0136834831995765e-06, + "loss": 0.1347, + "step": 1261 + }, + { + "epoch": 0.2868181818181818, + "grad_norm": 2.0034007303795214, + "learning_rate": 1.0133339258069791e-06, + "loss": 0.1304, + "step": 1262 + }, + { + "epoch": 0.28704545454545455, + "grad_norm": 4.059513989835615, + "learning_rate": 1.0129841704442339e-06, + "loss": 0.1258, + "step": 1263 + }, + { + "epoch": 0.2872727272727273, + "grad_norm": 2.697773150143577, + "learning_rate": 1.012634217289644e-06, + "loss": 0.1281, + "step": 1264 + }, + { + "epoch": 0.2875, + "grad_norm": 2.1431717067871614, + "learning_rate": 1.0122840665216133e-06, + "loss": 0.0529, + "step": 1265 + }, + { + "epoch": 0.2877272727272727, + "grad_norm": 1.9252213190376763, + "learning_rate": 1.0119337183186465e-06, + "loss": 0.0643, + "step": 1266 + }, + { + "epoch": 0.28795454545454546, + "grad_norm": 5.287838301843424, + "learning_rate": 1.0115831728593483e-06, + "loss": 0.1439, + "step": 1267 + }, + { + "epoch": 0.2881818181818182, + "grad_norm": 3.9942625080888914, + "learning_rate": 1.0112324303224255e-06, + "loss": 0.1687, + "step": 1268 + }, + { + "epoch": 0.2884090909090909, + "grad_norm": 1.630328096821266, + "learning_rate": 1.0108814908866837e-06, + "loss": 0.0676, + "step": 1269 + }, + { + "epoch": 0.28863636363636364, + "grad_norm": 2.4261839020020655, + "learning_rate": 1.0105303547310298e-06, + "loss": 0.0766, + "step": 1270 + }, + { + "epoch": 0.2888636363636364, + "grad_norm": 4.082639508258652, + "learning_rate": 1.0101790220344707e-06, + "loss": 0.134, + "step": 1271 + }, + { + "epoch": 0.28909090909090907, + "grad_norm": 2.5555338886974215, + "learning_rate": 1.0098274929761134e-06, + "loss": 0.1109, + "step": 1272 + }, + { + "epoch": 0.2893181818181818, + "grad_norm": 2.0386437593060625, + "learning_rate": 1.0094757677351658e-06, + "loss": 0.0953, + "step": 1273 + }, + { + "epoch": 0.28954545454545455, + "grad_norm": 2.5831261469842013, + "learning_rate": 1.0091238464909343e-06, + "loss": 0.1257, + "step": 1274 + }, + { + "epoch": 0.2897727272727273, + "grad_norm": 3.1706192495980847, + "learning_rate": 1.0087717294228266e-06, + "loss": 0.0688, + "step": 1275 + }, + { + "epoch": 0.29, + "grad_norm": 2.1024018355047303, + "learning_rate": 1.0084194167103499e-06, + "loss": 0.0572, + "step": 1276 + }, + { + "epoch": 0.2902272727272727, + "grad_norm": 1.9653217329841441, + "learning_rate": 1.0080669085331104e-06, + "loss": 0.0521, + "step": 1277 + }, + { + "epoch": 0.29045454545454547, + "grad_norm": 2.8707471196960914, + "learning_rate": 1.007714205070815e-06, + "loss": 0.0982, + "step": 1278 + }, + { + "epoch": 0.2906818181818182, + "grad_norm": 3.0946641020443244, + "learning_rate": 1.0073613065032694e-06, + "loss": 0.1745, + "step": 1279 + }, + { + "epoch": 0.2909090909090909, + "grad_norm": 1.8592979765340194, + "learning_rate": 1.007008213010379e-06, + "loss": 0.0534, + "step": 1280 + }, + { + "epoch": 0.29113636363636364, + "grad_norm": 2.7966396568169003, + "learning_rate": 1.0066549247721489e-06, + "loss": 0.1191, + "step": 1281 + }, + { + "epoch": 0.2913636363636364, + "grad_norm": 3.3853492938564367, + "learning_rate": 1.006301441968683e-06, + "loss": 0.095, + "step": 1282 + }, + { + "epoch": 0.29159090909090907, + "grad_norm": 4.076769932330385, + "learning_rate": 1.0059477647801845e-06, + "loss": 0.1355, + "step": 1283 + }, + { + "epoch": 0.2918181818181818, + "grad_norm": 6.350978090033833, + "learning_rate": 1.0055938933869557e-06, + "loss": 0.2321, + "step": 1284 + }, + { + "epoch": 0.29204545454545455, + "grad_norm": 2.056434734845043, + "learning_rate": 1.0052398279693984e-06, + "loss": 0.077, + "step": 1285 + }, + { + "epoch": 0.2922727272727273, + "grad_norm": 3.9726661941251007, + "learning_rate": 1.0048855687080122e-06, + "loss": 0.1254, + "step": 1286 + }, + { + "epoch": 0.2925, + "grad_norm": 2.6782372946599278, + "learning_rate": 1.0045311157833968e-06, + "loss": 0.0759, + "step": 1287 + }, + { + "epoch": 0.2927272727272727, + "grad_norm": 2.358174284500183, + "learning_rate": 1.0041764693762499e-06, + "loss": 0.0634, + "step": 1288 + }, + { + "epoch": 0.29295454545454547, + "grad_norm": 3.500819041003119, + "learning_rate": 1.0038216296673678e-06, + "loss": 0.0814, + "step": 1289 + }, + { + "epoch": 0.29318181818181815, + "grad_norm": 1.451160858485835, + "learning_rate": 1.0034665968376457e-06, + "loss": 0.0637, + "step": 1290 + }, + { + "epoch": 0.2934090909090909, + "grad_norm": 4.27351980821897, + "learning_rate": 1.0031113710680767e-06, + "loss": 0.1745, + "step": 1291 + }, + { + "epoch": 0.29363636363636364, + "grad_norm": 2.897386729854884, + "learning_rate": 1.002755952539753e-06, + "loss": 0.1385, + "step": 1292 + }, + { + "epoch": 0.2938636363636364, + "grad_norm": 5.681387027502958, + "learning_rate": 1.0024003414338645e-06, + "loss": 0.188, + "step": 1293 + }, + { + "epoch": 0.29409090909090907, + "grad_norm": 4.7832075595431, + "learning_rate": 1.0020445379316996e-06, + "loss": 0.1468, + "step": 1294 + }, + { + "epoch": 0.2943181818181818, + "grad_norm": 3.1781333091553443, + "learning_rate": 1.0016885422146445e-06, + "loss": 0.1205, + "step": 1295 + }, + { + "epoch": 0.29454545454545455, + "grad_norm": 3.4736478650088523, + "learning_rate": 1.0013323544641837e-06, + "loss": 0.1213, + "step": 1296 + }, + { + "epoch": 0.2947727272727273, + "grad_norm": 4.307066180616335, + "learning_rate": 1.0009759748618993e-06, + "loss": 0.0941, + "step": 1297 + }, + { + "epoch": 0.295, + "grad_norm": 3.085345004080549, + "learning_rate": 1.0006194035894714e-06, + "loss": 0.1307, + "step": 1298 + }, + { + "epoch": 0.2952272727272727, + "grad_norm": 2.3961264136845792, + "learning_rate": 1.0002626408286777e-06, + "loss": 0.0586, + "step": 1299 + }, + { + "epoch": 0.29545454545454547, + "grad_norm": 4.50349854061309, + "learning_rate": 9.999056867613935e-07, + "loss": 0.0968, + "step": 1300 + }, + { + "epoch": 0.29568181818181816, + "grad_norm": 5.739275206465198, + "learning_rate": 9.995485415695919e-07, + "loss": 0.1988, + "step": 1301 + }, + { + "epoch": 0.2959090909090909, + "grad_norm": 2.8350606427604568, + "learning_rate": 9.99191205435343e-07, + "loss": 0.0854, + "step": 1302 + }, + { + "epoch": 0.29613636363636364, + "grad_norm": 2.413404961831212, + "learning_rate": 9.988336785408147e-07, + "loss": 0.0874, + "step": 1303 + }, + { + "epoch": 0.2963636363636364, + "grad_norm": 2.819766861022598, + "learning_rate": 9.984759610682721e-07, + "loss": 0.1086, + "step": 1304 + }, + { + "epoch": 0.29659090909090907, + "grad_norm": 3.091898701103331, + "learning_rate": 9.981180532000768e-07, + "loss": 0.0773, + "step": 1305 + }, + { + "epoch": 0.2968181818181818, + "grad_norm": 1.9755269405483924, + "learning_rate": 9.97759955118688e-07, + "loss": 0.0738, + "step": 1306 + }, + { + "epoch": 0.29704545454545456, + "grad_norm": 4.815066569654823, + "learning_rate": 9.97401667006662e-07, + "loss": 0.2234, + "step": 1307 + }, + { + "epoch": 0.2972727272727273, + "grad_norm": 3.384862609986829, + "learning_rate": 9.970431890466516e-07, + "loss": 0.1257, + "step": 1308 + }, + { + "epoch": 0.2975, + "grad_norm": 2.090512668445467, + "learning_rate": 9.96684521421407e-07, + "loss": 0.1607, + "step": 1309 + }, + { + "epoch": 0.29772727272727273, + "grad_norm": 2.7607391816710667, + "learning_rate": 9.963256643137743e-07, + "loss": 0.0783, + "step": 1310 + }, + { + "epoch": 0.29795454545454547, + "grad_norm": 2.514783289891609, + "learning_rate": 9.959666179066965e-07, + "loss": 0.111, + "step": 1311 + }, + { + "epoch": 0.29818181818181816, + "grad_norm": 2.3302143991496465, + "learning_rate": 9.95607382383213e-07, + "loss": 0.0701, + "step": 1312 + }, + { + "epoch": 0.2984090909090909, + "grad_norm": 6.906378970175363, + "learning_rate": 9.952479579264605e-07, + "loss": 0.1666, + "step": 1313 + }, + { + "epoch": 0.29863636363636364, + "grad_norm": 2.670503484119994, + "learning_rate": 9.948883447196706e-07, + "loss": 0.1093, + "step": 1314 + }, + { + "epoch": 0.2988636363636364, + "grad_norm": 2.8377851738459, + "learning_rate": 9.945285429461722e-07, + "loss": 0.1058, + "step": 1315 + }, + { + "epoch": 0.2990909090909091, + "grad_norm": 4.1109147587713135, + "learning_rate": 9.941685527893894e-07, + "loss": 0.1201, + "step": 1316 + }, + { + "epoch": 0.2993181818181818, + "grad_norm": 2.250881748649241, + "learning_rate": 9.938083744328436e-07, + "loss": 0.0573, + "step": 1317 + }, + { + "epoch": 0.29954545454545456, + "grad_norm": 4.640548724396375, + "learning_rate": 9.934480080601507e-07, + "loss": 0.0921, + "step": 1318 + }, + { + "epoch": 0.29977272727272725, + "grad_norm": 5.101537006227957, + "learning_rate": 9.930874538550238e-07, + "loss": 0.1239, + "step": 1319 + }, + { + "epoch": 0.3, + "grad_norm": 2.2428227022754936, + "learning_rate": 9.92726712001271e-07, + "loss": 0.0685, + "step": 1320 + }, + { + "epoch": 0.30022727272727273, + "grad_norm": 2.608927752728315, + "learning_rate": 9.923657826827958e-07, + "loss": 0.1358, + "step": 1321 + }, + { + "epoch": 0.3004545454545455, + "grad_norm": 2.5306594659164174, + "learning_rate": 9.92004666083598e-07, + "loss": 0.1229, + "step": 1322 + }, + { + "epoch": 0.30068181818181816, + "grad_norm": 3.352945525983868, + "learning_rate": 9.916433623877722e-07, + "loss": 0.0996, + "step": 1323 + }, + { + "epoch": 0.3009090909090909, + "grad_norm": 3.102525995749043, + "learning_rate": 9.91281871779509e-07, + "loss": 0.1109, + "step": 1324 + }, + { + "epoch": 0.30113636363636365, + "grad_norm": 2.7570013768943995, + "learning_rate": 9.909201944430943e-07, + "loss": 0.1646, + "step": 1325 + }, + { + "epoch": 0.3013636363636364, + "grad_norm": 2.335141963091201, + "learning_rate": 9.905583305629082e-07, + "loss": 0.0995, + "step": 1326 + }, + { + "epoch": 0.3015909090909091, + "grad_norm": 2.3667896169519675, + "learning_rate": 9.90196280323427e-07, + "loss": 0.0686, + "step": 1327 + }, + { + "epoch": 0.3018181818181818, + "grad_norm": 2.738727574925565, + "learning_rate": 9.898340439092216e-07, + "loss": 0.1364, + "step": 1328 + }, + { + "epoch": 0.30204545454545456, + "grad_norm": 2.2436245531284396, + "learning_rate": 9.894716215049575e-07, + "loss": 0.1132, + "step": 1329 + }, + { + "epoch": 0.30227272727272725, + "grad_norm": 3.026914385327542, + "learning_rate": 9.891090132953957e-07, + "loss": 0.0641, + "step": 1330 + }, + { + "epoch": 0.3025, + "grad_norm": 2.669730530094395, + "learning_rate": 9.887462194653912e-07, + "loss": 0.1195, + "step": 1331 + }, + { + "epoch": 0.30272727272727273, + "grad_norm": 3.844498666894979, + "learning_rate": 9.883832401998942e-07, + "loss": 0.0816, + "step": 1332 + }, + { + "epoch": 0.3029545454545455, + "grad_norm": 3.8949770488671733, + "learning_rate": 9.880200756839492e-07, + "loss": 0.1161, + "step": 1333 + }, + { + "epoch": 0.30318181818181816, + "grad_norm": 3.860146661600503, + "learning_rate": 9.87656726102695e-07, + "loss": 0.0924, + "step": 1334 + }, + { + "epoch": 0.3034090909090909, + "grad_norm": 4.153037751527899, + "learning_rate": 9.87293191641365e-07, + "loss": 0.1108, + "step": 1335 + }, + { + "epoch": 0.30363636363636365, + "grad_norm": 1.7596143756689262, + "learning_rate": 9.869294724852868e-07, + "loss": 0.0553, + "step": 1336 + }, + { + "epoch": 0.3038636363636364, + "grad_norm": 4.595090775639966, + "learning_rate": 9.86565568819882e-07, + "loss": 0.1525, + "step": 1337 + }, + { + "epoch": 0.3040909090909091, + "grad_norm": 2.246412324562083, + "learning_rate": 9.86201480830666e-07, + "loss": 0.0954, + "step": 1338 + }, + { + "epoch": 0.3043181818181818, + "grad_norm": 2.5289532888770068, + "learning_rate": 9.858372087032493e-07, + "loss": 0.0769, + "step": 1339 + }, + { + "epoch": 0.30454545454545456, + "grad_norm": 2.775247530533837, + "learning_rate": 9.854727526233351e-07, + "loss": 0.0973, + "step": 1340 + }, + { + "epoch": 0.30477272727272725, + "grad_norm": 3.698511796377243, + "learning_rate": 9.851081127767206e-07, + "loss": 0.0948, + "step": 1341 + }, + { + "epoch": 0.305, + "grad_norm": 5.262938231680995, + "learning_rate": 9.847432893492972e-07, + "loss": 0.1238, + "step": 1342 + }, + { + "epoch": 0.30522727272727274, + "grad_norm": 3.007709265314297, + "learning_rate": 9.843782825270494e-07, + "loss": 0.1093, + "step": 1343 + }, + { + "epoch": 0.3054545454545455, + "grad_norm": 3.5344568780278607, + "learning_rate": 9.84013092496055e-07, + "loss": 0.206, + "step": 1344 + }, + { + "epoch": 0.30568181818181817, + "grad_norm": 2.0810873940303223, + "learning_rate": 9.83647719442486e-07, + "loss": 0.1052, + "step": 1345 + }, + { + "epoch": 0.3059090909090909, + "grad_norm": 3.493898323778603, + "learning_rate": 9.832821635526072e-07, + "loss": 0.1015, + "step": 1346 + }, + { + "epoch": 0.30613636363636365, + "grad_norm": 2.849223518766788, + "learning_rate": 9.829164250127761e-07, + "loss": 0.0969, + "step": 1347 + }, + { + "epoch": 0.30636363636363634, + "grad_norm": 3.4436065818787274, + "learning_rate": 9.82550504009445e-07, + "loss": 0.1092, + "step": 1348 + }, + { + "epoch": 0.3065909090909091, + "grad_norm": 2.859617007463162, + "learning_rate": 9.82184400729157e-07, + "loss": 0.0645, + "step": 1349 + }, + { + "epoch": 0.3068181818181818, + "grad_norm": 3.6870936147103928, + "learning_rate": 9.818181153585495e-07, + "loss": 0.0837, + "step": 1350 + }, + { + "epoch": 0.30704545454545457, + "grad_norm": 7.324757446337969, + "learning_rate": 9.814516480843526e-07, + "loss": 0.1716, + "step": 1351 + }, + { + "epoch": 0.30727272727272725, + "grad_norm": 3.554760539185807, + "learning_rate": 9.810849990933891e-07, + "loss": 0.1302, + "step": 1352 + }, + { + "epoch": 0.3075, + "grad_norm": 2.001362744341814, + "learning_rate": 9.80718168572574e-07, + "loss": 0.0861, + "step": 1353 + }, + { + "epoch": 0.30772727272727274, + "grad_norm": 4.183602958929192, + "learning_rate": 9.803511567089151e-07, + "loss": 0.123, + "step": 1354 + }, + { + "epoch": 0.3079545454545455, + "grad_norm": 1.9454036951080083, + "learning_rate": 9.799839636895129e-07, + "loss": 0.0687, + "step": 1355 + }, + { + "epoch": 0.30818181818181817, + "grad_norm": 2.3069121002130464, + "learning_rate": 9.796165897015602e-07, + "loss": 0.062, + "step": 1356 + }, + { + "epoch": 0.3084090909090909, + "grad_norm": 1.7146244970600306, + "learning_rate": 9.792490349323415e-07, + "loss": 0.0591, + "step": 1357 + }, + { + "epoch": 0.30863636363636365, + "grad_norm": 4.327930886171828, + "learning_rate": 9.788812995692342e-07, + "loss": 0.2081, + "step": 1358 + }, + { + "epoch": 0.30886363636363634, + "grad_norm": 2.659577894354376, + "learning_rate": 9.78513383799707e-07, + "loss": 0.1086, + "step": 1359 + }, + { + "epoch": 0.3090909090909091, + "grad_norm": 1.8894571411175314, + "learning_rate": 9.781452878113219e-07, + "loss": 0.1006, + "step": 1360 + }, + { + "epoch": 0.3093181818181818, + "grad_norm": 8.866968691242294, + "learning_rate": 9.77777011791731e-07, + "loss": 0.1793, + "step": 1361 + }, + { + "epoch": 0.30954545454545457, + "grad_norm": 3.2231187540926416, + "learning_rate": 9.774085559286793e-07, + "loss": 0.1156, + "step": 1362 + }, + { + "epoch": 0.30977272727272726, + "grad_norm": 4.870702400482233, + "learning_rate": 9.770399204100032e-07, + "loss": 0.1215, + "step": 1363 + }, + { + "epoch": 0.31, + "grad_norm": 4.096999842111716, + "learning_rate": 9.766711054236308e-07, + "loss": 0.1164, + "step": 1364 + }, + { + "epoch": 0.31022727272727274, + "grad_norm": 4.797311485215423, + "learning_rate": 9.763021111575816e-07, + "loss": 0.2656, + "step": 1365 + }, + { + "epoch": 0.3104545454545454, + "grad_norm": 2.3260642196445125, + "learning_rate": 9.75932937799967e-07, + "loss": 0.136, + "step": 1366 + }, + { + "epoch": 0.31068181818181817, + "grad_norm": 3.1257662539495694, + "learning_rate": 9.755635855389884e-07, + "loss": 0.0867, + "step": 1367 + }, + { + "epoch": 0.3109090909090909, + "grad_norm": 2.8028675899062048, + "learning_rate": 9.751940545629394e-07, + "loss": 0.0921, + "step": 1368 + }, + { + "epoch": 0.31113636363636366, + "grad_norm": 5.26677466215148, + "learning_rate": 9.748243450602049e-07, + "loss": 0.2107, + "step": 1369 + }, + { + "epoch": 0.31136363636363634, + "grad_norm": 2.0847358111922945, + "learning_rate": 9.7445445721926e-07, + "loss": 0.0411, + "step": 1370 + }, + { + "epoch": 0.3115909090909091, + "grad_norm": 3.102859471173873, + "learning_rate": 9.740843912286717e-07, + "loss": 0.127, + "step": 1371 + }, + { + "epoch": 0.3118181818181818, + "grad_norm": 5.513908370781446, + "learning_rate": 9.737141472770967e-07, + "loss": 0.1382, + "step": 1372 + }, + { + "epoch": 0.31204545454545457, + "grad_norm": 4.250271338315768, + "learning_rate": 9.733437255532836e-07, + "loss": 0.1104, + "step": 1373 + }, + { + "epoch": 0.31227272727272726, + "grad_norm": 4.86534460612227, + "learning_rate": 9.729731262460702e-07, + "loss": 0.14, + "step": 1374 + }, + { + "epoch": 0.3125, + "grad_norm": 2.827628610496685, + "learning_rate": 9.726023495443866e-07, + "loss": 0.0942, + "step": 1375 + }, + { + "epoch": 0.31272727272727274, + "grad_norm": 3.915517681794156, + "learning_rate": 9.722313956372515e-07, + "loss": 0.147, + "step": 1376 + }, + { + "epoch": 0.31295454545454543, + "grad_norm": 2.438547976976633, + "learning_rate": 9.718602647137754e-07, + "loss": 0.088, + "step": 1377 + }, + { + "epoch": 0.3131818181818182, + "grad_norm": 2.1239747906740227, + "learning_rate": 9.714889569631583e-07, + "loss": 0.0694, + "step": 1378 + }, + { + "epoch": 0.3134090909090909, + "grad_norm": 2.885329920848903, + "learning_rate": 9.711174725746904e-07, + "loss": 0.1031, + "step": 1379 + }, + { + "epoch": 0.31363636363636366, + "grad_norm": 3.1767533231776572, + "learning_rate": 9.707458117377525e-07, + "loss": 0.0919, + "step": 1380 + }, + { + "epoch": 0.31386363636363634, + "grad_norm": 2.111703129704044, + "learning_rate": 9.703739746418141e-07, + "loss": 0.1196, + "step": 1381 + }, + { + "epoch": 0.3140909090909091, + "grad_norm": 2.268776185380176, + "learning_rate": 9.700019614764364e-07, + "loss": 0.0955, + "step": 1382 + }, + { + "epoch": 0.31431818181818183, + "grad_norm": 2.848449494521636, + "learning_rate": 9.696297724312683e-07, + "loss": 0.0622, + "step": 1383 + }, + { + "epoch": 0.3145454545454546, + "grad_norm": 1.9859127983288927, + "learning_rate": 9.692574076960501e-07, + "loss": 0.0714, + "step": 1384 + }, + { + "epoch": 0.31477272727272726, + "grad_norm": 4.816710380153, + "learning_rate": 9.688848674606108e-07, + "loss": 0.165, + "step": 1385 + }, + { + "epoch": 0.315, + "grad_norm": 4.571502462731669, + "learning_rate": 9.68512151914869e-07, + "loss": 0.138, + "step": 1386 + }, + { + "epoch": 0.31522727272727274, + "grad_norm": 1.9987868327600835, + "learning_rate": 9.681392612488326e-07, + "loss": 0.0853, + "step": 1387 + }, + { + "epoch": 0.31545454545454543, + "grad_norm": 3.3458791332722226, + "learning_rate": 9.677661956525986e-07, + "loss": 0.0856, + "step": 1388 + }, + { + "epoch": 0.3156818181818182, + "grad_norm": 3.029216857231613, + "learning_rate": 9.673929553163541e-07, + "loss": 0.1262, + "step": 1389 + }, + { + "epoch": 0.3159090909090909, + "grad_norm": 3.092767266425702, + "learning_rate": 9.670195404303744e-07, + "loss": 0.0751, + "step": 1390 + }, + { + "epoch": 0.31613636363636366, + "grad_norm": 2.4932531777287865, + "learning_rate": 9.666459511850237e-07, + "loss": 0.0778, + "step": 1391 + }, + { + "epoch": 0.31636363636363635, + "grad_norm": 2.7513087882634, + "learning_rate": 9.662721877707557e-07, + "loss": 0.13, + "step": 1392 + }, + { + "epoch": 0.3165909090909091, + "grad_norm": 3.7649166889298566, + "learning_rate": 9.658982503781127e-07, + "loss": 0.114, + "step": 1393 + }, + { + "epoch": 0.31681818181818183, + "grad_norm": 2.7205142189127542, + "learning_rate": 9.655241391977251e-07, + "loss": 0.0931, + "step": 1394 + }, + { + "epoch": 0.3170454545454545, + "grad_norm": 2.312371255125462, + "learning_rate": 9.65149854420313e-07, + "loss": 0.1423, + "step": 1395 + }, + { + "epoch": 0.31727272727272726, + "grad_norm": 3.0010390930046777, + "learning_rate": 9.64775396236684e-07, + "loss": 0.0878, + "step": 1396 + }, + { + "epoch": 0.3175, + "grad_norm": 2.251418795647015, + "learning_rate": 9.644007648377347e-07, + "loss": 0.0831, + "step": 1397 + }, + { + "epoch": 0.31772727272727275, + "grad_norm": 3.3947639934950025, + "learning_rate": 9.640259604144493e-07, + "loss": 0.0673, + "step": 1398 + }, + { + "epoch": 0.31795454545454543, + "grad_norm": 4.071681937722883, + "learning_rate": 9.636509831579013e-07, + "loss": 0.178, + "step": 1399 + }, + { + "epoch": 0.3181818181818182, + "grad_norm": 2.4296920180727977, + "learning_rate": 9.632758332592514e-07, + "loss": 0.1487, + "step": 1400 + }, + { + "epoch": 0.3184090909090909, + "grad_norm": 3.8262418670712828, + "learning_rate": 9.629005109097486e-07, + "loss": 0.1597, + "step": 1401 + }, + { + "epoch": 0.31863636363636366, + "grad_norm": 2.926029492396776, + "learning_rate": 9.6252501630073e-07, + "loss": 0.1307, + "step": 1402 + }, + { + "epoch": 0.31886363636363635, + "grad_norm": 1.9148689387972835, + "learning_rate": 9.6214934962362e-07, + "loss": 0.0406, + "step": 1403 + }, + { + "epoch": 0.3190909090909091, + "grad_norm": 5.013807646332278, + "learning_rate": 9.617735110699317e-07, + "loss": 0.153, + "step": 1404 + }, + { + "epoch": 0.31931818181818183, + "grad_norm": 2.4149202638791776, + "learning_rate": 9.613975008312646e-07, + "loss": 0.1296, + "step": 1405 + }, + { + "epoch": 0.3195454545454545, + "grad_norm": 2.6546425176176864, + "learning_rate": 9.610213190993063e-07, + "loss": 0.1217, + "step": 1406 + }, + { + "epoch": 0.31977272727272726, + "grad_norm": 2.7240923162778143, + "learning_rate": 9.606449660658323e-07, + "loss": 0.109, + "step": 1407 + }, + { + "epoch": 0.32, + "grad_norm": 1.924813895251273, + "learning_rate": 9.602684419227047e-07, + "loss": 0.1006, + "step": 1408 + }, + { + "epoch": 0.32022727272727275, + "grad_norm": 2.6452452075254644, + "learning_rate": 9.59891746861873e-07, + "loss": 0.0827, + "step": 1409 + }, + { + "epoch": 0.32045454545454544, + "grad_norm": 3.4800253375337027, + "learning_rate": 9.59514881075374e-07, + "loss": 0.1314, + "step": 1410 + }, + { + "epoch": 0.3206818181818182, + "grad_norm": 3.220784019491107, + "learning_rate": 9.591378447553316e-07, + "loss": 0.1152, + "step": 1411 + }, + { + "epoch": 0.3209090909090909, + "grad_norm": 2.557183288971397, + "learning_rate": 9.587606380939565e-07, + "loss": 0.1119, + "step": 1412 + }, + { + "epoch": 0.3211363636363636, + "grad_norm": 2.3181766632216205, + "learning_rate": 9.58383261283546e-07, + "loss": 0.1213, + "step": 1413 + }, + { + "epoch": 0.32136363636363635, + "grad_norm": 3.793663443191867, + "learning_rate": 9.580057145164844e-07, + "loss": 0.0836, + "step": 1414 + }, + { + "epoch": 0.3215909090909091, + "grad_norm": 1.9864588224623836, + "learning_rate": 9.576279979852429e-07, + "loss": 0.0987, + "step": 1415 + }, + { + "epoch": 0.32181818181818184, + "grad_norm": 4.498614120384967, + "learning_rate": 9.572501118823787e-07, + "loss": 0.0971, + "step": 1416 + }, + { + "epoch": 0.3220454545454545, + "grad_norm": 2.25096350484539, + "learning_rate": 9.56872056400536e-07, + "loss": 0.0485, + "step": 1417 + }, + { + "epoch": 0.32227272727272727, + "grad_norm": 2.480624489681732, + "learning_rate": 9.564938317324447e-07, + "loss": 0.0995, + "step": 1418 + }, + { + "epoch": 0.3225, + "grad_norm": 4.30083155487595, + "learning_rate": 9.561154380709218e-07, + "loss": 0.1047, + "step": 1419 + }, + { + "epoch": 0.32272727272727275, + "grad_norm": 5.497513436828155, + "learning_rate": 9.557368756088691e-07, + "loss": 0.1434, + "step": 1420 + }, + { + "epoch": 0.32295454545454544, + "grad_norm": 2.9582884920539336, + "learning_rate": 9.553581445392762e-07, + "loss": 0.0993, + "step": 1421 + }, + { + "epoch": 0.3231818181818182, + "grad_norm": 3.1647501127238082, + "learning_rate": 9.549792450552176e-07, + "loss": 0.1666, + "step": 1422 + }, + { + "epoch": 0.3234090909090909, + "grad_norm": 2.619479852473425, + "learning_rate": 9.546001773498537e-07, + "loss": 0.0828, + "step": 1423 + }, + { + "epoch": 0.3236363636363636, + "grad_norm": 2.805199228173761, + "learning_rate": 9.542209416164305e-07, + "loss": 0.1179, + "step": 1424 + }, + { + "epoch": 0.32386363636363635, + "grad_norm": 5.083317352916822, + "learning_rate": 9.538415380482804e-07, + "loss": 0.1303, + "step": 1425 + }, + { + "epoch": 0.3240909090909091, + "grad_norm": 3.936938039347584, + "learning_rate": 9.534619668388206e-07, + "loss": 0.1783, + "step": 1426 + }, + { + "epoch": 0.32431818181818184, + "grad_norm": 2.5906088844372865, + "learning_rate": 9.530822281815543e-07, + "loss": 0.0763, + "step": 1427 + }, + { + "epoch": 0.3245454545454545, + "grad_norm": 4.301808289098417, + "learning_rate": 9.527023222700697e-07, + "loss": 0.1492, + "step": 1428 + }, + { + "epoch": 0.32477272727272727, + "grad_norm": 3.131633134268753, + "learning_rate": 9.523222492980405e-07, + "loss": 0.0773, + "step": 1429 + }, + { + "epoch": 0.325, + "grad_norm": 3.706575612529939, + "learning_rate": 9.519420094592252e-07, + "loss": 0.1124, + "step": 1430 + }, + { + "epoch": 0.32522727272727275, + "grad_norm": 3.7271347918612734, + "learning_rate": 9.51561602947468e-07, + "loss": 0.0977, + "step": 1431 + }, + { + "epoch": 0.32545454545454544, + "grad_norm": 3.0648177951534183, + "learning_rate": 9.511810299566976e-07, + "loss": 0.0674, + "step": 1432 + }, + { + "epoch": 0.3256818181818182, + "grad_norm": 2.402520575881115, + "learning_rate": 9.508002906809276e-07, + "loss": 0.0867, + "step": 1433 + }, + { + "epoch": 0.3259090909090909, + "grad_norm": 4.411951361343026, + "learning_rate": 9.504193853142564e-07, + "loss": 0.1094, + "step": 1434 + }, + { + "epoch": 0.3261363636363636, + "grad_norm": 4.180232337693173, + "learning_rate": 9.500383140508671e-07, + "loss": 0.1237, + "step": 1435 + }, + { + "epoch": 0.32636363636363636, + "grad_norm": 2.564634609934471, + "learning_rate": 9.496570770850274e-07, + "loss": 0.0583, + "step": 1436 + }, + { + "epoch": 0.3265909090909091, + "grad_norm": 2.194166096262771, + "learning_rate": 9.492756746110898e-07, + "loss": 0.0672, + "step": 1437 + }, + { + "epoch": 0.32681818181818184, + "grad_norm": 2.3228631322137097, + "learning_rate": 9.488941068234904e-07, + "loss": 0.1205, + "step": 1438 + }, + { + "epoch": 0.32704545454545453, + "grad_norm": 2.6160962176699694, + "learning_rate": 9.485123739167501e-07, + "loss": 0.1073, + "step": 1439 + }, + { + "epoch": 0.32727272727272727, + "grad_norm": 4.8836914353280285, + "learning_rate": 9.481304760854743e-07, + "loss": 0.123, + "step": 1440 + }, + { + "epoch": 0.3275, + "grad_norm": 2.934693144931681, + "learning_rate": 9.477484135243513e-07, + "loss": 0.079, + "step": 1441 + }, + { + "epoch": 0.3277272727272727, + "grad_norm": 5.198870831007414, + "learning_rate": 9.473661864281547e-07, + "loss": 0.2088, + "step": 1442 + }, + { + "epoch": 0.32795454545454544, + "grad_norm": 4.020780226162855, + "learning_rate": 9.469837949917415e-07, + "loss": 0.1465, + "step": 1443 + }, + { + "epoch": 0.3281818181818182, + "grad_norm": 3.9594277275941843, + "learning_rate": 9.466012394100519e-07, + "loss": 0.0881, + "step": 1444 + }, + { + "epoch": 0.32840909090909093, + "grad_norm": 3.381190499499163, + "learning_rate": 9.46218519878111e-07, + "loss": 0.111, + "step": 1445 + }, + { + "epoch": 0.3286363636363636, + "grad_norm": 3.6296328992575093, + "learning_rate": 9.458356365910259e-07, + "loss": 0.1083, + "step": 1446 + }, + { + "epoch": 0.32886363636363636, + "grad_norm": 2.071930092875363, + "learning_rate": 9.454525897439888e-07, + "loss": 0.1039, + "step": 1447 + }, + { + "epoch": 0.3290909090909091, + "grad_norm": 3.242989940735661, + "learning_rate": 9.450693795322742e-07, + "loss": 0.1021, + "step": 1448 + }, + { + "epoch": 0.32931818181818184, + "grad_norm": 2.3120709742252568, + "learning_rate": 9.446860061512402e-07, + "loss": 0.0886, + "step": 1449 + }, + { + "epoch": 0.32954545454545453, + "grad_norm": 3.356935661870599, + "learning_rate": 9.443024697963279e-07, + "loss": 0.0702, + "step": 1450 + }, + { + "epoch": 0.3297727272727273, + "grad_norm": 2.610850089285071, + "learning_rate": 9.439187706630625e-07, + "loss": 0.062, + "step": 1451 + }, + { + "epoch": 0.33, + "grad_norm": 2.6668770261706864, + "learning_rate": 9.435349089470502e-07, + "loss": 0.1365, + "step": 1452 + }, + { + "epoch": 0.3302272727272727, + "grad_norm": 1.2579359948867799, + "learning_rate": 9.431508848439822e-07, + "loss": 0.0408, + "step": 1453 + }, + { + "epoch": 0.33045454545454545, + "grad_norm": 2.836211695553417, + "learning_rate": 9.427666985496311e-07, + "loss": 0.125, + "step": 1454 + }, + { + "epoch": 0.3306818181818182, + "grad_norm": 3.584165648234853, + "learning_rate": 9.423823502598526e-07, + "loss": 0.1375, + "step": 1455 + }, + { + "epoch": 0.33090909090909093, + "grad_norm": 4.364390843471005, + "learning_rate": 9.419978401705849e-07, + "loss": 0.1076, + "step": 1456 + }, + { + "epoch": 0.3311363636363636, + "grad_norm": 2.9891527229813204, + "learning_rate": 9.41613168477849e-07, + "loss": 0.0661, + "step": 1457 + }, + { + "epoch": 0.33136363636363636, + "grad_norm": 6.681132779097644, + "learning_rate": 9.412283353777482e-07, + "loss": 0.1641, + "step": 1458 + }, + { + "epoch": 0.3315909090909091, + "grad_norm": 3.662263180898016, + "learning_rate": 9.408433410664675e-07, + "loss": 0.1088, + "step": 1459 + }, + { + "epoch": 0.33181818181818185, + "grad_norm": 2.162745004069301, + "learning_rate": 9.40458185740275e-07, + "loss": 0.0656, + "step": 1460 + }, + { + "epoch": 0.33204545454545453, + "grad_norm": 3.2028001563694755, + "learning_rate": 9.4007286959552e-07, + "loss": 0.0797, + "step": 1461 + }, + { + "epoch": 0.3322727272727273, + "grad_norm": 3.154978227312697, + "learning_rate": 9.396873928286346e-07, + "loss": 0.0703, + "step": 1462 + }, + { + "epoch": 0.3325, + "grad_norm": 1.9762089807276784, + "learning_rate": 9.393017556361321e-07, + "loss": 0.1165, + "step": 1463 + }, + { + "epoch": 0.3327272727272727, + "grad_norm": 2.5573353207625797, + "learning_rate": 9.389159582146079e-07, + "loss": 0.0694, + "step": 1464 + }, + { + "epoch": 0.33295454545454545, + "grad_norm": 1.8456277887258172, + "learning_rate": 9.385300007607392e-07, + "loss": 0.0586, + "step": 1465 + }, + { + "epoch": 0.3331818181818182, + "grad_norm": 4.316749845316916, + "learning_rate": 9.381438834712846e-07, + "loss": 0.111, + "step": 1466 + }, + { + "epoch": 0.33340909090909093, + "grad_norm": 4.27555410294357, + "learning_rate": 9.377576065430841e-07, + "loss": 0.0863, + "step": 1467 + }, + { + "epoch": 0.3336363636363636, + "grad_norm": 2.429506205380231, + "learning_rate": 9.373711701730593e-07, + "loss": 0.0811, + "step": 1468 + }, + { + "epoch": 0.33386363636363636, + "grad_norm": 1.9908014462792742, + "learning_rate": 9.369845745582129e-07, + "loss": 0.0879, + "step": 1469 + }, + { + "epoch": 0.3340909090909091, + "grad_norm": 1.8753311279634954, + "learning_rate": 9.365978198956289e-07, + "loss": 0.1071, + "step": 1470 + }, + { + "epoch": 0.3343181818181818, + "grad_norm": 2.2569063439952917, + "learning_rate": 9.362109063824725e-07, + "loss": 0.1251, + "step": 1471 + }, + { + "epoch": 0.33454545454545453, + "grad_norm": 2.8168504385725717, + "learning_rate": 9.358238342159898e-07, + "loss": 0.1179, + "step": 1472 + }, + { + "epoch": 0.3347727272727273, + "grad_norm": 3.4167019281691635, + "learning_rate": 9.354366035935071e-07, + "loss": 0.169, + "step": 1473 + }, + { + "epoch": 0.335, + "grad_norm": 3.874261358960311, + "learning_rate": 9.350492147124328e-07, + "loss": 0.1238, + "step": 1474 + }, + { + "epoch": 0.3352272727272727, + "grad_norm": 2.720347525910536, + "learning_rate": 9.346616677702548e-07, + "loss": 0.1892, + "step": 1475 + }, + { + "epoch": 0.33545454545454545, + "grad_norm": 3.157523512017096, + "learning_rate": 9.342739629645423e-07, + "loss": 0.1072, + "step": 1476 + }, + { + "epoch": 0.3356818181818182, + "grad_norm": 3.2418319284812656, + "learning_rate": 9.338861004929447e-07, + "loss": 0.0953, + "step": 1477 + }, + { + "epoch": 0.33590909090909093, + "grad_norm": 2.0388080040575822, + "learning_rate": 9.334980805531918e-07, + "loss": 0.0706, + "step": 1478 + }, + { + "epoch": 0.3361363636363636, + "grad_norm": 5.878955787517834, + "learning_rate": 9.331099033430935e-07, + "loss": 0.2978, + "step": 1479 + }, + { + "epoch": 0.33636363636363636, + "grad_norm": 2.084655209513429, + "learning_rate": 9.327215690605403e-07, + "loss": 0.0767, + "step": 1480 + }, + { + "epoch": 0.3365909090909091, + "grad_norm": 2.4338372052666886, + "learning_rate": 9.323330779035025e-07, + "loss": 0.0465, + "step": 1481 + }, + { + "epoch": 0.3368181818181818, + "grad_norm": 3.967549948405282, + "learning_rate": 9.319444300700301e-07, + "loss": 0.0832, + "step": 1482 + }, + { + "epoch": 0.33704545454545454, + "grad_norm": 1.8240919988740847, + "learning_rate": 9.315556257582536e-07, + "loss": 0.0656, + "step": 1483 + }, + { + "epoch": 0.3372727272727273, + "grad_norm": 7.182989560444609, + "learning_rate": 9.311666651663831e-07, + "loss": 0.159, + "step": 1484 + }, + { + "epoch": 0.3375, + "grad_norm": 4.171869617806305, + "learning_rate": 9.307775484927078e-07, + "loss": 0.1434, + "step": 1485 + }, + { + "epoch": 0.3377272727272727, + "grad_norm": 1.6878909949050034, + "learning_rate": 9.30388275935597e-07, + "loss": 0.0521, + "step": 1486 + }, + { + "epoch": 0.33795454545454545, + "grad_norm": 2.3233268577455894, + "learning_rate": 9.299988476934995e-07, + "loss": 0.1156, + "step": 1487 + }, + { + "epoch": 0.3381818181818182, + "grad_norm": 3.847748050259655, + "learning_rate": 9.29609263964943e-07, + "loss": 0.1912, + "step": 1488 + }, + { + "epoch": 0.3384090909090909, + "grad_norm": 4.705566289248712, + "learning_rate": 9.292195249485353e-07, + "loss": 0.1091, + "step": 1489 + }, + { + "epoch": 0.3386363636363636, + "grad_norm": 2.569926354335467, + "learning_rate": 9.288296308429626e-07, + "loss": 0.109, + "step": 1490 + }, + { + "epoch": 0.33886363636363637, + "grad_norm": 1.2560229501492568, + "learning_rate": 9.284395818469902e-07, + "loss": 0.0425, + "step": 1491 + }, + { + "epoch": 0.3390909090909091, + "grad_norm": 4.299862925599742, + "learning_rate": 9.280493781594628e-07, + "loss": 0.2312, + "step": 1492 + }, + { + "epoch": 0.3393181818181818, + "grad_norm": 2.5966874643237903, + "learning_rate": 9.276590199793037e-07, + "loss": 0.0977, + "step": 1493 + }, + { + "epoch": 0.33954545454545454, + "grad_norm": 2.3513595240093164, + "learning_rate": 9.272685075055152e-07, + "loss": 0.0883, + "step": 1494 + }, + { + "epoch": 0.3397727272727273, + "grad_norm": 2.5626551422994486, + "learning_rate": 9.268778409371778e-07, + "loss": 0.1384, + "step": 1495 + }, + { + "epoch": 0.34, + "grad_norm": 2.438965888873109, + "learning_rate": 9.264870204734508e-07, + "loss": 0.0942, + "step": 1496 + }, + { + "epoch": 0.3402272727272727, + "grad_norm": 3.714850341384429, + "learning_rate": 9.260960463135721e-07, + "loss": 0.0777, + "step": 1497 + }, + { + "epoch": 0.34045454545454545, + "grad_norm": 4.68888472820301, + "learning_rate": 9.257049186568579e-07, + "loss": 0.2231, + "step": 1498 + }, + { + "epoch": 0.3406818181818182, + "grad_norm": 5.053103843375375, + "learning_rate": 9.253136377027022e-07, + "loss": 0.1576, + "step": 1499 + }, + { + "epoch": 0.3409090909090909, + "grad_norm": 7.980093409080376, + "learning_rate": 9.249222036505779e-07, + "loss": 0.1871, + "step": 1500 + }, + { + "epoch": 0.3411363636363636, + "grad_norm": 6.030016520595248, + "learning_rate": 9.245306167000357e-07, + "loss": 0.1916, + "step": 1501 + }, + { + "epoch": 0.34136363636363637, + "grad_norm": 2.673892784875381, + "learning_rate": 9.241388770507036e-07, + "loss": 0.083, + "step": 1502 + }, + { + "epoch": 0.3415909090909091, + "grad_norm": 2.804849179949328, + "learning_rate": 9.237469849022881e-07, + "loss": 0.0799, + "step": 1503 + }, + { + "epoch": 0.3418181818181818, + "grad_norm": 4.2028127930217725, + "learning_rate": 9.233549404545737e-07, + "loss": 0.0892, + "step": 1504 + }, + { + "epoch": 0.34204545454545454, + "grad_norm": 1.9570633447632524, + "learning_rate": 9.229627439074219e-07, + "loss": 0.0587, + "step": 1505 + }, + { + "epoch": 0.3422727272727273, + "grad_norm": 2.2926045139718187, + "learning_rate": 9.225703954607718e-07, + "loss": 0.0625, + "step": 1506 + }, + { + "epoch": 0.3425, + "grad_norm": 1.4444366616291626, + "learning_rate": 9.221778953146405e-07, + "loss": 0.1108, + "step": 1507 + }, + { + "epoch": 0.3427272727272727, + "grad_norm": 2.6602160895546825, + "learning_rate": 9.217852436691216e-07, + "loss": 0.1166, + "step": 1508 + }, + { + "epoch": 0.34295454545454546, + "grad_norm": 1.9154516260261671, + "learning_rate": 9.213924407243868e-07, + "loss": 0.0737, + "step": 1509 + }, + { + "epoch": 0.3431818181818182, + "grad_norm": 4.2964516895671725, + "learning_rate": 9.209994866806846e-07, + "loss": 0.1373, + "step": 1510 + }, + { + "epoch": 0.3434090909090909, + "grad_norm": 3.719544088683467, + "learning_rate": 9.206063817383399e-07, + "loss": 0.2385, + "step": 1511 + }, + { + "epoch": 0.34363636363636363, + "grad_norm": 1.8603218739275673, + "learning_rate": 9.202131260977555e-07, + "loss": 0.0831, + "step": 1512 + }, + { + "epoch": 0.34386363636363637, + "grad_norm": 4.507682765243742, + "learning_rate": 9.198197199594105e-07, + "loss": 0.1203, + "step": 1513 + }, + { + "epoch": 0.3440909090909091, + "grad_norm": 3.439738119102419, + "learning_rate": 9.194261635238609e-07, + "loss": 0.0985, + "step": 1514 + }, + { + "epoch": 0.3443181818181818, + "grad_norm": 3.6017231699118732, + "learning_rate": 9.190324569917393e-07, + "loss": 0.1294, + "step": 1515 + }, + { + "epoch": 0.34454545454545454, + "grad_norm": 2.1065755494091487, + "learning_rate": 9.186386005637546e-07, + "loss": 0.0906, + "step": 1516 + }, + { + "epoch": 0.3447727272727273, + "grad_norm": 2.7276630348390576, + "learning_rate": 9.182445944406923e-07, + "loss": 0.0676, + "step": 1517 + }, + { + "epoch": 0.345, + "grad_norm": 3.233853074530074, + "learning_rate": 9.178504388234143e-07, + "loss": 0.0718, + "step": 1518 + }, + { + "epoch": 0.3452272727272727, + "grad_norm": 3.8653094724974295, + "learning_rate": 9.174561339128584e-07, + "loss": 0.1156, + "step": 1519 + }, + { + "epoch": 0.34545454545454546, + "grad_norm": 2.2528427251373047, + "learning_rate": 9.170616799100391e-07, + "loss": 0.0688, + "step": 1520 + }, + { + "epoch": 0.3456818181818182, + "grad_norm": 2.743683184963417, + "learning_rate": 9.166670770160461e-07, + "loss": 0.0838, + "step": 1521 + }, + { + "epoch": 0.3459090909090909, + "grad_norm": 4.1989335412838065, + "learning_rate": 9.162723254320458e-07, + "loss": 0.1822, + "step": 1522 + }, + { + "epoch": 0.34613636363636363, + "grad_norm": 3.7705897626026017, + "learning_rate": 9.158774253592798e-07, + "loss": 0.1292, + "step": 1523 + }, + { + "epoch": 0.3463636363636364, + "grad_norm": 2.349245428587693, + "learning_rate": 9.154823769990656e-07, + "loss": 0.0787, + "step": 1524 + }, + { + "epoch": 0.3465909090909091, + "grad_norm": 4.181324454575973, + "learning_rate": 9.150871805527965e-07, + "loss": 0.1936, + "step": 1525 + }, + { + "epoch": 0.3468181818181818, + "grad_norm": 2.1653871103074107, + "learning_rate": 9.146918362219409e-07, + "loss": 0.0767, + "step": 1526 + }, + { + "epoch": 0.34704545454545455, + "grad_norm": 0.7274135258297554, + "learning_rate": 9.142963442080432e-07, + "loss": 0.0198, + "step": 1527 + }, + { + "epoch": 0.3472727272727273, + "grad_norm": 2.0436339692678196, + "learning_rate": 9.139007047127222e-07, + "loss": 0.0758, + "step": 1528 + }, + { + "epoch": 0.3475, + "grad_norm": 4.881290917682833, + "learning_rate": 9.135049179376726e-07, + "loss": 0.1026, + "step": 1529 + }, + { + "epoch": 0.3477272727272727, + "grad_norm": 3.419262380496396, + "learning_rate": 9.131089840846641e-07, + "loss": 0.12, + "step": 1530 + }, + { + "epoch": 0.34795454545454546, + "grad_norm": 1.8431129679745129, + "learning_rate": 9.12712903355541e-07, + "loss": 0.0635, + "step": 1531 + }, + { + "epoch": 0.3481818181818182, + "grad_norm": 1.9111164505415756, + "learning_rate": 9.123166759522229e-07, + "loss": 0.0957, + "step": 1532 + }, + { + "epoch": 0.3484090909090909, + "grad_norm": 2.894043046521963, + "learning_rate": 9.119203020767038e-07, + "loss": 0.0672, + "step": 1533 + }, + { + "epoch": 0.34863636363636363, + "grad_norm": 3.1566977003939614, + "learning_rate": 9.11523781931053e-07, + "loss": 0.0752, + "step": 1534 + }, + { + "epoch": 0.3488636363636364, + "grad_norm": 2.0920092299511945, + "learning_rate": 9.111271157174131e-07, + "loss": 0.0798, + "step": 1535 + }, + { + "epoch": 0.3490909090909091, + "grad_norm": 2.14919400148947, + "learning_rate": 9.107303036380028e-07, + "loss": 0.0838, + "step": 1536 + }, + { + "epoch": 0.3493181818181818, + "grad_norm": 3.3235494454793955, + "learning_rate": 9.103333458951142e-07, + "loss": 0.0978, + "step": 1537 + }, + { + "epoch": 0.34954545454545455, + "grad_norm": 2.2903602579515034, + "learning_rate": 9.099362426911135e-07, + "loss": 0.065, + "step": 1538 + }, + { + "epoch": 0.3497727272727273, + "grad_norm": 2.5641819091315776, + "learning_rate": 9.095389942284416e-07, + "loss": 0.0703, + "step": 1539 + }, + { + "epoch": 0.35, + "grad_norm": 3.3531046517364027, + "learning_rate": 9.091416007096131e-07, + "loss": 0.1927, + "step": 1540 + }, + { + "epoch": 0.3502272727272727, + "grad_norm": 3.5446186403000737, + "learning_rate": 9.087440623372168e-07, + "loss": 0.1366, + "step": 1541 + }, + { + "epoch": 0.35045454545454546, + "grad_norm": 2.5889432255578706, + "learning_rate": 9.083463793139152e-07, + "loss": 0.0695, + "step": 1542 + }, + { + "epoch": 0.3506818181818182, + "grad_norm": 3.3514286880683875, + "learning_rate": 9.079485518424444e-07, + "loss": 0.0948, + "step": 1543 + }, + { + "epoch": 0.3509090909090909, + "grad_norm": 7.220557788861881, + "learning_rate": 9.075505801256145e-07, + "loss": 0.1844, + "step": 1544 + }, + { + "epoch": 0.35113636363636364, + "grad_norm": 2.9788095439266615, + "learning_rate": 9.071524643663089e-07, + "loss": 0.1329, + "step": 1545 + }, + { + "epoch": 0.3513636363636364, + "grad_norm": 4.490105587361098, + "learning_rate": 9.067542047674843e-07, + "loss": 0.1675, + "step": 1546 + }, + { + "epoch": 0.35159090909090907, + "grad_norm": 2.3213024281698686, + "learning_rate": 9.063558015321711e-07, + "loss": 0.0745, + "step": 1547 + }, + { + "epoch": 0.3518181818181818, + "grad_norm": 3.7773742178913876, + "learning_rate": 9.059572548634725e-07, + "loss": 0.0746, + "step": 1548 + }, + { + "epoch": 0.35204545454545455, + "grad_norm": 1.9858972242533819, + "learning_rate": 9.055585649645651e-07, + "loss": 0.1141, + "step": 1549 + }, + { + "epoch": 0.3522727272727273, + "grad_norm": 6.845851023554955, + "learning_rate": 9.051597320386986e-07, + "loss": 0.2527, + "step": 1550 + }, + { + "epoch": 0.3525, + "grad_norm": 2.7014886308370314, + "learning_rate": 9.047607562891954e-07, + "loss": 0.0629, + "step": 1551 + }, + { + "epoch": 0.3527272727272727, + "grad_norm": 3.0281163422656787, + "learning_rate": 9.043616379194508e-07, + "loss": 0.2042, + "step": 1552 + }, + { + "epoch": 0.35295454545454547, + "grad_norm": 3.401320663010264, + "learning_rate": 9.039623771329326e-07, + "loss": 0.1287, + "step": 1553 + }, + { + "epoch": 0.3531818181818182, + "grad_norm": 1.890006361942812, + "learning_rate": 9.035629741331817e-07, + "loss": 0.0499, + "step": 1554 + }, + { + "epoch": 0.3534090909090909, + "grad_norm": 2.1148710606242314, + "learning_rate": 9.031634291238109e-07, + "loss": 0.0618, + "step": 1555 + }, + { + "epoch": 0.35363636363636364, + "grad_norm": 4.2846075712438045, + "learning_rate": 9.027637423085061e-07, + "loss": 0.1662, + "step": 1556 + }, + { + "epoch": 0.3538636363636364, + "grad_norm": 4.700723696864227, + "learning_rate": 9.023639138910246e-07, + "loss": 0.1576, + "step": 1557 + }, + { + "epoch": 0.35409090909090907, + "grad_norm": 5.550057872708505, + "learning_rate": 9.019639440751964e-07, + "loss": 0.2103, + "step": 1558 + }, + { + "epoch": 0.3543181818181818, + "grad_norm": 3.3642697353794606, + "learning_rate": 9.015638330649239e-07, + "loss": 0.126, + "step": 1559 + }, + { + "epoch": 0.35454545454545455, + "grad_norm": 2.678669560505932, + "learning_rate": 9.011635810641809e-07, + "loss": 0.0865, + "step": 1560 + }, + { + "epoch": 0.3547727272727273, + "grad_norm": 3.442721365145265, + "learning_rate": 9.007631882770133e-07, + "loss": 0.0835, + "step": 1561 + }, + { + "epoch": 0.355, + "grad_norm": 2.5888348815225934, + "learning_rate": 9.00362654907539e-07, + "loss": 0.0694, + "step": 1562 + }, + { + "epoch": 0.3552272727272727, + "grad_norm": 2.1353961118751865, + "learning_rate": 8.99961981159947e-07, + "loss": 0.0484, + "step": 1563 + }, + { + "epoch": 0.35545454545454547, + "grad_norm": 3.174326792791628, + "learning_rate": 8.995611672384985e-07, + "loss": 0.0765, + "step": 1564 + }, + { + "epoch": 0.35568181818181815, + "grad_norm": 4.415685624250806, + "learning_rate": 8.991602133475258e-07, + "loss": 0.1435, + "step": 1565 + }, + { + "epoch": 0.3559090909090909, + "grad_norm": 1.6348105485261062, + "learning_rate": 8.987591196914325e-07, + "loss": 0.0416, + "step": 1566 + }, + { + "epoch": 0.35613636363636364, + "grad_norm": 5.076974538582742, + "learning_rate": 8.983578864746935e-07, + "loss": 0.1937, + "step": 1567 + }, + { + "epoch": 0.3563636363636364, + "grad_norm": 3.5373127240404125, + "learning_rate": 8.979565139018554e-07, + "loss": 0.0903, + "step": 1568 + }, + { + "epoch": 0.35659090909090907, + "grad_norm": 3.5736929245324967, + "learning_rate": 8.975550021775346e-07, + "loss": 0.0721, + "step": 1569 + }, + { + "epoch": 0.3568181818181818, + "grad_norm": 5.481147808430896, + "learning_rate": 8.9715335150642e-07, + "loss": 0.1405, + "step": 1570 + }, + { + "epoch": 0.35704545454545455, + "grad_norm": 2.895145109324995, + "learning_rate": 8.967515620932702e-07, + "loss": 0.1158, + "step": 1571 + }, + { + "epoch": 0.3572727272727273, + "grad_norm": 5.058150393249606, + "learning_rate": 8.963496341429145e-07, + "loss": 0.2211, + "step": 1572 + }, + { + "epoch": 0.3575, + "grad_norm": 3.5504133387948538, + "learning_rate": 8.959475678602539e-07, + "loss": 0.1548, + "step": 1573 + }, + { + "epoch": 0.3577272727272727, + "grad_norm": 2.476686602330689, + "learning_rate": 8.95545363450259e-07, + "loss": 0.0781, + "step": 1574 + }, + { + "epoch": 0.35795454545454547, + "grad_norm": 3.697753039819814, + "learning_rate": 8.951430211179704e-07, + "loss": 0.1174, + "step": 1575 + }, + { + "epoch": 0.35818181818181816, + "grad_norm": 4.343407161425239, + "learning_rate": 8.947405410685004e-07, + "loss": 0.1111, + "step": 1576 + }, + { + "epoch": 0.3584090909090909, + "grad_norm": 6.489264472236681, + "learning_rate": 8.943379235070307e-07, + "loss": 0.2051, + "step": 1577 + }, + { + "epoch": 0.35863636363636364, + "grad_norm": 5.1859093199122555, + "learning_rate": 8.939351686388127e-07, + "loss": 0.2386, + "step": 1578 + }, + { + "epoch": 0.3588636363636364, + "grad_norm": 3.8451786514671578, + "learning_rate": 8.935322766691684e-07, + "loss": 0.091, + "step": 1579 + }, + { + "epoch": 0.35909090909090907, + "grad_norm": 3.7122253383713213, + "learning_rate": 8.931292478034899e-07, + "loss": 0.1877, + "step": 1580 + }, + { + "epoch": 0.3593181818181818, + "grad_norm": 3.485489360094761, + "learning_rate": 8.927260822472384e-07, + "loss": 0.115, + "step": 1581 + }, + { + "epoch": 0.35954545454545456, + "grad_norm": 3.396612414652505, + "learning_rate": 8.92322780205945e-07, + "loss": 0.1189, + "step": 1582 + }, + { + "epoch": 0.3597727272727273, + "grad_norm": 2.7698695834976728, + "learning_rate": 8.91919341885211e-07, + "loss": 0.076, + "step": 1583 + }, + { + "epoch": 0.36, + "grad_norm": 3.0957555352223283, + "learning_rate": 8.915157674907062e-07, + "loss": 0.088, + "step": 1584 + }, + { + "epoch": 0.36022727272727273, + "grad_norm": 4.477335612662012, + "learning_rate": 8.911120572281705e-07, + "loss": 0.2952, + "step": 1585 + }, + { + "epoch": 0.36045454545454547, + "grad_norm": 2.7416381215136854, + "learning_rate": 8.907082113034126e-07, + "loss": 0.2169, + "step": 1586 + }, + { + "epoch": 0.36068181818181816, + "grad_norm": 2.036570382124373, + "learning_rate": 8.903042299223108e-07, + "loss": 0.0484, + "step": 1587 + }, + { + "epoch": 0.3609090909090909, + "grad_norm": 2.655039691183255, + "learning_rate": 8.89900113290812e-07, + "loss": 0.0665, + "step": 1588 + }, + { + "epoch": 0.36113636363636364, + "grad_norm": 3.7814284555566506, + "learning_rate": 8.894958616149324e-07, + "loss": 0.1533, + "step": 1589 + }, + { + "epoch": 0.3613636363636364, + "grad_norm": 2.412430933202243, + "learning_rate": 8.890914751007568e-07, + "loss": 0.1129, + "step": 1590 + }, + { + "epoch": 0.3615909090909091, + "grad_norm": 2.0226596327366257, + "learning_rate": 8.88686953954439e-07, + "loss": 0.0828, + "step": 1591 + }, + { + "epoch": 0.3618181818181818, + "grad_norm": 2.029038695905775, + "learning_rate": 8.882822983822014e-07, + "loss": 0.0844, + "step": 1592 + }, + { + "epoch": 0.36204545454545456, + "grad_norm": 1.9712667169297762, + "learning_rate": 8.878775085903344e-07, + "loss": 0.0848, + "step": 1593 + }, + { + "epoch": 0.36227272727272725, + "grad_norm": 3.2288099764129767, + "learning_rate": 8.874725847851976e-07, + "loss": 0.1266, + "step": 1594 + }, + { + "epoch": 0.3625, + "grad_norm": 4.761691508676201, + "learning_rate": 8.870675271732184e-07, + "loss": 0.1996, + "step": 1595 + }, + { + "epoch": 0.36272727272727273, + "grad_norm": 3.2011027781809434, + "learning_rate": 8.866623359608927e-07, + "loss": 0.1656, + "step": 1596 + }, + { + "epoch": 0.3629545454545455, + "grad_norm": 3.124822445344642, + "learning_rate": 8.862570113547843e-07, + "loss": 0.1633, + "step": 1597 + }, + { + "epoch": 0.36318181818181816, + "grad_norm": 4.297276526394933, + "learning_rate": 8.858515535615249e-07, + "loss": 0.2217, + "step": 1598 + }, + { + "epoch": 0.3634090909090909, + "grad_norm": 2.630443426197557, + "learning_rate": 8.854459627878146e-07, + "loss": 0.1, + "step": 1599 + }, + { + "epoch": 0.36363636363636365, + "grad_norm": 2.086367826972121, + "learning_rate": 8.85040239240421e-07, + "loss": 0.0544, + "step": 1600 + }, + { + "epoch": 0.3638636363636364, + "grad_norm": 4.220090249175366, + "learning_rate": 8.846343831261792e-07, + "loss": 0.1225, + "step": 1601 + }, + { + "epoch": 0.3640909090909091, + "grad_norm": 4.991405314267306, + "learning_rate": 8.842283946519918e-07, + "loss": 0.2854, + "step": 1602 + }, + { + "epoch": 0.3643181818181818, + "grad_norm": 2.553425273491599, + "learning_rate": 8.838222740248297e-07, + "loss": 0.0797, + "step": 1603 + }, + { + "epoch": 0.36454545454545456, + "grad_norm": 2.6465705767834415, + "learning_rate": 8.834160214517302e-07, + "loss": 0.0628, + "step": 1604 + }, + { + "epoch": 0.36477272727272725, + "grad_norm": 3.4247073453125774, + "learning_rate": 8.830096371397982e-07, + "loss": 0.1086, + "step": 1605 + }, + { + "epoch": 0.365, + "grad_norm": 1.0726631200597323, + "learning_rate": 8.826031212962064e-07, + "loss": 0.0468, + "step": 1606 + }, + { + "epoch": 0.36522727272727273, + "grad_norm": 2.2302444224916034, + "learning_rate": 8.821964741281932e-07, + "loss": 0.0526, + "step": 1607 + }, + { + "epoch": 0.3654545454545455, + "grad_norm": 5.878874055262201, + "learning_rate": 8.81789695843065e-07, + "loss": 0.1494, + "step": 1608 + }, + { + "epoch": 0.36568181818181816, + "grad_norm": 2.779297269369383, + "learning_rate": 8.81382786648195e-07, + "loss": 0.1423, + "step": 1609 + }, + { + "epoch": 0.3659090909090909, + "grad_norm": 3.5124855920715703, + "learning_rate": 8.809757467510227e-07, + "loss": 0.202, + "step": 1610 + }, + { + "epoch": 0.36613636363636365, + "grad_norm": 2.6108779995174194, + "learning_rate": 8.805685763590543e-07, + "loss": 0.0742, + "step": 1611 + }, + { + "epoch": 0.3663636363636364, + "grad_norm": 2.759996351422273, + "learning_rate": 8.80161275679863e-07, + "loss": 0.1652, + "step": 1612 + }, + { + "epoch": 0.3665909090909091, + "grad_norm": 1.3683184253730374, + "learning_rate": 8.797538449210878e-07, + "loss": 0.0963, + "step": 1613 + }, + { + "epoch": 0.3668181818181818, + "grad_norm": 2.4589143446643202, + "learning_rate": 8.793462842904344e-07, + "loss": 0.0918, + "step": 1614 + }, + { + "epoch": 0.36704545454545456, + "grad_norm": 2.5527884499809668, + "learning_rate": 8.789385939956747e-07, + "loss": 0.1009, + "step": 1615 + }, + { + "epoch": 0.36727272727272725, + "grad_norm": 3.2478891098489737, + "learning_rate": 8.785307742446464e-07, + "loss": 0.0736, + "step": 1616 + }, + { + "epoch": 0.3675, + "grad_norm": 2.932324167095727, + "learning_rate": 8.781228252452534e-07, + "loss": 0.0999, + "step": 1617 + }, + { + "epoch": 0.36772727272727274, + "grad_norm": 2.5771376733294127, + "learning_rate": 8.777147472054657e-07, + "loss": 0.0857, + "step": 1618 + }, + { + "epoch": 0.3679545454545455, + "grad_norm": 1.8452491285822252, + "learning_rate": 8.773065403333186e-07, + "loss": 0.0919, + "step": 1619 + }, + { + "epoch": 0.36818181818181817, + "grad_norm": 2.68738906385096, + "learning_rate": 8.768982048369136e-07, + "loss": 0.0582, + "step": 1620 + }, + { + "epoch": 0.3684090909090909, + "grad_norm": 4.666394270820283, + "learning_rate": 8.764897409244176e-07, + "loss": 0.146, + "step": 1621 + }, + { + "epoch": 0.36863636363636365, + "grad_norm": 2.2462719558895623, + "learning_rate": 8.760811488040624e-07, + "loss": 0.0705, + "step": 1622 + }, + { + "epoch": 0.36886363636363634, + "grad_norm": 6.822926367815532, + "learning_rate": 8.756724286841458e-07, + "loss": 0.1688, + "step": 1623 + }, + { + "epoch": 0.3690909090909091, + "grad_norm": 2.494361254041984, + "learning_rate": 8.752635807730314e-07, + "loss": 0.1172, + "step": 1624 + }, + { + "epoch": 0.3693181818181818, + "grad_norm": 3.7195255083610217, + "learning_rate": 8.748546052791462e-07, + "loss": 0.0947, + "step": 1625 + }, + { + "epoch": 0.36954545454545457, + "grad_norm": 2.423179296734131, + "learning_rate": 8.744455024109838e-07, + "loss": 0.0904, + "step": 1626 + }, + { + "epoch": 0.36977272727272725, + "grad_norm": 3.0919085084362656, + "learning_rate": 8.740362723771024e-07, + "loss": 0.1508, + "step": 1627 + }, + { + "epoch": 0.37, + "grad_norm": 3.6485317739043417, + "learning_rate": 8.736269153861246e-07, + "loss": 0.109, + "step": 1628 + }, + { + "epoch": 0.37022727272727274, + "grad_norm": 3.057513667664217, + "learning_rate": 8.732174316467379e-07, + "loss": 0.1384, + "step": 1629 + }, + { + "epoch": 0.3704545454545455, + "grad_norm": 3.5986238798923362, + "learning_rate": 8.728078213676948e-07, + "loss": 0.1669, + "step": 1630 + }, + { + "epoch": 0.37068181818181817, + "grad_norm": 5.230175156920944, + "learning_rate": 8.723980847578116e-07, + "loss": 0.125, + "step": 1631 + }, + { + "epoch": 0.3709090909090909, + "grad_norm": 3.564177215633902, + "learning_rate": 8.719882220259697e-07, + "loss": 0.0946, + "step": 1632 + }, + { + "epoch": 0.37113636363636365, + "grad_norm": 3.5975693769032144, + "learning_rate": 8.715782333811143e-07, + "loss": 0.0939, + "step": 1633 + }, + { + "epoch": 0.37136363636363634, + "grad_norm": 3.9518116842837383, + "learning_rate": 8.711681190322552e-07, + "loss": 0.0974, + "step": 1634 + }, + { + "epoch": 0.3715909090909091, + "grad_norm": 2.0581187517374766, + "learning_rate": 8.707578791884659e-07, + "loss": 0.0662, + "step": 1635 + }, + { + "epoch": 0.3718181818181818, + "grad_norm": 4.596097364658624, + "learning_rate": 8.703475140588841e-07, + "loss": 0.1905, + "step": 1636 + }, + { + "epoch": 0.37204545454545457, + "grad_norm": 3.5334468839601323, + "learning_rate": 8.699370238527112e-07, + "loss": 0.1103, + "step": 1637 + }, + { + "epoch": 0.37227272727272726, + "grad_norm": 2.1086696546692396, + "learning_rate": 8.695264087792127e-07, + "loss": 0.0575, + "step": 1638 + }, + { + "epoch": 0.3725, + "grad_norm": 4.583726087414378, + "learning_rate": 8.691156690477174e-07, + "loss": 0.1037, + "step": 1639 + }, + { + "epoch": 0.37272727272727274, + "grad_norm": 3.677559384340975, + "learning_rate": 8.687048048676178e-07, + "loss": 0.125, + "step": 1640 + }, + { + "epoch": 0.3729545454545454, + "grad_norm": 1.6763202099371792, + "learning_rate": 8.682938164483698e-07, + "loss": 0.0712, + "step": 1641 + }, + { + "epoch": 0.37318181818181817, + "grad_norm": 2.4032895592668955, + "learning_rate": 8.678827039994928e-07, + "loss": 0.0912, + "step": 1642 + }, + { + "epoch": 0.3734090909090909, + "grad_norm": 3.8898704028627513, + "learning_rate": 8.67471467730569e-07, + "loss": 0.2134, + "step": 1643 + }, + { + "epoch": 0.37363636363636366, + "grad_norm": 2.5883299835099773, + "learning_rate": 8.670601078512443e-07, + "loss": 0.087, + "step": 1644 + }, + { + "epoch": 0.37386363636363634, + "grad_norm": 3.1570459493152003, + "learning_rate": 8.666486245712271e-07, + "loss": 0.1078, + "step": 1645 + }, + { + "epoch": 0.3740909090909091, + "grad_norm": 4.131500771262186, + "learning_rate": 8.662370181002892e-07, + "loss": 0.1601, + "step": 1646 + }, + { + "epoch": 0.3743181818181818, + "grad_norm": 3.476432508445972, + "learning_rate": 8.658252886482649e-07, + "loss": 0.0894, + "step": 1647 + }, + { + "epoch": 0.37454545454545457, + "grad_norm": 5.692710119304813, + "learning_rate": 8.654134364250509e-07, + "loss": 0.099, + "step": 1648 + }, + { + "epoch": 0.37477272727272726, + "grad_norm": 2.7465449202327936, + "learning_rate": 8.650014616406072e-07, + "loss": 0.0839, + "step": 1649 + }, + { + "epoch": 0.375, + "grad_norm": 3.2322689489023846, + "learning_rate": 8.64589364504956e-07, + "loss": 0.1081, + "step": 1650 + }, + { + "epoch": 0.37522727272727274, + "grad_norm": 1.773317618690949, + "learning_rate": 8.641771452281812e-07, + "loss": 0.0412, + "step": 1651 + }, + { + "epoch": 0.37545454545454543, + "grad_norm": 2.9377370140232566, + "learning_rate": 8.637648040204299e-07, + "loss": 0.0788, + "step": 1652 + }, + { + "epoch": 0.3756818181818182, + "grad_norm": 2.7212258891911536, + "learning_rate": 8.633523410919114e-07, + "loss": 0.0696, + "step": 1653 + }, + { + "epoch": 0.3759090909090909, + "grad_norm": 2.573935660403796, + "learning_rate": 8.62939756652896e-07, + "loss": 0.1419, + "step": 1654 + }, + { + "epoch": 0.37613636363636366, + "grad_norm": 2.2710757163558624, + "learning_rate": 8.625270509137168e-07, + "loss": 0.0952, + "step": 1655 + }, + { + "epoch": 0.37636363636363634, + "grad_norm": 3.6258141759415285, + "learning_rate": 8.621142240847688e-07, + "loss": 0.2154, + "step": 1656 + }, + { + "epoch": 0.3765909090909091, + "grad_norm": 3.5157090522781664, + "learning_rate": 8.617012763765084e-07, + "loss": 0.1031, + "step": 1657 + }, + { + "epoch": 0.37681818181818183, + "grad_norm": 2.365926816939538, + "learning_rate": 8.612882079994534e-07, + "loss": 0.1385, + "step": 1658 + }, + { + "epoch": 0.3770454545454546, + "grad_norm": 3.132022621170078, + "learning_rate": 8.608750191641835e-07, + "loss": 0.1351, + "step": 1659 + }, + { + "epoch": 0.37727272727272726, + "grad_norm": 2.508992338476082, + "learning_rate": 8.604617100813399e-07, + "loss": 0.0932, + "step": 1660 + }, + { + "epoch": 0.3775, + "grad_norm": 3.1064988886806897, + "learning_rate": 8.600482809616249e-07, + "loss": 0.0776, + "step": 1661 + }, + { + "epoch": 0.37772727272727274, + "grad_norm": 2.0426882152912467, + "learning_rate": 8.59634732015802e-07, + "loss": 0.0606, + "step": 1662 + }, + { + "epoch": 0.37795454545454543, + "grad_norm": 2.554991716691398, + "learning_rate": 8.592210634546958e-07, + "loss": 0.0729, + "step": 1663 + }, + { + "epoch": 0.3781818181818182, + "grad_norm": 3.775995575411098, + "learning_rate": 8.588072754891917e-07, + "loss": 0.1305, + "step": 1664 + }, + { + "epoch": 0.3784090909090909, + "grad_norm": 3.021799954742129, + "learning_rate": 8.583933683302364e-07, + "loss": 0.1268, + "step": 1665 + }, + { + "epoch": 0.37863636363636366, + "grad_norm": 4.972982903437959, + "learning_rate": 8.579793421888369e-07, + "loss": 0.1676, + "step": 1666 + }, + { + "epoch": 0.37886363636363635, + "grad_norm": 2.9831543984124873, + "learning_rate": 8.575651972760613e-07, + "loss": 0.1084, + "step": 1667 + }, + { + "epoch": 0.3790909090909091, + "grad_norm": 1.9730062617480968, + "learning_rate": 8.57150933803038e-07, + "loss": 0.1095, + "step": 1668 + }, + { + "epoch": 0.37931818181818183, + "grad_norm": 3.7676769756442448, + "learning_rate": 8.567365519809556e-07, + "loss": 0.1278, + "step": 1669 + }, + { + "epoch": 0.3795454545454545, + "grad_norm": 2.617920179972844, + "learning_rate": 8.563220520210636e-07, + "loss": 0.0963, + "step": 1670 + }, + { + "epoch": 0.37977272727272726, + "grad_norm": 3.7644358970540965, + "learning_rate": 8.559074341346713e-07, + "loss": 0.1073, + "step": 1671 + }, + { + "epoch": 0.38, + "grad_norm": 1.3844443606138463, + "learning_rate": 8.554926985331481e-07, + "loss": 0.0569, + "step": 1672 + }, + { + "epoch": 0.38022727272727275, + "grad_norm": 1.7226191550665115, + "learning_rate": 8.550778454279239e-07, + "loss": 0.0744, + "step": 1673 + }, + { + "epoch": 0.38045454545454543, + "grad_norm": 1.621862353118295, + "learning_rate": 8.546628750304877e-07, + "loss": 0.0406, + "step": 1674 + }, + { + "epoch": 0.3806818181818182, + "grad_norm": 3.9436428277374835, + "learning_rate": 8.542477875523893e-07, + "loss": 0.0788, + "step": 1675 + }, + { + "epoch": 0.3809090909090909, + "grad_norm": 4.205721318972414, + "learning_rate": 8.538325832052373e-07, + "loss": 0.1098, + "step": 1676 + }, + { + "epoch": 0.38113636363636366, + "grad_norm": 3.4170903060004014, + "learning_rate": 8.534172622007e-07, + "loss": 0.145, + "step": 1677 + }, + { + "epoch": 0.38136363636363635, + "grad_norm": 8.203079749390382, + "learning_rate": 8.530018247505058e-07, + "loss": 0.1474, + "step": 1678 + }, + { + "epoch": 0.3815909090909091, + "grad_norm": 2.0884205921744634, + "learning_rate": 8.525862710664421e-07, + "loss": 0.0515, + "step": 1679 + }, + { + "epoch": 0.38181818181818183, + "grad_norm": 5.623830672837946, + "learning_rate": 8.521706013603551e-07, + "loss": 0.169, + "step": 1680 + }, + { + "epoch": 0.3820454545454545, + "grad_norm": 3.366987067469043, + "learning_rate": 8.517548158441506e-07, + "loss": 0.0932, + "step": 1681 + }, + { + "epoch": 0.38227272727272726, + "grad_norm": 3.7614011684821262, + "learning_rate": 8.51338914729794e-07, + "loss": 0.0977, + "step": 1682 + }, + { + "epoch": 0.3825, + "grad_norm": 1.9645949161981535, + "learning_rate": 8.509228982293083e-07, + "loss": 0.0591, + "step": 1683 + }, + { + "epoch": 0.38272727272727275, + "grad_norm": 1.708932082237944, + "learning_rate": 8.505067665547761e-07, + "loss": 0.0677, + "step": 1684 + }, + { + "epoch": 0.38295454545454544, + "grad_norm": 2.9757115065852062, + "learning_rate": 8.500905199183393e-07, + "loss": 0.1147, + "step": 1685 + }, + { + "epoch": 0.3831818181818182, + "grad_norm": 3.280950499019517, + "learning_rate": 8.496741585321971e-07, + "loss": 0.0895, + "step": 1686 + }, + { + "epoch": 0.3834090909090909, + "grad_norm": 3.426798952151292, + "learning_rate": 8.492576826086083e-07, + "loss": 0.0801, + "step": 1687 + }, + { + "epoch": 0.3836363636363636, + "grad_norm": 3.0718358883583843, + "learning_rate": 8.488410923598893e-07, + "loss": 0.077, + "step": 1688 + }, + { + "epoch": 0.38386363636363635, + "grad_norm": 1.8998280944738741, + "learning_rate": 8.484243879984154e-07, + "loss": 0.0558, + "step": 1689 + }, + { + "epoch": 0.3840909090909091, + "grad_norm": 3.9776503181046703, + "learning_rate": 8.480075697366197e-07, + "loss": 0.188, + "step": 1690 + }, + { + "epoch": 0.38431818181818184, + "grad_norm": 2.427359084935202, + "learning_rate": 8.475906377869936e-07, + "loss": 0.0584, + "step": 1691 + }, + { + "epoch": 0.3845454545454545, + "grad_norm": 3.328930863048826, + "learning_rate": 8.471735923620861e-07, + "loss": 0.1099, + "step": 1692 + }, + { + "epoch": 0.38477272727272727, + "grad_norm": 3.442438612466405, + "learning_rate": 8.467564336745046e-07, + "loss": 0.1958, + "step": 1693 + }, + { + "epoch": 0.385, + "grad_norm": 3.846124145553009, + "learning_rate": 8.463391619369136e-07, + "loss": 0.1437, + "step": 1694 + }, + { + "epoch": 0.38522727272727275, + "grad_norm": 4.959151734491105, + "learning_rate": 8.459217773620358e-07, + "loss": 0.1196, + "step": 1695 + }, + { + "epoch": 0.38545454545454544, + "grad_norm": 1.7804919777362223, + "learning_rate": 8.45504280162651e-07, + "loss": 0.0556, + "step": 1696 + }, + { + "epoch": 0.3856818181818182, + "grad_norm": 4.270156747470037, + "learning_rate": 8.45086670551597e-07, + "loss": 0.1135, + "step": 1697 + }, + { + "epoch": 0.3859090909090909, + "grad_norm": 5.183345254813083, + "learning_rate": 8.44668948741768e-07, + "loss": 0.2944, + "step": 1698 + }, + { + "epoch": 0.3861363636363636, + "grad_norm": 2.7570699940781007, + "learning_rate": 8.442511149461159e-07, + "loss": 0.1028, + "step": 1699 + }, + { + "epoch": 0.38636363636363635, + "grad_norm": 2.064980316815718, + "learning_rate": 8.438331693776504e-07, + "loss": 0.0629, + "step": 1700 + }, + { + "epoch": 0.3865909090909091, + "grad_norm": 3.3118142887057056, + "learning_rate": 8.434151122494366e-07, + "loss": 0.1854, + "step": 1701 + }, + { + "epoch": 0.38681818181818184, + "grad_norm": 3.2618511476743186, + "learning_rate": 8.429969437745977e-07, + "loss": 0.0891, + "step": 1702 + }, + { + "epoch": 0.3870454545454545, + "grad_norm": 2.1268874821378168, + "learning_rate": 8.425786641663135e-07, + "loss": 0.1008, + "step": 1703 + }, + { + "epoch": 0.38727272727272727, + "grad_norm": 2.2571555240157464, + "learning_rate": 8.421602736378201e-07, + "loss": 0.081, + "step": 1704 + }, + { + "epoch": 0.3875, + "grad_norm": 3.156677125515753, + "learning_rate": 8.417417724024101e-07, + "loss": 0.1152, + "step": 1705 + }, + { + "epoch": 0.38772727272727275, + "grad_norm": 3.721341749384125, + "learning_rate": 8.413231606734332e-07, + "loss": 0.0859, + "step": 1706 + }, + { + "epoch": 0.38795454545454544, + "grad_norm": 2.8918340686584765, + "learning_rate": 8.409044386642945e-07, + "loss": 0.1552, + "step": 1707 + }, + { + "epoch": 0.3881818181818182, + "grad_norm": 2.570815036669092, + "learning_rate": 8.404856065884564e-07, + "loss": 0.0831, + "step": 1708 + }, + { + "epoch": 0.3884090909090909, + "grad_norm": 2.753002107687662, + "learning_rate": 8.400666646594362e-07, + "loss": 0.1131, + "step": 1709 + }, + { + "epoch": 0.3886363636363636, + "grad_norm": 2.934394332583303, + "learning_rate": 8.39647613090808e-07, + "loss": 0.1431, + "step": 1710 + }, + { + "epoch": 0.38886363636363636, + "grad_norm": 4.1855091094117025, + "learning_rate": 8.392284520962017e-07, + "loss": 0.1478, + "step": 1711 + }, + { + "epoch": 0.3890909090909091, + "grad_norm": 1.940058468767866, + "learning_rate": 8.388091818893029e-07, + "loss": 0.0685, + "step": 1712 + }, + { + "epoch": 0.38931818181818184, + "grad_norm": 1.7692872069339232, + "learning_rate": 8.383898026838527e-07, + "loss": 0.0509, + "step": 1713 + }, + { + "epoch": 0.38954545454545453, + "grad_norm": 4.437262572953942, + "learning_rate": 8.379703146936482e-07, + "loss": 0.167, + "step": 1714 + }, + { + "epoch": 0.38977272727272727, + "grad_norm": 2.937599682709242, + "learning_rate": 8.375507181325415e-07, + "loss": 0.0693, + "step": 1715 + }, + { + "epoch": 0.39, + "grad_norm": 4.575214715712697, + "learning_rate": 8.371310132144401e-07, + "loss": 0.1054, + "step": 1716 + }, + { + "epoch": 0.3902272727272727, + "grad_norm": 3.109110652397384, + "learning_rate": 8.367112001533073e-07, + "loss": 0.1449, + "step": 1717 + }, + { + "epoch": 0.39045454545454544, + "grad_norm": 3.144965199910514, + "learning_rate": 8.362912791631607e-07, + "loss": 0.1163, + "step": 1718 + }, + { + "epoch": 0.3906818181818182, + "grad_norm": 4.342600618409988, + "learning_rate": 8.358712504580735e-07, + "loss": 0.1056, + "step": 1719 + }, + { + "epoch": 0.39090909090909093, + "grad_norm": 2.726252648887146, + "learning_rate": 8.354511142521738e-07, + "loss": 0.1245, + "step": 1720 + }, + { + "epoch": 0.3911363636363636, + "grad_norm": 2.034084705824678, + "learning_rate": 8.350308707596438e-07, + "loss": 0.0847, + "step": 1721 + }, + { + "epoch": 0.39136363636363636, + "grad_norm": 4.121118033826629, + "learning_rate": 8.346105201947217e-07, + "loss": 0.1246, + "step": 1722 + }, + { + "epoch": 0.3915909090909091, + "grad_norm": 3.8569053337565857, + "learning_rate": 8.34190062771699e-07, + "loss": 0.1387, + "step": 1723 + }, + { + "epoch": 0.39181818181818184, + "grad_norm": 7.8100664566272515, + "learning_rate": 8.337694987049225e-07, + "loss": 0.2184, + "step": 1724 + }, + { + "epoch": 0.39204545454545453, + "grad_norm": 3.7697398546166436, + "learning_rate": 8.333488282087926e-07, + "loss": 0.2178, + "step": 1725 + }, + { + "epoch": 0.3922727272727273, + "grad_norm": 4.916518308563101, + "learning_rate": 8.329280514977652e-07, + "loss": 0.1468, + "step": 1726 + }, + { + "epoch": 0.3925, + "grad_norm": 2.383202471306213, + "learning_rate": 8.325071687863487e-07, + "loss": 0.144, + "step": 1727 + }, + { + "epoch": 0.3927272727272727, + "grad_norm": 3.695862531054123, + "learning_rate": 8.320861802891068e-07, + "loss": 0.1288, + "step": 1728 + }, + { + "epoch": 0.39295454545454545, + "grad_norm": 3.879424295886134, + "learning_rate": 8.316650862206572e-07, + "loss": 0.153, + "step": 1729 + }, + { + "epoch": 0.3931818181818182, + "grad_norm": 3.4597070198598945, + "learning_rate": 8.312438867956702e-07, + "loss": 0.1084, + "step": 1730 + }, + { + "epoch": 0.39340909090909093, + "grad_norm": 4.764074181607729, + "learning_rate": 8.308225822288709e-07, + "loss": 0.1368, + "step": 1731 + }, + { + "epoch": 0.3936363636363636, + "grad_norm": 1.6969151030581575, + "learning_rate": 8.304011727350377e-07, + "loss": 0.124, + "step": 1732 + }, + { + "epoch": 0.39386363636363636, + "grad_norm": 1.8938684089757047, + "learning_rate": 8.299796585290024e-07, + "loss": 0.0666, + "step": 1733 + }, + { + "epoch": 0.3940909090909091, + "grad_norm": 3.9786897178827227, + "learning_rate": 8.295580398256503e-07, + "loss": 0.17, + "step": 1734 + }, + { + "epoch": 0.39431818181818185, + "grad_norm": 2.4644243106396155, + "learning_rate": 8.291363168399198e-07, + "loss": 0.1566, + "step": 1735 + }, + { + "epoch": 0.39454545454545453, + "grad_norm": 4.2349275405389415, + "learning_rate": 8.287144897868027e-07, + "loss": 0.1675, + "step": 1736 + }, + { + "epoch": 0.3947727272727273, + "grad_norm": 4.287962250873789, + "learning_rate": 8.282925588813437e-07, + "loss": 0.1862, + "step": 1737 + }, + { + "epoch": 0.395, + "grad_norm": 4.395607406813464, + "learning_rate": 8.278705243386404e-07, + "loss": 0.1546, + "step": 1738 + }, + { + "epoch": 0.3952272727272727, + "grad_norm": 4.335197020508404, + "learning_rate": 8.274483863738435e-07, + "loss": 0.0962, + "step": 1739 + }, + { + "epoch": 0.39545454545454545, + "grad_norm": 2.521248165110261, + "learning_rate": 8.270261452021559e-07, + "loss": 0.0661, + "step": 1740 + }, + { + "epoch": 0.3956818181818182, + "grad_norm": 3.390258272130442, + "learning_rate": 8.266038010388339e-07, + "loss": 0.0846, + "step": 1741 + }, + { + "epoch": 0.39590909090909093, + "grad_norm": 5.110509550519997, + "learning_rate": 8.261813540991853e-07, + "loss": 0.1433, + "step": 1742 + }, + { + "epoch": 0.3961363636363636, + "grad_norm": 4.017617153792596, + "learning_rate": 8.257588045985712e-07, + "loss": 0.104, + "step": 1743 + }, + { + "epoch": 0.39636363636363636, + "grad_norm": 1.5399949566774278, + "learning_rate": 8.253361527524047e-07, + "loss": 0.0843, + "step": 1744 + }, + { + "epoch": 0.3965909090909091, + "grad_norm": 2.524566374718052, + "learning_rate": 8.249133987761507e-07, + "loss": 0.0532, + "step": 1745 + }, + { + "epoch": 0.3968181818181818, + "grad_norm": 1.9451266139206465, + "learning_rate": 8.244905428853267e-07, + "loss": 0.0903, + "step": 1746 + }, + { + "epoch": 0.39704545454545453, + "grad_norm": 3.1837539903782353, + "learning_rate": 8.240675852955019e-07, + "loss": 0.111, + "step": 1747 + }, + { + "epoch": 0.3972727272727273, + "grad_norm": 3.029784144030318, + "learning_rate": 8.236445262222972e-07, + "loss": 0.1924, + "step": 1748 + }, + { + "epoch": 0.3975, + "grad_norm": 3.7358278327328045, + "learning_rate": 8.232213658813856e-07, + "loss": 0.1835, + "step": 1749 + }, + { + "epoch": 0.3977272727272727, + "grad_norm": 4.75220287419906, + "learning_rate": 8.227981044884914e-07, + "loss": 0.2004, + "step": 1750 + }, + { + "epoch": 0.39795454545454545, + "grad_norm": 2.646906147274965, + "learning_rate": 8.223747422593907e-07, + "loss": 0.1016, + "step": 1751 + }, + { + "epoch": 0.3981818181818182, + "grad_norm": 3.537154104860749, + "learning_rate": 8.219512794099108e-07, + "loss": 0.0988, + "step": 1752 + }, + { + "epoch": 0.39840909090909093, + "grad_norm": 2.01548630737095, + "learning_rate": 8.215277161559301e-07, + "loss": 0.1203, + "step": 1753 + }, + { + "epoch": 0.3986363636363636, + "grad_norm": 1.9506095719241674, + "learning_rate": 8.211040527133787e-07, + "loss": 0.0906, + "step": 1754 + }, + { + "epoch": 0.39886363636363636, + "grad_norm": 3.873003673041192, + "learning_rate": 8.206802892982376e-07, + "loss": 0.1373, + "step": 1755 + }, + { + "epoch": 0.3990909090909091, + "grad_norm": 2.5963211837426776, + "learning_rate": 8.202564261265385e-07, + "loss": 0.0887, + "step": 1756 + }, + { + "epoch": 0.3993181818181818, + "grad_norm": 4.565233601278384, + "learning_rate": 8.19832463414364e-07, + "loss": 0.1529, + "step": 1757 + }, + { + "epoch": 0.39954545454545454, + "grad_norm": 3.2067558915074517, + "learning_rate": 8.194084013778478e-07, + "loss": 0.1247, + "step": 1758 + }, + { + "epoch": 0.3997727272727273, + "grad_norm": 5.11236333031081, + "learning_rate": 8.189842402331741e-07, + "loss": 0.3073, + "step": 1759 + }, + { + "epoch": 0.4, + "grad_norm": 3.3862587410864866, + "learning_rate": 8.185599801965772e-07, + "loss": 0.0937, + "step": 1760 + }, + { + "epoch": 0.4002272727272727, + "grad_norm": 2.9809621433237847, + "learning_rate": 8.181356214843422e-07, + "loss": 0.1091, + "step": 1761 + }, + { + "epoch": 0.40045454545454545, + "grad_norm": 4.54064326886375, + "learning_rate": 8.177111643128048e-07, + "loss": 0.1371, + "step": 1762 + }, + { + "epoch": 0.4006818181818182, + "grad_norm": 4.074927515831305, + "learning_rate": 8.172866088983501e-07, + "loss": 0.1284, + "step": 1763 + }, + { + "epoch": 0.4009090909090909, + "grad_norm": 2.784694191014518, + "learning_rate": 8.168619554574142e-07, + "loss": 0.1521, + "step": 1764 + }, + { + "epoch": 0.4011363636363636, + "grad_norm": 4.017422987923594, + "learning_rate": 8.164372042064825e-07, + "loss": 0.1383, + "step": 1765 + }, + { + "epoch": 0.40136363636363637, + "grad_norm": 4.487815716771649, + "learning_rate": 8.160123553620901e-07, + "loss": 0.2045, + "step": 1766 + }, + { + "epoch": 0.4015909090909091, + "grad_norm": 2.052625729327693, + "learning_rate": 8.155874091408228e-07, + "loss": 0.0893, + "step": 1767 + }, + { + "epoch": 0.4018181818181818, + "grad_norm": 2.9093459262520454, + "learning_rate": 8.15162365759315e-07, + "loss": 0.1586, + "step": 1768 + }, + { + "epoch": 0.40204545454545454, + "grad_norm": 2.2872778692075677, + "learning_rate": 8.147372254342514e-07, + "loss": 0.0639, + "step": 1769 + }, + { + "epoch": 0.4022727272727273, + "grad_norm": 4.112328440513925, + "learning_rate": 8.143119883823658e-07, + "loss": 0.1429, + "step": 1770 + }, + { + "epoch": 0.4025, + "grad_norm": 3.1376070994132976, + "learning_rate": 8.138866548204412e-07, + "loss": 0.1789, + "step": 1771 + }, + { + "epoch": 0.4027272727272727, + "grad_norm": 2.7012042996992855, + "learning_rate": 8.134612249653098e-07, + "loss": 0.0995, + "step": 1772 + }, + { + "epoch": 0.40295454545454545, + "grad_norm": 2.499560273528047, + "learning_rate": 8.130356990338535e-07, + "loss": 0.1374, + "step": 1773 + }, + { + "epoch": 0.4031818181818182, + "grad_norm": 3.158829859742393, + "learning_rate": 8.126100772430021e-07, + "loss": 0.1179, + "step": 1774 + }, + { + "epoch": 0.4034090909090909, + "grad_norm": 5.547536840789584, + "learning_rate": 8.121843598097351e-07, + "loss": 0.1329, + "step": 1775 + }, + { + "epoch": 0.4036363636363636, + "grad_norm": 3.800693641770729, + "learning_rate": 8.117585469510808e-07, + "loss": 0.1135, + "step": 1776 + }, + { + "epoch": 0.40386363636363637, + "grad_norm": 2.293984183447976, + "learning_rate": 8.113326388841154e-07, + "loss": 0.0895, + "step": 1777 + }, + { + "epoch": 0.4040909090909091, + "grad_norm": 5.856693978071074, + "learning_rate": 8.109066358259642e-07, + "loss": 0.3035, + "step": 1778 + }, + { + "epoch": 0.4043181818181818, + "grad_norm": 1.5137240812707746, + "learning_rate": 8.104805379938013e-07, + "loss": 0.0397, + "step": 1779 + }, + { + "epoch": 0.40454545454545454, + "grad_norm": 2.3802096075464094, + "learning_rate": 8.100543456048482e-07, + "loss": 0.0792, + "step": 1780 + }, + { + "epoch": 0.4047727272727273, + "grad_norm": 5.306297383488865, + "learning_rate": 8.096280588763751e-07, + "loss": 0.2464, + "step": 1781 + }, + { + "epoch": 0.405, + "grad_norm": 1.4552064241036378, + "learning_rate": 8.092016780257003e-07, + "loss": 0.1, + "step": 1782 + }, + { + "epoch": 0.4052272727272727, + "grad_norm": 2.652583692502112, + "learning_rate": 8.0877520327019e-07, + "loss": 0.1129, + "step": 1783 + }, + { + "epoch": 0.40545454545454546, + "grad_norm": 2.4724201215279984, + "learning_rate": 8.083486348272584e-07, + "loss": 0.1058, + "step": 1784 + }, + { + "epoch": 0.4056818181818182, + "grad_norm": 3.3661892009861023, + "learning_rate": 8.079219729143672e-07, + "loss": 0.1331, + "step": 1785 + }, + { + "epoch": 0.4059090909090909, + "grad_norm": 3.9969492875004837, + "learning_rate": 8.07495217749026e-07, + "loss": 0.1874, + "step": 1786 + }, + { + "epoch": 0.40613636363636363, + "grad_norm": 5.070791665215432, + "learning_rate": 8.070683695487919e-07, + "loss": 0.2143, + "step": 1787 + }, + { + "epoch": 0.40636363636363637, + "grad_norm": 4.3964442316063765, + "learning_rate": 8.066414285312693e-07, + "loss": 0.1338, + "step": 1788 + }, + { + "epoch": 0.4065909090909091, + "grad_norm": 4.547728600711837, + "learning_rate": 8.062143949141099e-07, + "loss": 0.1357, + "step": 1789 + }, + { + "epoch": 0.4068181818181818, + "grad_norm": 2.739298524946307, + "learning_rate": 8.057872689150129e-07, + "loss": 0.0821, + "step": 1790 + }, + { + "epoch": 0.40704545454545454, + "grad_norm": 2.5897561780492118, + "learning_rate": 8.053600507517244e-07, + "loss": 0.1153, + "step": 1791 + }, + { + "epoch": 0.4072727272727273, + "grad_norm": 2.6428583589890784, + "learning_rate": 8.049327406420371e-07, + "loss": 0.0899, + "step": 1792 + }, + { + "epoch": 0.4075, + "grad_norm": 2.16322245527167, + "learning_rate": 8.045053388037916e-07, + "loss": 0.1564, + "step": 1793 + }, + { + "epoch": 0.4077272727272727, + "grad_norm": 5.2414060092445345, + "learning_rate": 8.040778454548741e-07, + "loss": 0.2859, + "step": 1794 + }, + { + "epoch": 0.40795454545454546, + "grad_norm": 3.8817035854092206, + "learning_rate": 8.036502608132181e-07, + "loss": 0.1328, + "step": 1795 + }, + { + "epoch": 0.4081818181818182, + "grad_norm": 3.180796606630757, + "learning_rate": 8.032225850968037e-07, + "loss": 0.0884, + "step": 1796 + }, + { + "epoch": 0.4084090909090909, + "grad_norm": 2.435906911918572, + "learning_rate": 8.027948185236568e-07, + "loss": 0.0636, + "step": 1797 + }, + { + "epoch": 0.40863636363636363, + "grad_norm": 3.289508820279937, + "learning_rate": 8.023669613118507e-07, + "loss": 0.0847, + "step": 1798 + }, + { + "epoch": 0.4088636363636364, + "grad_norm": 5.433876315790335, + "learning_rate": 8.019390136795039e-07, + "loss": 0.1731, + "step": 1799 + }, + { + "epoch": 0.4090909090909091, + "grad_norm": 3.7945903300342243, + "learning_rate": 8.01510975844781e-07, + "loss": 0.1398, + "step": 1800 + }, + { + "epoch": 0.4093181818181818, + "grad_norm": 1.982257202373055, + "learning_rate": 8.010828480258937e-07, + "loss": 0.0822, + "step": 1801 + }, + { + "epoch": 0.40954545454545455, + "grad_norm": 2.087662581908163, + "learning_rate": 8.006546304410984e-07, + "loss": 0.069, + "step": 1802 + }, + { + "epoch": 0.4097727272727273, + "grad_norm": 3.5559117665614184, + "learning_rate": 8.002263233086975e-07, + "loss": 0.0909, + "step": 1803 + }, + { + "epoch": 0.41, + "grad_norm": 3.544127382602151, + "learning_rate": 7.997979268470396e-07, + "loss": 0.2315, + "step": 1804 + }, + { + "epoch": 0.4102272727272727, + "grad_norm": 2.8641710070386464, + "learning_rate": 7.993694412745184e-07, + "loss": 0.1277, + "step": 1805 + }, + { + "epoch": 0.41045454545454546, + "grad_norm": 5.108302290737701, + "learning_rate": 7.98940866809573e-07, + "loss": 0.1654, + "step": 1806 + }, + { + "epoch": 0.4106818181818182, + "grad_norm": 3.297381345132546, + "learning_rate": 7.985122036706879e-07, + "loss": 0.1064, + "step": 1807 + }, + { + "epoch": 0.4109090909090909, + "grad_norm": 4.0288096049828175, + "learning_rate": 7.980834520763929e-07, + "loss": 0.1116, + "step": 1808 + }, + { + "epoch": 0.41113636363636363, + "grad_norm": 2.5796971814029126, + "learning_rate": 7.976546122452627e-07, + "loss": 0.0856, + "step": 1809 + }, + { + "epoch": 0.4113636363636364, + "grad_norm": 3.05452232558922, + "learning_rate": 7.972256843959172e-07, + "loss": 0.0891, + "step": 1810 + }, + { + "epoch": 0.4115909090909091, + "grad_norm": 3.8914332442618966, + "learning_rate": 7.967966687470211e-07, + "loss": 0.1823, + "step": 1811 + }, + { + "epoch": 0.4118181818181818, + "grad_norm": 3.6448778044002403, + "learning_rate": 7.963675655172839e-07, + "loss": 0.1064, + "step": 1812 + }, + { + "epoch": 0.41204545454545455, + "grad_norm": 2.6505986970584425, + "learning_rate": 7.959383749254595e-07, + "loss": 0.1179, + "step": 1813 + }, + { + "epoch": 0.4122727272727273, + "grad_norm": 4.314639456492938, + "learning_rate": 7.955090971903464e-07, + "loss": 0.1932, + "step": 1814 + }, + { + "epoch": 0.4125, + "grad_norm": 2.862668967107881, + "learning_rate": 7.95079732530788e-07, + "loss": 0.0706, + "step": 1815 + }, + { + "epoch": 0.4127272727272727, + "grad_norm": 3.1174876047127733, + "learning_rate": 7.946502811656714e-07, + "loss": 0.1159, + "step": 1816 + }, + { + "epoch": 0.41295454545454546, + "grad_norm": 2.688025384474168, + "learning_rate": 7.942207433139284e-07, + "loss": 0.0916, + "step": 1817 + }, + { + "epoch": 0.4131818181818182, + "grad_norm": 3.2029339679697264, + "learning_rate": 7.937911191945342e-07, + "loss": 0.2062, + "step": 1818 + }, + { + "epoch": 0.4134090909090909, + "grad_norm": 2.933062255176414, + "learning_rate": 7.933614090265089e-07, + "loss": 0.0643, + "step": 1819 + }, + { + "epoch": 0.41363636363636364, + "grad_norm": 3.6687283282914627, + "learning_rate": 7.929316130289158e-07, + "loss": 0.0764, + "step": 1820 + }, + { + "epoch": 0.4138636363636364, + "grad_norm": 2.5936666967129445, + "learning_rate": 7.925017314208619e-07, + "loss": 0.113, + "step": 1821 + }, + { + "epoch": 0.41409090909090907, + "grad_norm": 2.838047292902675, + "learning_rate": 7.920717644214986e-07, + "loss": 0.1111, + "step": 1822 + }, + { + "epoch": 0.4143181818181818, + "grad_norm": 3.9669566452318783, + "learning_rate": 7.916417122500199e-07, + "loss": 0.0912, + "step": 1823 + }, + { + "epoch": 0.41454545454545455, + "grad_norm": 3.4797273221521223, + "learning_rate": 7.912115751256637e-07, + "loss": 0.1237, + "step": 1824 + }, + { + "epoch": 0.4147727272727273, + "grad_norm": 3.7125738833611917, + "learning_rate": 7.907813532677111e-07, + "loss": 0.1108, + "step": 1825 + }, + { + "epoch": 0.415, + "grad_norm": 3.4850473206556267, + "learning_rate": 7.90351046895487e-07, + "loss": 0.1232, + "step": 1826 + }, + { + "epoch": 0.4152272727272727, + "grad_norm": 3.6658384002159448, + "learning_rate": 7.899206562283582e-07, + "loss": 0.1668, + "step": 1827 + }, + { + "epoch": 0.41545454545454547, + "grad_norm": 2.6246937982798912, + "learning_rate": 7.894901814857353e-07, + "loss": 0.1267, + "step": 1828 + }, + { + "epoch": 0.4156818181818182, + "grad_norm": 4.212699649434891, + "learning_rate": 7.890596228870716e-07, + "loss": 0.2207, + "step": 1829 + }, + { + "epoch": 0.4159090909090909, + "grad_norm": 1.4323772363715073, + "learning_rate": 7.88628980651863e-07, + "loss": 0.068, + "step": 1830 + }, + { + "epoch": 0.41613636363636364, + "grad_norm": 3.3215630004905616, + "learning_rate": 7.881982549996485e-07, + "loss": 0.2339, + "step": 1831 + }, + { + "epoch": 0.4163636363636364, + "grad_norm": 2.9294971967663654, + "learning_rate": 7.87767446150009e-07, + "loss": 0.0754, + "step": 1832 + }, + { + "epoch": 0.41659090909090907, + "grad_norm": 3.2184965201938875, + "learning_rate": 7.873365543225678e-07, + "loss": 0.107, + "step": 1833 + }, + { + "epoch": 0.4168181818181818, + "grad_norm": 4.9347956191912665, + "learning_rate": 7.869055797369915e-07, + "loss": 0.1186, + "step": 1834 + }, + { + "epoch": 0.41704545454545455, + "grad_norm": 4.005693852947852, + "learning_rate": 7.864745226129877e-07, + "loss": 0.1179, + "step": 1835 + }, + { + "epoch": 0.4172727272727273, + "grad_norm": 2.1764719452250234, + "learning_rate": 7.860433831703066e-07, + "loss": 0.0808, + "step": 1836 + }, + { + "epoch": 0.4175, + "grad_norm": 2.556737654050165, + "learning_rate": 7.856121616287403e-07, + "loss": 0.0992, + "step": 1837 + }, + { + "epoch": 0.4177272727272727, + "grad_norm": 3.9663325484823746, + "learning_rate": 7.85180858208123e-07, + "loss": 0.0695, + "step": 1838 + }, + { + "epoch": 0.41795454545454547, + "grad_norm": 2.9551558773000908, + "learning_rate": 7.847494731283302e-07, + "loss": 0.0899, + "step": 1839 + }, + { + "epoch": 0.41818181818181815, + "grad_norm": 2.612607061405107, + "learning_rate": 7.843180066092795e-07, + "loss": 0.0735, + "step": 1840 + }, + { + "epoch": 0.4184090909090909, + "grad_norm": 4.33069270106527, + "learning_rate": 7.838864588709296e-07, + "loss": 0.1584, + "step": 1841 + }, + { + "epoch": 0.41863636363636364, + "grad_norm": 1.8177938912895009, + "learning_rate": 7.834548301332805e-07, + "loss": 0.0545, + "step": 1842 + }, + { + "epoch": 0.4188636363636364, + "grad_norm": 2.1084416134002613, + "learning_rate": 7.830231206163742e-07, + "loss": 0.0639, + "step": 1843 + }, + { + "epoch": 0.41909090909090907, + "grad_norm": 2.093219849943601, + "learning_rate": 7.825913305402932e-07, + "loss": 0.0562, + "step": 1844 + }, + { + "epoch": 0.4193181818181818, + "grad_norm": 4.10196434724138, + "learning_rate": 7.821594601251617e-07, + "loss": 0.1107, + "step": 1845 + }, + { + "epoch": 0.41954545454545455, + "grad_norm": 3.2372838531632104, + "learning_rate": 7.817275095911439e-07, + "loss": 0.1251, + "step": 1846 + }, + { + "epoch": 0.4197727272727273, + "grad_norm": 1.8236948635920744, + "learning_rate": 7.812954791584458e-07, + "loss": 0.0579, + "step": 1847 + }, + { + "epoch": 0.42, + "grad_norm": 3.031682671002225, + "learning_rate": 7.808633690473136e-07, + "loss": 0.1406, + "step": 1848 + }, + { + "epoch": 0.4202272727272727, + "grad_norm": 2.5039280506359676, + "learning_rate": 7.804311794780345e-07, + "loss": 0.056, + "step": 1849 + }, + { + "epoch": 0.42045454545454547, + "grad_norm": 2.782046914562772, + "learning_rate": 7.799989106709354e-07, + "loss": 0.0862, + "step": 1850 + }, + { + "epoch": 0.42068181818181816, + "grad_norm": 2.5672422630177483, + "learning_rate": 7.795665628463847e-07, + "loss": 0.0929, + "step": 1851 + }, + { + "epoch": 0.4209090909090909, + "grad_norm": 3.6496039762810852, + "learning_rate": 7.791341362247906e-07, + "loss": 0.1045, + "step": 1852 + }, + { + "epoch": 0.42113636363636364, + "grad_norm": 2.6145820919999343, + "learning_rate": 7.78701631026601e-07, + "loss": 0.1034, + "step": 1853 + }, + { + "epoch": 0.4213636363636364, + "grad_norm": 2.8104676099789363, + "learning_rate": 7.782690474723045e-07, + "loss": 0.1352, + "step": 1854 + }, + { + "epoch": 0.42159090909090907, + "grad_norm": 1.622880070293311, + "learning_rate": 7.778363857824295e-07, + "loss": 0.0865, + "step": 1855 + }, + { + "epoch": 0.4218181818181818, + "grad_norm": 2.9918290778437626, + "learning_rate": 7.774036461775442e-07, + "loss": 0.1645, + "step": 1856 + }, + { + "epoch": 0.42204545454545456, + "grad_norm": 4.055245497181557, + "learning_rate": 7.769708288782562e-07, + "loss": 0.1459, + "step": 1857 + }, + { + "epoch": 0.4222727272727273, + "grad_norm": 2.235991604005269, + "learning_rate": 7.765379341052132e-07, + "loss": 0.0692, + "step": 1858 + }, + { + "epoch": 0.4225, + "grad_norm": 1.4089577228322532, + "learning_rate": 7.761049620791019e-07, + "loss": 0.0382, + "step": 1859 + }, + { + "epoch": 0.42272727272727273, + "grad_norm": 2.8038603094362586, + "learning_rate": 7.756719130206493e-07, + "loss": 0.1738, + "step": 1860 + }, + { + "epoch": 0.42295454545454547, + "grad_norm": 3.4307297219507134, + "learning_rate": 7.752387871506203e-07, + "loss": 0.1763, + "step": 1861 + }, + { + "epoch": 0.42318181818181816, + "grad_norm": 3.3709726961544138, + "learning_rate": 7.748055846898201e-07, + "loss": 0.0891, + "step": 1862 + }, + { + "epoch": 0.4234090909090909, + "grad_norm": 3.0893853002123874, + "learning_rate": 7.743723058590926e-07, + "loss": 0.142, + "step": 1863 + }, + { + "epoch": 0.42363636363636364, + "grad_norm": 2.363154536723354, + "learning_rate": 7.739389508793204e-07, + "loss": 0.1474, + "step": 1864 + }, + { + "epoch": 0.4238636363636364, + "grad_norm": 2.657549964912422, + "learning_rate": 7.73505519971425e-07, + "loss": 0.0832, + "step": 1865 + }, + { + "epoch": 0.4240909090909091, + "grad_norm": 3.3106374249507002, + "learning_rate": 7.730720133563671e-07, + "loss": 0.0895, + "step": 1866 + }, + { + "epoch": 0.4243181818181818, + "grad_norm": 4.05172229411086, + "learning_rate": 7.726384312551452e-07, + "loss": 0.2057, + "step": 1867 + }, + { + "epoch": 0.42454545454545456, + "grad_norm": 2.5587437466815666, + "learning_rate": 7.722047738887967e-07, + "loss": 0.0849, + "step": 1868 + }, + { + "epoch": 0.42477272727272725, + "grad_norm": 2.93788375602779, + "learning_rate": 7.717710414783974e-07, + "loss": 0.1484, + "step": 1869 + }, + { + "epoch": 0.425, + "grad_norm": 3.1142145365344716, + "learning_rate": 7.713372342450616e-07, + "loss": 0.1244, + "step": 1870 + }, + { + "epoch": 0.42522727272727273, + "grad_norm": 2.2415768227037507, + "learning_rate": 7.709033524099411e-07, + "loss": 0.0469, + "step": 1871 + }, + { + "epoch": 0.4254545454545455, + "grad_norm": 2.336644791080754, + "learning_rate": 7.704693961942262e-07, + "loss": 0.0694, + "step": 1872 + }, + { + "epoch": 0.42568181818181816, + "grad_norm": 2.6268152489734233, + "learning_rate": 7.700353658191448e-07, + "loss": 0.0635, + "step": 1873 + }, + { + "epoch": 0.4259090909090909, + "grad_norm": 2.1891616558842264, + "learning_rate": 7.696012615059631e-07, + "loss": 0.1118, + "step": 1874 + }, + { + "epoch": 0.42613636363636365, + "grad_norm": 3.452963675882406, + "learning_rate": 7.691670834759846e-07, + "loss": 0.1298, + "step": 1875 + }, + { + "epoch": 0.4263636363636364, + "grad_norm": 1.8381029387708185, + "learning_rate": 7.687328319505505e-07, + "loss": 0.0872, + "step": 1876 + }, + { + "epoch": 0.4265909090909091, + "grad_norm": 2.9416224405115803, + "learning_rate": 7.682985071510392e-07, + "loss": 0.1359, + "step": 1877 + }, + { + "epoch": 0.4268181818181818, + "grad_norm": 2.7423043759443297, + "learning_rate": 7.678641092988673e-07, + "loss": 0.1104, + "step": 1878 + }, + { + "epoch": 0.42704545454545456, + "grad_norm": 2.468658758714957, + "learning_rate": 7.674296386154873e-07, + "loss": 0.1303, + "step": 1879 + }, + { + "epoch": 0.42727272727272725, + "grad_norm": 4.378880066172616, + "learning_rate": 7.6699509532239e-07, + "loss": 0.1902, + "step": 1880 + }, + { + "epoch": 0.4275, + "grad_norm": 3.4647298324208866, + "learning_rate": 7.665604796411029e-07, + "loss": 0.0782, + "step": 1881 + }, + { + "epoch": 0.42772727272727273, + "grad_norm": 2.5768365705605816, + "learning_rate": 7.661257917931899e-07, + "loss": 0.0629, + "step": 1882 + }, + { + "epoch": 0.4279545454545455, + "grad_norm": 2.670192933932198, + "learning_rate": 7.656910320002522e-07, + "loss": 0.0772, + "step": 1883 + }, + { + "epoch": 0.42818181818181816, + "grad_norm": 1.8561115391904432, + "learning_rate": 7.65256200483928e-07, + "loss": 0.0553, + "step": 1884 + }, + { + "epoch": 0.4284090909090909, + "grad_norm": 2.7587384589336548, + "learning_rate": 7.64821297465891e-07, + "loss": 0.1404, + "step": 1885 + }, + { + "epoch": 0.42863636363636365, + "grad_norm": 3.0424455197386315, + "learning_rate": 7.643863231678523e-07, + "loss": 0.1531, + "step": 1886 + }, + { + "epoch": 0.4288636363636364, + "grad_norm": 3.16176965530844, + "learning_rate": 7.639512778115589e-07, + "loss": 0.0825, + "step": 1887 + }, + { + "epoch": 0.4290909090909091, + "grad_norm": 3.6599159401770263, + "learning_rate": 7.635161616187941e-07, + "loss": 0.0844, + "step": 1888 + }, + { + "epoch": 0.4293181818181818, + "grad_norm": 3.3389733742276055, + "learning_rate": 7.630809748113777e-07, + "loss": 0.1672, + "step": 1889 + }, + { + "epoch": 0.42954545454545456, + "grad_norm": 3.5708378608915528, + "learning_rate": 7.626457176111647e-07, + "loss": 0.1068, + "step": 1890 + }, + { + "epoch": 0.42977272727272725, + "grad_norm": 3.9818629958111798, + "learning_rate": 7.622103902400466e-07, + "loss": 0.0776, + "step": 1891 + }, + { + "epoch": 0.43, + "grad_norm": 3.709557486160077, + "learning_rate": 7.617749929199506e-07, + "loss": 0.1409, + "step": 1892 + }, + { + "epoch": 0.43022727272727274, + "grad_norm": 3.994134019852253, + "learning_rate": 7.613395258728392e-07, + "loss": 0.1927, + "step": 1893 + }, + { + "epoch": 0.4304545454545455, + "grad_norm": 4.478895141027428, + "learning_rate": 7.60903989320711e-07, + "loss": 0.1691, + "step": 1894 + }, + { + "epoch": 0.43068181818181817, + "grad_norm": 2.1703543794744555, + "learning_rate": 7.604683834855995e-07, + "loss": 0.1067, + "step": 1895 + }, + { + "epoch": 0.4309090909090909, + "grad_norm": 2.9998681302216235, + "learning_rate": 7.600327085895742e-07, + "loss": 0.0994, + "step": 1896 + }, + { + "epoch": 0.43113636363636365, + "grad_norm": 2.4766981179641863, + "learning_rate": 7.595969648547388e-07, + "loss": 0.1003, + "step": 1897 + }, + { + "epoch": 0.43136363636363634, + "grad_norm": 2.9766627549680362, + "learning_rate": 7.591611525032328e-07, + "loss": 0.1631, + "step": 1898 + }, + { + "epoch": 0.4315909090909091, + "grad_norm": 1.9872165270425672, + "learning_rate": 7.587252717572312e-07, + "loss": 0.1207, + "step": 1899 + }, + { + "epoch": 0.4318181818181818, + "grad_norm": 3.729284317113322, + "learning_rate": 7.582893228389423e-07, + "loss": 0.0966, + "step": 1900 + }, + { + "epoch": 0.43204545454545457, + "grad_norm": 3.0877578403235173, + "learning_rate": 7.578533059706104e-07, + "loss": 0.0805, + "step": 1901 + }, + { + "epoch": 0.43227272727272725, + "grad_norm": 3.076427475339341, + "learning_rate": 7.574172213745144e-07, + "loss": 0.058, + "step": 1902 + }, + { + "epoch": 0.4325, + "grad_norm": 2.993601617017678, + "learning_rate": 7.569810692729672e-07, + "loss": 0.1067, + "step": 1903 + }, + { + "epoch": 0.43272727272727274, + "grad_norm": 2.734134065139685, + "learning_rate": 7.56544849888316e-07, + "loss": 0.0552, + "step": 1904 + }, + { + "epoch": 0.4329545454545455, + "grad_norm": 2.56067065326671, + "learning_rate": 7.561085634429434e-07, + "loss": 0.0877, + "step": 1905 + }, + { + "epoch": 0.43318181818181817, + "grad_norm": 2.101829666683981, + "learning_rate": 7.556722101592648e-07, + "loss": 0.1543, + "step": 1906 + }, + { + "epoch": 0.4334090909090909, + "grad_norm": 3.363973342879091, + "learning_rate": 7.552357902597306e-07, + "loss": 0.0825, + "step": 1907 + }, + { + "epoch": 0.43363636363636365, + "grad_norm": 1.709057982119929, + "learning_rate": 7.547993039668248e-07, + "loss": 0.0837, + "step": 1908 + }, + { + "epoch": 0.43386363636363634, + "grad_norm": 1.5323404964261174, + "learning_rate": 7.543627515030652e-07, + "loss": 0.1325, + "step": 1909 + }, + { + "epoch": 0.4340909090909091, + "grad_norm": 3.586841530734463, + "learning_rate": 7.539261330910038e-07, + "loss": 0.0858, + "step": 1910 + }, + { + "epoch": 0.4343181818181818, + "grad_norm": 3.68513566553729, + "learning_rate": 7.534894489532255e-07, + "loss": 0.1969, + "step": 1911 + }, + { + "epoch": 0.43454545454545457, + "grad_norm": 4.450901419212549, + "learning_rate": 7.530526993123491e-07, + "loss": 0.2157, + "step": 1912 + }, + { + "epoch": 0.43477272727272726, + "grad_norm": 2.986979569231672, + "learning_rate": 7.526158843910271e-07, + "loss": 0.1324, + "step": 1913 + }, + { + "epoch": 0.435, + "grad_norm": 3.3564117474932247, + "learning_rate": 7.521790044119447e-07, + "loss": 0.1648, + "step": 1914 + }, + { + "epoch": 0.43522727272727274, + "grad_norm": 3.1060176770481056, + "learning_rate": 7.517420595978203e-07, + "loss": 0.094, + "step": 1915 + }, + { + "epoch": 0.4354545454545454, + "grad_norm": 2.2598571104537766, + "learning_rate": 7.513050501714061e-07, + "loss": 0.0677, + "step": 1916 + }, + { + "epoch": 0.43568181818181817, + "grad_norm": 2.825300903967002, + "learning_rate": 7.508679763554865e-07, + "loss": 0.1411, + "step": 1917 + }, + { + "epoch": 0.4359090909090909, + "grad_norm": 3.15523670717897, + "learning_rate": 7.504308383728785e-07, + "loss": 0.1399, + "step": 1918 + }, + { + "epoch": 0.43613636363636366, + "grad_norm": 1.419032076283524, + "learning_rate": 7.499936364464328e-07, + "loss": 0.036, + "step": 1919 + }, + { + "epoch": 0.43636363636363634, + "grad_norm": 2.273405520342511, + "learning_rate": 7.495563707990318e-07, + "loss": 0.0647, + "step": 1920 + }, + { + "epoch": 0.4365909090909091, + "grad_norm": 2.610381698884995, + "learning_rate": 7.491190416535909e-07, + "loss": 0.0717, + "step": 1921 + }, + { + "epoch": 0.4368181818181818, + "grad_norm": 1.8569445763191812, + "learning_rate": 7.486816492330579e-07, + "loss": 0.0834, + "step": 1922 + }, + { + "epoch": 0.43704545454545457, + "grad_norm": 3.3172591928800563, + "learning_rate": 7.482441937604121e-07, + "loss": 0.1318, + "step": 1923 + }, + { + "epoch": 0.43727272727272726, + "grad_norm": 3.1728141793853952, + "learning_rate": 7.478066754586659e-07, + "loss": 0.1019, + "step": 1924 + }, + { + "epoch": 0.4375, + "grad_norm": 3.8574629502011573, + "learning_rate": 7.473690945508636e-07, + "loss": 0.2033, + "step": 1925 + }, + { + "epoch": 0.43772727272727274, + "grad_norm": 3.3040285633676265, + "learning_rate": 7.469314512600804e-07, + "loss": 0.113, + "step": 1926 + }, + { + "epoch": 0.43795454545454543, + "grad_norm": 2.525550014406599, + "learning_rate": 7.464937458094243e-07, + "loss": 0.0766, + "step": 1927 + }, + { + "epoch": 0.4381818181818182, + "grad_norm": 3.8575642266429235, + "learning_rate": 7.460559784220351e-07, + "loss": 0.1496, + "step": 1928 + }, + { + "epoch": 0.4384090909090909, + "grad_norm": 3.439873280092042, + "learning_rate": 7.456181493210836e-07, + "loss": 0.0652, + "step": 1929 + }, + { + "epoch": 0.43863636363636366, + "grad_norm": 7.377569730509127, + "learning_rate": 7.451802587297721e-07, + "loss": 0.2322, + "step": 1930 + }, + { + "epoch": 0.43886363636363634, + "grad_norm": 4.033767386298218, + "learning_rate": 7.447423068713347e-07, + "loss": 0.0904, + "step": 1931 + }, + { + "epoch": 0.4390909090909091, + "grad_norm": 2.197960137204769, + "learning_rate": 7.443042939690362e-07, + "loss": 0.0822, + "step": 1932 + }, + { + "epoch": 0.43931818181818183, + "grad_norm": 2.098570449988918, + "learning_rate": 7.438662202461728e-07, + "loss": 0.0693, + "step": 1933 + }, + { + "epoch": 0.4395454545454546, + "grad_norm": 1.4188862736504249, + "learning_rate": 7.43428085926072e-07, + "loss": 0.0471, + "step": 1934 + }, + { + "epoch": 0.43977272727272726, + "grad_norm": 4.274836208635331, + "learning_rate": 7.429898912320916e-07, + "loss": 0.1575, + "step": 1935 + }, + { + "epoch": 0.44, + "grad_norm": 3.7379201450139803, + "learning_rate": 7.425516363876204e-07, + "loss": 0.1019, + "step": 1936 + }, + { + "epoch": 0.44022727272727274, + "grad_norm": 4.410805522643766, + "learning_rate": 7.421133216160781e-07, + "loss": 0.1568, + "step": 1937 + }, + { + "epoch": 0.44045454545454543, + "grad_norm": 4.235697608919579, + "learning_rate": 7.416749471409144e-07, + "loss": 0.1933, + "step": 1938 + }, + { + "epoch": 0.4406818181818182, + "grad_norm": 3.430071584769345, + "learning_rate": 7.412365131856103e-07, + "loss": 0.155, + "step": 1939 + }, + { + "epoch": 0.4409090909090909, + "grad_norm": 4.4849097921584296, + "learning_rate": 7.407980199736762e-07, + "loss": 0.0957, + "step": 1940 + }, + { + "epoch": 0.44113636363636366, + "grad_norm": 4.195950267323048, + "learning_rate": 7.403594677286534e-07, + "loss": 0.0808, + "step": 1941 + }, + { + "epoch": 0.44136363636363635, + "grad_norm": 2.291407757184764, + "learning_rate": 7.39920856674113e-07, + "loss": 0.0469, + "step": 1942 + }, + { + "epoch": 0.4415909090909091, + "grad_norm": 2.2317646034425254, + "learning_rate": 7.394821870336561e-07, + "loss": 0.1096, + "step": 1943 + }, + { + "epoch": 0.44181818181818183, + "grad_norm": 4.88504809913934, + "learning_rate": 7.390434590309135e-07, + "loss": 0.1796, + "step": 1944 + }, + { + "epoch": 0.4420454545454545, + "grad_norm": 4.242609115735363, + "learning_rate": 7.386046728895459e-07, + "loss": 0.1354, + "step": 1945 + }, + { + "epoch": 0.44227272727272726, + "grad_norm": 5.584342137032149, + "learning_rate": 7.38165828833244e-07, + "loss": 0.1318, + "step": 1946 + }, + { + "epoch": 0.4425, + "grad_norm": 3.957940405624805, + "learning_rate": 7.377269270857274e-07, + "loss": 0.144, + "step": 1947 + }, + { + "epoch": 0.44272727272727275, + "grad_norm": 1.773451563395401, + "learning_rate": 7.372879678707455e-07, + "loss": 0.0657, + "step": 1948 + }, + { + "epoch": 0.44295454545454543, + "grad_norm": 5.1875285963004005, + "learning_rate": 7.368489514120768e-07, + "loss": 0.1317, + "step": 1949 + }, + { + "epoch": 0.4431818181818182, + "grad_norm": 2.708960945168458, + "learning_rate": 7.364098779335294e-07, + "loss": 0.0816, + "step": 1950 + }, + { + "epoch": 0.4434090909090909, + "grad_norm": 4.074020388603923, + "learning_rate": 7.359707476589398e-07, + "loss": 0.1325, + "step": 1951 + }, + { + "epoch": 0.44363636363636366, + "grad_norm": 3.4211860951660586, + "learning_rate": 7.355315608121736e-07, + "loss": 0.1879, + "step": 1952 + }, + { + "epoch": 0.44386363636363635, + "grad_norm": 3.9821661186237267, + "learning_rate": 7.35092317617126e-07, + "loss": 0.0994, + "step": 1953 + }, + { + "epoch": 0.4440909090909091, + "grad_norm": 3.757989455395455, + "learning_rate": 7.346530182977203e-07, + "loss": 0.1578, + "step": 1954 + }, + { + "epoch": 0.44431818181818183, + "grad_norm": 4.716518363899776, + "learning_rate": 7.342136630779082e-07, + "loss": 0.1879, + "step": 1955 + }, + { + "epoch": 0.4445454545454545, + "grad_norm": 2.7019068749970336, + "learning_rate": 7.3377425218167e-07, + "loss": 0.0963, + "step": 1956 + }, + { + "epoch": 0.44477272727272726, + "grad_norm": 4.352905981712314, + "learning_rate": 7.333347858330153e-07, + "loss": 0.1111, + "step": 1957 + }, + { + "epoch": 0.445, + "grad_norm": 3.5185202925324384, + "learning_rate": 7.328952642559804e-07, + "loss": 0.1481, + "step": 1958 + }, + { + "epoch": 0.44522727272727275, + "grad_norm": 3.3892949842669085, + "learning_rate": 7.32455687674631e-07, + "loss": 0.1107, + "step": 1959 + }, + { + "epoch": 0.44545454545454544, + "grad_norm": 3.5343419604157207, + "learning_rate": 7.320160563130605e-07, + "loss": 0.103, + "step": 1960 + }, + { + "epoch": 0.4456818181818182, + "grad_norm": 4.7725844595742695, + "learning_rate": 7.3157637039539e-07, + "loss": 0.1475, + "step": 1961 + }, + { + "epoch": 0.4459090909090909, + "grad_norm": 2.7503505248271205, + "learning_rate": 7.311366301457684e-07, + "loss": 0.0852, + "step": 1962 + }, + { + "epoch": 0.4461363636363636, + "grad_norm": 3.354865658777263, + "learning_rate": 7.306968357883729e-07, + "loss": 0.1738, + "step": 1963 + }, + { + "epoch": 0.44636363636363635, + "grad_norm": 2.995177306328317, + "learning_rate": 7.302569875474076e-07, + "loss": 0.1651, + "step": 1964 + }, + { + "epoch": 0.4465909090909091, + "grad_norm": 1.9765597184536288, + "learning_rate": 7.298170856471042e-07, + "loss": 0.0915, + "step": 1965 + }, + { + "epoch": 0.44681818181818184, + "grad_norm": 2.2437988113240577, + "learning_rate": 7.29377130311722e-07, + "loss": 0.0749, + "step": 1966 + }, + { + "epoch": 0.4470454545454545, + "grad_norm": 4.478868039330648, + "learning_rate": 7.289371217655473e-07, + "loss": 0.0984, + "step": 1967 + }, + { + "epoch": 0.44727272727272727, + "grad_norm": 2.880835860253152, + "learning_rate": 7.28497060232894e-07, + "loss": 0.0929, + "step": 1968 + }, + { + "epoch": 0.4475, + "grad_norm": 2.425948563998928, + "learning_rate": 7.280569459381022e-07, + "loss": 0.1205, + "step": 1969 + }, + { + "epoch": 0.44772727272727275, + "grad_norm": 2.3423747325627016, + "learning_rate": 7.276167791055394e-07, + "loss": 0.1095, + "step": 1970 + }, + { + "epoch": 0.44795454545454544, + "grad_norm": 3.3072575594309988, + "learning_rate": 7.271765599596001e-07, + "loss": 0.1704, + "step": 1971 + }, + { + "epoch": 0.4481818181818182, + "grad_norm": 2.2165940851963772, + "learning_rate": 7.267362887247052e-07, + "loss": 0.0709, + "step": 1972 + }, + { + "epoch": 0.4484090909090909, + "grad_norm": 2.3590088339466937, + "learning_rate": 7.262959656253019e-07, + "loss": 0.0966, + "step": 1973 + }, + { + "epoch": 0.4486363636363636, + "grad_norm": 2.071652265598975, + "learning_rate": 7.258555908858641e-07, + "loss": 0.1109, + "step": 1974 + }, + { + "epoch": 0.44886363636363635, + "grad_norm": 3.3567538403390307, + "learning_rate": 7.254151647308924e-07, + "loss": 0.0863, + "step": 1975 + }, + { + "epoch": 0.4490909090909091, + "grad_norm": 3.5489063695179803, + "learning_rate": 7.249746873849127e-07, + "loss": 0.1166, + "step": 1976 + }, + { + "epoch": 0.44931818181818184, + "grad_norm": 5.231426091255909, + "learning_rate": 7.245341590724779e-07, + "loss": 0.2066, + "step": 1977 + }, + { + "epoch": 0.4495454545454545, + "grad_norm": 1.8929381314355, + "learning_rate": 7.240935800181664e-07, + "loss": 0.0808, + "step": 1978 + }, + { + "epoch": 0.44977272727272727, + "grad_norm": 2.7502672581889827, + "learning_rate": 7.236529504465828e-07, + "loss": 0.1196, + "step": 1979 + }, + { + "epoch": 0.45, + "grad_norm": 3.514608565685813, + "learning_rate": 7.232122705823567e-07, + "loss": 0.1368, + "step": 1980 + }, + { + "epoch": 0.45022727272727275, + "grad_norm": 4.889371736954598, + "learning_rate": 7.227715406501444e-07, + "loss": 0.2511, + "step": 1981 + }, + { + "epoch": 0.45045454545454544, + "grad_norm": 2.3776616529438632, + "learning_rate": 7.223307608746269e-07, + "loss": 0.1312, + "step": 1982 + }, + { + "epoch": 0.4506818181818182, + "grad_norm": 3.8886303072660753, + "learning_rate": 7.21889931480511e-07, + "loss": 0.228, + "step": 1983 + }, + { + "epoch": 0.4509090909090909, + "grad_norm": 2.458454804741036, + "learning_rate": 7.214490526925289e-07, + "loss": 0.0647, + "step": 1984 + }, + { + "epoch": 0.4511363636363636, + "grad_norm": 2.093943914949851, + "learning_rate": 7.210081247354372e-07, + "loss": 0.0825, + "step": 1985 + }, + { + "epoch": 0.45136363636363636, + "grad_norm": 1.4138151734932016, + "learning_rate": 7.205671478340189e-07, + "loss": 0.0753, + "step": 1986 + }, + { + "epoch": 0.4515909090909091, + "grad_norm": 3.0472798461106283, + "learning_rate": 7.201261222130807e-07, + "loss": 0.146, + "step": 1987 + }, + { + "epoch": 0.45181818181818184, + "grad_norm": 4.2442548345864735, + "learning_rate": 7.196850480974546e-07, + "loss": 0.1561, + "step": 1988 + }, + { + "epoch": 0.45204545454545453, + "grad_norm": 4.137333645987156, + "learning_rate": 7.192439257119976e-07, + "loss": 0.0955, + "step": 1989 + }, + { + "epoch": 0.45227272727272727, + "grad_norm": 4.146736421143346, + "learning_rate": 7.188027552815912e-07, + "loss": 0.1796, + "step": 1990 + }, + { + "epoch": 0.4525, + "grad_norm": 3.6165792620041315, + "learning_rate": 7.183615370311407e-07, + "loss": 0.2019, + "step": 1991 + }, + { + "epoch": 0.4527272727272727, + "grad_norm": 3.216537347596979, + "learning_rate": 7.179202711855771e-07, + "loss": 0.1543, + "step": 1992 + }, + { + "epoch": 0.45295454545454544, + "grad_norm": 3.5978623035450887, + "learning_rate": 7.174789579698542e-07, + "loss": 0.0928, + "step": 1993 + }, + { + "epoch": 0.4531818181818182, + "grad_norm": 4.691142345266573, + "learning_rate": 7.170375976089509e-07, + "loss": 0.0849, + "step": 1994 + }, + { + "epoch": 0.45340909090909093, + "grad_norm": 2.150284603620602, + "learning_rate": 7.1659619032787e-07, + "loss": 0.0848, + "step": 1995 + }, + { + "epoch": 0.4536363636363636, + "grad_norm": 5.998076110034539, + "learning_rate": 7.16154736351638e-07, + "loss": 0.1219, + "step": 1996 + }, + { + "epoch": 0.45386363636363636, + "grad_norm": 3.392655818942902, + "learning_rate": 7.157132359053056e-07, + "loss": 0.1256, + "step": 1997 + }, + { + "epoch": 0.4540909090909091, + "grad_norm": 3.094923624616438, + "learning_rate": 7.152716892139464e-07, + "loss": 0.1081, + "step": 1998 + }, + { + "epoch": 0.45431818181818184, + "grad_norm": 2.2856797140974052, + "learning_rate": 7.148300965026582e-07, + "loss": 0.1108, + "step": 1999 + }, + { + "epoch": 0.45454545454545453, + "grad_norm": 2.984254710691812, + "learning_rate": 7.143884579965624e-07, + "loss": 0.1703, + "step": 2000 + }, + { + "epoch": 0.4547727272727273, + "grad_norm": 1.9675564765905496, + "learning_rate": 7.139467739208034e-07, + "loss": 0.0757, + "step": 2001 + }, + { + "epoch": 0.455, + "grad_norm": 3.1556497397696797, + "learning_rate": 7.135050445005487e-07, + "loss": 0.154, + "step": 2002 + }, + { + "epoch": 0.4552272727272727, + "grad_norm": 3.5087030817493248, + "learning_rate": 7.130632699609892e-07, + "loss": 0.1179, + "step": 2003 + }, + { + "epoch": 0.45545454545454545, + "grad_norm": 3.231793947568177, + "learning_rate": 7.12621450527339e-07, + "loss": 0.1859, + "step": 2004 + }, + { + "epoch": 0.4556818181818182, + "grad_norm": 2.7045530114128433, + "learning_rate": 7.121795864248345e-07, + "loss": 0.1125, + "step": 2005 + }, + { + "epoch": 0.45590909090909093, + "grad_norm": 4.01844635531317, + "learning_rate": 7.117376778787353e-07, + "loss": 0.1365, + "step": 2006 + }, + { + "epoch": 0.4561363636363636, + "grad_norm": 2.571006039855259, + "learning_rate": 7.112957251143236e-07, + "loss": 0.1467, + "step": 2007 + }, + { + "epoch": 0.45636363636363636, + "grad_norm": 4.115391477893308, + "learning_rate": 7.10853728356904e-07, + "loss": 0.1525, + "step": 2008 + }, + { + "epoch": 0.4565909090909091, + "grad_norm": 3.005755320855063, + "learning_rate": 7.104116878318037e-07, + "loss": 0.0722, + "step": 2009 + }, + { + "epoch": 0.45681818181818185, + "grad_norm": 3.0532159184814103, + "learning_rate": 7.09969603764372e-07, + "loss": 0.1085, + "step": 2010 + }, + { + "epoch": 0.45704545454545453, + "grad_norm": 3.9908091018262515, + "learning_rate": 7.095274763799805e-07, + "loss": 0.1069, + "step": 2011 + }, + { + "epoch": 0.4572727272727273, + "grad_norm": 2.4503139530869333, + "learning_rate": 7.090853059040234e-07, + "loss": 0.0819, + "step": 2012 + }, + { + "epoch": 0.4575, + "grad_norm": 2.44189944584534, + "learning_rate": 7.086430925619157e-07, + "loss": 0.1101, + "step": 2013 + }, + { + "epoch": 0.4577272727272727, + "grad_norm": 2.8310859676480753, + "learning_rate": 7.082008365790953e-07, + "loss": 0.0591, + "step": 2014 + }, + { + "epoch": 0.45795454545454545, + "grad_norm": 2.38760072771867, + "learning_rate": 7.077585381810213e-07, + "loss": 0.137, + "step": 2015 + }, + { + "epoch": 0.4581818181818182, + "grad_norm": 2.9432431041977867, + "learning_rate": 7.073161975931748e-07, + "loss": 0.1753, + "step": 2016 + }, + { + "epoch": 0.45840909090909093, + "grad_norm": 2.303940840857091, + "learning_rate": 7.068738150410582e-07, + "loss": 0.0796, + "step": 2017 + }, + { + "epoch": 0.4586363636363636, + "grad_norm": 2.479456100946905, + "learning_rate": 7.064313907501949e-07, + "loss": 0.1104, + "step": 2018 + }, + { + "epoch": 0.45886363636363636, + "grad_norm": 2.814976025494183, + "learning_rate": 7.059889249461303e-07, + "loss": 0.0922, + "step": 2019 + }, + { + "epoch": 0.4590909090909091, + "grad_norm": 2.9161153995939704, + "learning_rate": 7.055464178544306e-07, + "loss": 0.1256, + "step": 2020 + }, + { + "epoch": 0.4593181818181818, + "grad_norm": 6.287695672925511, + "learning_rate": 7.051038697006832e-07, + "loss": 0.1283, + "step": 2021 + }, + { + "epoch": 0.45954545454545453, + "grad_norm": 4.655677891513784, + "learning_rate": 7.04661280710496e-07, + "loss": 0.126, + "step": 2022 + }, + { + "epoch": 0.4597727272727273, + "grad_norm": 2.6661141951304406, + "learning_rate": 7.042186511094981e-07, + "loss": 0.132, + "step": 2023 + }, + { + "epoch": 0.46, + "grad_norm": 1.4433328147006652, + "learning_rate": 7.037759811233395e-07, + "loss": 0.0504, + "step": 2024 + }, + { + "epoch": 0.4602272727272727, + "grad_norm": 1.229401041424278, + "learning_rate": 7.033332709776902e-07, + "loss": 0.0687, + "step": 2025 + }, + { + "epoch": 0.46045454545454545, + "grad_norm": 6.152030377991928, + "learning_rate": 7.028905208982413e-07, + "loss": 0.1633, + "step": 2026 + }, + { + "epoch": 0.4606818181818182, + "grad_norm": 4.416516366536607, + "learning_rate": 7.024477311107038e-07, + "loss": 0.1203, + "step": 2027 + }, + { + "epoch": 0.46090909090909093, + "grad_norm": 2.632944674575189, + "learning_rate": 7.02004901840809e-07, + "loss": 0.1116, + "step": 2028 + }, + { + "epoch": 0.4611363636363636, + "grad_norm": 3.4769215343528965, + "learning_rate": 7.015620333143085e-07, + "loss": 0.1514, + "step": 2029 + }, + { + "epoch": 0.46136363636363636, + "grad_norm": 2.6547948736046947, + "learning_rate": 7.011191257569739e-07, + "loss": 0.108, + "step": 2030 + }, + { + "epoch": 0.4615909090909091, + "grad_norm": 2.5189486740005718, + "learning_rate": 7.006761793945968e-07, + "loss": 0.0874, + "step": 2031 + }, + { + "epoch": 0.4618181818181818, + "grad_norm": 2.89312740526475, + "learning_rate": 7.00233194452988e-07, + "loss": 0.1147, + "step": 2032 + }, + { + "epoch": 0.46204545454545454, + "grad_norm": 2.892842014163498, + "learning_rate": 6.997901711579787e-07, + "loss": 0.0929, + "step": 2033 + }, + { + "epoch": 0.4622727272727273, + "grad_norm": 2.3974393136317014, + "learning_rate": 6.993471097354193e-07, + "loss": 0.1083, + "step": 2034 + }, + { + "epoch": 0.4625, + "grad_norm": 3.9351651333307114, + "learning_rate": 6.989040104111794e-07, + "loss": 0.0845, + "step": 2035 + }, + { + "epoch": 0.4627272727272727, + "grad_norm": 2.8296539069184377, + "learning_rate": 6.984608734111486e-07, + "loss": 0.1192, + "step": 2036 + }, + { + "epoch": 0.46295454545454545, + "grad_norm": 1.7984842321255332, + "learning_rate": 6.980176989612352e-07, + "loss": 0.0747, + "step": 2037 + }, + { + "epoch": 0.4631818181818182, + "grad_norm": 3.3171716642017017, + "learning_rate": 6.975744872873663e-07, + "loss": 0.1127, + "step": 2038 + }, + { + "epoch": 0.4634090909090909, + "grad_norm": 2.7197649659123777, + "learning_rate": 6.97131238615489e-07, + "loss": 0.1127, + "step": 2039 + }, + { + "epoch": 0.4636363636363636, + "grad_norm": 2.5014590833836374, + "learning_rate": 6.966879531715681e-07, + "loss": 0.0813, + "step": 2040 + }, + { + "epoch": 0.46386363636363637, + "grad_norm": 2.8731632250608876, + "learning_rate": 6.962446311815879e-07, + "loss": 0.0756, + "step": 2041 + }, + { + "epoch": 0.4640909090909091, + "grad_norm": 1.6955575856590688, + "learning_rate": 6.958012728715513e-07, + "loss": 0.0882, + "step": 2042 + }, + { + "epoch": 0.4643181818181818, + "grad_norm": 3.8397572558582604, + "learning_rate": 6.953578784674791e-07, + "loss": 0.1222, + "step": 2043 + }, + { + "epoch": 0.46454545454545454, + "grad_norm": 4.733512700417062, + "learning_rate": 6.949144481954114e-07, + "loss": 0.1846, + "step": 2044 + }, + { + "epoch": 0.4647727272727273, + "grad_norm": 2.397104637403568, + "learning_rate": 6.944709822814058e-07, + "loss": 0.0988, + "step": 2045 + }, + { + "epoch": 0.465, + "grad_norm": 5.665827509491553, + "learning_rate": 6.940274809515384e-07, + "loss": 0.1529, + "step": 2046 + }, + { + "epoch": 0.4652272727272727, + "grad_norm": 5.978093840063041, + "learning_rate": 6.935839444319033e-07, + "loss": 0.1434, + "step": 2047 + }, + { + "epoch": 0.46545454545454545, + "grad_norm": 4.54860810193348, + "learning_rate": 6.931403729486128e-07, + "loss": 0.1217, + "step": 2048 + }, + { + "epoch": 0.4656818181818182, + "grad_norm": 1.9143195463384695, + "learning_rate": 6.926967667277967e-07, + "loss": 0.0873, + "step": 2049 + }, + { + "epoch": 0.4659090909090909, + "grad_norm": 2.4728576950770838, + "learning_rate": 6.922531259956026e-07, + "loss": 0.0901, + "step": 2050 + }, + { + "epoch": 0.4661363636363636, + "grad_norm": 8.400275735859179, + "learning_rate": 6.918094509781957e-07, + "loss": 0.1534, + "step": 2051 + }, + { + "epoch": 0.46636363636363637, + "grad_norm": 3.2022878142861404, + "learning_rate": 6.913657419017585e-07, + "loss": 0.0872, + "step": 2052 + }, + { + "epoch": 0.4665909090909091, + "grad_norm": 2.9973081321749846, + "learning_rate": 6.909219989924912e-07, + "loss": 0.1615, + "step": 2053 + }, + { + "epoch": 0.4668181818181818, + "grad_norm": 5.484364404038541, + "learning_rate": 6.904782224766112e-07, + "loss": 0.1326, + "step": 2054 + }, + { + "epoch": 0.46704545454545454, + "grad_norm": 2.412488839744118, + "learning_rate": 6.900344125803529e-07, + "loss": 0.1254, + "step": 2055 + }, + { + "epoch": 0.4672727272727273, + "grad_norm": 3.8015078172452426, + "learning_rate": 6.895905695299676e-07, + "loss": 0.1671, + "step": 2056 + }, + { + "epoch": 0.4675, + "grad_norm": 4.807035271477219, + "learning_rate": 6.891466935517237e-07, + "loss": 0.1311, + "step": 2057 + }, + { + "epoch": 0.4677272727272727, + "grad_norm": 6.173904480251301, + "learning_rate": 6.887027848719063e-07, + "loss": 0.2713, + "step": 2058 + }, + { + "epoch": 0.46795454545454546, + "grad_norm": 5.538899931576732, + "learning_rate": 6.882588437168173e-07, + "loss": 0.3015, + "step": 2059 + }, + { + "epoch": 0.4681818181818182, + "grad_norm": 4.141065642957912, + "learning_rate": 6.878148703127751e-07, + "loss": 0.2077, + "step": 2060 + }, + { + "epoch": 0.4684090909090909, + "grad_norm": 3.444959741766656, + "learning_rate": 6.873708648861143e-07, + "loss": 0.1077, + "step": 2061 + }, + { + "epoch": 0.46863636363636363, + "grad_norm": 5.594141106033254, + "learning_rate": 6.869268276631862e-07, + "loss": 0.1602, + "step": 2062 + }, + { + "epoch": 0.46886363636363637, + "grad_norm": 2.9704984801461483, + "learning_rate": 6.86482758870358e-07, + "loss": 0.1661, + "step": 2063 + }, + { + "epoch": 0.4690909090909091, + "grad_norm": 3.8865798197052945, + "learning_rate": 6.86038658734013e-07, + "loss": 0.1933, + "step": 2064 + }, + { + "epoch": 0.4693181818181818, + "grad_norm": 2.7264873523770032, + "learning_rate": 6.85594527480551e-07, + "loss": 0.0974, + "step": 2065 + }, + { + "epoch": 0.46954545454545454, + "grad_norm": 3.36693135232514, + "learning_rate": 6.851503653363871e-07, + "loss": 0.146, + "step": 2066 + }, + { + "epoch": 0.4697727272727273, + "grad_norm": 4.496389978116932, + "learning_rate": 6.847061725279521e-07, + "loss": 0.1276, + "step": 2067 + }, + { + "epoch": 0.47, + "grad_norm": 3.9018175447809367, + "learning_rate": 6.84261949281693e-07, + "loss": 0.1196, + "step": 2068 + }, + { + "epoch": 0.4702272727272727, + "grad_norm": 3.14915637068765, + "learning_rate": 6.838176958240716e-07, + "loss": 0.1279, + "step": 2069 + }, + { + "epoch": 0.47045454545454546, + "grad_norm": 1.690019322335705, + "learning_rate": 6.833734123815657e-07, + "loss": 0.0524, + "step": 2070 + }, + { + "epoch": 0.4706818181818182, + "grad_norm": 2.0966045091280012, + "learning_rate": 6.829290991806681e-07, + "loss": 0.0984, + "step": 2071 + }, + { + "epoch": 0.4709090909090909, + "grad_norm": 5.12658315515599, + "learning_rate": 6.824847564478868e-07, + "loss": 0.198, + "step": 2072 + }, + { + "epoch": 0.47113636363636363, + "grad_norm": 2.3806970232050375, + "learning_rate": 6.820403844097449e-07, + "loss": 0.128, + "step": 2073 + }, + { + "epoch": 0.4713636363636364, + "grad_norm": 2.403450494790023, + "learning_rate": 6.815959832927805e-07, + "loss": 0.0558, + "step": 2074 + }, + { + "epoch": 0.4715909090909091, + "grad_norm": 1.880219787772445, + "learning_rate": 6.811515533235463e-07, + "loss": 0.0843, + "step": 2075 + }, + { + "epoch": 0.4718181818181818, + "grad_norm": 2.852530739209598, + "learning_rate": 6.8070709472861e-07, + "loss": 0.0717, + "step": 2076 + }, + { + "epoch": 0.47204545454545455, + "grad_norm": 3.1458457831259445, + "learning_rate": 6.802626077345538e-07, + "loss": 0.1189, + "step": 2077 + }, + { + "epoch": 0.4722727272727273, + "grad_norm": 3.1821512826645995, + "learning_rate": 6.798180925679738e-07, + "loss": 0.081, + "step": 2078 + }, + { + "epoch": 0.4725, + "grad_norm": 4.424054830702027, + "learning_rate": 6.793735494554815e-07, + "loss": 0.1509, + "step": 2079 + }, + { + "epoch": 0.4727272727272727, + "grad_norm": 5.33178031469734, + "learning_rate": 6.789289786237023e-07, + "loss": 0.2108, + "step": 2080 + }, + { + "epoch": 0.47295454545454546, + "grad_norm": 3.213597353536753, + "learning_rate": 6.78484380299275e-07, + "loss": 0.1381, + "step": 2081 + }, + { + "epoch": 0.4731818181818182, + "grad_norm": 2.6825283049918043, + "learning_rate": 6.780397547088531e-07, + "loss": 0.2204, + "step": 2082 + }, + { + "epoch": 0.4734090909090909, + "grad_norm": 2.1558982301712595, + "learning_rate": 6.775951020791043e-07, + "loss": 0.104, + "step": 2083 + }, + { + "epoch": 0.47363636363636363, + "grad_norm": 2.8352012091456005, + "learning_rate": 6.771504226367093e-07, + "loss": 0.1287, + "step": 2084 + }, + { + "epoch": 0.4738636363636364, + "grad_norm": 4.570322714197513, + "learning_rate": 6.767057166083628e-07, + "loss": 0.1331, + "step": 2085 + }, + { + "epoch": 0.4740909090909091, + "grad_norm": 2.994910039601439, + "learning_rate": 6.762609842207733e-07, + "loss": 0.1988, + "step": 2086 + }, + { + "epoch": 0.4743181818181818, + "grad_norm": 2.7136001737552213, + "learning_rate": 6.758162257006623e-07, + "loss": 0.0835, + "step": 2087 + }, + { + "epoch": 0.47454545454545455, + "grad_norm": 3.189183047622881, + "learning_rate": 6.753714412747652e-07, + "loss": 0.2402, + "step": 2088 + }, + { + "epoch": 0.4747727272727273, + "grad_norm": 4.091897595734396, + "learning_rate": 6.749266311698298e-07, + "loss": 0.0894, + "step": 2089 + }, + { + "epoch": 0.475, + "grad_norm": 3.6366083331661034, + "learning_rate": 6.744817956126176e-07, + "loss": 0.0898, + "step": 2090 + }, + { + "epoch": 0.4752272727272727, + "grad_norm": 2.585550501688436, + "learning_rate": 6.740369348299033e-07, + "loss": 0.1367, + "step": 2091 + }, + { + "epoch": 0.47545454545454546, + "grad_norm": 2.0315193163671528, + "learning_rate": 6.735920490484734e-07, + "loss": 0.0781, + "step": 2092 + }, + { + "epoch": 0.4756818181818182, + "grad_norm": 3.1338049298887607, + "learning_rate": 6.731471384951283e-07, + "loss": 0.1484, + "step": 2093 + }, + { + "epoch": 0.4759090909090909, + "grad_norm": 2.75201891219274, + "learning_rate": 6.727022033966804e-07, + "loss": 0.0942, + "step": 2094 + }, + { + "epoch": 0.47613636363636364, + "grad_norm": 2.689110106875561, + "learning_rate": 6.722572439799549e-07, + "loss": 0.1317, + "step": 2095 + }, + { + "epoch": 0.4763636363636364, + "grad_norm": 2.8210313253357393, + "learning_rate": 6.718122604717891e-07, + "loss": 0.0988, + "step": 2096 + }, + { + "epoch": 0.47659090909090907, + "grad_norm": 5.110850963296442, + "learning_rate": 6.713672530990327e-07, + "loss": 0.1987, + "step": 2097 + }, + { + "epoch": 0.4768181818181818, + "grad_norm": 2.8527554283455285, + "learning_rate": 6.709222220885478e-07, + "loss": 0.0673, + "step": 2098 + }, + { + "epoch": 0.47704545454545455, + "grad_norm": 1.8159114370553175, + "learning_rate": 6.704771676672081e-07, + "loss": 0.0997, + "step": 2099 + }, + { + "epoch": 0.4772727272727273, + "grad_norm": 5.185954043935928, + "learning_rate": 6.700320900618997e-07, + "loss": 0.1892, + "step": 2100 + }, + { + "epoch": 0.4775, + "grad_norm": 2.5244676010883764, + "learning_rate": 6.695869894995202e-07, + "loss": 0.0682, + "step": 2101 + }, + { + "epoch": 0.4777272727272727, + "grad_norm": 3.129029808784798, + "learning_rate": 6.691418662069793e-07, + "loss": 0.0719, + "step": 2102 + }, + { + "epoch": 0.47795454545454547, + "grad_norm": 2.7873914529785497, + "learning_rate": 6.686967204111976e-07, + "loss": 0.1241, + "step": 2103 + }, + { + "epoch": 0.4781818181818182, + "grad_norm": 4.85721153848579, + "learning_rate": 6.682515523391078e-07, + "loss": 0.1438, + "step": 2104 + }, + { + "epoch": 0.4784090909090909, + "grad_norm": 3.2889349177629916, + "learning_rate": 6.678063622176536e-07, + "loss": 0.1004, + "step": 2105 + }, + { + "epoch": 0.47863636363636364, + "grad_norm": 1.888608101302673, + "learning_rate": 6.673611502737902e-07, + "loss": 0.0772, + "step": 2106 + }, + { + "epoch": 0.4788636363636364, + "grad_norm": 3.548256402789989, + "learning_rate": 6.669159167344839e-07, + "loss": 0.1076, + "step": 2107 + }, + { + "epoch": 0.47909090909090907, + "grad_norm": 2.7815325733829255, + "learning_rate": 6.664706618267115e-07, + "loss": 0.1213, + "step": 2108 + }, + { + "epoch": 0.4793181818181818, + "grad_norm": 2.8903105128351037, + "learning_rate": 6.660253857774616e-07, + "loss": 0.1484, + "step": 2109 + }, + { + "epoch": 0.47954545454545455, + "grad_norm": 3.38835063204251, + "learning_rate": 6.655800888137327e-07, + "loss": 0.163, + "step": 2110 + }, + { + "epoch": 0.4797727272727273, + "grad_norm": 3.982010426886424, + "learning_rate": 6.651347711625344e-07, + "loss": 0.1041, + "step": 2111 + }, + { + "epoch": 0.48, + "grad_norm": 2.1915966303777275, + "learning_rate": 6.646894330508872e-07, + "loss": 0.0597, + "step": 2112 + }, + { + "epoch": 0.4802272727272727, + "grad_norm": 3.0042905564921987, + "learning_rate": 6.642440747058209e-07, + "loss": 0.1227, + "step": 2113 + }, + { + "epoch": 0.48045454545454547, + "grad_norm": 1.9780043694789033, + "learning_rate": 6.637986963543769e-07, + "loss": 0.0725, + "step": 2114 + }, + { + "epoch": 0.48068181818181815, + "grad_norm": 2.818155223024504, + "learning_rate": 6.63353298223606e-07, + "loss": 0.1051, + "step": 2115 + }, + { + "epoch": 0.4809090909090909, + "grad_norm": 2.12801464160873, + "learning_rate": 6.629078805405695e-07, + "loss": 0.0695, + "step": 2116 + }, + { + "epoch": 0.48113636363636364, + "grad_norm": 2.6311663416075897, + "learning_rate": 6.62462443532338e-07, + "loss": 0.0778, + "step": 2117 + }, + { + "epoch": 0.4813636363636364, + "grad_norm": 3.5204010632219904, + "learning_rate": 6.62016987425993e-07, + "loss": 0.2315, + "step": 2118 + }, + { + "epoch": 0.48159090909090907, + "grad_norm": 2.502490528096051, + "learning_rate": 6.615715124486248e-07, + "loss": 0.082, + "step": 2119 + }, + { + "epoch": 0.4818181818181818, + "grad_norm": 3.560798978025333, + "learning_rate": 6.61126018827334e-07, + "loss": 0.204, + "step": 2120 + }, + { + "epoch": 0.48204545454545455, + "grad_norm": 1.5461111751336367, + "learning_rate": 6.606805067892301e-07, + "loss": 0.0632, + "step": 2121 + }, + { + "epoch": 0.4822727272727273, + "grad_norm": 3.0933981160810844, + "learning_rate": 6.602349765614325e-07, + "loss": 0.1179, + "step": 2122 + }, + { + "epoch": 0.4825, + "grad_norm": 1.6434368640027537, + "learning_rate": 6.597894283710692e-07, + "loss": 0.0746, + "step": 2123 + }, + { + "epoch": 0.4827272727272727, + "grad_norm": 2.3586663668164416, + "learning_rate": 6.593438624452787e-07, + "loss": 0.0976, + "step": 2124 + }, + { + "epoch": 0.48295454545454547, + "grad_norm": 3.0032483137520125, + "learning_rate": 6.588982790112068e-07, + "loss": 0.0922, + "step": 2125 + }, + { + "epoch": 0.48318181818181816, + "grad_norm": 3.197307969532937, + "learning_rate": 6.584526782960095e-07, + "loss": 0.138, + "step": 2126 + }, + { + "epoch": 0.4834090909090909, + "grad_norm": 1.674658424639185, + "learning_rate": 6.580070605268513e-07, + "loss": 0.0844, + "step": 2127 + }, + { + "epoch": 0.48363636363636364, + "grad_norm": 2.4126707544257058, + "learning_rate": 6.575614259309048e-07, + "loss": 0.1434, + "step": 2128 + }, + { + "epoch": 0.4838636363636364, + "grad_norm": 2.57355469816842, + "learning_rate": 6.571157747353519e-07, + "loss": 0.1425, + "step": 2129 + }, + { + "epoch": 0.48409090909090907, + "grad_norm": 3.0936373425219035, + "learning_rate": 6.56670107167383e-07, + "loss": 0.0724, + "step": 2130 + }, + { + "epoch": 0.4843181818181818, + "grad_norm": 3.4872894113374886, + "learning_rate": 6.562244234541964e-07, + "loss": 0.1157, + "step": 2131 + }, + { + "epoch": 0.48454545454545456, + "grad_norm": 3.779370883704926, + "learning_rate": 6.557787238229984e-07, + "loss": 0.1266, + "step": 2132 + }, + { + "epoch": 0.4847727272727273, + "grad_norm": 4.510151055954135, + "learning_rate": 6.553330085010044e-07, + "loss": 0.1057, + "step": 2133 + }, + { + "epoch": 0.485, + "grad_norm": 2.9442471313811476, + "learning_rate": 6.548872777154369e-07, + "loss": 0.0891, + "step": 2134 + }, + { + "epoch": 0.48522727272727273, + "grad_norm": 3.4283099089195654, + "learning_rate": 6.544415316935267e-07, + "loss": 0.0746, + "step": 2135 + }, + { + "epoch": 0.48545454545454547, + "grad_norm": 2.85784889507271, + "learning_rate": 6.539957706625123e-07, + "loss": 0.1327, + "step": 2136 + }, + { + "epoch": 0.48568181818181816, + "grad_norm": 1.9628106642263428, + "learning_rate": 6.535499948496396e-07, + "loss": 0.066, + "step": 2137 + }, + { + "epoch": 0.4859090909090909, + "grad_norm": 1.5954807343833117, + "learning_rate": 6.531042044821625e-07, + "loss": 0.1006, + "step": 2138 + }, + { + "epoch": 0.48613636363636364, + "grad_norm": 4.295424201870693, + "learning_rate": 6.526583997873422e-07, + "loss": 0.1619, + "step": 2139 + }, + { + "epoch": 0.4863636363636364, + "grad_norm": 3.2710995045263798, + "learning_rate": 6.522125809924465e-07, + "loss": 0.081, + "step": 2140 + }, + { + "epoch": 0.4865909090909091, + "grad_norm": 3.90603732129082, + "learning_rate": 6.517667483247517e-07, + "loss": 0.2034, + "step": 2141 + }, + { + "epoch": 0.4868181818181818, + "grad_norm": 3.2086688884724306, + "learning_rate": 6.513209020115399e-07, + "loss": 0.0734, + "step": 2142 + }, + { + "epoch": 0.48704545454545456, + "grad_norm": 2.9591075872034316, + "learning_rate": 6.508750422801006e-07, + "loss": 0.1012, + "step": 2143 + }, + { + "epoch": 0.48727272727272725, + "grad_norm": 2.3850384618654403, + "learning_rate": 6.504291693577309e-07, + "loss": 0.0982, + "step": 2144 + }, + { + "epoch": 0.4875, + "grad_norm": 2.624339392151581, + "learning_rate": 6.499832834717335e-07, + "loss": 0.0981, + "step": 2145 + }, + { + "epoch": 0.48772727272727273, + "grad_norm": 5.470813324740111, + "learning_rate": 6.495373848494179e-07, + "loss": 0.2063, + "step": 2146 + }, + { + "epoch": 0.4879545454545455, + "grad_norm": 3.1393762735744914, + "learning_rate": 6.49091473718101e-07, + "loss": 0.1565, + "step": 2147 + }, + { + "epoch": 0.48818181818181816, + "grad_norm": 3.491967358759376, + "learning_rate": 6.48645550305105e-07, + "loss": 0.1502, + "step": 2148 + }, + { + "epoch": 0.4884090909090909, + "grad_norm": 2.1174130631647463, + "learning_rate": 6.48199614837759e-07, + "loss": 0.1348, + "step": 2149 + }, + { + "epoch": 0.48863636363636365, + "grad_norm": 3.918636116072268, + "learning_rate": 6.477536675433979e-07, + "loss": 0.1669, + "step": 2150 + }, + { + "epoch": 0.4888636363636364, + "grad_norm": 2.753398480259231, + "learning_rate": 6.473077086493629e-07, + "loss": 0.094, + "step": 2151 + }, + { + "epoch": 0.4890909090909091, + "grad_norm": 3.682621748188102, + "learning_rate": 6.468617383830007e-07, + "loss": 0.1432, + "step": 2152 + }, + { + "epoch": 0.4893181818181818, + "grad_norm": 2.4285929901095606, + "learning_rate": 6.464157569716648e-07, + "loss": 0.0839, + "step": 2153 + }, + { + "epoch": 0.48954545454545456, + "grad_norm": 2.8971248300668915, + "learning_rate": 6.459697646427128e-07, + "loss": 0.1205, + "step": 2154 + }, + { + "epoch": 0.48977272727272725, + "grad_norm": 2.0095575654951627, + "learning_rate": 6.455237616235091e-07, + "loss": 0.0675, + "step": 2155 + }, + { + "epoch": 0.49, + "grad_norm": 2.7624648801554272, + "learning_rate": 6.450777481414236e-07, + "loss": 0.1065, + "step": 2156 + }, + { + "epoch": 0.49022727272727273, + "grad_norm": 3.9742768403801394, + "learning_rate": 6.446317244238302e-07, + "loss": 0.2142, + "step": 2157 + }, + { + "epoch": 0.4904545454545455, + "grad_norm": 3.61652561503901, + "learning_rate": 6.441856906981096e-07, + "loss": 0.1195, + "step": 2158 + }, + { + "epoch": 0.49068181818181816, + "grad_norm": 2.843838283481522, + "learning_rate": 6.437396471916469e-07, + "loss": 0.0703, + "step": 2159 + }, + { + "epoch": 0.4909090909090909, + "grad_norm": 2.2815427488316744, + "learning_rate": 6.432935941318322e-07, + "loss": 0.0963, + "step": 2160 + }, + { + "epoch": 0.49113636363636365, + "grad_norm": 2.883361160981839, + "learning_rate": 6.428475317460601e-07, + "loss": 0.0697, + "step": 2161 + }, + { + "epoch": 0.4913636363636364, + "grad_norm": 3.8048261759039073, + "learning_rate": 6.424014602617309e-07, + "loss": 0.1041, + "step": 2162 + }, + { + "epoch": 0.4915909090909091, + "grad_norm": 3.317390583201811, + "learning_rate": 6.419553799062486e-07, + "loss": 0.2336, + "step": 2163 + }, + { + "epoch": 0.4918181818181818, + "grad_norm": 2.2219974551810644, + "learning_rate": 6.415092909070224e-07, + "loss": 0.1576, + "step": 2164 + }, + { + "epoch": 0.49204545454545456, + "grad_norm": 2.920305460546785, + "learning_rate": 6.410631934914653e-07, + "loss": 0.1361, + "step": 2165 + }, + { + "epoch": 0.49227272727272725, + "grad_norm": 5.047135229009562, + "learning_rate": 6.406170878869951e-07, + "loss": 0.1704, + "step": 2166 + }, + { + "epoch": 0.4925, + "grad_norm": 3.6963168127589237, + "learning_rate": 6.401709743210335e-07, + "loss": 0.0861, + "step": 2167 + }, + { + "epoch": 0.49272727272727274, + "grad_norm": 3.5830480901678987, + "learning_rate": 6.397248530210065e-07, + "loss": 0.1247, + "step": 2168 + }, + { + "epoch": 0.4929545454545455, + "grad_norm": 2.0898496952675956, + "learning_rate": 6.392787242143437e-07, + "loss": 0.115, + "step": 2169 + }, + { + "epoch": 0.49318181818181817, + "grad_norm": 2.5146973059842472, + "learning_rate": 6.388325881284785e-07, + "loss": 0.0784, + "step": 2170 + }, + { + "epoch": 0.4934090909090909, + "grad_norm": 5.929838512624532, + "learning_rate": 6.383864449908489e-07, + "loss": 0.1448, + "step": 2171 + }, + { + "epoch": 0.49363636363636365, + "grad_norm": 2.8200677767383473, + "learning_rate": 6.379402950288952e-07, + "loss": 0.0914, + "step": 2172 + }, + { + "epoch": 0.49386363636363634, + "grad_norm": 3.0955883916855695, + "learning_rate": 6.374941384700618e-07, + "loss": 0.1825, + "step": 2173 + }, + { + "epoch": 0.4940909090909091, + "grad_norm": 5.8987664978651075, + "learning_rate": 6.37047975541797e-07, + "loss": 0.1622, + "step": 2174 + }, + { + "epoch": 0.4943181818181818, + "grad_norm": 2.847009679375902, + "learning_rate": 6.366018064715512e-07, + "loss": 0.1023, + "step": 2175 + }, + { + "epoch": 0.49454545454545457, + "grad_norm": 3.5296100459347826, + "learning_rate": 6.361556314867789e-07, + "loss": 0.1279, + "step": 2176 + }, + { + "epoch": 0.49477272727272725, + "grad_norm": 2.434618814363439, + "learning_rate": 6.35709450814937e-07, + "loss": 0.0839, + "step": 2177 + }, + { + "epoch": 0.495, + "grad_norm": 2.44826116964108, + "learning_rate": 6.352632646834858e-07, + "loss": 0.1555, + "step": 2178 + }, + { + "epoch": 0.49522727272727274, + "grad_norm": 2.8771449237470907, + "learning_rate": 6.34817073319888e-07, + "loss": 0.1591, + "step": 2179 + }, + { + "epoch": 0.4954545454545455, + "grad_norm": 6.233442844594177, + "learning_rate": 6.34370876951609e-07, + "loss": 0.2542, + "step": 2180 + }, + { + "epoch": 0.49568181818181817, + "grad_norm": 3.2787670131648112, + "learning_rate": 6.339246758061171e-07, + "loss": 0.1972, + "step": 2181 + }, + { + "epoch": 0.4959090909090909, + "grad_norm": 3.0628637815405297, + "learning_rate": 6.334784701108826e-07, + "loss": 0.1527, + "step": 2182 + }, + { + "epoch": 0.49613636363636365, + "grad_norm": 2.607514205568518, + "learning_rate": 6.330322600933784e-07, + "loss": 0.0764, + "step": 2183 + }, + { + "epoch": 0.49636363636363634, + "grad_norm": 3.024586398591082, + "learning_rate": 6.325860459810793e-07, + "loss": 0.1768, + "step": 2184 + }, + { + "epoch": 0.4965909090909091, + "grad_norm": 2.242524300666679, + "learning_rate": 6.321398280014627e-07, + "loss": 0.0991, + "step": 2185 + }, + { + "epoch": 0.4968181818181818, + "grad_norm": 3.7236629231065472, + "learning_rate": 6.316936063820075e-07, + "loss": 0.0928, + "step": 2186 + }, + { + "epoch": 0.49704545454545457, + "grad_norm": 2.760123860013934, + "learning_rate": 6.312473813501946e-07, + "loss": 0.1111, + "step": 2187 + }, + { + "epoch": 0.49727272727272726, + "grad_norm": 2.684208180835582, + "learning_rate": 6.308011531335068e-07, + "loss": 0.0652, + "step": 2188 + }, + { + "epoch": 0.4975, + "grad_norm": 3.7967085808456686, + "learning_rate": 6.303549219594282e-07, + "loss": 0.1401, + "step": 2189 + }, + { + "epoch": 0.49772727272727274, + "grad_norm": 3.0557858743235635, + "learning_rate": 6.299086880554446e-07, + "loss": 0.1031, + "step": 2190 + }, + { + "epoch": 0.4979545454545454, + "grad_norm": 3.5792016910891857, + "learning_rate": 6.294624516490435e-07, + "loss": 0.1504, + "step": 2191 + }, + { + "epoch": 0.49818181818181817, + "grad_norm": 3.516819899020428, + "learning_rate": 6.290162129677131e-07, + "loss": 0.1057, + "step": 2192 + }, + { + "epoch": 0.4984090909090909, + "grad_norm": 4.883357104403011, + "learning_rate": 6.285699722389428e-07, + "loss": 0.1722, + "step": 2193 + }, + { + "epoch": 0.49863636363636366, + "grad_norm": 3.1931306240438624, + "learning_rate": 6.281237296902237e-07, + "loss": 0.1194, + "step": 2194 + }, + { + "epoch": 0.49886363636363634, + "grad_norm": 3.766049537754116, + "learning_rate": 6.276774855490472e-07, + "loss": 0.0977, + "step": 2195 + }, + { + "epoch": 0.4990909090909091, + "grad_norm": 0.9209399183939365, + "learning_rate": 6.272312400429057e-07, + "loss": 0.0522, + "step": 2196 + }, + { + "epoch": 0.4993181818181818, + "grad_norm": 4.119749660457485, + "learning_rate": 6.267849933992926e-07, + "loss": 0.1928, + "step": 2197 + }, + { + "epoch": 0.49954545454545457, + "grad_norm": 2.2236637091015767, + "learning_rate": 6.263387458457009e-07, + "loss": 0.062, + "step": 2198 + }, + { + "epoch": 0.49977272727272726, + "grad_norm": 2.000225030815468, + "learning_rate": 6.258924976096252e-07, + "loss": 0.1034, + "step": 2199 + }, + { + "epoch": 0.5, + "grad_norm": 2.6275094025983554, + "learning_rate": 6.254462489185601e-07, + "loss": 0.1727, + "step": 2200 + }, + { + "epoch": 0.5002272727272727, + "grad_norm": 2.235467648922331, + "learning_rate": 6.25e-07, + "loss": 0.1334, + "step": 2201 + }, + { + "epoch": 0.5004545454545455, + "grad_norm": 3.7322215204997957, + "learning_rate": 6.2455375108144e-07, + "loss": 0.2151, + "step": 2202 + }, + { + "epoch": 0.5006818181818182, + "grad_norm": 4.105782275033072, + "learning_rate": 6.241075023903749e-07, + "loss": 0.0927, + "step": 2203 + }, + { + "epoch": 0.5009090909090909, + "grad_norm": 3.1912910575192, + "learning_rate": 6.236612541542992e-07, + "loss": 0.1335, + "step": 2204 + }, + { + "epoch": 0.5011363636363636, + "grad_norm": 3.4455882850976303, + "learning_rate": 6.232150066007079e-07, + "loss": 0.1109, + "step": 2205 + }, + { + "epoch": 0.5013636363636363, + "grad_norm": 2.8629719948675616, + "learning_rate": 6.227687599570946e-07, + "loss": 0.0984, + "step": 2206 + }, + { + "epoch": 0.5015909090909091, + "grad_norm": 2.3593548828544226, + "learning_rate": 6.223225144509529e-07, + "loss": 0.0463, + "step": 2207 + }, + { + "epoch": 0.5018181818181818, + "grad_norm": 4.482656649392185, + "learning_rate": 6.218762703097764e-07, + "loss": 0.1156, + "step": 2208 + }, + { + "epoch": 0.5020454545454546, + "grad_norm": 1.7809176626391987, + "learning_rate": 6.214300277610574e-07, + "loss": 0.0503, + "step": 2209 + }, + { + "epoch": 0.5022727272727273, + "grad_norm": 3.2999156272381316, + "learning_rate": 6.209837870322871e-07, + "loss": 0.1748, + "step": 2210 + }, + { + "epoch": 0.5025, + "grad_norm": 2.557444436309367, + "learning_rate": 6.205375483509566e-07, + "loss": 0.1332, + "step": 2211 + }, + { + "epoch": 0.5027272727272727, + "grad_norm": 3.922349961140257, + "learning_rate": 6.200913119445557e-07, + "loss": 0.1068, + "step": 2212 + }, + { + "epoch": 0.5029545454545454, + "grad_norm": 2.9992848701328634, + "learning_rate": 6.196450780405721e-07, + "loss": 0.0832, + "step": 2213 + }, + { + "epoch": 0.5031818181818182, + "grad_norm": 3.83699466865425, + "learning_rate": 6.191988468664934e-07, + "loss": 0.1881, + "step": 2214 + }, + { + "epoch": 0.5034090909090909, + "grad_norm": 2.668054225877761, + "learning_rate": 6.187526186498055e-07, + "loss": 0.1455, + "step": 2215 + }, + { + "epoch": 0.5036363636363637, + "grad_norm": 4.080653773834415, + "learning_rate": 6.183063936179927e-07, + "loss": 0.097, + "step": 2216 + }, + { + "epoch": 0.5038636363636364, + "grad_norm": 2.6478244293429105, + "learning_rate": 6.178601719985375e-07, + "loss": 0.0603, + "step": 2217 + }, + { + "epoch": 0.5040909090909091, + "grad_norm": 3.460402571255276, + "learning_rate": 6.174139540189207e-07, + "loss": 0.1833, + "step": 2218 + }, + { + "epoch": 0.5043181818181818, + "grad_norm": 2.6065677219153387, + "learning_rate": 6.169677399066217e-07, + "loss": 0.1381, + "step": 2219 + }, + { + "epoch": 0.5045454545454545, + "grad_norm": 2.7175147293916395, + "learning_rate": 6.165215298891176e-07, + "loss": 0.0746, + "step": 2220 + }, + { + "epoch": 0.5047727272727273, + "grad_norm": 2.4147035308698594, + "learning_rate": 6.160753241938832e-07, + "loss": 0.1808, + "step": 2221 + }, + { + "epoch": 0.505, + "grad_norm": 3.2743451379328774, + "learning_rate": 6.156291230483911e-07, + "loss": 0.095, + "step": 2222 + }, + { + "epoch": 0.5052272727272727, + "grad_norm": 2.488338337269088, + "learning_rate": 6.151829266801122e-07, + "loss": 0.1246, + "step": 2223 + }, + { + "epoch": 0.5054545454545455, + "grad_norm": 3.9222062752624214, + "learning_rate": 6.147367353165143e-07, + "loss": 0.0927, + "step": 2224 + }, + { + "epoch": 0.5056818181818182, + "grad_norm": 2.157641280870416, + "learning_rate": 6.14290549185063e-07, + "loss": 0.0626, + "step": 2225 + }, + { + "epoch": 0.5059090909090909, + "grad_norm": 3.9707925542560076, + "learning_rate": 6.138443685132212e-07, + "loss": 0.1014, + "step": 2226 + }, + { + "epoch": 0.5061363636363636, + "grad_norm": 2.81366366711272, + "learning_rate": 6.133981935284491e-07, + "loss": 0.0811, + "step": 2227 + }, + { + "epoch": 0.5063636363636363, + "grad_norm": 2.40351949170894, + "learning_rate": 6.129520244582033e-07, + "loss": 0.0965, + "step": 2228 + }, + { + "epoch": 0.5065909090909091, + "grad_norm": 2.5591674513776295, + "learning_rate": 6.125058615299384e-07, + "loss": 0.1096, + "step": 2229 + }, + { + "epoch": 0.5068181818181818, + "grad_norm": 2.321717623954591, + "learning_rate": 6.12059704971105e-07, + "loss": 0.0594, + "step": 2230 + }, + { + "epoch": 0.5070454545454546, + "grad_norm": 1.5205694640370642, + "learning_rate": 6.116135550091513e-07, + "loss": 0.0778, + "step": 2231 + }, + { + "epoch": 0.5072727272727273, + "grad_norm": 2.1860412304358987, + "learning_rate": 6.111674118715215e-07, + "loss": 0.1104, + "step": 2232 + }, + { + "epoch": 0.5075, + "grad_norm": 2.4348364633903077, + "learning_rate": 6.107212757856564e-07, + "loss": 0.0984, + "step": 2233 + }, + { + "epoch": 0.5077272727272727, + "grad_norm": 4.584314756949201, + "learning_rate": 6.102751469789937e-07, + "loss": 0.2036, + "step": 2234 + }, + { + "epoch": 0.5079545454545454, + "grad_norm": 2.7851342464620963, + "learning_rate": 6.098290256789666e-07, + "loss": 0.1083, + "step": 2235 + }, + { + "epoch": 0.5081818181818182, + "grad_norm": 2.402481070187496, + "learning_rate": 6.093829121130052e-07, + "loss": 0.0982, + "step": 2236 + }, + { + "epoch": 0.5084090909090909, + "grad_norm": 4.036858795442769, + "learning_rate": 6.089368065085349e-07, + "loss": 0.1198, + "step": 2237 + }, + { + "epoch": 0.5086363636363637, + "grad_norm": 13.101284242282366, + "learning_rate": 6.084907090929778e-07, + "loss": 0.155, + "step": 2238 + }, + { + "epoch": 0.5088636363636364, + "grad_norm": 3.0025650713926626, + "learning_rate": 6.080446200937514e-07, + "loss": 0.0838, + "step": 2239 + }, + { + "epoch": 0.509090909090909, + "grad_norm": 3.6073760750354875, + "learning_rate": 6.075985397382691e-07, + "loss": 0.124, + "step": 2240 + }, + { + "epoch": 0.5093181818181818, + "grad_norm": 3.025570838064598, + "learning_rate": 6.0715246825394e-07, + "loss": 0.0786, + "step": 2241 + }, + { + "epoch": 0.5095454545454545, + "grad_norm": 3.1481493933795592, + "learning_rate": 6.067064058681681e-07, + "loss": 0.1543, + "step": 2242 + }, + { + "epoch": 0.5097727272727273, + "grad_norm": 3.243091807444454, + "learning_rate": 6.062603528083532e-07, + "loss": 0.064, + "step": 2243 + }, + { + "epoch": 0.51, + "grad_norm": 3.5505456581849373, + "learning_rate": 6.058143093018905e-07, + "loss": 0.0878, + "step": 2244 + }, + { + "epoch": 0.5102272727272728, + "grad_norm": 3.8100637682602163, + "learning_rate": 6.053682755761699e-07, + "loss": 0.0984, + "step": 2245 + }, + { + "epoch": 0.5104545454545455, + "grad_norm": 3.2250526313012937, + "learning_rate": 6.049222518585767e-07, + "loss": 0.1836, + "step": 2246 + }, + { + "epoch": 0.5106818181818182, + "grad_norm": 3.466204057035892, + "learning_rate": 6.044762383764909e-07, + "loss": 0.1407, + "step": 2247 + }, + { + "epoch": 0.5109090909090909, + "grad_norm": 2.3720959099601906, + "learning_rate": 6.040302353572872e-07, + "loss": 0.0744, + "step": 2248 + }, + { + "epoch": 0.5111363636363636, + "grad_norm": 6.233904075539728, + "learning_rate": 6.035842430283355e-07, + "loss": 0.1723, + "step": 2249 + }, + { + "epoch": 0.5113636363636364, + "grad_norm": 3.5392478067544024, + "learning_rate": 6.031382616169994e-07, + "loss": 0.2615, + "step": 2250 + }, + { + "epoch": 0.5115909090909091, + "grad_norm": 1.95605491924954, + "learning_rate": 6.026922913506373e-07, + "loss": 0.0704, + "step": 2251 + }, + { + "epoch": 0.5118181818181818, + "grad_norm": 4.280754029554871, + "learning_rate": 6.022463324566022e-07, + "loss": 0.2278, + "step": 2252 + }, + { + "epoch": 0.5120454545454546, + "grad_norm": 3.1590677883210647, + "learning_rate": 6.018003851622411e-07, + "loss": 0.1125, + "step": 2253 + }, + { + "epoch": 0.5122727272727273, + "grad_norm": 2.756946088766311, + "learning_rate": 6.01354449694895e-07, + "loss": 0.101, + "step": 2254 + }, + { + "epoch": 0.5125, + "grad_norm": 3.937338110701692, + "learning_rate": 6.009085262818992e-07, + "loss": 0.0843, + "step": 2255 + }, + { + "epoch": 0.5127272727272727, + "grad_norm": 2.3806656114761435, + "learning_rate": 6.004626151505824e-07, + "loss": 0.0689, + "step": 2256 + }, + { + "epoch": 0.5129545454545454, + "grad_norm": 5.172442753106651, + "learning_rate": 6.000167165282669e-07, + "loss": 0.1253, + "step": 2257 + }, + { + "epoch": 0.5131818181818182, + "grad_norm": 2.274382136551756, + "learning_rate": 5.995708306422693e-07, + "loss": 0.0917, + "step": 2258 + }, + { + "epoch": 0.5134090909090909, + "grad_norm": 4.439726544815456, + "learning_rate": 5.991249577198995e-07, + "loss": 0.1241, + "step": 2259 + }, + { + "epoch": 0.5136363636363637, + "grad_norm": 2.357638131776029, + "learning_rate": 5.986790979884603e-07, + "loss": 0.0872, + "step": 2260 + }, + { + "epoch": 0.5138636363636364, + "grad_norm": 2.6465154920772385, + "learning_rate": 5.982332516752486e-07, + "loss": 0.067, + "step": 2261 + }, + { + "epoch": 0.514090909090909, + "grad_norm": 4.875325385896511, + "learning_rate": 5.977874190075534e-07, + "loss": 0.1214, + "step": 2262 + }, + { + "epoch": 0.5143181818181818, + "grad_norm": 1.543301315849031, + "learning_rate": 5.973416002126581e-07, + "loss": 0.0541, + "step": 2263 + }, + { + "epoch": 0.5145454545454545, + "grad_norm": 3.898179151524051, + "learning_rate": 5.968957955178376e-07, + "loss": 0.0958, + "step": 2264 + }, + { + "epoch": 0.5147727272727273, + "grad_norm": 3.2960389674008344, + "learning_rate": 5.964500051503605e-07, + "loss": 0.1181, + "step": 2265 + }, + { + "epoch": 0.515, + "grad_norm": 3.3933061135749805, + "learning_rate": 5.960042293374879e-07, + "loss": 0.1793, + "step": 2266 + }, + { + "epoch": 0.5152272727272728, + "grad_norm": 4.277231003610205, + "learning_rate": 5.955584683064734e-07, + "loss": 0.1731, + "step": 2267 + }, + { + "epoch": 0.5154545454545455, + "grad_norm": 1.9552431043601877, + "learning_rate": 5.951127222845632e-07, + "loss": 0.0539, + "step": 2268 + }, + { + "epoch": 0.5156818181818181, + "grad_norm": 3.0586506475549546, + "learning_rate": 5.946669914989956e-07, + "loss": 0.1554, + "step": 2269 + }, + { + "epoch": 0.5159090909090909, + "grad_norm": 2.9403953531053824, + "learning_rate": 5.942212761770018e-07, + "loss": 0.0734, + "step": 2270 + }, + { + "epoch": 0.5161363636363636, + "grad_norm": 3.5960696869459037, + "learning_rate": 5.93775576545804e-07, + "loss": 0.0947, + "step": 2271 + }, + { + "epoch": 0.5163636363636364, + "grad_norm": 3.23692537698556, + "learning_rate": 5.933298928326171e-07, + "loss": 0.111, + "step": 2272 + }, + { + "epoch": 0.5165909090909091, + "grad_norm": 1.9659743288958036, + "learning_rate": 5.928842252646482e-07, + "loss": 0.0837, + "step": 2273 + }, + { + "epoch": 0.5168181818181818, + "grad_norm": 3.7675594961148433, + "learning_rate": 5.924385740690953e-07, + "loss": 0.1045, + "step": 2274 + }, + { + "epoch": 0.5170454545454546, + "grad_norm": 2.3010538818050503, + "learning_rate": 5.919929394731488e-07, + "loss": 0.0514, + "step": 2275 + }, + { + "epoch": 0.5172727272727272, + "grad_norm": 2.8425887206993696, + "learning_rate": 5.915473217039906e-07, + "loss": 0.1528, + "step": 2276 + }, + { + "epoch": 0.5175, + "grad_norm": 5.1069841311056665, + "learning_rate": 5.911017209887935e-07, + "loss": 0.1396, + "step": 2277 + }, + { + "epoch": 0.5177272727272727, + "grad_norm": 2.5505311227590672, + "learning_rate": 5.906561375547216e-07, + "loss": 0.092, + "step": 2278 + }, + { + "epoch": 0.5179545454545454, + "grad_norm": 5.212775290875958, + "learning_rate": 5.902105716289309e-07, + "loss": 0.1891, + "step": 2279 + }, + { + "epoch": 0.5181818181818182, + "grad_norm": 4.8590006207454435, + "learning_rate": 5.897650234385678e-07, + "loss": 0.216, + "step": 2280 + }, + { + "epoch": 0.5184090909090909, + "grad_norm": 3.8684069301252206, + "learning_rate": 5.8931949321077e-07, + "loss": 0.0948, + "step": 2281 + }, + { + "epoch": 0.5186363636363637, + "grad_norm": 3.7210565574189185, + "learning_rate": 5.888739811726662e-07, + "loss": 0.2134, + "step": 2282 + }, + { + "epoch": 0.5188636363636364, + "grad_norm": 3.2549386448690476, + "learning_rate": 5.884284875513752e-07, + "loss": 0.0785, + "step": 2283 + }, + { + "epoch": 0.519090909090909, + "grad_norm": 4.302815977210092, + "learning_rate": 5.879830125740073e-07, + "loss": 0.1282, + "step": 2284 + }, + { + "epoch": 0.5193181818181818, + "grad_norm": 4.083538261653371, + "learning_rate": 5.875375564676623e-07, + "loss": 0.1552, + "step": 2285 + }, + { + "epoch": 0.5195454545454545, + "grad_norm": 2.785107121448421, + "learning_rate": 5.870921194594309e-07, + "loss": 0.1473, + "step": 2286 + }, + { + "epoch": 0.5197727272727273, + "grad_norm": 2.303635625028287, + "learning_rate": 5.866467017763941e-07, + "loss": 0.095, + "step": 2287 + }, + { + "epoch": 0.52, + "grad_norm": 1.958857622068759, + "learning_rate": 5.862013036456233e-07, + "loss": 0.0661, + "step": 2288 + }, + { + "epoch": 0.5202272727272728, + "grad_norm": 2.922397419799541, + "learning_rate": 5.857559252941792e-07, + "loss": 0.0974, + "step": 2289 + }, + { + "epoch": 0.5204545454545455, + "grad_norm": 2.914748104974442, + "learning_rate": 5.85310566949113e-07, + "loss": 0.1145, + "step": 2290 + }, + { + "epoch": 0.5206818181818181, + "grad_norm": 3.567369019097331, + "learning_rate": 5.848652288374655e-07, + "loss": 0.0877, + "step": 2291 + }, + { + "epoch": 0.5209090909090909, + "grad_norm": 2.1646549694671053, + "learning_rate": 5.844199111862675e-07, + "loss": 0.0641, + "step": 2292 + }, + { + "epoch": 0.5211363636363636, + "grad_norm": 3.1709300922529744, + "learning_rate": 5.839746142225386e-07, + "loss": 0.1003, + "step": 2293 + }, + { + "epoch": 0.5213636363636364, + "grad_norm": 3.0660387552436092, + "learning_rate": 5.835293381732887e-07, + "loss": 0.0961, + "step": 2294 + }, + { + "epoch": 0.5215909090909091, + "grad_norm": 3.3080373262232685, + "learning_rate": 5.830840832655163e-07, + "loss": 0.1207, + "step": 2295 + }, + { + "epoch": 0.5218181818181818, + "grad_norm": 3.6634071525533014, + "learning_rate": 5.826388497262099e-07, + "loss": 0.156, + "step": 2296 + }, + { + "epoch": 0.5220454545454546, + "grad_norm": 2.2668242195289743, + "learning_rate": 5.821936377823464e-07, + "loss": 0.1265, + "step": 2297 + }, + { + "epoch": 0.5222727272727272, + "grad_norm": 2.844224739101693, + "learning_rate": 5.817484476608923e-07, + "loss": 0.1659, + "step": 2298 + }, + { + "epoch": 0.5225, + "grad_norm": 4.378221417094891, + "learning_rate": 5.813032795888027e-07, + "loss": 0.1319, + "step": 2299 + }, + { + "epoch": 0.5227272727272727, + "grad_norm": 2.7072871705981343, + "learning_rate": 5.80858133793021e-07, + "loss": 0.0965, + "step": 2300 + }, + { + "epoch": 0.5229545454545454, + "grad_norm": 5.13940888587471, + "learning_rate": 5.804130105004799e-07, + "loss": 0.1555, + "step": 2301 + }, + { + "epoch": 0.5231818181818182, + "grad_norm": 2.998387853092376, + "learning_rate": 5.799679099381005e-07, + "loss": 0.1003, + "step": 2302 + }, + { + "epoch": 0.5234090909090909, + "grad_norm": 2.0456400902758722, + "learning_rate": 5.795228323327921e-07, + "loss": 0.0602, + "step": 2303 + }, + { + "epoch": 0.5236363636363637, + "grad_norm": 2.5908984611498456, + "learning_rate": 5.790777779114524e-07, + "loss": 0.0745, + "step": 2304 + }, + { + "epoch": 0.5238636363636363, + "grad_norm": 1.8199882573294075, + "learning_rate": 5.786327469009675e-07, + "loss": 0.1056, + "step": 2305 + }, + { + "epoch": 0.524090909090909, + "grad_norm": 4.204329196522644, + "learning_rate": 5.781877395282112e-07, + "loss": 0.2489, + "step": 2306 + }, + { + "epoch": 0.5243181818181818, + "grad_norm": 1.8920137416981344, + "learning_rate": 5.777427560200454e-07, + "loss": 0.1153, + "step": 2307 + }, + { + "epoch": 0.5245454545454545, + "grad_norm": 2.634396332343288, + "learning_rate": 5.772977966033197e-07, + "loss": 0.0778, + "step": 2308 + }, + { + "epoch": 0.5247727272727273, + "grad_norm": 1.8731486887235567, + "learning_rate": 5.768528615048718e-07, + "loss": 0.0891, + "step": 2309 + }, + { + "epoch": 0.525, + "grad_norm": 3.3568226823842338, + "learning_rate": 5.764079509515267e-07, + "loss": 0.1837, + "step": 2310 + }, + { + "epoch": 0.5252272727272728, + "grad_norm": 2.3389832882905215, + "learning_rate": 5.759630651700971e-07, + "loss": 0.0777, + "step": 2311 + }, + { + "epoch": 0.5254545454545455, + "grad_norm": 4.067961400485023, + "learning_rate": 5.755182043873824e-07, + "loss": 0.1463, + "step": 2312 + }, + { + "epoch": 0.5256818181818181, + "grad_norm": 4.351141674871935, + "learning_rate": 5.750733688301703e-07, + "loss": 0.095, + "step": 2313 + }, + { + "epoch": 0.5259090909090909, + "grad_norm": 3.717047520429562, + "learning_rate": 5.746285587252351e-07, + "loss": 0.2185, + "step": 2314 + }, + { + "epoch": 0.5261363636363636, + "grad_norm": 3.4378658119295213, + "learning_rate": 5.741837742993379e-07, + "loss": 0.0691, + "step": 2315 + }, + { + "epoch": 0.5263636363636364, + "grad_norm": 2.065369027043132, + "learning_rate": 5.737390157792269e-07, + "loss": 0.116, + "step": 2316 + }, + { + "epoch": 0.5265909090909091, + "grad_norm": 2.8809373665996496, + "learning_rate": 5.732942833916374e-07, + "loss": 0.1233, + "step": 2317 + }, + { + "epoch": 0.5268181818181819, + "grad_norm": 3.3350865685038635, + "learning_rate": 5.72849577363291e-07, + "loss": 0.1027, + "step": 2318 + }, + { + "epoch": 0.5270454545454546, + "grad_norm": 2.578577881126313, + "learning_rate": 5.724048979208958e-07, + "loss": 0.1493, + "step": 2319 + }, + { + "epoch": 0.5272727272727272, + "grad_norm": 4.689661410731659, + "learning_rate": 5.719602452911469e-07, + "loss": 0.1178, + "step": 2320 + }, + { + "epoch": 0.5275, + "grad_norm": 3.0420137407781884, + "learning_rate": 5.715156197007253e-07, + "loss": 0.1264, + "step": 2321 + }, + { + "epoch": 0.5277272727272727, + "grad_norm": 4.752967980093874, + "learning_rate": 5.71071021376298e-07, + "loss": 0.1194, + "step": 2322 + }, + { + "epoch": 0.5279545454545455, + "grad_norm": 4.205501355993421, + "learning_rate": 5.706264505445186e-07, + "loss": 0.1649, + "step": 2323 + }, + { + "epoch": 0.5281818181818182, + "grad_norm": 1.651792458393887, + "learning_rate": 5.701819074320264e-07, + "loss": 0.091, + "step": 2324 + }, + { + "epoch": 0.5284090909090909, + "grad_norm": 3.7044668713689544, + "learning_rate": 5.697373922654464e-07, + "loss": 0.2169, + "step": 2325 + }, + { + "epoch": 0.5286363636363637, + "grad_norm": 3.20625359128848, + "learning_rate": 5.692929052713901e-07, + "loss": 0.0912, + "step": 2326 + }, + { + "epoch": 0.5288636363636363, + "grad_norm": 2.502343086630452, + "learning_rate": 5.688484466764537e-07, + "loss": 0.1248, + "step": 2327 + }, + { + "epoch": 0.5290909090909091, + "grad_norm": 3.344797493896496, + "learning_rate": 5.684040167072197e-07, + "loss": 0.1356, + "step": 2328 + }, + { + "epoch": 0.5293181818181818, + "grad_norm": 2.081519671317204, + "learning_rate": 5.679596155902553e-07, + "loss": 0.0919, + "step": 2329 + }, + { + "epoch": 0.5295454545454545, + "grad_norm": 3.9949984273040244, + "learning_rate": 5.675152435521133e-07, + "loss": 0.1043, + "step": 2330 + }, + { + "epoch": 0.5297727272727273, + "grad_norm": 3.2983645767933374, + "learning_rate": 5.67070900819332e-07, + "loss": 0.1499, + "step": 2331 + }, + { + "epoch": 0.53, + "grad_norm": 1.9281526165382132, + "learning_rate": 5.666265876184345e-07, + "loss": 0.1099, + "step": 2332 + }, + { + "epoch": 0.5302272727272728, + "grad_norm": 2.316721521166445, + "learning_rate": 5.661823041759286e-07, + "loss": 0.0714, + "step": 2333 + }, + { + "epoch": 0.5304545454545454, + "grad_norm": 3.8799662129895287, + "learning_rate": 5.657380507183073e-07, + "loss": 0.1234, + "step": 2334 + }, + { + "epoch": 0.5306818181818181, + "grad_norm": 2.6065268366643135, + "learning_rate": 5.652938274720482e-07, + "loss": 0.0789, + "step": 2335 + }, + { + "epoch": 0.5309090909090909, + "grad_norm": 2.4351420434675934, + "learning_rate": 5.648496346636132e-07, + "loss": 0.1407, + "step": 2336 + }, + { + "epoch": 0.5311363636363636, + "grad_norm": 2.6902868399220123, + "learning_rate": 5.644054725194492e-07, + "loss": 0.1164, + "step": 2337 + }, + { + "epoch": 0.5313636363636364, + "grad_norm": 4.1966868522270175, + "learning_rate": 5.639613412659872e-07, + "loss": 0.1154, + "step": 2338 + }, + { + "epoch": 0.5315909090909091, + "grad_norm": 2.080301109115863, + "learning_rate": 5.635172411296423e-07, + "loss": 0.0611, + "step": 2339 + }, + { + "epoch": 0.5318181818181819, + "grad_norm": 3.1743166271591026, + "learning_rate": 5.630731723368141e-07, + "loss": 0.1492, + "step": 2340 + }, + { + "epoch": 0.5320454545454546, + "grad_norm": 3.8685747171268345, + "learning_rate": 5.626291351138858e-07, + "loss": 0.228, + "step": 2341 + }, + { + "epoch": 0.5322727272727272, + "grad_norm": 6.398569613326352, + "learning_rate": 5.62185129687225e-07, + "loss": 0.1147, + "step": 2342 + }, + { + "epoch": 0.5325, + "grad_norm": 3.556717932837829, + "learning_rate": 5.617411562831829e-07, + "loss": 0.0921, + "step": 2343 + }, + { + "epoch": 0.5327272727272727, + "grad_norm": 2.374266271103528, + "learning_rate": 5.612972151280939e-07, + "loss": 0.1043, + "step": 2344 + }, + { + "epoch": 0.5329545454545455, + "grad_norm": 2.2846992412411855, + "learning_rate": 5.608533064482764e-07, + "loss": 0.1315, + "step": 2345 + }, + { + "epoch": 0.5331818181818182, + "grad_norm": 3.15364723235417, + "learning_rate": 5.604094304700326e-07, + "loss": 0.1355, + "step": 2346 + }, + { + "epoch": 0.5334090909090909, + "grad_norm": 3.0168571484504305, + "learning_rate": 5.599655874196473e-07, + "loss": 0.0803, + "step": 2347 + }, + { + "epoch": 0.5336363636363637, + "grad_norm": 2.4923692349232627, + "learning_rate": 5.595217775233889e-07, + "loss": 0.0799, + "step": 2348 + }, + { + "epoch": 0.5338636363636363, + "grad_norm": 2.7006086801923286, + "learning_rate": 5.590780010075088e-07, + "loss": 0.1637, + "step": 2349 + }, + { + "epoch": 0.5340909090909091, + "grad_norm": 2.579567708487795, + "learning_rate": 5.586342580982419e-07, + "loss": 0.151, + "step": 2350 + }, + { + "epoch": 0.5343181818181818, + "grad_norm": 4.547491617626686, + "learning_rate": 5.581905490218047e-07, + "loss": 0.1444, + "step": 2351 + }, + { + "epoch": 0.5345454545454545, + "grad_norm": 1.97263768339454, + "learning_rate": 5.577468740043978e-07, + "loss": 0.1088, + "step": 2352 + }, + { + "epoch": 0.5347727272727273, + "grad_norm": 3.0327583379957845, + "learning_rate": 5.573032332722035e-07, + "loss": 0.0838, + "step": 2353 + }, + { + "epoch": 0.535, + "grad_norm": 4.544480925416853, + "learning_rate": 5.568596270513873e-07, + "loss": 0.2223, + "step": 2354 + }, + { + "epoch": 0.5352272727272728, + "grad_norm": 4.585216509874686, + "learning_rate": 5.564160555680967e-07, + "loss": 0.1078, + "step": 2355 + }, + { + "epoch": 0.5354545454545454, + "grad_norm": 2.7035601557262092, + "learning_rate": 5.559725190484617e-07, + "loss": 0.1267, + "step": 2356 + }, + { + "epoch": 0.5356818181818181, + "grad_norm": 3.32857034034132, + "learning_rate": 5.555290177185945e-07, + "loss": 0.1072, + "step": 2357 + }, + { + "epoch": 0.5359090909090909, + "grad_norm": 1.5575189582097693, + "learning_rate": 5.550855518045889e-07, + "loss": 0.0668, + "step": 2358 + }, + { + "epoch": 0.5361363636363636, + "grad_norm": 2.714368863947764, + "learning_rate": 5.54642121532521e-07, + "loss": 0.1723, + "step": 2359 + }, + { + "epoch": 0.5363636363636364, + "grad_norm": 2.5478882483497856, + "learning_rate": 5.541987271284489e-07, + "loss": 0.1128, + "step": 2360 + }, + { + "epoch": 0.5365909090909091, + "grad_norm": 3.368030619248257, + "learning_rate": 5.537553688184122e-07, + "loss": 0.1663, + "step": 2361 + }, + { + "epoch": 0.5368181818181819, + "grad_norm": 2.8233571664802017, + "learning_rate": 5.53312046828432e-07, + "loss": 0.0867, + "step": 2362 + }, + { + "epoch": 0.5370454545454545, + "grad_norm": 1.6226630622785922, + "learning_rate": 5.528687613845112e-07, + "loss": 0.0997, + "step": 2363 + }, + { + "epoch": 0.5372727272727272, + "grad_norm": 3.7768178758848383, + "learning_rate": 5.52425512712634e-07, + "loss": 0.1011, + "step": 2364 + }, + { + "epoch": 0.5375, + "grad_norm": 4.673510362945767, + "learning_rate": 5.519823010387651e-07, + "loss": 0.1297, + "step": 2365 + }, + { + "epoch": 0.5377272727272727, + "grad_norm": 2.387101841437461, + "learning_rate": 5.515391265888516e-07, + "loss": 0.1032, + "step": 2366 + }, + { + "epoch": 0.5379545454545455, + "grad_norm": 1.416362107471065, + "learning_rate": 5.510959895888207e-07, + "loss": 0.0717, + "step": 2367 + }, + { + "epoch": 0.5381818181818182, + "grad_norm": 2.418591809508207, + "learning_rate": 5.50652890264581e-07, + "loss": 0.1153, + "step": 2368 + }, + { + "epoch": 0.538409090909091, + "grad_norm": 0.8575399138505985, + "learning_rate": 5.502098288420215e-07, + "loss": 0.0395, + "step": 2369 + }, + { + "epoch": 0.5386363636363637, + "grad_norm": 4.778967077971868, + "learning_rate": 5.497668055470121e-07, + "loss": 0.2393, + "step": 2370 + }, + { + "epoch": 0.5388636363636363, + "grad_norm": 4.382496324299509, + "learning_rate": 5.493238206054032e-07, + "loss": 0.1913, + "step": 2371 + }, + { + "epoch": 0.5390909090909091, + "grad_norm": 3.445477291602498, + "learning_rate": 5.488808742430262e-07, + "loss": 0.0616, + "step": 2372 + }, + { + "epoch": 0.5393181818181818, + "grad_norm": 4.244063053050491, + "learning_rate": 5.484379666856917e-07, + "loss": 0.1117, + "step": 2373 + }, + { + "epoch": 0.5395454545454546, + "grad_norm": 3.196258908233854, + "learning_rate": 5.479950981591911e-07, + "loss": 0.2086, + "step": 2374 + }, + { + "epoch": 0.5397727272727273, + "grad_norm": 3.661188088849851, + "learning_rate": 5.475522688892963e-07, + "loss": 0.1119, + "step": 2375 + }, + { + "epoch": 0.54, + "grad_norm": 2.4384136365009397, + "learning_rate": 5.471094791017588e-07, + "loss": 0.0742, + "step": 2376 + }, + { + "epoch": 0.5402272727272728, + "grad_norm": 3.991915481336656, + "learning_rate": 5.466667290223098e-07, + "loss": 0.0922, + "step": 2377 + }, + { + "epoch": 0.5404545454545454, + "grad_norm": 4.037347374173697, + "learning_rate": 5.462240188766607e-07, + "loss": 0.19, + "step": 2378 + }, + { + "epoch": 0.5406818181818182, + "grad_norm": 4.319974249719954, + "learning_rate": 5.457813488905022e-07, + "loss": 0.1594, + "step": 2379 + }, + { + "epoch": 0.5409090909090909, + "grad_norm": 4.095593215965782, + "learning_rate": 5.453387192895044e-07, + "loss": 0.1094, + "step": 2380 + }, + { + "epoch": 0.5411363636363636, + "grad_norm": 5.316689684092535, + "learning_rate": 5.448961302993171e-07, + "loss": 0.2008, + "step": 2381 + }, + { + "epoch": 0.5413636363636364, + "grad_norm": 3.5697875731052795, + "learning_rate": 5.444535821455696e-07, + "loss": 0.1409, + "step": 2382 + }, + { + "epoch": 0.5415909090909091, + "grad_norm": 2.860855547761017, + "learning_rate": 5.440110750538698e-07, + "loss": 0.0757, + "step": 2383 + }, + { + "epoch": 0.5418181818181819, + "grad_norm": 3.8549085593299846, + "learning_rate": 5.435686092498053e-07, + "loss": 0.149, + "step": 2384 + }, + { + "epoch": 0.5420454545454545, + "grad_norm": 2.5890717000332186, + "learning_rate": 5.431261849589419e-07, + "loss": 0.1528, + "step": 2385 + }, + { + "epoch": 0.5422727272727272, + "grad_norm": 2.9987917752279247, + "learning_rate": 5.426838024068254e-07, + "loss": 0.0949, + "step": 2386 + }, + { + "epoch": 0.5425, + "grad_norm": 3.577192160851525, + "learning_rate": 5.422414618189788e-07, + "loss": 0.2325, + "step": 2387 + }, + { + "epoch": 0.5427272727272727, + "grad_norm": 2.7367447855347242, + "learning_rate": 5.417991634209049e-07, + "loss": 0.1549, + "step": 2388 + }, + { + "epoch": 0.5429545454545455, + "grad_norm": 3.272572316481582, + "learning_rate": 5.413569074380845e-07, + "loss": 0.1276, + "step": 2389 + }, + { + "epoch": 0.5431818181818182, + "grad_norm": 4.160987775642646, + "learning_rate": 5.409146940959768e-07, + "loss": 0.1257, + "step": 2390 + }, + { + "epoch": 0.543409090909091, + "grad_norm": 2.0523255658820228, + "learning_rate": 5.404725236200195e-07, + "loss": 0.0919, + "step": 2391 + }, + { + "epoch": 0.5436363636363636, + "grad_norm": 2.3620155287916482, + "learning_rate": 5.40030396235628e-07, + "loss": 0.0928, + "step": 2392 + }, + { + "epoch": 0.5438636363636363, + "grad_norm": 1.8656827031205279, + "learning_rate": 5.395883121681965e-07, + "loss": 0.1086, + "step": 2393 + }, + { + "epoch": 0.5440909090909091, + "grad_norm": 2.2350912642240583, + "learning_rate": 5.391462716430963e-07, + "loss": 0.0761, + "step": 2394 + }, + { + "epoch": 0.5443181818181818, + "grad_norm": 3.239975731637355, + "learning_rate": 5.387042748856766e-07, + "loss": 0.1012, + "step": 2395 + }, + { + "epoch": 0.5445454545454546, + "grad_norm": 4.904964581523957, + "learning_rate": 5.382623221212648e-07, + "loss": 0.1121, + "step": 2396 + }, + { + "epoch": 0.5447727272727273, + "grad_norm": 3.22951283079793, + "learning_rate": 5.378204135751656e-07, + "loss": 0.0973, + "step": 2397 + }, + { + "epoch": 0.545, + "grad_norm": 2.2401168213663216, + "learning_rate": 5.373785494726612e-07, + "loss": 0.1003, + "step": 2398 + }, + { + "epoch": 0.5452272727272728, + "grad_norm": 4.293618719170746, + "learning_rate": 5.369367300390109e-07, + "loss": 0.1892, + "step": 2399 + }, + { + "epoch": 0.5454545454545454, + "grad_norm": 1.9623170865688295, + "learning_rate": 5.364949554994514e-07, + "loss": 0.0788, + "step": 2400 + }, + { + "epoch": 0.5456818181818182, + "grad_norm": 2.138754536446163, + "learning_rate": 5.360532260791969e-07, + "loss": 0.0805, + "step": 2401 + }, + { + "epoch": 0.5459090909090909, + "grad_norm": 4.073968707636006, + "learning_rate": 5.356115420034378e-07, + "loss": 0.0964, + "step": 2402 + }, + { + "epoch": 0.5461363636363636, + "grad_norm": 2.2693695454525824, + "learning_rate": 5.35169903497342e-07, + "loss": 0.0999, + "step": 2403 + }, + { + "epoch": 0.5463636363636364, + "grad_norm": 2.819775214556877, + "learning_rate": 5.347283107860538e-07, + "loss": 0.0985, + "step": 2404 + }, + { + "epoch": 0.5465909090909091, + "grad_norm": 4.119708957741094, + "learning_rate": 5.342867640946948e-07, + "loss": 0.1904, + "step": 2405 + }, + { + "epoch": 0.5468181818181819, + "grad_norm": 3.472150559794801, + "learning_rate": 5.33845263648362e-07, + "loss": 0.0866, + "step": 2406 + }, + { + "epoch": 0.5470454545454545, + "grad_norm": 4.329854949502893, + "learning_rate": 5.334038096721301e-07, + "loss": 0.0866, + "step": 2407 + }, + { + "epoch": 0.5472727272727272, + "grad_norm": 4.631840520722702, + "learning_rate": 5.329624023910494e-07, + "loss": 0.1992, + "step": 2408 + }, + { + "epoch": 0.5475, + "grad_norm": 5.534465676222028, + "learning_rate": 5.325210420301461e-07, + "loss": 0.1882, + "step": 2409 + }, + { + "epoch": 0.5477272727272727, + "grad_norm": 3.406347334290032, + "learning_rate": 5.320797288144231e-07, + "loss": 0.0785, + "step": 2410 + }, + { + "epoch": 0.5479545454545455, + "grad_norm": 3.9713794170871117, + "learning_rate": 5.316384629688594e-07, + "loss": 0.1083, + "step": 2411 + }, + { + "epoch": 0.5481818181818182, + "grad_norm": 2.6187052204753902, + "learning_rate": 5.31197244718409e-07, + "loss": 0.115, + "step": 2412 + }, + { + "epoch": 0.548409090909091, + "grad_norm": 3.5743324879773914, + "learning_rate": 5.307560742880025e-07, + "loss": 0.1167, + "step": 2413 + }, + { + "epoch": 0.5486363636363636, + "grad_norm": 3.1653107144607624, + "learning_rate": 5.303149519025455e-07, + "loss": 0.1707, + "step": 2414 + }, + { + "epoch": 0.5488636363636363, + "grad_norm": 2.3423049256851547, + "learning_rate": 5.298738777869198e-07, + "loss": 0.0798, + "step": 2415 + }, + { + "epoch": 0.5490909090909091, + "grad_norm": 3.2418456005038325, + "learning_rate": 5.294328521659814e-07, + "loss": 0.11, + "step": 2416 + }, + { + "epoch": 0.5493181818181818, + "grad_norm": 5.318523050798828, + "learning_rate": 5.28991875264563e-07, + "loss": 0.3042, + "step": 2417 + }, + { + "epoch": 0.5495454545454546, + "grad_norm": 4.318837209593635, + "learning_rate": 5.285509473074714e-07, + "loss": 0.2197, + "step": 2418 + }, + { + "epoch": 0.5497727272727273, + "grad_norm": 4.685279770940064, + "learning_rate": 5.281100685194891e-07, + "loss": 0.097, + "step": 2419 + }, + { + "epoch": 0.55, + "grad_norm": 4.070147794366691, + "learning_rate": 5.276692391253732e-07, + "loss": 0.1935, + "step": 2420 + }, + { + "epoch": 0.5502272727272727, + "grad_norm": 2.506986901736058, + "learning_rate": 5.272284593498556e-07, + "loss": 0.1725, + "step": 2421 + }, + { + "epoch": 0.5504545454545454, + "grad_norm": 4.566805633168652, + "learning_rate": 5.267877294176435e-07, + "loss": 0.1171, + "step": 2422 + }, + { + "epoch": 0.5506818181818182, + "grad_norm": 3.978155447219618, + "learning_rate": 5.263470495534176e-07, + "loss": 0.1654, + "step": 2423 + }, + { + "epoch": 0.5509090909090909, + "grad_norm": 2.539035736507674, + "learning_rate": 5.259064199818337e-07, + "loss": 0.0647, + "step": 2424 + }, + { + "epoch": 0.5511363636363636, + "grad_norm": 2.687626999045767, + "learning_rate": 5.254658409275223e-07, + "loss": 0.0758, + "step": 2425 + }, + { + "epoch": 0.5513636363636364, + "grad_norm": 3.806749371508724, + "learning_rate": 5.250253126150875e-07, + "loss": 0.117, + "step": 2426 + }, + { + "epoch": 0.5515909090909091, + "grad_norm": 2.524481456531903, + "learning_rate": 5.245848352691077e-07, + "loss": 0.1494, + "step": 2427 + }, + { + "epoch": 0.5518181818181818, + "grad_norm": 2.1379262706751407, + "learning_rate": 5.24144409114136e-07, + "loss": 0.2032, + "step": 2428 + }, + { + "epoch": 0.5520454545454545, + "grad_norm": 3.0490835329307546, + "learning_rate": 5.237040343746986e-07, + "loss": 0.132, + "step": 2429 + }, + { + "epoch": 0.5522727272727272, + "grad_norm": 2.644380330008828, + "learning_rate": 5.232637112752951e-07, + "loss": 0.0895, + "step": 2430 + }, + { + "epoch": 0.5525, + "grad_norm": 3.354521844250964, + "learning_rate": 5.228234400404002e-07, + "loss": 0.1155, + "step": 2431 + }, + { + "epoch": 0.5527272727272727, + "grad_norm": 4.418748544680731, + "learning_rate": 5.223832208944607e-07, + "loss": 0.1106, + "step": 2432 + }, + { + "epoch": 0.5529545454545455, + "grad_norm": 4.65422639162849, + "learning_rate": 5.21943054061898e-07, + "loss": 0.0886, + "step": 2433 + }, + { + "epoch": 0.5531818181818182, + "grad_norm": 3.1415274009626253, + "learning_rate": 5.215029397671064e-07, + "loss": 0.0893, + "step": 2434 + }, + { + "epoch": 0.553409090909091, + "grad_norm": 2.055485137253343, + "learning_rate": 5.210628782344528e-07, + "loss": 0.1386, + "step": 2435 + }, + { + "epoch": 0.5536363636363636, + "grad_norm": 3.8688413002522637, + "learning_rate": 5.206228696882782e-07, + "loss": 0.1589, + "step": 2436 + }, + { + "epoch": 0.5538636363636363, + "grad_norm": 2.1466104766693284, + "learning_rate": 5.201829143528962e-07, + "loss": 0.0838, + "step": 2437 + }, + { + "epoch": 0.5540909090909091, + "grad_norm": 4.4015493826913685, + "learning_rate": 5.197430124525927e-07, + "loss": 0.1196, + "step": 2438 + }, + { + "epoch": 0.5543181818181818, + "grad_norm": 4.9135023884321845, + "learning_rate": 5.193031642116273e-07, + "loss": 0.1144, + "step": 2439 + }, + { + "epoch": 0.5545454545454546, + "grad_norm": 4.239795055556384, + "learning_rate": 5.188633698542317e-07, + "loss": 0.105, + "step": 2440 + }, + { + "epoch": 0.5547727272727273, + "grad_norm": 2.672473764807448, + "learning_rate": 5.184236296046101e-07, + "loss": 0.0767, + "step": 2441 + }, + { + "epoch": 0.555, + "grad_norm": 3.392514923985101, + "learning_rate": 5.179839436869396e-07, + "loss": 0.1475, + "step": 2442 + }, + { + "epoch": 0.5552272727272727, + "grad_norm": 5.672705230397529, + "learning_rate": 5.17544312325369e-07, + "loss": 0.1492, + "step": 2443 + }, + { + "epoch": 0.5554545454545454, + "grad_norm": 2.705903112947714, + "learning_rate": 5.171047357440199e-07, + "loss": 0.168, + "step": 2444 + }, + { + "epoch": 0.5556818181818182, + "grad_norm": 1.7883398209023293, + "learning_rate": 5.16665214166985e-07, + "loss": 0.0914, + "step": 2445 + }, + { + "epoch": 0.5559090909090909, + "grad_norm": 1.995383094568805, + "learning_rate": 5.162257478183301e-07, + "loss": 0.0908, + "step": 2446 + }, + { + "epoch": 0.5561363636363637, + "grad_norm": 4.637594953464127, + "learning_rate": 5.15786336922092e-07, + "loss": 0.1213, + "step": 2447 + }, + { + "epoch": 0.5563636363636364, + "grad_norm": 3.124856125327974, + "learning_rate": 5.153469817022798e-07, + "loss": 0.1975, + "step": 2448 + }, + { + "epoch": 0.5565909090909091, + "grad_norm": 4.888808661671313, + "learning_rate": 5.14907682382874e-07, + "loss": 0.1165, + "step": 2449 + }, + { + "epoch": 0.5568181818181818, + "grad_norm": 2.274597477228122, + "learning_rate": 5.144684391878264e-07, + "loss": 0.162, + "step": 2450 + }, + { + "epoch": 0.5570454545454545, + "grad_norm": 2.304547481934206, + "learning_rate": 5.140292523410606e-07, + "loss": 0.0894, + "step": 2451 + }, + { + "epoch": 0.5572727272727273, + "grad_norm": 3.258430177573711, + "learning_rate": 5.13590122066471e-07, + "loss": 0.0609, + "step": 2452 + }, + { + "epoch": 0.5575, + "grad_norm": 2.5848771761196607, + "learning_rate": 5.131510485879235e-07, + "loss": 0.0893, + "step": 2453 + }, + { + "epoch": 0.5577272727272727, + "grad_norm": 4.5221233632393565, + "learning_rate": 5.127120321292547e-07, + "loss": 0.1834, + "step": 2454 + }, + { + "epoch": 0.5579545454545455, + "grad_norm": 2.8499071311078406, + "learning_rate": 5.122730729142729e-07, + "loss": 0.1267, + "step": 2455 + }, + { + "epoch": 0.5581818181818182, + "grad_norm": 1.8775789129981284, + "learning_rate": 5.118341711667562e-07, + "loss": 0.0559, + "step": 2456 + }, + { + "epoch": 0.5584090909090909, + "grad_norm": 3.1234935862395465, + "learning_rate": 5.113953271104542e-07, + "loss": 0.0941, + "step": 2457 + }, + { + "epoch": 0.5586363636363636, + "grad_norm": 1.7607408284245654, + "learning_rate": 5.10956540969087e-07, + "loss": 0.0485, + "step": 2458 + }, + { + "epoch": 0.5588636363636363, + "grad_norm": 2.8872252842829753, + "learning_rate": 5.105178129663443e-07, + "loss": 0.1263, + "step": 2459 + }, + { + "epoch": 0.5590909090909091, + "grad_norm": 3.459889639820743, + "learning_rate": 5.100791433258871e-07, + "loss": 0.1691, + "step": 2460 + }, + { + "epoch": 0.5593181818181818, + "grad_norm": 4.9167445614081, + "learning_rate": 5.096405322713467e-07, + "loss": 0.1691, + "step": 2461 + }, + { + "epoch": 0.5595454545454546, + "grad_norm": 3.950558148189518, + "learning_rate": 5.092019800263239e-07, + "loss": 0.1936, + "step": 2462 + }, + { + "epoch": 0.5597727272727273, + "grad_norm": 2.723030618904908, + "learning_rate": 5.0876348681439e-07, + "loss": 0.1401, + "step": 2463 + }, + { + "epoch": 0.56, + "grad_norm": 2.738280567900871, + "learning_rate": 5.083250528590857e-07, + "loss": 0.0902, + "step": 2464 + }, + { + "epoch": 0.5602272727272727, + "grad_norm": 2.4714386262169032, + "learning_rate": 5.07886678383922e-07, + "loss": 0.1179, + "step": 2465 + }, + { + "epoch": 0.5604545454545454, + "grad_norm": 1.652988589398669, + "learning_rate": 5.074483636123798e-07, + "loss": 0.0683, + "step": 2466 + }, + { + "epoch": 0.5606818181818182, + "grad_norm": 4.303504508185148, + "learning_rate": 5.070101087679087e-07, + "loss": 0.133, + "step": 2467 + }, + { + "epoch": 0.5609090909090909, + "grad_norm": 2.9677836752773548, + "learning_rate": 5.065719140739281e-07, + "loss": 0.0688, + "step": 2468 + }, + { + "epoch": 0.5611363636363637, + "grad_norm": 2.7467835564727645, + "learning_rate": 5.061337797538273e-07, + "loss": 0.0717, + "step": 2469 + }, + { + "epoch": 0.5613636363636364, + "grad_norm": 1.9886710249173187, + "learning_rate": 5.056957060309639e-07, + "loss": 0.0702, + "step": 2470 + }, + { + "epoch": 0.5615909090909091, + "grad_norm": 2.849953386417299, + "learning_rate": 5.052576931286654e-07, + "loss": 0.1024, + "step": 2471 + }, + { + "epoch": 0.5618181818181818, + "grad_norm": 2.342338736300014, + "learning_rate": 5.04819741270228e-07, + "loss": 0.0818, + "step": 2472 + }, + { + "epoch": 0.5620454545454545, + "grad_norm": 2.6416132616276213, + "learning_rate": 5.043818506789168e-07, + "loss": 0.0591, + "step": 2473 + }, + { + "epoch": 0.5622727272727273, + "grad_norm": 6.057293862677206, + "learning_rate": 5.039440215779651e-07, + "loss": 0.1675, + "step": 2474 + }, + { + "epoch": 0.5625, + "grad_norm": 3.133666399209017, + "learning_rate": 5.035062541905759e-07, + "loss": 0.1189, + "step": 2475 + }, + { + "epoch": 0.5627272727272727, + "grad_norm": 2.2310941812771214, + "learning_rate": 5.0306854873992e-07, + "loss": 0.1275, + "step": 2476 + }, + { + "epoch": 0.5629545454545455, + "grad_norm": 6.236531653035315, + "learning_rate": 5.026309054491367e-07, + "loss": 0.171, + "step": 2477 + }, + { + "epoch": 0.5631818181818182, + "grad_norm": 4.431174153716426, + "learning_rate": 5.021933245413342e-07, + "loss": 0.0923, + "step": 2478 + }, + { + "epoch": 0.5634090909090909, + "grad_norm": 2.926672017387125, + "learning_rate": 5.017558062395879e-07, + "loss": 0.1895, + "step": 2479 + }, + { + "epoch": 0.5636363636363636, + "grad_norm": 1.8447739800703142, + "learning_rate": 5.013183507669424e-07, + "loss": 0.1411, + "step": 2480 + }, + { + "epoch": 0.5638636363636363, + "grad_norm": 2.5778218741421055, + "learning_rate": 5.008809583464093e-07, + "loss": 0.1216, + "step": 2481 + }, + { + "epoch": 0.5640909090909091, + "grad_norm": 3.6294406699417583, + "learning_rate": 5.004436292009684e-07, + "loss": 0.1564, + "step": 2482 + }, + { + "epoch": 0.5643181818181818, + "grad_norm": 4.64838748191868, + "learning_rate": 5.000063635535673e-07, + "loss": 0.101, + "step": 2483 + }, + { + "epoch": 0.5645454545454546, + "grad_norm": 2.8081463364675305, + "learning_rate": 4.995691616271217e-07, + "loss": 0.1497, + "step": 2484 + }, + { + "epoch": 0.5647727272727273, + "grad_norm": 2.465746155436621, + "learning_rate": 4.991320236445138e-07, + "loss": 0.097, + "step": 2485 + }, + { + "epoch": 0.565, + "grad_norm": 2.802283870822658, + "learning_rate": 4.986949498285941e-07, + "loss": 0.1238, + "step": 2486 + }, + { + "epoch": 0.5652272727272727, + "grad_norm": 2.9379500319478997, + "learning_rate": 4.982579404021799e-07, + "loss": 0.1468, + "step": 2487 + }, + { + "epoch": 0.5654545454545454, + "grad_norm": 2.4015233103742726, + "learning_rate": 4.978209955880556e-07, + "loss": 0.1185, + "step": 2488 + }, + { + "epoch": 0.5656818181818182, + "grad_norm": 1.5174694882602615, + "learning_rate": 4.97384115608973e-07, + "loss": 0.1111, + "step": 2489 + }, + { + "epoch": 0.5659090909090909, + "grad_norm": 4.402056400889555, + "learning_rate": 4.96947300687651e-07, + "loss": 0.1043, + "step": 2490 + }, + { + "epoch": 0.5661363636363637, + "grad_norm": 3.553266048170793, + "learning_rate": 4.965105510467747e-07, + "loss": 0.1228, + "step": 2491 + }, + { + "epoch": 0.5663636363636364, + "grad_norm": 2.415150867453801, + "learning_rate": 4.960738669089964e-07, + "loss": 0.1269, + "step": 2492 + }, + { + "epoch": 0.5665909090909091, + "grad_norm": 3.2210973294833853, + "learning_rate": 4.956372484969348e-07, + "loss": 0.1774, + "step": 2493 + }, + { + "epoch": 0.5668181818181818, + "grad_norm": 4.619862621790018, + "learning_rate": 4.952006960331752e-07, + "loss": 0.2296, + "step": 2494 + }, + { + "epoch": 0.5670454545454545, + "grad_norm": 5.22153879993606, + "learning_rate": 4.947642097402696e-07, + "loss": 0.2501, + "step": 2495 + }, + { + "epoch": 0.5672727272727273, + "grad_norm": 2.1424594876020713, + "learning_rate": 4.943277898407354e-07, + "loss": 0.1235, + "step": 2496 + }, + { + "epoch": 0.5675, + "grad_norm": 4.274871728241211, + "learning_rate": 4.938914365570568e-07, + "loss": 0.1145, + "step": 2497 + }, + { + "epoch": 0.5677272727272727, + "grad_norm": 4.24267516823951, + "learning_rate": 4.934551501116841e-07, + "loss": 0.0981, + "step": 2498 + }, + { + "epoch": 0.5679545454545455, + "grad_norm": 3.806649269322541, + "learning_rate": 4.930189307270331e-07, + "loss": 0.1235, + "step": 2499 + }, + { + "epoch": 0.5681818181818182, + "grad_norm": 5.348348026617713, + "learning_rate": 4.925827786254857e-07, + "loss": 0.1663, + "step": 2500 + }, + { + "epoch": 0.5684090909090909, + "grad_norm": 4.816558876834777, + "learning_rate": 4.921466940293896e-07, + "loss": 0.1112, + "step": 2501 + }, + { + "epoch": 0.5686363636363636, + "grad_norm": 2.756100099272847, + "learning_rate": 4.91710677161058e-07, + "loss": 0.0688, + "step": 2502 + }, + { + "epoch": 0.5688636363636363, + "grad_norm": 5.040426703868668, + "learning_rate": 4.912747282427691e-07, + "loss": 0.2185, + "step": 2503 + }, + { + "epoch": 0.5690909090909091, + "grad_norm": 2.557524893613585, + "learning_rate": 4.908388474967673e-07, + "loss": 0.1092, + "step": 2504 + }, + { + "epoch": 0.5693181818181818, + "grad_norm": 4.329921061646522, + "learning_rate": 4.904030351452614e-07, + "loss": 0.2359, + "step": 2505 + }, + { + "epoch": 0.5695454545454546, + "grad_norm": 2.098576744754703, + "learning_rate": 4.89967291410426e-07, + "loss": 0.0717, + "step": 2506 + }, + { + "epoch": 0.5697727272727273, + "grad_norm": 3.583016809117724, + "learning_rate": 4.895316165144005e-07, + "loss": 0.1576, + "step": 2507 + }, + { + "epoch": 0.57, + "grad_norm": 3.021847849593515, + "learning_rate": 4.890960106792891e-07, + "loss": 0.1564, + "step": 2508 + }, + { + "epoch": 0.5702272727272727, + "grad_norm": 2.8509589492688234, + "learning_rate": 4.886604741271611e-07, + "loss": 0.0979, + "step": 2509 + }, + { + "epoch": 0.5704545454545454, + "grad_norm": 3.225583094931911, + "learning_rate": 4.882250070800497e-07, + "loss": 0.1249, + "step": 2510 + }, + { + "epoch": 0.5706818181818182, + "grad_norm": 3.3327298054403416, + "learning_rate": 4.877896097599536e-07, + "loss": 0.1835, + "step": 2511 + }, + { + "epoch": 0.5709090909090909, + "grad_norm": 3.677627207136625, + "learning_rate": 4.873542823888355e-07, + "loss": 0.0935, + "step": 2512 + }, + { + "epoch": 0.5711363636363637, + "grad_norm": 3.1476001913014646, + "learning_rate": 4.869190251886226e-07, + "loss": 0.1108, + "step": 2513 + }, + { + "epoch": 0.5713636363636364, + "grad_norm": 1.8357597857281702, + "learning_rate": 4.86483838381206e-07, + "loss": 0.0858, + "step": 2514 + }, + { + "epoch": 0.571590909090909, + "grad_norm": 2.521952908742896, + "learning_rate": 4.860487221884411e-07, + "loss": 0.1496, + "step": 2515 + }, + { + "epoch": 0.5718181818181818, + "grad_norm": 3.1001793128060635, + "learning_rate": 4.856136768321479e-07, + "loss": 0.1551, + "step": 2516 + }, + { + "epoch": 0.5720454545454545, + "grad_norm": 1.9695021489079239, + "learning_rate": 4.851787025341093e-07, + "loss": 0.0799, + "step": 2517 + }, + { + "epoch": 0.5722727272727273, + "grad_norm": 3.1950430057507804, + "learning_rate": 4.847437995160722e-07, + "loss": 0.1224, + "step": 2518 + }, + { + "epoch": 0.5725, + "grad_norm": 4.587220226893224, + "learning_rate": 4.843089679997479e-07, + "loss": 0.2329, + "step": 2519 + }, + { + "epoch": 0.5727272727272728, + "grad_norm": 2.8062399466185797, + "learning_rate": 4.838742082068103e-07, + "loss": 0.1547, + "step": 2520 + }, + { + "epoch": 0.5729545454545455, + "grad_norm": 2.6711318849455616, + "learning_rate": 4.834395203588972e-07, + "loss": 0.1048, + "step": 2521 + }, + { + "epoch": 0.5731818181818182, + "grad_norm": 3.8780307998003942, + "learning_rate": 4.8300490467761e-07, + "loss": 0.1128, + "step": 2522 + }, + { + "epoch": 0.5734090909090909, + "grad_norm": 4.560736742380297, + "learning_rate": 4.825703613845127e-07, + "loss": 0.1092, + "step": 2523 + }, + { + "epoch": 0.5736363636363636, + "grad_norm": 2.8901825484600923, + "learning_rate": 4.821358907011331e-07, + "loss": 0.0976, + "step": 2524 + }, + { + "epoch": 0.5738636363636364, + "grad_norm": 2.907759183402684, + "learning_rate": 4.81701492848961e-07, + "loss": 0.1297, + "step": 2525 + }, + { + "epoch": 0.5740909090909091, + "grad_norm": 2.5706020826114053, + "learning_rate": 4.812671680494497e-07, + "loss": 0.0926, + "step": 2526 + }, + { + "epoch": 0.5743181818181818, + "grad_norm": 2.72447664824071, + "learning_rate": 4.808329165240155e-07, + "loss": 0.121, + "step": 2527 + }, + { + "epoch": 0.5745454545454546, + "grad_norm": 2.0157450550663243, + "learning_rate": 4.803987384940371e-07, + "loss": 0.0904, + "step": 2528 + }, + { + "epoch": 0.5747727272727273, + "grad_norm": 2.3917122112795632, + "learning_rate": 4.799646341808553e-07, + "loss": 0.1135, + "step": 2529 + }, + { + "epoch": 0.575, + "grad_norm": 3.008524913188212, + "learning_rate": 4.795306038057742e-07, + "loss": 0.0833, + "step": 2530 + }, + { + "epoch": 0.5752272727272727, + "grad_norm": 3.1241749104991148, + "learning_rate": 4.790966475900594e-07, + "loss": 0.0966, + "step": 2531 + }, + { + "epoch": 0.5754545454545454, + "grad_norm": 1.9105520267260785, + "learning_rate": 4.786627657549388e-07, + "loss": 0.0729, + "step": 2532 + }, + { + "epoch": 0.5756818181818182, + "grad_norm": 4.606280636942503, + "learning_rate": 4.782289585216027e-07, + "loss": 0.1445, + "step": 2533 + }, + { + "epoch": 0.5759090909090909, + "grad_norm": 2.8981751642156315, + "learning_rate": 4.777952261112036e-07, + "loss": 0.129, + "step": 2534 + }, + { + "epoch": 0.5761363636363637, + "grad_norm": 3.0123383037005453, + "learning_rate": 4.77361568744855e-07, + "loss": 0.1369, + "step": 2535 + }, + { + "epoch": 0.5763636363636364, + "grad_norm": 2.726390960369975, + "learning_rate": 4.769279866436331e-07, + "loss": 0.0847, + "step": 2536 + }, + { + "epoch": 0.576590909090909, + "grad_norm": 2.566205345478119, + "learning_rate": 4.764944800285749e-07, + "loss": 0.0707, + "step": 2537 + }, + { + "epoch": 0.5768181818181818, + "grad_norm": 2.7445289617611186, + "learning_rate": 4.7606104912067984e-07, + "loss": 0.0972, + "step": 2538 + }, + { + "epoch": 0.5770454545454545, + "grad_norm": 3.045607224337679, + "learning_rate": 4.756276941409075e-07, + "loss": 0.1743, + "step": 2539 + }, + { + "epoch": 0.5772727272727273, + "grad_norm": 4.443242405996137, + "learning_rate": 4.7519441531018e-07, + "loss": 0.1032, + "step": 2540 + }, + { + "epoch": 0.5775, + "grad_norm": 5.0743724441769436, + "learning_rate": 4.7476121284937976e-07, + "loss": 0.2585, + "step": 2541 + }, + { + "epoch": 0.5777272727272728, + "grad_norm": 3.0077936694948235, + "learning_rate": 4.7432808697935096e-07, + "loss": 0.1736, + "step": 2542 + }, + { + "epoch": 0.5779545454545455, + "grad_norm": 3.5322958410060394, + "learning_rate": 4.7389503792089806e-07, + "loss": 0.1656, + "step": 2543 + }, + { + "epoch": 0.5781818181818181, + "grad_norm": 3.28110890041884, + "learning_rate": 4.7346206589478695e-07, + "loss": 0.0791, + "step": 2544 + }, + { + "epoch": 0.5784090909090909, + "grad_norm": 2.982602048594925, + "learning_rate": 4.7302917112174405e-07, + "loss": 0.0867, + "step": 2545 + }, + { + "epoch": 0.5786363636363636, + "grad_norm": 2.5245952993731953, + "learning_rate": 4.7259635382245617e-07, + "loss": 0.1227, + "step": 2546 + }, + { + "epoch": 0.5788636363636364, + "grad_norm": 2.504198300902271, + "learning_rate": 4.721636142175706e-07, + "loss": 0.0699, + "step": 2547 + }, + { + "epoch": 0.5790909090909091, + "grad_norm": 2.1350158342580015, + "learning_rate": 4.717309525276957e-07, + "loss": 0.1287, + "step": 2548 + }, + { + "epoch": 0.5793181818181818, + "grad_norm": 2.440314771730277, + "learning_rate": 4.7129836897339915e-07, + "loss": 0.0928, + "step": 2549 + }, + { + "epoch": 0.5795454545454546, + "grad_norm": 3.208887751256706, + "learning_rate": 4.7086586377520954e-07, + "loss": 0.0755, + "step": 2550 + }, + { + "epoch": 0.5797727272727272, + "grad_norm": 3.158071203210638, + "learning_rate": 4.7043343715361534e-07, + "loss": 0.1817, + "step": 2551 + }, + { + "epoch": 0.58, + "grad_norm": 2.2701095802560856, + "learning_rate": 4.700010893290649e-07, + "loss": 0.0989, + "step": 2552 + }, + { + "epoch": 0.5802272727272727, + "grad_norm": 3.414567542275143, + "learning_rate": 4.695688205219659e-07, + "loss": 0.1554, + "step": 2553 + }, + { + "epoch": 0.5804545454545454, + "grad_norm": 3.8361681519222173, + "learning_rate": 4.691366309526867e-07, + "loss": 0.1372, + "step": 2554 + }, + { + "epoch": 0.5806818181818182, + "grad_norm": 3.3841132090848784, + "learning_rate": 4.6870452084155444e-07, + "loss": 0.1504, + "step": 2555 + }, + { + "epoch": 0.5809090909090909, + "grad_norm": 3.0654142115291156, + "learning_rate": 4.682724904088562e-07, + "loss": 0.1837, + "step": 2556 + }, + { + "epoch": 0.5811363636363637, + "grad_norm": 1.3421104429501804, + "learning_rate": 4.6784053987483853e-07, + "loss": 0.0451, + "step": 2557 + }, + { + "epoch": 0.5813636363636364, + "grad_norm": 2.088416722802629, + "learning_rate": 4.674086694597067e-07, + "loss": 0.1335, + "step": 2558 + }, + { + "epoch": 0.581590909090909, + "grad_norm": 1.7253635066175261, + "learning_rate": 4.6697687938362585e-07, + "loss": 0.0674, + "step": 2559 + }, + { + "epoch": 0.5818181818181818, + "grad_norm": 3.3186556596477863, + "learning_rate": 4.665451698667198e-07, + "loss": 0.1873, + "step": 2560 + }, + { + "epoch": 0.5820454545454545, + "grad_norm": 3.9875398672606863, + "learning_rate": 4.6611354112907084e-07, + "loss": 0.106, + "step": 2561 + }, + { + "epoch": 0.5822727272727273, + "grad_norm": 2.9691270619431602, + "learning_rate": 4.656819933907207e-07, + "loss": 0.096, + "step": 2562 + }, + { + "epoch": 0.5825, + "grad_norm": 4.141367842390959, + "learning_rate": 4.6525052687167e-07, + "loss": 0.2127, + "step": 2563 + }, + { + "epoch": 0.5827272727272728, + "grad_norm": 2.341461437785818, + "learning_rate": 4.648191417918771e-07, + "loss": 0.1461, + "step": 2564 + }, + { + "epoch": 0.5829545454545455, + "grad_norm": 3.7320086368517416, + "learning_rate": 4.643878383712598e-07, + "loss": 0.0945, + "step": 2565 + }, + { + "epoch": 0.5831818181818181, + "grad_norm": 2.9952846187992814, + "learning_rate": 4.639566168296936e-07, + "loss": 0.0902, + "step": 2566 + }, + { + "epoch": 0.5834090909090909, + "grad_norm": 2.5612204359673054, + "learning_rate": 4.6352547738701266e-07, + "loss": 0.1329, + "step": 2567 + }, + { + "epoch": 0.5836363636363636, + "grad_norm": 4.257588726703706, + "learning_rate": 4.6309442026300875e-07, + "loss": 0.0908, + "step": 2568 + }, + { + "epoch": 0.5838636363636364, + "grad_norm": 3.357621144347301, + "learning_rate": 4.626634456774323e-07, + "loss": 0.2162, + "step": 2569 + }, + { + "epoch": 0.5840909090909091, + "grad_norm": 4.697260275225164, + "learning_rate": 4.6223255384999124e-07, + "loss": 0.2351, + "step": 2570 + }, + { + "epoch": 0.5843181818181818, + "grad_norm": 2.249521620868953, + "learning_rate": 4.6180174500035173e-07, + "loss": 0.1256, + "step": 2571 + }, + { + "epoch": 0.5845454545454546, + "grad_norm": 3.7199342502474484, + "learning_rate": 4.61371019348137e-07, + "loss": 0.1553, + "step": 2572 + }, + { + "epoch": 0.5847727272727272, + "grad_norm": 2.441261082356622, + "learning_rate": 4.6094037711292843e-07, + "loss": 0.1314, + "step": 2573 + }, + { + "epoch": 0.585, + "grad_norm": 1.3869172797140783, + "learning_rate": 4.6050981851426495e-07, + "loss": 0.1053, + "step": 2574 + }, + { + "epoch": 0.5852272727272727, + "grad_norm": 3.443225144059737, + "learning_rate": 4.600793437716421e-07, + "loss": 0.0841, + "step": 2575 + }, + { + "epoch": 0.5854545454545454, + "grad_norm": 3.63539172848414, + "learning_rate": 4.5964895310451325e-07, + "loss": 0.233, + "step": 2576 + }, + { + "epoch": 0.5856818181818182, + "grad_norm": 4.088598314397255, + "learning_rate": 4.59218646732289e-07, + "loss": 0.1706, + "step": 2577 + }, + { + "epoch": 0.5859090909090909, + "grad_norm": 2.6357236782359945, + "learning_rate": 4.5878842487433657e-07, + "loss": 0.0722, + "step": 2578 + }, + { + "epoch": 0.5861363636363637, + "grad_norm": 5.403586180788968, + "learning_rate": 4.5835828774998035e-07, + "loss": 0.1473, + "step": 2579 + }, + { + "epoch": 0.5863636363636363, + "grad_norm": 4.4276642096674355, + "learning_rate": 4.5792823557850167e-07, + "loss": 0.2346, + "step": 2580 + }, + { + "epoch": 0.586590909090909, + "grad_norm": 3.7513673987969756, + "learning_rate": 4.5749826857913843e-07, + "loss": 0.1148, + "step": 2581 + }, + { + "epoch": 0.5868181818181818, + "grad_norm": 2.6228998308247617, + "learning_rate": 4.570683869710846e-07, + "loss": 0.06, + "step": 2582 + }, + { + "epoch": 0.5870454545454545, + "grad_norm": 2.529132331123581, + "learning_rate": 4.5663859097349127e-07, + "loss": 0.1355, + "step": 2583 + }, + { + "epoch": 0.5872727272727273, + "grad_norm": 4.206955552510056, + "learning_rate": 4.562088808054659e-07, + "loss": 0.12, + "step": 2584 + }, + { + "epoch": 0.5875, + "grad_norm": 3.58653672978892, + "learning_rate": 4.557792566860718e-07, + "loss": 0.13, + "step": 2585 + }, + { + "epoch": 0.5877272727272728, + "grad_norm": 4.364773561570753, + "learning_rate": 4.5534971883432866e-07, + "loss": 0.1296, + "step": 2586 + }, + { + "epoch": 0.5879545454545455, + "grad_norm": 1.5723344186901094, + "learning_rate": 4.54920267469212e-07, + "loss": 0.1134, + "step": 2587 + }, + { + "epoch": 0.5881818181818181, + "grad_norm": 1.969687249667709, + "learning_rate": 4.544909028096535e-07, + "loss": 0.0627, + "step": 2588 + }, + { + "epoch": 0.5884090909090909, + "grad_norm": 3.676227662314279, + "learning_rate": 4.540616250745407e-07, + "loss": 0.2498, + "step": 2589 + }, + { + "epoch": 0.5886363636363636, + "grad_norm": 3.3526848975924253, + "learning_rate": 4.536324344827163e-07, + "loss": 0.2177, + "step": 2590 + }, + { + "epoch": 0.5888636363636364, + "grad_norm": 3.2981237863069492, + "learning_rate": 4.532033312529789e-07, + "loss": 0.1352, + "step": 2591 + }, + { + "epoch": 0.5890909090909091, + "grad_norm": 3.1950260055344297, + "learning_rate": 4.527743156040829e-07, + "loss": 0.1949, + "step": 2592 + }, + { + "epoch": 0.5893181818181819, + "grad_norm": 3.1924584211075495, + "learning_rate": 4.5234538775473737e-07, + "loss": 0.0845, + "step": 2593 + }, + { + "epoch": 0.5895454545454546, + "grad_norm": 2.632721729352918, + "learning_rate": 4.519165479236072e-07, + "loss": 0.0766, + "step": 2594 + }, + { + "epoch": 0.5897727272727272, + "grad_norm": 2.4625438966796187, + "learning_rate": 4.5148779632931225e-07, + "loss": 0.08, + "step": 2595 + }, + { + "epoch": 0.59, + "grad_norm": 2.7923140665212007, + "learning_rate": 4.510591331904274e-07, + "loss": 0.0828, + "step": 2596 + }, + { + "epoch": 0.5902272727272727, + "grad_norm": 3.4572546490077563, + "learning_rate": 4.5063055872548183e-07, + "loss": 0.1391, + "step": 2597 + }, + { + "epoch": 0.5904545454545455, + "grad_norm": 3.577512067836921, + "learning_rate": 4.5020207315296065e-07, + "loss": 0.0878, + "step": 2598 + }, + { + "epoch": 0.5906818181818182, + "grad_norm": 3.918413240329226, + "learning_rate": 4.497736766913026e-07, + "loss": 0.1643, + "step": 2599 + }, + { + "epoch": 0.5909090909090909, + "grad_norm": 3.1240649682824344, + "learning_rate": 4.493453695589018e-07, + "loss": 0.1486, + "step": 2600 + }, + { + "epoch": 0.5911363636363637, + "grad_norm": 1.8794219470005211, + "learning_rate": 4.4891715197410643e-07, + "loss": 0.1034, + "step": 2601 + }, + { + "epoch": 0.5913636363636363, + "grad_norm": 2.951290214681131, + "learning_rate": 4.4848902415521897e-07, + "loss": 0.1196, + "step": 2602 + }, + { + "epoch": 0.5915909090909091, + "grad_norm": 2.9285025095897708, + "learning_rate": 4.4806098632049655e-07, + "loss": 0.1121, + "step": 2603 + }, + { + "epoch": 0.5918181818181818, + "grad_norm": 3.283615741077938, + "learning_rate": 4.476330386881496e-07, + "loss": 0.1439, + "step": 2604 + }, + { + "epoch": 0.5920454545454545, + "grad_norm": 4.123881126628777, + "learning_rate": 4.4720518147634325e-07, + "loss": 0.1464, + "step": 2605 + }, + { + "epoch": 0.5922727272727273, + "grad_norm": 3.3729810898961543, + "learning_rate": 4.4677741490319655e-07, + "loss": 0.1917, + "step": 2606 + }, + { + "epoch": 0.5925, + "grad_norm": 3.546081551688971, + "learning_rate": 4.4634973918678207e-07, + "loss": 0.2152, + "step": 2607 + }, + { + "epoch": 0.5927272727272728, + "grad_norm": 3.543128482122255, + "learning_rate": 4.4592215454512604e-07, + "loss": 0.1397, + "step": 2608 + }, + { + "epoch": 0.5929545454545454, + "grad_norm": 2.5554842691263686, + "learning_rate": 4.4549466119620864e-07, + "loss": 0.1371, + "step": 2609 + }, + { + "epoch": 0.5931818181818181, + "grad_norm": 3.656930427409667, + "learning_rate": 4.450672593579632e-07, + "loss": 0.1016, + "step": 2610 + }, + { + "epoch": 0.5934090909090909, + "grad_norm": 2.4474894104936333, + "learning_rate": 4.44639949248276e-07, + "loss": 0.1203, + "step": 2611 + }, + { + "epoch": 0.5936363636363636, + "grad_norm": 4.319909043466572, + "learning_rate": 4.4421273108498725e-07, + "loss": 0.1854, + "step": 2612 + }, + { + "epoch": 0.5938636363636364, + "grad_norm": 4.213657641178036, + "learning_rate": 4.4378560508589033e-07, + "loss": 0.1346, + "step": 2613 + }, + { + "epoch": 0.5940909090909091, + "grad_norm": 2.502038027593956, + "learning_rate": 4.433585714687309e-07, + "loss": 0.0811, + "step": 2614 + }, + { + "epoch": 0.5943181818181819, + "grad_norm": 2.7506194573777782, + "learning_rate": 4.429316304512083e-07, + "loss": 0.0833, + "step": 2615 + }, + { + "epoch": 0.5945454545454546, + "grad_norm": 2.4249797722687116, + "learning_rate": 4.4250478225097404e-07, + "loss": 0.075, + "step": 2616 + }, + { + "epoch": 0.5947727272727272, + "grad_norm": 2.81903156244215, + "learning_rate": 4.420780270856328e-07, + "loss": 0.1066, + "step": 2617 + }, + { + "epoch": 0.595, + "grad_norm": 2.480467536966794, + "learning_rate": 4.416513651727418e-07, + "loss": 0.1319, + "step": 2618 + }, + { + "epoch": 0.5952272727272727, + "grad_norm": 3.3462233305395284, + "learning_rate": 4.4122479672981016e-07, + "loss": 0.1641, + "step": 2619 + }, + { + "epoch": 0.5954545454545455, + "grad_norm": 3.8452066324640874, + "learning_rate": 4.407983219742998e-07, + "loss": 0.1371, + "step": 2620 + }, + { + "epoch": 0.5956818181818182, + "grad_norm": 2.5076015826103646, + "learning_rate": 4.4037194112362506e-07, + "loss": 0.0779, + "step": 2621 + }, + { + "epoch": 0.5959090909090909, + "grad_norm": 5.035067286805649, + "learning_rate": 4.399456543951519e-07, + "loss": 0.1867, + "step": 2622 + }, + { + "epoch": 0.5961363636363637, + "grad_norm": 2.936675993303821, + "learning_rate": 4.3951946200619865e-07, + "loss": 0.1002, + "step": 2623 + }, + { + "epoch": 0.5963636363636363, + "grad_norm": 2.7839602825630645, + "learning_rate": 4.3909336417403565e-07, + "loss": 0.1313, + "step": 2624 + }, + { + "epoch": 0.5965909090909091, + "grad_norm": 2.4164524829023777, + "learning_rate": 4.3866736111588486e-07, + "loss": 0.0919, + "step": 2625 + }, + { + "epoch": 0.5968181818181818, + "grad_norm": 2.2743169274686768, + "learning_rate": 4.3824145304891945e-07, + "loss": 0.1133, + "step": 2626 + }, + { + "epoch": 0.5970454545454545, + "grad_norm": 3.067240195581444, + "learning_rate": 4.378156401902651e-07, + "loss": 0.0795, + "step": 2627 + }, + { + "epoch": 0.5972727272727273, + "grad_norm": 5.396789918627366, + "learning_rate": 4.373899227569981e-07, + "loss": 0.1792, + "step": 2628 + }, + { + "epoch": 0.5975, + "grad_norm": 3.684509110421856, + "learning_rate": 4.3696430096614674e-07, + "loss": 0.0967, + "step": 2629 + }, + { + "epoch": 0.5977272727272728, + "grad_norm": 2.4032550856380888, + "learning_rate": 4.3653877503469025e-07, + "loss": 0.101, + "step": 2630 + }, + { + "epoch": 0.5979545454545454, + "grad_norm": 3.645959666822541, + "learning_rate": 4.3611334517955893e-07, + "loss": 0.1399, + "step": 2631 + }, + { + "epoch": 0.5981818181818181, + "grad_norm": 2.8828127720856975, + "learning_rate": 4.3568801161763445e-07, + "loss": 0.1043, + "step": 2632 + }, + { + "epoch": 0.5984090909090909, + "grad_norm": 3.908916078783658, + "learning_rate": 4.352627745657488e-07, + "loss": 0.1444, + "step": 2633 + }, + { + "epoch": 0.5986363636363636, + "grad_norm": 4.094487004855471, + "learning_rate": 4.3483763424068514e-07, + "loss": 0.1751, + "step": 2634 + }, + { + "epoch": 0.5988636363636364, + "grad_norm": 2.71646388697831, + "learning_rate": 4.3441259085917745e-07, + "loss": 0.1078, + "step": 2635 + }, + { + "epoch": 0.5990909090909091, + "grad_norm": 3.2995979021421546, + "learning_rate": 4.3398764463791e-07, + "loss": 0.12, + "step": 2636 + }, + { + "epoch": 0.5993181818181819, + "grad_norm": 1.7312700080891827, + "learning_rate": 4.335627957935178e-07, + "loss": 0.0831, + "step": 2637 + }, + { + "epoch": 0.5995454545454545, + "grad_norm": 2.5773571720295467, + "learning_rate": 4.33138044542586e-07, + "loss": 0.2587, + "step": 2638 + }, + { + "epoch": 0.5997727272727272, + "grad_norm": 2.1812811078613934, + "learning_rate": 4.3271339110165013e-07, + "loss": 0.073, + "step": 2639 + }, + { + "epoch": 0.6, + "grad_norm": 1.4404274378782835, + "learning_rate": 4.322888356871955e-07, + "loss": 0.0735, + "step": 2640 + }, + { + "epoch": 0.6002272727272727, + "grad_norm": 2.940821157959936, + "learning_rate": 4.3186437851565794e-07, + "loss": 0.0812, + "step": 2641 + }, + { + "epoch": 0.6004545454545455, + "grad_norm": 5.822334046962591, + "learning_rate": 4.314400198034231e-07, + "loss": 0.1419, + "step": 2642 + }, + { + "epoch": 0.6006818181818182, + "grad_norm": 3.0438948914785957, + "learning_rate": 4.3101575976682615e-07, + "loss": 0.1389, + "step": 2643 + }, + { + "epoch": 0.600909090909091, + "grad_norm": 2.7206941992794884, + "learning_rate": 4.305915986221524e-07, + "loss": 0.1523, + "step": 2644 + }, + { + "epoch": 0.6011363636363637, + "grad_norm": 2.8658771443437465, + "learning_rate": 4.3016753658563615e-07, + "loss": 0.1267, + "step": 2645 + }, + { + "epoch": 0.6013636363636363, + "grad_norm": 1.138538678915411, + "learning_rate": 4.297435738734616e-07, + "loss": 0.069, + "step": 2646 + }, + { + "epoch": 0.6015909090909091, + "grad_norm": 3.2523295489251693, + "learning_rate": 4.2931971070176266e-07, + "loss": 0.1165, + "step": 2647 + }, + { + "epoch": 0.6018181818181818, + "grad_norm": 4.573663695083289, + "learning_rate": 4.288959472866215e-07, + "loss": 0.207, + "step": 2648 + }, + { + "epoch": 0.6020454545454546, + "grad_norm": 2.473973269037138, + "learning_rate": 4.284722838440701e-07, + "loss": 0.1022, + "step": 2649 + }, + { + "epoch": 0.6022727272727273, + "grad_norm": 2.8433445190526943, + "learning_rate": 4.280487205900894e-07, + "loss": 0.0906, + "step": 2650 + }, + { + "epoch": 0.6025, + "grad_norm": 3.0366063445443965, + "learning_rate": 4.2762525774060953e-07, + "loss": 0.0831, + "step": 2651 + }, + { + "epoch": 0.6027272727272728, + "grad_norm": 1.8214755367073945, + "learning_rate": 4.272018955115087e-07, + "loss": 0.0819, + "step": 2652 + }, + { + "epoch": 0.6029545454545454, + "grad_norm": 3.866380134013444, + "learning_rate": 4.2677863411861456e-07, + "loss": 0.0967, + "step": 2653 + }, + { + "epoch": 0.6031818181818182, + "grad_norm": 2.8892799200432773, + "learning_rate": 4.2635547377770315e-07, + "loss": 0.0965, + "step": 2654 + }, + { + "epoch": 0.6034090909090909, + "grad_norm": 3.6294510629148835, + "learning_rate": 4.259324147044984e-07, + "loss": 0.1821, + "step": 2655 + }, + { + "epoch": 0.6036363636363636, + "grad_norm": 2.805639479344981, + "learning_rate": 4.255094571146735e-07, + "loss": 0.0791, + "step": 2656 + }, + { + "epoch": 0.6038636363636364, + "grad_norm": 2.4921308297363804, + "learning_rate": 4.250866012238495e-07, + "loss": 0.0914, + "step": 2657 + }, + { + "epoch": 0.6040909090909091, + "grad_norm": 1.5291169134448748, + "learning_rate": 4.2466384724759543e-07, + "loss": 0.0898, + "step": 2658 + }, + { + "epoch": 0.6043181818181819, + "grad_norm": 3.90814601808056, + "learning_rate": 4.242411954014289e-07, + "loss": 0.3141, + "step": 2659 + }, + { + "epoch": 0.6045454545454545, + "grad_norm": 2.7983279228074935, + "learning_rate": 4.238186459008148e-07, + "loss": 0.1188, + "step": 2660 + }, + { + "epoch": 0.6047727272727272, + "grad_norm": 2.1734194941911937, + "learning_rate": 4.233961989611664e-07, + "loss": 0.0701, + "step": 2661 + }, + { + "epoch": 0.605, + "grad_norm": 2.202187827070499, + "learning_rate": 4.229738547978442e-07, + "loss": 0.1274, + "step": 2662 + }, + { + "epoch": 0.6052272727272727, + "grad_norm": 3.096724436792838, + "learning_rate": 4.225516136261568e-07, + "loss": 0.1614, + "step": 2663 + }, + { + "epoch": 0.6054545454545455, + "grad_norm": 3.767398899101661, + "learning_rate": 4.221294756613596e-07, + "loss": 0.2128, + "step": 2664 + }, + { + "epoch": 0.6056818181818182, + "grad_norm": 1.3068359858346696, + "learning_rate": 4.217074411186564e-07, + "loss": 0.0755, + "step": 2665 + }, + { + "epoch": 0.605909090909091, + "grad_norm": 3.149543820634907, + "learning_rate": 4.212855102131974e-07, + "loss": 0.1876, + "step": 2666 + }, + { + "epoch": 0.6061363636363636, + "grad_norm": 3.050867331323404, + "learning_rate": 4.2086368316008017e-07, + "loss": 0.1528, + "step": 2667 + }, + { + "epoch": 0.6063636363636363, + "grad_norm": 3.5888615708931213, + "learning_rate": 4.204419601743499e-07, + "loss": 0.1336, + "step": 2668 + }, + { + "epoch": 0.6065909090909091, + "grad_norm": 4.528777132852598, + "learning_rate": 4.200203414709979e-07, + "loss": 0.2232, + "step": 2669 + }, + { + "epoch": 0.6068181818181818, + "grad_norm": 2.3869573267152746, + "learning_rate": 4.1959882726496245e-07, + "loss": 0.1516, + "step": 2670 + }, + { + "epoch": 0.6070454545454546, + "grad_norm": 3.0119648123590053, + "learning_rate": 4.1917741777112934e-07, + "loss": 0.1565, + "step": 2671 + }, + { + "epoch": 0.6072727272727273, + "grad_norm": 3.951084611357951, + "learning_rate": 4.1875611320433e-07, + "loss": 0.1049, + "step": 2672 + }, + { + "epoch": 0.6075, + "grad_norm": 3.217220571636264, + "learning_rate": 4.18334913779343e-07, + "loss": 0.1922, + "step": 2673 + }, + { + "epoch": 0.6077272727272728, + "grad_norm": 2.8089215317279628, + "learning_rate": 4.179138197108931e-07, + "loss": 0.2009, + "step": 2674 + }, + { + "epoch": 0.6079545454545454, + "grad_norm": 1.2060346593441855, + "learning_rate": 4.1749283121365137e-07, + "loss": 0.0895, + "step": 2675 + }, + { + "epoch": 0.6081818181818182, + "grad_norm": 3.159144286119468, + "learning_rate": 4.1707194850223515e-07, + "loss": 0.1105, + "step": 2676 + }, + { + "epoch": 0.6084090909090909, + "grad_norm": 3.571106714944236, + "learning_rate": 4.166511717912076e-07, + "loss": 0.1344, + "step": 2677 + }, + { + "epoch": 0.6086363636363636, + "grad_norm": 3.226865528795964, + "learning_rate": 4.162305012950778e-07, + "loss": 0.2004, + "step": 2678 + }, + { + "epoch": 0.6088636363636364, + "grad_norm": 3.532188270546148, + "learning_rate": 4.158099372283011e-07, + "loss": 0.0928, + "step": 2679 + }, + { + "epoch": 0.6090909090909091, + "grad_norm": 3.5460214626593074, + "learning_rate": 4.153894798052785e-07, + "loss": 0.1148, + "step": 2680 + }, + { + "epoch": 0.6093181818181819, + "grad_norm": 3.08851575986529, + "learning_rate": 4.149691292403561e-07, + "loss": 0.1214, + "step": 2681 + }, + { + "epoch": 0.6095454545454545, + "grad_norm": 4.044458197294237, + "learning_rate": 4.145488857478265e-07, + "loss": 0.1823, + "step": 2682 + }, + { + "epoch": 0.6097727272727272, + "grad_norm": 2.2013644131843235, + "learning_rate": 4.1412874954192684e-07, + "loss": 0.0677, + "step": 2683 + }, + { + "epoch": 0.61, + "grad_norm": 4.188627672712374, + "learning_rate": 4.1370872083683957e-07, + "loss": 0.1026, + "step": 2684 + }, + { + "epoch": 0.6102272727272727, + "grad_norm": 2.0722158168841998, + "learning_rate": 4.132887998466929e-07, + "loss": 0.0683, + "step": 2685 + }, + { + "epoch": 0.6104545454545455, + "grad_norm": 1.274823599794254, + "learning_rate": 4.128689867855601e-07, + "loss": 0.0619, + "step": 2686 + }, + { + "epoch": 0.6106818181818182, + "grad_norm": 4.653513318507322, + "learning_rate": 4.124492818674587e-07, + "loss": 0.1216, + "step": 2687 + }, + { + "epoch": 0.610909090909091, + "grad_norm": 2.2435454939912596, + "learning_rate": 4.12029685306352e-07, + "loss": 0.1049, + "step": 2688 + }, + { + "epoch": 0.6111363636363636, + "grad_norm": 3.55478209994119, + "learning_rate": 4.1161019731614736e-07, + "loss": 0.0947, + "step": 2689 + }, + { + "epoch": 0.6113636363636363, + "grad_norm": 2.3065017352920307, + "learning_rate": 4.1119081811069743e-07, + "loss": 0.1029, + "step": 2690 + }, + { + "epoch": 0.6115909090909091, + "grad_norm": 2.6286153949980524, + "learning_rate": 4.107715479037985e-07, + "loss": 0.1221, + "step": 2691 + }, + { + "epoch": 0.6118181818181818, + "grad_norm": 1.3421084951071496, + "learning_rate": 4.103523869091923e-07, + "loss": 0.0621, + "step": 2692 + }, + { + "epoch": 0.6120454545454546, + "grad_norm": 4.555481753826526, + "learning_rate": 4.0993333534056404e-07, + "loss": 0.2981, + "step": 2693 + }, + { + "epoch": 0.6122727272727273, + "grad_norm": 5.645083041990199, + "learning_rate": 4.095143934115439e-07, + "loss": 0.1609, + "step": 2694 + }, + { + "epoch": 0.6125, + "grad_norm": 3.6895143549599574, + "learning_rate": 4.090955613357055e-07, + "loss": 0.1521, + "step": 2695 + }, + { + "epoch": 0.6127272727272727, + "grad_norm": 4.7586238028456265, + "learning_rate": 4.086768393265668e-07, + "loss": 0.1167, + "step": 2696 + }, + { + "epoch": 0.6129545454545454, + "grad_norm": 3.9966367032941204, + "learning_rate": 4.0825822759759006e-07, + "loss": 0.1404, + "step": 2697 + }, + { + "epoch": 0.6131818181818182, + "grad_norm": 2.5113271979206484, + "learning_rate": 4.078397263621802e-07, + "loss": 0.1022, + "step": 2698 + }, + { + "epoch": 0.6134090909090909, + "grad_norm": 3.6475280115774993, + "learning_rate": 4.074213358336867e-07, + "loss": 0.1701, + "step": 2699 + }, + { + "epoch": 0.6136363636363636, + "grad_norm": 1.311048338491556, + "learning_rate": 4.070030562254025e-07, + "loss": 0.0947, + "step": 2700 + }, + { + "epoch": 0.6138636363636364, + "grad_norm": 3.5906313003744827, + "learning_rate": 4.0658488775056365e-07, + "loss": 0.1478, + "step": 2701 + }, + { + "epoch": 0.6140909090909091, + "grad_norm": 2.736828944832799, + "learning_rate": 4.0616683062234987e-07, + "loss": 0.1558, + "step": 2702 + }, + { + "epoch": 0.6143181818181818, + "grad_norm": 2.884396689370658, + "learning_rate": 4.05748885053884e-07, + "loss": 0.2059, + "step": 2703 + }, + { + "epoch": 0.6145454545454545, + "grad_norm": 3.4067074450984665, + "learning_rate": 4.0533105125823244e-07, + "loss": 0.1298, + "step": 2704 + }, + { + "epoch": 0.6147727272727272, + "grad_norm": 1.3538978857827164, + "learning_rate": 4.049133294484033e-07, + "loss": 0.0445, + "step": 2705 + }, + { + "epoch": 0.615, + "grad_norm": 3.1766984861514582, + "learning_rate": 4.044957198373492e-07, + "loss": 0.081, + "step": 2706 + }, + { + "epoch": 0.6152272727272727, + "grad_norm": 1.67179167254242, + "learning_rate": 4.040782226379644e-07, + "loss": 0.0641, + "step": 2707 + }, + { + "epoch": 0.6154545454545455, + "grad_norm": 3.0683183990563343, + "learning_rate": 4.036608380630865e-07, + "loss": 0.0934, + "step": 2708 + }, + { + "epoch": 0.6156818181818182, + "grad_norm": 2.756933167550968, + "learning_rate": 4.032435663254956e-07, + "loss": 0.0993, + "step": 2709 + }, + { + "epoch": 0.615909090909091, + "grad_norm": 5.02268295665848, + "learning_rate": 4.02826407637914e-07, + "loss": 0.2669, + "step": 2710 + }, + { + "epoch": 0.6161363636363636, + "grad_norm": 3.372115627740141, + "learning_rate": 4.024093622130067e-07, + "loss": 0.1901, + "step": 2711 + }, + { + "epoch": 0.6163636363636363, + "grad_norm": 2.8392820202091986, + "learning_rate": 4.0199243026338065e-07, + "loss": 0.0907, + "step": 2712 + }, + { + "epoch": 0.6165909090909091, + "grad_norm": 3.3380295224133683, + "learning_rate": 4.0157561200158484e-07, + "loss": 0.099, + "step": 2713 + }, + { + "epoch": 0.6168181818181818, + "grad_norm": 2.0725529586665936, + "learning_rate": 4.011589076401109e-07, + "loss": 0.0903, + "step": 2714 + }, + { + "epoch": 0.6170454545454546, + "grad_norm": 4.585225706476802, + "learning_rate": 4.0074231739139197e-07, + "loss": 0.2776, + "step": 2715 + }, + { + "epoch": 0.6172727272727273, + "grad_norm": 3.1454859906798247, + "learning_rate": 4.0032584146780305e-07, + "loss": 0.0722, + "step": 2716 + }, + { + "epoch": 0.6175, + "grad_norm": 3.181350389219599, + "learning_rate": 3.9990948008166097e-07, + "loss": 0.1865, + "step": 2717 + }, + { + "epoch": 0.6177272727272727, + "grad_norm": 4.475820807783762, + "learning_rate": 3.994932334452239e-07, + "loss": 0.1037, + "step": 2718 + }, + { + "epoch": 0.6179545454545454, + "grad_norm": 3.149265959415668, + "learning_rate": 3.990771017706921e-07, + "loss": 0.0861, + "step": 2719 + }, + { + "epoch": 0.6181818181818182, + "grad_norm": 1.5458532265611595, + "learning_rate": 3.986610852702064e-07, + "loss": 0.1053, + "step": 2720 + }, + { + "epoch": 0.6184090909090909, + "grad_norm": 3.133902509009137, + "learning_rate": 3.9824518415584947e-07, + "loss": 0.0941, + "step": 2721 + }, + { + "epoch": 0.6186363636363637, + "grad_norm": 5.124248021705049, + "learning_rate": 3.978293986396452e-07, + "loss": 0.15, + "step": 2722 + }, + { + "epoch": 0.6188636363636364, + "grad_norm": 2.0996598799292983, + "learning_rate": 3.974137289335581e-07, + "loss": 0.0944, + "step": 2723 + }, + { + "epoch": 0.6190909090909091, + "grad_norm": 2.8394799829246407, + "learning_rate": 3.9699817524949426e-07, + "loss": 0.1719, + "step": 2724 + }, + { + "epoch": 0.6193181818181818, + "grad_norm": 2.2873699948937642, + "learning_rate": 3.965827377993e-07, + "loss": 0.0728, + "step": 2725 + }, + { + "epoch": 0.6195454545454545, + "grad_norm": 5.605730338179308, + "learning_rate": 3.961674167947631e-07, + "loss": 0.1208, + "step": 2726 + }, + { + "epoch": 0.6197727272727273, + "grad_norm": 3.150306951714443, + "learning_rate": 3.9575221244761097e-07, + "loss": 0.0868, + "step": 2727 + }, + { + "epoch": 0.62, + "grad_norm": 3.343326138916391, + "learning_rate": 3.9533712496951247e-07, + "loss": 0.164, + "step": 2728 + }, + { + "epoch": 0.6202272727272727, + "grad_norm": 3.908174891735918, + "learning_rate": 3.9492215457207634e-07, + "loss": 0.1831, + "step": 2729 + }, + { + "epoch": 0.6204545454545455, + "grad_norm": 1.7766895371413043, + "learning_rate": 3.9450730146685206e-07, + "loss": 0.0877, + "step": 2730 + }, + { + "epoch": 0.6206818181818182, + "grad_norm": 5.021409712949322, + "learning_rate": 3.9409256586532885e-07, + "loss": 0.2834, + "step": 2731 + }, + { + "epoch": 0.6209090909090909, + "grad_norm": 7.06780739510605, + "learning_rate": 3.936779479789366e-07, + "loss": 0.1579, + "step": 2732 + }, + { + "epoch": 0.6211363636363636, + "grad_norm": 2.062437511067936, + "learning_rate": 3.9326344801904477e-07, + "loss": 0.0611, + "step": 2733 + }, + { + "epoch": 0.6213636363636363, + "grad_norm": 3.961094910834728, + "learning_rate": 3.9284906619696235e-07, + "loss": 0.1219, + "step": 2734 + }, + { + "epoch": 0.6215909090909091, + "grad_norm": 3.4558677502376893, + "learning_rate": 3.9243480272393887e-07, + "loss": 0.1254, + "step": 2735 + }, + { + "epoch": 0.6218181818181818, + "grad_norm": 2.397888263977544, + "learning_rate": 3.920206578111633e-07, + "loss": 0.0597, + "step": 2736 + }, + { + "epoch": 0.6220454545454546, + "grad_norm": 3.2869699597899715, + "learning_rate": 3.916066316697638e-07, + "loss": 0.1775, + "step": 2737 + }, + { + "epoch": 0.6222727272727273, + "grad_norm": 3.465460088846853, + "learning_rate": 3.9119272451080854e-07, + "loss": 0.1225, + "step": 2738 + }, + { + "epoch": 0.6225, + "grad_norm": 2.5985487502571947, + "learning_rate": 3.9077893654530437e-07, + "loss": 0.1537, + "step": 2739 + }, + { + "epoch": 0.6227272727272727, + "grad_norm": 2.0225429273915387, + "learning_rate": 3.90365267984198e-07, + "loss": 0.073, + "step": 2740 + }, + { + "epoch": 0.6229545454545454, + "grad_norm": 1.891246658693488, + "learning_rate": 3.8995171903837526e-07, + "loss": 0.0749, + "step": 2741 + }, + { + "epoch": 0.6231818181818182, + "grad_norm": 2.963582218339465, + "learning_rate": 3.8953828991866024e-07, + "loss": 0.0775, + "step": 2742 + }, + { + "epoch": 0.6234090909090909, + "grad_norm": 3.3985643844537985, + "learning_rate": 3.8912498083581664e-07, + "loss": 0.1446, + "step": 2743 + }, + { + "epoch": 0.6236363636363637, + "grad_norm": 4.158283960969452, + "learning_rate": 3.887117920005469e-07, + "loss": 0.1573, + "step": 2744 + }, + { + "epoch": 0.6238636363636364, + "grad_norm": 2.940859867646085, + "learning_rate": 3.882987236234918e-07, + "loss": 0.1476, + "step": 2745 + }, + { + "epoch": 0.6240909090909091, + "grad_norm": 4.384849136601549, + "learning_rate": 3.8788577591523116e-07, + "loss": 0.1048, + "step": 2746 + }, + { + "epoch": 0.6243181818181818, + "grad_norm": 4.151887339741491, + "learning_rate": 3.874729490862832e-07, + "loss": 0.1586, + "step": 2747 + }, + { + "epoch": 0.6245454545454545, + "grad_norm": 3.6120762924965595, + "learning_rate": 3.8706024334710426e-07, + "loss": 0.0966, + "step": 2748 + }, + { + "epoch": 0.6247727272727273, + "grad_norm": 2.4069636651701978, + "learning_rate": 3.8664765890808884e-07, + "loss": 0.1232, + "step": 2749 + }, + { + "epoch": 0.625, + "grad_norm": 2.5899234130186954, + "learning_rate": 3.8623519597957013e-07, + "loss": 0.1552, + "step": 2750 + }, + { + "epoch": 0.6252272727272727, + "grad_norm": 2.9387927525154884, + "learning_rate": 3.8582285477181894e-07, + "loss": 0.0767, + "step": 2751 + }, + { + "epoch": 0.6254545454545455, + "grad_norm": 3.430941223654626, + "learning_rate": 3.8541063549504426e-07, + "loss": 0.0854, + "step": 2752 + }, + { + "epoch": 0.6256818181818182, + "grad_norm": 4.017683126216733, + "learning_rate": 3.849985383593929e-07, + "loss": 0.1208, + "step": 2753 + }, + { + "epoch": 0.6259090909090909, + "grad_norm": 2.1790094788399794, + "learning_rate": 3.8458656357494913e-07, + "loss": 0.0483, + "step": 2754 + }, + { + "epoch": 0.6261363636363636, + "grad_norm": 2.1036398829454646, + "learning_rate": 3.8417471135173543e-07, + "loss": 0.0684, + "step": 2755 + }, + { + "epoch": 0.6263636363636363, + "grad_norm": 3.9832691009326378, + "learning_rate": 3.837629818997111e-07, + "loss": 0.2192, + "step": 2756 + }, + { + "epoch": 0.6265909090909091, + "grad_norm": 2.0643388873492756, + "learning_rate": 3.8335137542877307e-07, + "loss": 0.0737, + "step": 2757 + }, + { + "epoch": 0.6268181818181818, + "grad_norm": 3.567851973288697, + "learning_rate": 3.829398921487559e-07, + "loss": 0.1192, + "step": 2758 + }, + { + "epoch": 0.6270454545454546, + "grad_norm": 2.8744564001626474, + "learning_rate": 3.825285322694312e-07, + "loss": 0.1292, + "step": 2759 + }, + { + "epoch": 0.6272727272727273, + "grad_norm": 1.7886117066789569, + "learning_rate": 3.821172960005074e-07, + "loss": 0.1172, + "step": 2760 + }, + { + "epoch": 0.6275, + "grad_norm": 3.3206126902616697, + "learning_rate": 3.817061835516304e-07, + "loss": 0.1123, + "step": 2761 + }, + { + "epoch": 0.6277272727272727, + "grad_norm": 3.039078272011314, + "learning_rate": 3.8129519513238256e-07, + "loss": 0.1051, + "step": 2762 + }, + { + "epoch": 0.6279545454545454, + "grad_norm": 3.8588636099528637, + "learning_rate": 3.8088433095228287e-07, + "loss": 0.0756, + "step": 2763 + }, + { + "epoch": 0.6281818181818182, + "grad_norm": 2.4818309135516663, + "learning_rate": 3.804735912207874e-07, + "loss": 0.1598, + "step": 2764 + }, + { + "epoch": 0.6284090909090909, + "grad_norm": 2.804011369631731, + "learning_rate": 3.80062976147289e-07, + "loss": 0.1278, + "step": 2765 + }, + { + "epoch": 0.6286363636363637, + "grad_norm": 2.0809320248575958, + "learning_rate": 3.79652485941116e-07, + "loss": 0.0897, + "step": 2766 + }, + { + "epoch": 0.6288636363636364, + "grad_norm": 3.3651823079778906, + "learning_rate": 3.7924212081153434e-07, + "loss": 0.0995, + "step": 2767 + }, + { + "epoch": 0.6290909090909091, + "grad_norm": 1.7472191933926633, + "learning_rate": 3.7883188096774494e-07, + "loss": 0.117, + "step": 2768 + }, + { + "epoch": 0.6293181818181818, + "grad_norm": 3.4682601878826977, + "learning_rate": 3.7842176661888565e-07, + "loss": 0.1012, + "step": 2769 + }, + { + "epoch": 0.6295454545454545, + "grad_norm": 2.8495564865418337, + "learning_rate": 3.780117779740305e-07, + "loss": 0.0721, + "step": 2770 + }, + { + "epoch": 0.6297727272727273, + "grad_norm": 2.879886056796494, + "learning_rate": 3.7760191524218864e-07, + "loss": 0.1049, + "step": 2771 + }, + { + "epoch": 0.63, + "grad_norm": 2.8959323834268806, + "learning_rate": 3.7719217863230545e-07, + "loss": 0.1082, + "step": 2772 + }, + { + "epoch": 0.6302272727272727, + "grad_norm": 3.219899601439955, + "learning_rate": 3.767825683532622e-07, + "loss": 0.0938, + "step": 2773 + }, + { + "epoch": 0.6304545454545455, + "grad_norm": 4.240586911879049, + "learning_rate": 3.7637308461387553e-07, + "loss": 0.0979, + "step": 2774 + }, + { + "epoch": 0.6306818181818182, + "grad_norm": 5.561414717429338, + "learning_rate": 3.7596372762289766e-07, + "loss": 0.1657, + "step": 2775 + }, + { + "epoch": 0.6309090909090909, + "grad_norm": 3.901434734755181, + "learning_rate": 3.755544975890162e-07, + "loss": 0.1063, + "step": 2776 + }, + { + "epoch": 0.6311363636363636, + "grad_norm": 4.215557378552669, + "learning_rate": 3.751453947208541e-07, + "loss": 0.0997, + "step": 2777 + }, + { + "epoch": 0.6313636363636363, + "grad_norm": 3.6082667480906574, + "learning_rate": 3.7473641922696895e-07, + "loss": 0.2252, + "step": 2778 + }, + { + "epoch": 0.6315909090909091, + "grad_norm": 4.723510379234352, + "learning_rate": 3.743275713158542e-07, + "loss": 0.0893, + "step": 2779 + }, + { + "epoch": 0.6318181818181818, + "grad_norm": 2.617876782243539, + "learning_rate": 3.739188511959378e-07, + "loss": 0.1019, + "step": 2780 + }, + { + "epoch": 0.6320454545454546, + "grad_norm": 4.567485135176131, + "learning_rate": 3.7351025907558266e-07, + "loss": 0.1838, + "step": 2781 + }, + { + "epoch": 0.6322727272727273, + "grad_norm": 2.7070294268112485, + "learning_rate": 3.7310179516308653e-07, + "loss": 0.1552, + "step": 2782 + }, + { + "epoch": 0.6325, + "grad_norm": 2.270599756650333, + "learning_rate": 3.7269345966668135e-07, + "loss": 0.0961, + "step": 2783 + }, + { + "epoch": 0.6327272727272727, + "grad_norm": 3.492846714472299, + "learning_rate": 3.7228525279453447e-07, + "loss": 0.1585, + "step": 2784 + }, + { + "epoch": 0.6329545454545454, + "grad_norm": 2.414487883844887, + "learning_rate": 3.7187717475474673e-07, + "loss": 0.1688, + "step": 2785 + }, + { + "epoch": 0.6331818181818182, + "grad_norm": 2.7741005267934016, + "learning_rate": 3.7146922575535383e-07, + "loss": 0.121, + "step": 2786 + }, + { + "epoch": 0.6334090909090909, + "grad_norm": 3.0728909242563467, + "learning_rate": 3.710614060043255e-07, + "loss": 0.0897, + "step": 2787 + }, + { + "epoch": 0.6336363636363637, + "grad_norm": 2.9165788336492913, + "learning_rate": 3.706537157095658e-07, + "loss": 0.1093, + "step": 2788 + }, + { + "epoch": 0.6338636363636364, + "grad_norm": 2.8937399248515168, + "learning_rate": 3.702461550789123e-07, + "loss": 0.0783, + "step": 2789 + }, + { + "epoch": 0.634090909090909, + "grad_norm": 4.013708519632064, + "learning_rate": 3.6983872432013707e-07, + "loss": 0.133, + "step": 2790 + }, + { + "epoch": 0.6343181818181818, + "grad_norm": 2.5088466042443653, + "learning_rate": 3.694314236409459e-07, + "loss": 0.068, + "step": 2791 + }, + { + "epoch": 0.6345454545454545, + "grad_norm": 2.540325635981351, + "learning_rate": 3.690242532489777e-07, + "loss": 0.1164, + "step": 2792 + }, + { + "epoch": 0.6347727272727273, + "grad_norm": 2.4019016525040158, + "learning_rate": 3.6861721335180526e-07, + "loss": 0.1229, + "step": 2793 + }, + { + "epoch": 0.635, + "grad_norm": 2.9859892212686523, + "learning_rate": 3.682103041569352e-07, + "loss": 0.1152, + "step": 2794 + }, + { + "epoch": 0.6352272727272728, + "grad_norm": 2.8429555321480375, + "learning_rate": 3.6780352587180705e-07, + "loss": 0.1755, + "step": 2795 + }, + { + "epoch": 0.6354545454545455, + "grad_norm": 4.077812367764054, + "learning_rate": 3.673968787037938e-07, + "loss": 0.204, + "step": 2796 + }, + { + "epoch": 0.6356818181818182, + "grad_norm": 5.476535509737428, + "learning_rate": 3.669903628602018e-07, + "loss": 0.181, + "step": 2797 + }, + { + "epoch": 0.6359090909090909, + "grad_norm": 2.5824447203673886, + "learning_rate": 3.6658397854826984e-07, + "loss": 0.1081, + "step": 2798 + }, + { + "epoch": 0.6361363636363636, + "grad_norm": 2.619781045293545, + "learning_rate": 3.661777259751705e-07, + "loss": 0.1067, + "step": 2799 + }, + { + "epoch": 0.6363636363636364, + "grad_norm": 4.680472326102747, + "learning_rate": 3.657716053480084e-07, + "loss": 0.3032, + "step": 2800 + }, + { + "epoch": 0.6365909090909091, + "grad_norm": 2.943630214307541, + "learning_rate": 3.653656168738211e-07, + "loss": 0.1076, + "step": 2801 + }, + { + "epoch": 0.6368181818181818, + "grad_norm": 3.5447925040716974, + "learning_rate": 3.649597607595791e-07, + "loss": 0.092, + "step": 2802 + }, + { + "epoch": 0.6370454545454546, + "grad_norm": 4.305167461226075, + "learning_rate": 3.6455403721218546e-07, + "loss": 0.1688, + "step": 2803 + }, + { + "epoch": 0.6372727272727273, + "grad_norm": 1.2394342286886386, + "learning_rate": 3.641484464384752e-07, + "loss": 0.0509, + "step": 2804 + }, + { + "epoch": 0.6375, + "grad_norm": 2.4868092083466284, + "learning_rate": 3.637429886452158e-07, + "loss": 0.1138, + "step": 2805 + }, + { + "epoch": 0.6377272727272727, + "grad_norm": 2.8487691465709926, + "learning_rate": 3.6333766403910764e-07, + "loss": 0.1212, + "step": 2806 + }, + { + "epoch": 0.6379545454545454, + "grad_norm": 3.56016589567312, + "learning_rate": 3.6293247282678193e-07, + "loss": 0.1192, + "step": 2807 + }, + { + "epoch": 0.6381818181818182, + "grad_norm": 2.4491109380620384, + "learning_rate": 3.625274152148026e-07, + "loss": 0.087, + "step": 2808 + }, + { + "epoch": 0.6384090909090909, + "grad_norm": 3.6504158849261414, + "learning_rate": 3.621224914096658e-07, + "loss": 0.1084, + "step": 2809 + }, + { + "epoch": 0.6386363636363637, + "grad_norm": 2.7131235239527474, + "learning_rate": 3.617177016177989e-07, + "loss": 0.1736, + "step": 2810 + }, + { + "epoch": 0.6388636363636364, + "grad_norm": 4.372590811882181, + "learning_rate": 3.6131304604556093e-07, + "loss": 0.2304, + "step": 2811 + }, + { + "epoch": 0.639090909090909, + "grad_norm": 2.314755751148958, + "learning_rate": 3.6090852489924317e-07, + "loss": 0.2025, + "step": 2812 + }, + { + "epoch": 0.6393181818181818, + "grad_norm": 4.25838810179443, + "learning_rate": 3.6050413838506796e-07, + "loss": 0.1597, + "step": 2813 + }, + { + "epoch": 0.6395454545454545, + "grad_norm": 3.2718846930127867, + "learning_rate": 3.6009988670918814e-07, + "loss": 0.1286, + "step": 2814 + }, + { + "epoch": 0.6397727272727273, + "grad_norm": 4.211646293903218, + "learning_rate": 3.596957700776894e-07, + "loss": 0.1107, + "step": 2815 + }, + { + "epoch": 0.64, + "grad_norm": 4.099784768285115, + "learning_rate": 3.5929178869658757e-07, + "loss": 0.113, + "step": 2816 + }, + { + "epoch": 0.6402272727272728, + "grad_norm": 3.327707779764585, + "learning_rate": 3.588879427718296e-07, + "loss": 0.1253, + "step": 2817 + }, + { + "epoch": 0.6404545454545455, + "grad_norm": 3.287525518424316, + "learning_rate": 3.584842325092938e-07, + "loss": 0.0677, + "step": 2818 + }, + { + "epoch": 0.6406818181818181, + "grad_norm": 2.467764525339016, + "learning_rate": 3.5808065811478916e-07, + "loss": 0.0944, + "step": 2819 + }, + { + "epoch": 0.6409090909090909, + "grad_norm": 4.0639408345713335, + "learning_rate": 3.5767721979405504e-07, + "loss": 0.154, + "step": 2820 + }, + { + "epoch": 0.6411363636363636, + "grad_norm": 2.507372992540373, + "learning_rate": 3.5727391775276193e-07, + "loss": 0.0987, + "step": 2821 + }, + { + "epoch": 0.6413636363636364, + "grad_norm": 3.0598520805270626, + "learning_rate": 3.5687075219651034e-07, + "loss": 0.0879, + "step": 2822 + }, + { + "epoch": 0.6415909090909091, + "grad_norm": 2.5572429839051494, + "learning_rate": 3.5646772333083155e-07, + "loss": 0.1535, + "step": 2823 + }, + { + "epoch": 0.6418181818181818, + "grad_norm": 3.3596201796450114, + "learning_rate": 3.5606483136118753e-07, + "loss": 0.1383, + "step": 2824 + }, + { + "epoch": 0.6420454545454546, + "grad_norm": 2.26421781208023, + "learning_rate": 3.5566207649296953e-07, + "loss": 0.0962, + "step": 2825 + }, + { + "epoch": 0.6422727272727272, + "grad_norm": 2.4724260165100898, + "learning_rate": 3.552594589314995e-07, + "loss": 0.1086, + "step": 2826 + }, + { + "epoch": 0.6425, + "grad_norm": 7.218668335917696, + "learning_rate": 3.5485697888202985e-07, + "loss": 0.1983, + "step": 2827 + }, + { + "epoch": 0.6427272727272727, + "grad_norm": 3.190412309842298, + "learning_rate": 3.5445463654974157e-07, + "loss": 0.2425, + "step": 2828 + }, + { + "epoch": 0.6429545454545454, + "grad_norm": 1.5757660655460755, + "learning_rate": 3.540524321397462e-07, + "loss": 0.0741, + "step": 2829 + }, + { + "epoch": 0.6431818181818182, + "grad_norm": 2.669323451001057, + "learning_rate": 3.536503658570856e-07, + "loss": 0.0519, + "step": 2830 + }, + { + "epoch": 0.6434090909090909, + "grad_norm": 2.7451667037674814, + "learning_rate": 3.532484379067301e-07, + "loss": 0.0684, + "step": 2831 + }, + { + "epoch": 0.6436363636363637, + "grad_norm": 2.891941565840157, + "learning_rate": 3.5284664849358e-07, + "loss": 0.0867, + "step": 2832 + }, + { + "epoch": 0.6438636363636364, + "grad_norm": 5.532342043926671, + "learning_rate": 3.524449978224654e-07, + "loss": 0.127, + "step": 2833 + }, + { + "epoch": 0.644090909090909, + "grad_norm": 2.7879574502408757, + "learning_rate": 3.520434860981449e-07, + "loss": 0.1326, + "step": 2834 + }, + { + "epoch": 0.6443181818181818, + "grad_norm": 3.8568289999957277, + "learning_rate": 3.516421135253066e-07, + "loss": 0.1409, + "step": 2835 + }, + { + "epoch": 0.6445454545454545, + "grad_norm": 3.6708665398541958, + "learning_rate": 3.5124088030856783e-07, + "loss": 0.0856, + "step": 2836 + }, + { + "epoch": 0.6447727272727273, + "grad_norm": 2.2655867134751495, + "learning_rate": 3.508397866524746e-07, + "loss": 0.1145, + "step": 2837 + }, + { + "epoch": 0.645, + "grad_norm": 3.475980316437763, + "learning_rate": 3.504388327615016e-07, + "loss": 0.1548, + "step": 2838 + }, + { + "epoch": 0.6452272727272728, + "grad_norm": 3.8806472910524064, + "learning_rate": 3.50038018840053e-07, + "loss": 0.2024, + "step": 2839 + }, + { + "epoch": 0.6454545454545455, + "grad_norm": 2.9362770744811675, + "learning_rate": 3.4963734509246117e-07, + "loss": 0.2153, + "step": 2840 + }, + { + "epoch": 0.6456818181818181, + "grad_norm": 3.259656933693064, + "learning_rate": 3.4923681172298656e-07, + "loss": 0.0912, + "step": 2841 + }, + { + "epoch": 0.6459090909090909, + "grad_norm": 2.13774610860356, + "learning_rate": 3.4883641893581933e-07, + "loss": 0.0855, + "step": 2842 + }, + { + "epoch": 0.6461363636363636, + "grad_norm": 4.1229835170107005, + "learning_rate": 3.484361669350764e-07, + "loss": 0.2333, + "step": 2843 + }, + { + "epoch": 0.6463636363636364, + "grad_norm": 2.767727211552175, + "learning_rate": 3.480360559248036e-07, + "loss": 0.1031, + "step": 2844 + }, + { + "epoch": 0.6465909090909091, + "grad_norm": 2.5395989144351496, + "learning_rate": 3.4763608610897566e-07, + "loss": 0.0769, + "step": 2845 + }, + { + "epoch": 0.6468181818181818, + "grad_norm": 4.53764930828373, + "learning_rate": 3.472362576914942e-07, + "loss": 0.1031, + "step": 2846 + }, + { + "epoch": 0.6470454545454546, + "grad_norm": 1.012002793111704, + "learning_rate": 3.46836570876189e-07, + "loss": 0.0389, + "step": 2847 + }, + { + "epoch": 0.6472727272727272, + "grad_norm": 5.4135513556797985, + "learning_rate": 3.464370258668183e-07, + "loss": 0.1271, + "step": 2848 + }, + { + "epoch": 0.6475, + "grad_norm": 2.7659131076592853, + "learning_rate": 3.460376228670677e-07, + "loss": 0.2158, + "step": 2849 + }, + { + "epoch": 0.6477272727272727, + "grad_norm": 5.112126550305872, + "learning_rate": 3.456383620805494e-07, + "loss": 0.1639, + "step": 2850 + }, + { + "epoch": 0.6479545454545454, + "grad_norm": 1.9467876185356563, + "learning_rate": 3.4523924371080474e-07, + "loss": 0.0747, + "step": 2851 + }, + { + "epoch": 0.6481818181818182, + "grad_norm": 2.680320061663129, + "learning_rate": 3.448402679613016e-07, + "loss": 0.075, + "step": 2852 + }, + { + "epoch": 0.6484090909090909, + "grad_norm": 2.2194237895596243, + "learning_rate": 3.444414350354348e-07, + "loss": 0.1583, + "step": 2853 + }, + { + "epoch": 0.6486363636363637, + "grad_norm": 1.9664113007788955, + "learning_rate": 3.440427451365276e-07, + "loss": 0.126, + "step": 2854 + }, + { + "epoch": 0.6488636363636363, + "grad_norm": 4.9799449606071455, + "learning_rate": 3.436441984678291e-07, + "loss": 0.1989, + "step": 2855 + }, + { + "epoch": 0.649090909090909, + "grad_norm": 2.9642907686810736, + "learning_rate": 3.432457952325159e-07, + "loss": 0.1274, + "step": 2856 + }, + { + "epoch": 0.6493181818181818, + "grad_norm": 3.419967551961556, + "learning_rate": 3.428475356336913e-07, + "loss": 0.1398, + "step": 2857 + }, + { + "epoch": 0.6495454545454545, + "grad_norm": 1.8320157868028617, + "learning_rate": 3.424494198743857e-07, + "loss": 0.1262, + "step": 2858 + }, + { + "epoch": 0.6497727272727273, + "grad_norm": 2.1776434553081176, + "learning_rate": 3.420514481575556e-07, + "loss": 0.1377, + "step": 2859 + }, + { + "epoch": 0.65, + "grad_norm": 2.8463013658084972, + "learning_rate": 3.4165362068608493e-07, + "loss": 0.1998, + "step": 2860 + }, + { + "epoch": 0.6502272727272728, + "grad_norm": 2.479638890149881, + "learning_rate": 3.4125593766278333e-07, + "loss": 0.1504, + "step": 2861 + }, + { + "epoch": 0.6504545454545455, + "grad_norm": 3.838779017844332, + "learning_rate": 3.4085839929038705e-07, + "loss": 0.1903, + "step": 2862 + }, + { + "epoch": 0.6506818181818181, + "grad_norm": 2.689955488659383, + "learning_rate": 3.404610057715584e-07, + "loss": 0.1948, + "step": 2863 + }, + { + "epoch": 0.6509090909090909, + "grad_norm": 2.824836999337908, + "learning_rate": 3.400637573088868e-07, + "loss": 0.1641, + "step": 2864 + }, + { + "epoch": 0.6511363636363636, + "grad_norm": 2.899711623384414, + "learning_rate": 3.39666654104886e-07, + "loss": 0.0797, + "step": 2865 + }, + { + "epoch": 0.6513636363636364, + "grad_norm": 2.5257556886008423, + "learning_rate": 3.392696963619972e-07, + "loss": 0.1372, + "step": 2866 + }, + { + "epoch": 0.6515909090909091, + "grad_norm": 2.6067257423946617, + "learning_rate": 3.3887288428258697e-07, + "loss": 0.077, + "step": 2867 + }, + { + "epoch": 0.6518181818181819, + "grad_norm": 3.520342813128682, + "learning_rate": 3.384762180689474e-07, + "loss": 0.1917, + "step": 2868 + }, + { + "epoch": 0.6520454545454546, + "grad_norm": 2.787220917076655, + "learning_rate": 3.380796979232961e-07, + "loss": 0.0873, + "step": 2869 + }, + { + "epoch": 0.6522727272727272, + "grad_norm": 5.319529066222058, + "learning_rate": 3.3768332404777724e-07, + "loss": 0.2554, + "step": 2870 + }, + { + "epoch": 0.6525, + "grad_norm": 2.6701330369129925, + "learning_rate": 3.372870966444591e-07, + "loss": 0.0733, + "step": 2871 + }, + { + "epoch": 0.6527272727272727, + "grad_norm": 4.1173072777170505, + "learning_rate": 3.3689101591533617e-07, + "loss": 0.1261, + "step": 2872 + }, + { + "epoch": 0.6529545454545455, + "grad_norm": 3.3998032457904377, + "learning_rate": 3.364950820623276e-07, + "loss": 0.1737, + "step": 2873 + }, + { + "epoch": 0.6531818181818182, + "grad_norm": 1.8080012756746988, + "learning_rate": 3.360992952872781e-07, + "loss": 0.1135, + "step": 2874 + }, + { + "epoch": 0.6534090909090909, + "grad_norm": 3.780868860727512, + "learning_rate": 3.35703655791957e-07, + "loss": 0.13, + "step": 2875 + }, + { + "epoch": 0.6536363636363637, + "grad_norm": 2.4982626848144625, + "learning_rate": 3.353081637780591e-07, + "loss": 0.1473, + "step": 2876 + }, + { + "epoch": 0.6538636363636363, + "grad_norm": 5.981263797029599, + "learning_rate": 3.3491281944720373e-07, + "loss": 0.1416, + "step": 2877 + }, + { + "epoch": 0.6540909090909091, + "grad_norm": 3.182172386280394, + "learning_rate": 3.345176230009346e-07, + "loss": 0.1041, + "step": 2878 + }, + { + "epoch": 0.6543181818181818, + "grad_norm": 3.7677093575730116, + "learning_rate": 3.341225746407205e-07, + "loss": 0.1295, + "step": 2879 + }, + { + "epoch": 0.6545454545454545, + "grad_norm": 3.0093824328581857, + "learning_rate": 3.337276745679545e-07, + "loss": 0.0811, + "step": 2880 + }, + { + "epoch": 0.6547727272727273, + "grad_norm": 4.8095804985838395, + "learning_rate": 3.3333292298395396e-07, + "loss": 0.1961, + "step": 2881 + }, + { + "epoch": 0.655, + "grad_norm": 2.8651425310223697, + "learning_rate": 3.3293832008996105e-07, + "loss": 0.109, + "step": 2882 + }, + { + "epoch": 0.6552272727272728, + "grad_norm": 3.811397728221008, + "learning_rate": 3.325438660871418e-07, + "loss": 0.1048, + "step": 2883 + }, + { + "epoch": 0.6554545454545454, + "grad_norm": 2.90211874247013, + "learning_rate": 3.321495611765858e-07, + "loss": 0.0867, + "step": 2884 + }, + { + "epoch": 0.6556818181818181, + "grad_norm": 3.882871711775285, + "learning_rate": 3.317554055593081e-07, + "loss": 0.1629, + "step": 2885 + }, + { + "epoch": 0.6559090909090909, + "grad_norm": 5.540832086839885, + "learning_rate": 3.313613994362456e-07, + "loss": 0.3073, + "step": 2886 + }, + { + "epoch": 0.6561363636363636, + "grad_norm": 2.6914585853074344, + "learning_rate": 3.309675430082609e-07, + "loss": 0.1138, + "step": 2887 + }, + { + "epoch": 0.6563636363636364, + "grad_norm": 2.7775176602589227, + "learning_rate": 3.305738364761392e-07, + "loss": 0.2427, + "step": 2888 + }, + { + "epoch": 0.6565909090909091, + "grad_norm": 2.593995784012559, + "learning_rate": 3.301802800405897e-07, + "loss": 0.1101, + "step": 2889 + }, + { + "epoch": 0.6568181818181819, + "grad_norm": 3.5857252096482553, + "learning_rate": 3.2978687390224454e-07, + "loss": 0.0968, + "step": 2890 + }, + { + "epoch": 0.6570454545454546, + "grad_norm": 3.476364039111219, + "learning_rate": 3.293936182616602e-07, + "loss": 0.2497, + "step": 2891 + }, + { + "epoch": 0.6572727272727272, + "grad_norm": 5.225249598854452, + "learning_rate": 3.290005133193157e-07, + "loss": 0.1202, + "step": 2892 + }, + { + "epoch": 0.6575, + "grad_norm": 3.324704240146493, + "learning_rate": 3.2860755927561326e-07, + "loss": 0.0877, + "step": 2893 + }, + { + "epoch": 0.6577272727272727, + "grad_norm": 7.0786196629355524, + "learning_rate": 3.2821475633087854e-07, + "loss": 0.1688, + "step": 2894 + }, + { + "epoch": 0.6579545454545455, + "grad_norm": 1.8449456930938073, + "learning_rate": 3.278221046853598e-07, + "loss": 0.0754, + "step": 2895 + }, + { + "epoch": 0.6581818181818182, + "grad_norm": 4.702583723136906, + "learning_rate": 3.274296045392282e-07, + "loss": 0.1099, + "step": 2896 + }, + { + "epoch": 0.6584090909090909, + "grad_norm": 3.2468054726271194, + "learning_rate": 3.2703725609257826e-07, + "loss": 0.1008, + "step": 2897 + }, + { + "epoch": 0.6586363636363637, + "grad_norm": 4.783930473671077, + "learning_rate": 3.266450595454264e-07, + "loss": 0.1153, + "step": 2898 + }, + { + "epoch": 0.6588636363636363, + "grad_norm": 2.81031993446317, + "learning_rate": 3.2625301509771174e-07, + "loss": 0.1307, + "step": 2899 + }, + { + "epoch": 0.6590909090909091, + "grad_norm": 3.452814858655389, + "learning_rate": 3.258611229492967e-07, + "loss": 0.2316, + "step": 2900 + }, + { + "epoch": 0.6593181818181818, + "grad_norm": 2.5100408234852156, + "learning_rate": 3.2546938329996475e-07, + "loss": 0.0866, + "step": 2901 + }, + { + "epoch": 0.6595454545454545, + "grad_norm": 2.523989346436583, + "learning_rate": 3.2507779634942216e-07, + "loss": 0.0957, + "step": 2902 + }, + { + "epoch": 0.6597727272727273, + "grad_norm": 3.0728237717038005, + "learning_rate": 3.246863622972979e-07, + "loss": 0.1722, + "step": 2903 + }, + { + "epoch": 0.66, + "grad_norm": 4.261347916132232, + "learning_rate": 3.2429508134314237e-07, + "loss": 0.1457, + "step": 2904 + }, + { + "epoch": 0.6602272727272728, + "grad_norm": 2.450912222867795, + "learning_rate": 3.239039536864279e-07, + "loss": 0.1213, + "step": 2905 + }, + { + "epoch": 0.6604545454545454, + "grad_norm": 3.2974765270194304, + "learning_rate": 3.235129795265492e-07, + "loss": 0.1059, + "step": 2906 + }, + { + "epoch": 0.6606818181818181, + "grad_norm": 2.9008972802892625, + "learning_rate": 3.231221590628226e-07, + "loss": 0.1438, + "step": 2907 + }, + { + "epoch": 0.6609090909090909, + "grad_norm": 4.535508190410969, + "learning_rate": 3.227314924944849e-07, + "loss": 0.1224, + "step": 2908 + }, + { + "epoch": 0.6611363636363636, + "grad_norm": 2.653517698760632, + "learning_rate": 3.223409800206964e-07, + "loss": 0.1079, + "step": 2909 + }, + { + "epoch": 0.6613636363636364, + "grad_norm": 4.04638226747324, + "learning_rate": 3.2195062184053745e-07, + "loss": 0.0918, + "step": 2910 + }, + { + "epoch": 0.6615909090909091, + "grad_norm": 3.0298929509881667, + "learning_rate": 3.2156041815300995e-07, + "loss": 0.085, + "step": 2911 + }, + { + "epoch": 0.6618181818181819, + "grad_norm": 3.424000960240399, + "learning_rate": 3.2117036915703766e-07, + "loss": 0.091, + "step": 2912 + }, + { + "epoch": 0.6620454545454545, + "grad_norm": 3.568972609015836, + "learning_rate": 3.207804750514649e-07, + "loss": 0.2076, + "step": 2913 + }, + { + "epoch": 0.6622727272727272, + "grad_norm": 3.128696939219207, + "learning_rate": 3.2039073603505715e-07, + "loss": 0.1376, + "step": 2914 + }, + { + "epoch": 0.6625, + "grad_norm": 3.5722123083562867, + "learning_rate": 3.200011523065009e-07, + "loss": 0.1996, + "step": 2915 + }, + { + "epoch": 0.6627272727272727, + "grad_norm": 2.973927445697834, + "learning_rate": 3.196117240644034e-07, + "loss": 0.1963, + "step": 2916 + }, + { + "epoch": 0.6629545454545455, + "grad_norm": 5.675210170054643, + "learning_rate": 3.192224515072924e-07, + "loss": 0.2874, + "step": 2917 + }, + { + "epoch": 0.6631818181818182, + "grad_norm": 2.6045850226504204, + "learning_rate": 3.1883333483361714e-07, + "loss": 0.1619, + "step": 2918 + }, + { + "epoch": 0.663409090909091, + "grad_norm": 3.5927255223683434, + "learning_rate": 3.1844437424174645e-07, + "loss": 0.2113, + "step": 2919 + }, + { + "epoch": 0.6636363636363637, + "grad_norm": 3.05170501427222, + "learning_rate": 3.1805556992996975e-07, + "loss": 0.1014, + "step": 2920 + }, + { + "epoch": 0.6638636363636363, + "grad_norm": 2.129206008176917, + "learning_rate": 3.1766692209649756e-07, + "loss": 0.1151, + "step": 2921 + }, + { + "epoch": 0.6640909090909091, + "grad_norm": 2.4314409101478307, + "learning_rate": 3.1727843093946e-07, + "loss": 0.0864, + "step": 2922 + }, + { + "epoch": 0.6643181818181818, + "grad_norm": 1.1554410882425596, + "learning_rate": 3.168900966569066e-07, + "loss": 0.0826, + "step": 2923 + }, + { + "epoch": 0.6645454545454546, + "grad_norm": 4.336750552490636, + "learning_rate": 3.1650191944680843e-07, + "loss": 0.161, + "step": 2924 + }, + { + "epoch": 0.6647727272727273, + "grad_norm": 3.369076964052286, + "learning_rate": 3.1611389950705547e-07, + "loss": 0.0738, + "step": 2925 + }, + { + "epoch": 0.665, + "grad_norm": 3.268128686210817, + "learning_rate": 3.157260370354577e-07, + "loss": 0.1916, + "step": 2926 + }, + { + "epoch": 0.6652272727272728, + "grad_norm": 2.684609876430554, + "learning_rate": 3.153383322297453e-07, + "loss": 0.1917, + "step": 2927 + }, + { + "epoch": 0.6654545454545454, + "grad_norm": 3.0821672258794863, + "learning_rate": 3.1495078528756746e-07, + "loss": 0.1478, + "step": 2928 + }, + { + "epoch": 0.6656818181818182, + "grad_norm": 2.7277145147606388, + "learning_rate": 3.145633964064931e-07, + "loss": 0.076, + "step": 2929 + }, + { + "epoch": 0.6659090909090909, + "grad_norm": 2.523729805485878, + "learning_rate": 3.141761657840106e-07, + "loss": 0.127, + "step": 2930 + }, + { + "epoch": 0.6661363636363636, + "grad_norm": 3.361671471054959, + "learning_rate": 3.1378909361752773e-07, + "loss": 0.1449, + "step": 2931 + }, + { + "epoch": 0.6663636363636364, + "grad_norm": 2.967515742051219, + "learning_rate": 3.134021801043711e-07, + "loss": 0.0988, + "step": 2932 + }, + { + "epoch": 0.6665909090909091, + "grad_norm": 3.3708290469901248, + "learning_rate": 3.130154254417872e-07, + "loss": 0.1801, + "step": 2933 + }, + { + "epoch": 0.6668181818181819, + "grad_norm": 2.710020984029405, + "learning_rate": 3.1262882982694085e-07, + "loss": 0.1031, + "step": 2934 + }, + { + "epoch": 0.6670454545454545, + "grad_norm": 5.617860645669326, + "learning_rate": 3.122423934569162e-07, + "loss": 0.1359, + "step": 2935 + }, + { + "epoch": 0.6672727272727272, + "grad_norm": 2.6080365474472367, + "learning_rate": 3.118561165287157e-07, + "loss": 0.0752, + "step": 2936 + }, + { + "epoch": 0.6675, + "grad_norm": 2.5015725184649926, + "learning_rate": 3.1146999923926104e-07, + "loss": 0.0794, + "step": 2937 + }, + { + "epoch": 0.6677272727272727, + "grad_norm": 2.5933023839725977, + "learning_rate": 3.110840417853922e-07, + "loss": 0.1103, + "step": 2938 + }, + { + "epoch": 0.6679545454545455, + "grad_norm": 2.692170577024804, + "learning_rate": 3.106982443638681e-07, + "loss": 0.1082, + "step": 2939 + }, + { + "epoch": 0.6681818181818182, + "grad_norm": 2.440354464351015, + "learning_rate": 3.103126071713657e-07, + "loss": 0.1531, + "step": 2940 + }, + { + "epoch": 0.668409090909091, + "grad_norm": 3.389987437799601, + "learning_rate": 3.099271304044802e-07, + "loss": 0.0993, + "step": 2941 + }, + { + "epoch": 0.6686363636363636, + "grad_norm": 3.23052006241942, + "learning_rate": 3.0954181425972515e-07, + "loss": 0.2171, + "step": 2942 + }, + { + "epoch": 0.6688636363636363, + "grad_norm": 2.7658555039918475, + "learning_rate": 3.091566589335328e-07, + "loss": 0.0785, + "step": 2943 + }, + { + "epoch": 0.6690909090909091, + "grad_norm": 3.377845518677326, + "learning_rate": 3.0877166462225203e-07, + "loss": 0.1709, + "step": 2944 + }, + { + "epoch": 0.6693181818181818, + "grad_norm": 3.1035614265842884, + "learning_rate": 3.0838683152215115e-07, + "loss": 0.1673, + "step": 2945 + }, + { + "epoch": 0.6695454545454546, + "grad_norm": 3.0523372709738372, + "learning_rate": 3.0800215982941535e-07, + "loss": 0.1305, + "step": 2946 + }, + { + "epoch": 0.6697727272727273, + "grad_norm": 2.4713262818342585, + "learning_rate": 3.0761764974014786e-07, + "loss": 0.0835, + "step": 2947 + }, + { + "epoch": 0.67, + "grad_norm": 3.0027727348693776, + "learning_rate": 3.072333014503691e-07, + "loss": 0.1295, + "step": 2948 + }, + { + "epoch": 0.6702272727272728, + "grad_norm": 5.258844871320412, + "learning_rate": 3.0684911515601795e-07, + "loss": 0.1879, + "step": 2949 + }, + { + "epoch": 0.6704545454545454, + "grad_norm": 2.357750910195413, + "learning_rate": 3.064650910529499e-07, + "loss": 0.2266, + "step": 2950 + }, + { + "epoch": 0.6706818181818182, + "grad_norm": 2.93894290867349, + "learning_rate": 3.0608122933693786e-07, + "loss": 0.1299, + "step": 2951 + }, + { + "epoch": 0.6709090909090909, + "grad_norm": 2.923995636155885, + "learning_rate": 3.056975302036722e-07, + "loss": 0.0695, + "step": 2952 + }, + { + "epoch": 0.6711363636363636, + "grad_norm": 4.089254177998647, + "learning_rate": 3.0531399384876016e-07, + "loss": 0.28, + "step": 2953 + }, + { + "epoch": 0.6713636363636364, + "grad_norm": 3.9436381508298384, + "learning_rate": 3.0493062046772597e-07, + "loss": 0.2901, + "step": 2954 + }, + { + "epoch": 0.6715909090909091, + "grad_norm": 2.9681685844684895, + "learning_rate": 3.0454741025601135e-07, + "loss": 0.1197, + "step": 2955 + }, + { + "epoch": 0.6718181818181819, + "grad_norm": 2.8518243728632253, + "learning_rate": 3.0416436340897423e-07, + "loss": 0.1606, + "step": 2956 + }, + { + "epoch": 0.6720454545454545, + "grad_norm": 2.459132514194548, + "learning_rate": 3.037814801218892e-07, + "loss": 0.1238, + "step": 2957 + }, + { + "epoch": 0.6722727272727272, + "grad_norm": 3.0044997705339327, + "learning_rate": 3.033987605899483e-07, + "loss": 0.128, + "step": 2958 + }, + { + "epoch": 0.6725, + "grad_norm": 2.8355943726045014, + "learning_rate": 3.030162050082587e-07, + "loss": 0.0553, + "step": 2959 + }, + { + "epoch": 0.6727272727272727, + "grad_norm": 3.007424355864169, + "learning_rate": 3.026338135718454e-07, + "loss": 0.1218, + "step": 2960 + }, + { + "epoch": 0.6729545454545455, + "grad_norm": 4.6993006809104, + "learning_rate": 3.022515864756489e-07, + "loss": 0.1631, + "step": 2961 + }, + { + "epoch": 0.6731818181818182, + "grad_norm": 3.3883319729695276, + "learning_rate": 3.0186952391452613e-07, + "loss": 0.0945, + "step": 2962 + }, + { + "epoch": 0.673409090909091, + "grad_norm": 3.3582368752910083, + "learning_rate": 3.0148762608324984e-07, + "loss": 0.1086, + "step": 2963 + }, + { + "epoch": 0.6736363636363636, + "grad_norm": 3.405523920788209, + "learning_rate": 3.011058931765097e-07, + "loss": 0.165, + "step": 2964 + }, + { + "epoch": 0.6738636363636363, + "grad_norm": 1.8937252953778612, + "learning_rate": 3.0072432538891033e-07, + "loss": 0.1011, + "step": 2965 + }, + { + "epoch": 0.6740909090909091, + "grad_norm": 5.224425322367585, + "learning_rate": 3.003429229149726e-07, + "loss": 0.1964, + "step": 2966 + }, + { + "epoch": 0.6743181818181818, + "grad_norm": 2.4120370906068493, + "learning_rate": 2.9996168594913305e-07, + "loss": 0.1881, + "step": 2967 + }, + { + "epoch": 0.6745454545454546, + "grad_norm": 4.55159790302191, + "learning_rate": 2.995806146857439e-07, + "loss": 0.1182, + "step": 2968 + }, + { + "epoch": 0.6747727272727273, + "grad_norm": 5.33670332810165, + "learning_rate": 2.991997093190725e-07, + "loss": 0.141, + "step": 2969 + }, + { + "epoch": 0.675, + "grad_norm": 2.6808615255803963, + "learning_rate": 2.9881897004330244e-07, + "loss": 0.209, + "step": 2970 + }, + { + "epoch": 0.6752272727272727, + "grad_norm": 2.7919414951999046, + "learning_rate": 2.98438397052532e-07, + "loss": 0.1805, + "step": 2971 + }, + { + "epoch": 0.6754545454545454, + "grad_norm": 3.35640942763399, + "learning_rate": 2.9805799054077487e-07, + "loss": 0.2477, + "step": 2972 + }, + { + "epoch": 0.6756818181818182, + "grad_norm": 4.871710325023895, + "learning_rate": 2.976777507019597e-07, + "loss": 0.1248, + "step": 2973 + }, + { + "epoch": 0.6759090909090909, + "grad_norm": 4.068976632248796, + "learning_rate": 2.972976777299306e-07, + "loss": 0.1711, + "step": 2974 + }, + { + "epoch": 0.6761363636363636, + "grad_norm": 2.055250943253886, + "learning_rate": 2.969177718184458e-07, + "loss": 0.0991, + "step": 2975 + }, + { + "epoch": 0.6763636363636364, + "grad_norm": 5.332119749241589, + "learning_rate": 2.965380331611795e-07, + "loss": 0.308, + "step": 2976 + }, + { + "epoch": 0.6765909090909091, + "grad_norm": 1.8123177569807865, + "learning_rate": 2.961584619517198e-07, + "loss": 0.08, + "step": 2977 + }, + { + "epoch": 0.6768181818181818, + "grad_norm": 1.812264959156743, + "learning_rate": 2.957790583835695e-07, + "loss": 0.0771, + "step": 2978 + }, + { + "epoch": 0.6770454545454545, + "grad_norm": 2.3791803956123316, + "learning_rate": 2.953998226501467e-07, + "loss": 0.1308, + "step": 2979 + }, + { + "epoch": 0.6772727272727272, + "grad_norm": 3.0770273323893074, + "learning_rate": 2.950207549447827e-07, + "loss": 0.192, + "step": 2980 + }, + { + "epoch": 0.6775, + "grad_norm": 3.366317758355903, + "learning_rate": 2.9464185546072383e-07, + "loss": 0.0766, + "step": 2981 + }, + { + "epoch": 0.6777272727272727, + "grad_norm": 3.92637574164343, + "learning_rate": 2.94263124391131e-07, + "loss": 0.2035, + "step": 2982 + }, + { + "epoch": 0.6779545454545455, + "grad_norm": 2.712451480974468, + "learning_rate": 2.938845619290787e-07, + "loss": 0.0736, + "step": 2983 + }, + { + "epoch": 0.6781818181818182, + "grad_norm": 2.9797679475869847, + "learning_rate": 2.935061682675553e-07, + "loss": 0.1283, + "step": 2984 + }, + { + "epoch": 0.678409090909091, + "grad_norm": 3.3307293261076243, + "learning_rate": 2.931279435994642e-07, + "loss": 0.1381, + "step": 2985 + }, + { + "epoch": 0.6786363636363636, + "grad_norm": 3.337120808130155, + "learning_rate": 2.927498881176214e-07, + "loss": 0.1298, + "step": 2986 + }, + { + "epoch": 0.6788636363636363, + "grad_norm": 3.4261348320094016, + "learning_rate": 2.923720020147573e-07, + "loss": 0.0929, + "step": 2987 + }, + { + "epoch": 0.6790909090909091, + "grad_norm": 3.0060096841321164, + "learning_rate": 2.9199428548351583e-07, + "loss": 0.2234, + "step": 2988 + }, + { + "epoch": 0.6793181818181818, + "grad_norm": 2.6016534178857045, + "learning_rate": 2.916167387164544e-07, + "loss": 0.0753, + "step": 2989 + }, + { + "epoch": 0.6795454545454546, + "grad_norm": 2.4124979547167906, + "learning_rate": 2.912393619060437e-07, + "loss": 0.0852, + "step": 2990 + }, + { + "epoch": 0.6797727272727273, + "grad_norm": 3.1474309562856346, + "learning_rate": 2.908621552446685e-07, + "loss": 0.0905, + "step": 2991 + }, + { + "epoch": 0.68, + "grad_norm": 2.406067282143954, + "learning_rate": 2.904851189246261e-07, + "loss": 0.1271, + "step": 2992 + }, + { + "epoch": 0.6802272727272727, + "grad_norm": 4.829611076993295, + "learning_rate": 2.90108253138127e-07, + "loss": 0.1688, + "step": 2993 + }, + { + "epoch": 0.6804545454545454, + "grad_norm": 4.316270696892667, + "learning_rate": 2.897315580772956e-07, + "loss": 0.0996, + "step": 2994 + }, + { + "epoch": 0.6806818181818182, + "grad_norm": 4.4517485546623705, + "learning_rate": 2.8935503393416805e-07, + "loss": 0.1287, + "step": 2995 + }, + { + "epoch": 0.6809090909090909, + "grad_norm": 2.3988486312253374, + "learning_rate": 2.889786809006937e-07, + "loss": 0.1234, + "step": 2996 + }, + { + "epoch": 0.6811363636363637, + "grad_norm": 2.8664436669318194, + "learning_rate": 2.886024991687357e-07, + "loss": 0.0817, + "step": 2997 + }, + { + "epoch": 0.6813636363636364, + "grad_norm": 2.927993463362845, + "learning_rate": 2.882264889300686e-07, + "loss": 0.1266, + "step": 2998 + }, + { + "epoch": 0.6815909090909091, + "grad_norm": 2.790384074442319, + "learning_rate": 2.878506503763799e-07, + "loss": 0.1238, + "step": 2999 + }, + { + "epoch": 0.6818181818181818, + "grad_norm": 2.490069187196264, + "learning_rate": 2.874749836992701e-07, + "loss": 0.1344, + "step": 3000 + }, + { + "epoch": 0.6820454545454545, + "grad_norm": 2.663340064024958, + "learning_rate": 2.8709948909025175e-07, + "loss": 0.0749, + "step": 3001 + }, + { + "epoch": 0.6822727272727273, + "grad_norm": 2.7122271620124456, + "learning_rate": 2.8672416674074883e-07, + "loss": 0.0977, + "step": 3002 + }, + { + "epoch": 0.6825, + "grad_norm": 2.9922039897253185, + "learning_rate": 2.86349016842099e-07, + "loss": 0.1733, + "step": 3003 + }, + { + "epoch": 0.6827272727272727, + "grad_norm": 3.363068570535257, + "learning_rate": 2.859740395855509e-07, + "loss": 0.1642, + "step": 3004 + }, + { + "epoch": 0.6829545454545455, + "grad_norm": 5.0118952630817315, + "learning_rate": 2.855992351622655e-07, + "loss": 0.1953, + "step": 3005 + }, + { + "epoch": 0.6831818181818182, + "grad_norm": 2.3355549889079117, + "learning_rate": 2.8522460376331605e-07, + "loss": 0.07, + "step": 3006 + }, + { + "epoch": 0.6834090909090909, + "grad_norm": 5.096645996467816, + "learning_rate": 2.848501455796871e-07, + "loss": 0.1315, + "step": 3007 + }, + { + "epoch": 0.6836363636363636, + "grad_norm": 3.989147274467974, + "learning_rate": 2.844758608022749e-07, + "loss": 0.211, + "step": 3008 + }, + { + "epoch": 0.6838636363636363, + "grad_norm": 2.463826750894484, + "learning_rate": 2.8410174962188757e-07, + "loss": 0.1461, + "step": 3009 + }, + { + "epoch": 0.6840909090909091, + "grad_norm": 2.4224540388578317, + "learning_rate": 2.8372781222924446e-07, + "loss": 0.103, + "step": 3010 + }, + { + "epoch": 0.6843181818181818, + "grad_norm": 3.124963037934886, + "learning_rate": 2.833540488149764e-07, + "loss": 0.1225, + "step": 3011 + }, + { + "epoch": 0.6845454545454546, + "grad_norm": 2.160146228934725, + "learning_rate": 2.8298045956962577e-07, + "loss": 0.094, + "step": 3012 + }, + { + "epoch": 0.6847727272727273, + "grad_norm": 3.814993976333651, + "learning_rate": 2.82607044683646e-07, + "loss": 0.1217, + "step": 3013 + }, + { + "epoch": 0.685, + "grad_norm": 1.833404389930112, + "learning_rate": 2.822338043474015e-07, + "loss": 0.0763, + "step": 3014 + }, + { + "epoch": 0.6852272727272727, + "grad_norm": 2.822720934298718, + "learning_rate": 2.818607387511676e-07, + "loss": 0.1275, + "step": 3015 + }, + { + "epoch": 0.6854545454545454, + "grad_norm": 3.074320733715311, + "learning_rate": 2.8148784808513146e-07, + "loss": 0.18, + "step": 3016 + }, + { + "epoch": 0.6856818181818182, + "grad_norm": 2.109650795690194, + "learning_rate": 2.811151325393893e-07, + "loss": 0.1719, + "step": 3017 + }, + { + "epoch": 0.6859090909090909, + "grad_norm": 3.784610618667962, + "learning_rate": 2.8074259230395003e-07, + "loss": 0.1594, + "step": 3018 + }, + { + "epoch": 0.6861363636363637, + "grad_norm": 3.7737426776929452, + "learning_rate": 2.803702275687318e-07, + "loss": 0.1836, + "step": 3019 + }, + { + "epoch": 0.6863636363636364, + "grad_norm": 3.099129391930774, + "learning_rate": 2.79998038523564e-07, + "loss": 0.165, + "step": 3020 + }, + { + "epoch": 0.6865909090909091, + "grad_norm": 2.6060433768203857, + "learning_rate": 2.7962602535818583e-07, + "loss": 0.0625, + "step": 3021 + }, + { + "epoch": 0.6868181818181818, + "grad_norm": 3.751087719361687, + "learning_rate": 2.792541882622477e-07, + "loss": 0.1835, + "step": 3022 + }, + { + "epoch": 0.6870454545454545, + "grad_norm": 2.982097653972222, + "learning_rate": 2.7888252742530967e-07, + "loss": 0.0834, + "step": 3023 + }, + { + "epoch": 0.6872727272727273, + "grad_norm": 4.356948199234834, + "learning_rate": 2.785110430368419e-07, + "loss": 0.0972, + "step": 3024 + }, + { + "epoch": 0.6875, + "grad_norm": 3.0955485026548746, + "learning_rate": 2.781397352862249e-07, + "loss": 0.0847, + "step": 3025 + }, + { + "epoch": 0.6877272727272727, + "grad_norm": 3.8477714439739668, + "learning_rate": 2.777686043627488e-07, + "loss": 0.0827, + "step": 3026 + }, + { + "epoch": 0.6879545454545455, + "grad_norm": 2.359496579608214, + "learning_rate": 2.773976504556137e-07, + "loss": 0.1745, + "step": 3027 + }, + { + "epoch": 0.6881818181818182, + "grad_norm": 3.7913516382465002, + "learning_rate": 2.7702687375392986e-07, + "loss": 0.1171, + "step": 3028 + }, + { + "epoch": 0.6884090909090909, + "grad_norm": 2.8274456842905695, + "learning_rate": 2.7665627444671673e-07, + "loss": 0.127, + "step": 3029 + }, + { + "epoch": 0.6886363636363636, + "grad_norm": 2.204697820223689, + "learning_rate": 2.7628585272290343e-07, + "loss": 0.1353, + "step": 3030 + }, + { + "epoch": 0.6888636363636363, + "grad_norm": 4.18055791594559, + "learning_rate": 2.7591560877132864e-07, + "loss": 0.2882, + "step": 3031 + }, + { + "epoch": 0.6890909090909091, + "grad_norm": 3.031114981791598, + "learning_rate": 2.7554554278074e-07, + "loss": 0.1712, + "step": 3032 + }, + { + "epoch": 0.6893181818181818, + "grad_norm": 4.349786957963655, + "learning_rate": 2.7517565493979535e-07, + "loss": 0.0792, + "step": 3033 + }, + { + "epoch": 0.6895454545454546, + "grad_norm": 2.6363010102537077, + "learning_rate": 2.748059454370608e-07, + "loss": 0.0823, + "step": 3034 + }, + { + "epoch": 0.6897727272727273, + "grad_norm": 2.658404306995582, + "learning_rate": 2.7443641446101206e-07, + "loss": 0.1018, + "step": 3035 + }, + { + "epoch": 0.69, + "grad_norm": 4.374885484037958, + "learning_rate": 2.7406706220003324e-07, + "loss": 0.1853, + "step": 3036 + }, + { + "epoch": 0.6902272727272727, + "grad_norm": 3.3392946347325605, + "learning_rate": 2.7369788884241856e-07, + "loss": 0.091, + "step": 3037 + }, + { + "epoch": 0.6904545454545454, + "grad_norm": 3.875590277944042, + "learning_rate": 2.7332889457636927e-07, + "loss": 0.1028, + "step": 3038 + }, + { + "epoch": 0.6906818181818182, + "grad_norm": 2.6833563356003403, + "learning_rate": 2.7296007958999704e-07, + "loss": 0.1695, + "step": 3039 + }, + { + "epoch": 0.6909090909090909, + "grad_norm": 2.515383041003403, + "learning_rate": 2.7259144407132105e-07, + "loss": 0.0873, + "step": 3040 + }, + { + "epoch": 0.6911363636363637, + "grad_norm": 3.2748116504902467, + "learning_rate": 2.7222298820826944e-07, + "loss": 0.1351, + "step": 3041 + }, + { + "epoch": 0.6913636363636364, + "grad_norm": 3.16923739019792, + "learning_rate": 2.718547121886783e-07, + "loss": 0.0982, + "step": 3042 + }, + { + "epoch": 0.6915909090909091, + "grad_norm": 4.308361308947658, + "learning_rate": 2.714866162002929e-07, + "loss": 0.1305, + "step": 3043 + }, + { + "epoch": 0.6918181818181818, + "grad_norm": 2.9434245862746846, + "learning_rate": 2.71118700430766e-07, + "loss": 0.1546, + "step": 3044 + }, + { + "epoch": 0.6920454545454545, + "grad_norm": 3.5418345770461808, + "learning_rate": 2.707509650676587e-07, + "loss": 0.1571, + "step": 3045 + }, + { + "epoch": 0.6922727272727273, + "grad_norm": 3.1493952272264463, + "learning_rate": 2.703834102984401e-07, + "loss": 0.2528, + "step": 3046 + }, + { + "epoch": 0.6925, + "grad_norm": 3.962787434702102, + "learning_rate": 2.7001603631048734e-07, + "loss": 0.2068, + "step": 3047 + }, + { + "epoch": 0.6927272727272727, + "grad_norm": 2.1296210506452695, + "learning_rate": 2.69648843291085e-07, + "loss": 0.0875, + "step": 3048 + }, + { + "epoch": 0.6929545454545455, + "grad_norm": 2.6530610700983104, + "learning_rate": 2.692818314274263e-07, + "loss": 0.1896, + "step": 3049 + }, + { + "epoch": 0.6931818181818182, + "grad_norm": 3.4642534305401944, + "learning_rate": 2.689150009066112e-07, + "loss": 0.1827, + "step": 3050 + }, + { + "epoch": 0.6934090909090909, + "grad_norm": 2.7328220920636324, + "learning_rate": 2.685483519156474e-07, + "loss": 0.094, + "step": 3051 + }, + { + "epoch": 0.6936363636363636, + "grad_norm": 3.015254561306412, + "learning_rate": 2.6818188464145076e-07, + "loss": 0.1034, + "step": 3052 + }, + { + "epoch": 0.6938636363636363, + "grad_norm": 2.735465212473927, + "learning_rate": 2.6781559927084344e-07, + "loss": 0.1363, + "step": 3053 + }, + { + "epoch": 0.6940909090909091, + "grad_norm": 4.8212273098657406, + "learning_rate": 2.6744949599055527e-07, + "loss": 0.1329, + "step": 3054 + }, + { + "epoch": 0.6943181818181818, + "grad_norm": 2.331989841817571, + "learning_rate": 2.6708357498722386e-07, + "loss": 0.198, + "step": 3055 + }, + { + "epoch": 0.6945454545454546, + "grad_norm": 2.9757198001666088, + "learning_rate": 2.6671783644739315e-07, + "loss": 0.0849, + "step": 3056 + }, + { + "epoch": 0.6947727272727273, + "grad_norm": 1.8421313074943542, + "learning_rate": 2.663522805575141e-07, + "loss": 0.0818, + "step": 3057 + }, + { + "epoch": 0.695, + "grad_norm": 3.689381944743677, + "learning_rate": 2.6598690750394513e-07, + "loss": 0.0908, + "step": 3058 + }, + { + "epoch": 0.6952272727272727, + "grad_norm": 4.3487412041245355, + "learning_rate": 2.656217174729511e-07, + "loss": 0.1451, + "step": 3059 + }, + { + "epoch": 0.6954545454545454, + "grad_norm": 2.3026453618599243, + "learning_rate": 2.65256710650703e-07, + "loss": 0.1526, + "step": 3060 + }, + { + "epoch": 0.6956818181818182, + "grad_norm": 1.7771644877556876, + "learning_rate": 2.6489188722327953e-07, + "loss": 0.1103, + "step": 3061 + }, + { + "epoch": 0.6959090909090909, + "grad_norm": 3.237226806576797, + "learning_rate": 2.645272473766651e-07, + "loss": 0.1034, + "step": 3062 + }, + { + "epoch": 0.6961363636363637, + "grad_norm": 3.0004444266022103, + "learning_rate": 2.6416279129675065e-07, + "loss": 0.1952, + "step": 3063 + }, + { + "epoch": 0.6963636363636364, + "grad_norm": 1.7238990687068765, + "learning_rate": 2.637985191693339e-07, + "loss": 0.1213, + "step": 3064 + }, + { + "epoch": 0.696590909090909, + "grad_norm": 3.100282979906189, + "learning_rate": 2.634344311801183e-07, + "loss": 0.2645, + "step": 3065 + }, + { + "epoch": 0.6968181818181818, + "grad_norm": 2.440070955005536, + "learning_rate": 2.6307052751471343e-07, + "loss": 0.1419, + "step": 3066 + }, + { + "epoch": 0.6970454545454545, + "grad_norm": 2.3458243256493163, + "learning_rate": 2.6270680835863525e-07, + "loss": 0.1073, + "step": 3067 + }, + { + "epoch": 0.6972727272727273, + "grad_norm": 2.3096823923169203, + "learning_rate": 2.6234327389730527e-07, + "loss": 0.1, + "step": 3068 + }, + { + "epoch": 0.6975, + "grad_norm": 2.9771264596591056, + "learning_rate": 2.619799243160509e-07, + "loss": 0.1196, + "step": 3069 + }, + { + "epoch": 0.6977272727272728, + "grad_norm": 3.4300656514321144, + "learning_rate": 2.616167598001059e-07, + "loss": 0.1409, + "step": 3070 + }, + { + "epoch": 0.6979545454545455, + "grad_norm": 3.3260389443788254, + "learning_rate": 2.6125378053460896e-07, + "loss": 0.1489, + "step": 3071 + }, + { + "epoch": 0.6981818181818182, + "grad_norm": 3.632672256293292, + "learning_rate": 2.608909867046044e-07, + "loss": 0.0703, + "step": 3072 + }, + { + "epoch": 0.6984090909090909, + "grad_norm": 2.4593875445660864, + "learning_rate": 2.6052837849504254e-07, + "loss": 0.0576, + "step": 3073 + }, + { + "epoch": 0.6986363636363636, + "grad_norm": 3.189427086444043, + "learning_rate": 2.6016595609077877e-07, + "loss": 0.0946, + "step": 3074 + }, + { + "epoch": 0.6988636363636364, + "grad_norm": 2.013002390769941, + "learning_rate": 2.5980371967657305e-07, + "loss": 0.1463, + "step": 3075 + }, + { + "epoch": 0.6990909090909091, + "grad_norm": 1.424618610363252, + "learning_rate": 2.594416694370919e-07, + "loss": 0.0995, + "step": 3076 + }, + { + "epoch": 0.6993181818181818, + "grad_norm": 1.992385123985492, + "learning_rate": 2.590798055569059e-07, + "loss": 0.1655, + "step": 3077 + }, + { + "epoch": 0.6995454545454546, + "grad_norm": 1.9262394082620917, + "learning_rate": 2.587181282204908e-07, + "loss": 0.074, + "step": 3078 + }, + { + "epoch": 0.6997727272727273, + "grad_norm": 3.4224870657863056, + "learning_rate": 2.5835663761222776e-07, + "loss": 0.1866, + "step": 3079 + }, + { + "epoch": 0.7, + "grad_norm": 2.666422923891436, + "learning_rate": 2.5799533391640216e-07, + "loss": 0.0834, + "step": 3080 + }, + { + "epoch": 0.7002272727272727, + "grad_norm": 2.9313802994757845, + "learning_rate": 2.5763421731720437e-07, + "loss": 0.0894, + "step": 3081 + }, + { + "epoch": 0.7004545454545454, + "grad_norm": 2.649989408968624, + "learning_rate": 2.572732879987293e-07, + "loss": 0.0669, + "step": 3082 + }, + { + "epoch": 0.7006818181818182, + "grad_norm": 3.788010368848363, + "learning_rate": 2.569125461449763e-07, + "loss": 0.1927, + "step": 3083 + }, + { + "epoch": 0.7009090909090909, + "grad_norm": 2.3023896023358956, + "learning_rate": 2.5655199193984915e-07, + "loss": 0.0987, + "step": 3084 + }, + { + "epoch": 0.7011363636363637, + "grad_norm": 3.433849067974874, + "learning_rate": 2.5619162556715657e-07, + "loss": 0.1071, + "step": 3085 + }, + { + "epoch": 0.7013636363636364, + "grad_norm": 2.6100182934521743, + "learning_rate": 2.558314472106107e-07, + "loss": 0.0876, + "step": 3086 + }, + { + "epoch": 0.701590909090909, + "grad_norm": 2.767983260267979, + "learning_rate": 2.554714570538282e-07, + "loss": 0.1109, + "step": 3087 + }, + { + "epoch": 0.7018181818181818, + "grad_norm": 3.1857854763295843, + "learning_rate": 2.551116552803296e-07, + "loss": 0.0807, + "step": 3088 + }, + { + "epoch": 0.7020454545454545, + "grad_norm": 2.630457698883106, + "learning_rate": 2.5475204207353985e-07, + "loss": 0.1691, + "step": 3089 + }, + { + "epoch": 0.7022727272727273, + "grad_norm": 2.3825107824177962, + "learning_rate": 2.543926176167869e-07, + "loss": 0.098, + "step": 3090 + }, + { + "epoch": 0.7025, + "grad_norm": 3.626446226360652, + "learning_rate": 2.540333820933038e-07, + "loss": 0.1362, + "step": 3091 + }, + { + "epoch": 0.7027272727272728, + "grad_norm": 2.547119707314556, + "learning_rate": 2.53674335686226e-07, + "loss": 0.1171, + "step": 3092 + }, + { + "epoch": 0.7029545454545455, + "grad_norm": 3.8418372769092675, + "learning_rate": 2.5331547857859323e-07, + "loss": 0.2895, + "step": 3093 + }, + { + "epoch": 0.7031818181818181, + "grad_norm": 4.9702228069990095, + "learning_rate": 2.529568109533483e-07, + "loss": 0.2002, + "step": 3094 + }, + { + "epoch": 0.7034090909090909, + "grad_norm": 4.699232674311827, + "learning_rate": 2.525983329933383e-07, + "loss": 0.1288, + "step": 3095 + }, + { + "epoch": 0.7036363636363636, + "grad_norm": 1.909039932107359, + "learning_rate": 2.522400448813122e-07, + "loss": 0.0758, + "step": 3096 + }, + { + "epoch": 0.7038636363636364, + "grad_norm": 2.74307795657431, + "learning_rate": 2.518819467999235e-07, + "loss": 0.1261, + "step": 3097 + }, + { + "epoch": 0.7040909090909091, + "grad_norm": 3.7040441024437163, + "learning_rate": 2.515240389317282e-07, + "loss": 0.2183, + "step": 3098 + }, + { + "epoch": 0.7043181818181818, + "grad_norm": 4.269811114311849, + "learning_rate": 2.511663214591852e-07, + "loss": 0.1386, + "step": 3099 + }, + { + "epoch": 0.7045454545454546, + "grad_norm": 3.2964523000356927, + "learning_rate": 2.5080879456465696e-07, + "loss": 0.1672, + "step": 3100 + }, + { + "epoch": 0.7047727272727272, + "grad_norm": 3.7013325041123797, + "learning_rate": 2.5045145843040816e-07, + "loss": 0.0941, + "step": 3101 + }, + { + "epoch": 0.705, + "grad_norm": 3.671691629135017, + "learning_rate": 2.500943132386066e-07, + "loss": 0.112, + "step": 3102 + }, + { + "epoch": 0.7052272727272727, + "grad_norm": 2.380353188732128, + "learning_rate": 2.497373591713226e-07, + "loss": 0.0958, + "step": 3103 + }, + { + "epoch": 0.7054545454545454, + "grad_norm": 2.6547253006332294, + "learning_rate": 2.4938059641052894e-07, + "loss": 0.0714, + "step": 3104 + }, + { + "epoch": 0.7056818181818182, + "grad_norm": 1.9915953565068825, + "learning_rate": 2.490240251381008e-07, + "loss": 0.1103, + "step": 3105 + }, + { + "epoch": 0.7059090909090909, + "grad_norm": 3.3770662950109673, + "learning_rate": 2.4866764553581647e-07, + "loss": 0.2524, + "step": 3106 + }, + { + "epoch": 0.7061363636363637, + "grad_norm": 4.0908154541715716, + "learning_rate": 2.4831145778535563e-07, + "loss": 0.1326, + "step": 3107 + }, + { + "epoch": 0.7063636363636364, + "grad_norm": 3.419513189158273, + "learning_rate": 2.479554620683006e-07, + "loss": 0.0762, + "step": 3108 + }, + { + "epoch": 0.706590909090909, + "grad_norm": 2.262979721659555, + "learning_rate": 2.4759965856613554e-07, + "loss": 0.1205, + "step": 3109 + }, + { + "epoch": 0.7068181818181818, + "grad_norm": 3.731003112389471, + "learning_rate": 2.4724404746024737e-07, + "loss": 0.1182, + "step": 3110 + }, + { + "epoch": 0.7070454545454545, + "grad_norm": 3.224272329767851, + "learning_rate": 2.468886289319235e-07, + "loss": 0.1742, + "step": 3111 + }, + { + "epoch": 0.7072727272727273, + "grad_norm": 4.953487925491954, + "learning_rate": 2.465334031623546e-07, + "loss": 0.1406, + "step": 3112 + }, + { + "epoch": 0.7075, + "grad_norm": 3.2904935676378395, + "learning_rate": 2.461783703326324e-07, + "loss": 0.1339, + "step": 3113 + }, + { + "epoch": 0.7077272727272728, + "grad_norm": 2.9857036633187604, + "learning_rate": 2.458235306237503e-07, + "loss": 0.1552, + "step": 3114 + }, + { + "epoch": 0.7079545454545455, + "grad_norm": 2.311864183125545, + "learning_rate": 2.454688842166032e-07, + "loss": 0.0854, + "step": 3115 + }, + { + "epoch": 0.7081818181818181, + "grad_norm": 2.0742598804906867, + "learning_rate": 2.451144312919878e-07, + "loss": 0.0652, + "step": 3116 + }, + { + "epoch": 0.7084090909090909, + "grad_norm": 4.074681156824652, + "learning_rate": 2.4476017203060184e-07, + "loss": 0.0991, + "step": 3117 + }, + { + "epoch": 0.7086363636363636, + "grad_norm": 4.789885981711295, + "learning_rate": 2.444061066130444e-07, + "loss": 0.2818, + "step": 3118 + }, + { + "epoch": 0.7088636363636364, + "grad_norm": 2.7540166288010064, + "learning_rate": 2.4405223521981577e-07, + "loss": 0.1027, + "step": 3119 + }, + { + "epoch": 0.7090909090909091, + "grad_norm": 3.278105038168268, + "learning_rate": 2.4369855803131735e-07, + "loss": 0.1922, + "step": 3120 + }, + { + "epoch": 0.7093181818181818, + "grad_norm": 2.388772623504477, + "learning_rate": 2.433450752278511e-07, + "loss": 0.0515, + "step": 3121 + }, + { + "epoch": 0.7095454545454546, + "grad_norm": 3.082102773437323, + "learning_rate": 2.4299178698962105e-07, + "loss": 0.1405, + "step": 3122 + }, + { + "epoch": 0.7097727272727272, + "grad_norm": 4.859286684713354, + "learning_rate": 2.4263869349673075e-07, + "loss": 0.1643, + "step": 3123 + }, + { + "epoch": 0.71, + "grad_norm": 3.916904237754569, + "learning_rate": 2.4228579492918527e-07, + "loss": 0.1371, + "step": 3124 + }, + { + "epoch": 0.7102272727272727, + "grad_norm": 4.516952137332514, + "learning_rate": 2.4193309146688984e-07, + "loss": 0.1387, + "step": 3125 + }, + { + "epoch": 0.7104545454545454, + "grad_norm": 2.651598104887895, + "learning_rate": 2.4158058328965045e-07, + "loss": 0.1626, + "step": 3126 + }, + { + "epoch": 0.7106818181818182, + "grad_norm": 3.1920935807633892, + "learning_rate": 2.4122827057717346e-07, + "loss": 0.166, + "step": 3127 + }, + { + "epoch": 0.7109090909090909, + "grad_norm": 1.94940474293319, + "learning_rate": 2.4087615350906584e-07, + "loss": 0.1488, + "step": 3128 + }, + { + "epoch": 0.7111363636363637, + "grad_norm": 1.257486565882867, + "learning_rate": 2.405242322648346e-07, + "loss": 0.0599, + "step": 3129 + }, + { + "epoch": 0.7113636363636363, + "grad_norm": 2.2349933527971717, + "learning_rate": 2.4017250702388644e-07, + "loss": 0.1105, + "step": 3130 + }, + { + "epoch": 0.711590909090909, + "grad_norm": 4.437889128066467, + "learning_rate": 2.3982097796552966e-07, + "loss": 0.1061, + "step": 3131 + }, + { + "epoch": 0.7118181818181818, + "grad_norm": 4.3430869906290095, + "learning_rate": 2.3946964526897054e-07, + "loss": 0.2056, + "step": 3132 + }, + { + "epoch": 0.7120454545454545, + "grad_norm": 2.423108877820905, + "learning_rate": 2.391185091133164e-07, + "loss": 0.1672, + "step": 3133 + }, + { + "epoch": 0.7122727272727273, + "grad_norm": 2.430502901281486, + "learning_rate": 2.387675696775747e-07, + "loss": 0.1467, + "step": 3134 + }, + { + "epoch": 0.7125, + "grad_norm": 2.1993937756923003, + "learning_rate": 2.3841682714065172e-07, + "loss": 0.1102, + "step": 3135 + }, + { + "epoch": 0.7127272727272728, + "grad_norm": 3.976848850981414, + "learning_rate": 2.3806628168135372e-07, + "loss": 0.2147, + "step": 3136 + }, + { + "epoch": 0.7129545454545455, + "grad_norm": 2.6175145111062363, + "learning_rate": 2.3771593347838685e-07, + "loss": 0.0828, + "step": 3137 + }, + { + "epoch": 0.7131818181818181, + "grad_norm": 1.0519645289603246, + "learning_rate": 2.3736578271035615e-07, + "loss": 0.0583, + "step": 3138 + }, + { + "epoch": 0.7134090909090909, + "grad_norm": 3.7211280931198343, + "learning_rate": 2.3701582955576632e-07, + "loss": 0.1422, + "step": 3139 + }, + { + "epoch": 0.7136363636363636, + "grad_norm": 2.009956600659122, + "learning_rate": 2.3666607419302114e-07, + "loss": 0.119, + "step": 3140 + }, + { + "epoch": 0.7138636363636364, + "grad_norm": 3.847844382049412, + "learning_rate": 2.3631651680042374e-07, + "loss": 0.2588, + "step": 3141 + }, + { + "epoch": 0.7140909090909091, + "grad_norm": 3.0806730906149755, + "learning_rate": 2.3596715755617602e-07, + "loss": 0.1761, + "step": 3142 + }, + { + "epoch": 0.7143181818181819, + "grad_norm": 3.194164222086085, + "learning_rate": 2.3561799663837945e-07, + "loss": 0.2433, + "step": 3143 + }, + { + "epoch": 0.7145454545454546, + "grad_norm": 4.4804649561085075, + "learning_rate": 2.352690342250339e-07, + "loss": 0.3062, + "step": 3144 + }, + { + "epoch": 0.7147727272727272, + "grad_norm": 3.426129394093195, + "learning_rate": 2.3492027049403788e-07, + "loss": 0.1969, + "step": 3145 + }, + { + "epoch": 0.715, + "grad_norm": 2.5488965716845238, + "learning_rate": 2.3457170562318957e-07, + "loss": 0.0912, + "step": 3146 + }, + { + "epoch": 0.7152272727272727, + "grad_norm": 3.436960123754485, + "learning_rate": 2.342233397901845e-07, + "loss": 0.2274, + "step": 3147 + }, + { + "epoch": 0.7154545454545455, + "grad_norm": 2.854136157736162, + "learning_rate": 2.3387517317261723e-07, + "loss": 0.109, + "step": 3148 + }, + { + "epoch": 0.7156818181818182, + "grad_norm": 4.736255815891444, + "learning_rate": 2.335272059479815e-07, + "loss": 0.2054, + "step": 3149 + }, + { + "epoch": 0.7159090909090909, + "grad_norm": 2.1773046133507226, + "learning_rate": 2.3317943829366833e-07, + "loss": 0.0624, + "step": 3150 + }, + { + "epoch": 0.7161363636363637, + "grad_norm": 3.921086116333874, + "learning_rate": 2.3283187038696746e-07, + "loss": 0.0972, + "step": 3151 + }, + { + "epoch": 0.7163636363636363, + "grad_norm": 2.3921633718892603, + "learning_rate": 2.3248450240506704e-07, + "loss": 0.0963, + "step": 3152 + }, + { + "epoch": 0.7165909090909091, + "grad_norm": 3.7837887495366913, + "learning_rate": 2.3213733452505307e-07, + "loss": 0.1048, + "step": 3153 + }, + { + "epoch": 0.7168181818181818, + "grad_norm": 3.226311824655517, + "learning_rate": 2.317903669239089e-07, + "loss": 0.1267, + "step": 3154 + }, + { + "epoch": 0.7170454545454545, + "grad_norm": 2.5641987030290903, + "learning_rate": 2.3144359977851708e-07, + "loss": 0.1883, + "step": 3155 + }, + { + "epoch": 0.7172727272727273, + "grad_norm": 1.6684547885412793, + "learning_rate": 2.3109703326565696e-07, + "loss": 0.0873, + "step": 3156 + }, + { + "epoch": 0.7175, + "grad_norm": 2.94386731839221, + "learning_rate": 2.3075066756200573e-07, + "loss": 0.1962, + "step": 3157 + }, + { + "epoch": 0.7177272727272728, + "grad_norm": 3.983180786696984, + "learning_rate": 2.3040450284413895e-07, + "loss": 0.0935, + "step": 3158 + }, + { + "epoch": 0.7179545454545454, + "grad_norm": 2.8476965046248655, + "learning_rate": 2.3005853928852886e-07, + "loss": 0.1379, + "step": 3159 + }, + { + "epoch": 0.7181818181818181, + "grad_norm": 4.500576916887576, + "learning_rate": 2.2971277707154543e-07, + "loss": 0.126, + "step": 3160 + }, + { + "epoch": 0.7184090909090909, + "grad_norm": 3.5749801235823386, + "learning_rate": 2.2936721636945609e-07, + "loss": 0.1235, + "step": 3161 + }, + { + "epoch": 0.7186363636363636, + "grad_norm": 2.4556385669539607, + "learning_rate": 2.2902185735842555e-07, + "loss": 0.1309, + "step": 3162 + }, + { + "epoch": 0.7188636363636364, + "grad_norm": 2.570424844465438, + "learning_rate": 2.2867670021451522e-07, + "loss": 0.0721, + "step": 3163 + }, + { + "epoch": 0.7190909090909091, + "grad_norm": 3.2337278588219935, + "learning_rate": 2.2833174511368462e-07, + "loss": 0.0743, + "step": 3164 + }, + { + "epoch": 0.7193181818181819, + "grad_norm": 4.104429444115039, + "learning_rate": 2.2798699223178928e-07, + "loss": 0.1365, + "step": 3165 + }, + { + "epoch": 0.7195454545454546, + "grad_norm": 3.9016604120484906, + "learning_rate": 2.2764244174458202e-07, + "loss": 0.1135, + "step": 3166 + }, + { + "epoch": 0.7197727272727272, + "grad_norm": 3.4339178198751386, + "learning_rate": 2.2729809382771244e-07, + "loss": 0.1316, + "step": 3167 + }, + { + "epoch": 0.72, + "grad_norm": 4.204925092825667, + "learning_rate": 2.269539486567273e-07, + "loss": 0.3052, + "step": 3168 + }, + { + "epoch": 0.7202272727272727, + "grad_norm": 3.641907254650571, + "learning_rate": 2.2661000640706893e-07, + "loss": 0.125, + "step": 3169 + }, + { + "epoch": 0.7204545454545455, + "grad_norm": 3.648694283224154, + "learning_rate": 2.2626626725407747e-07, + "loss": 0.2052, + "step": 3170 + }, + { + "epoch": 0.7206818181818182, + "grad_norm": 4.7954771758295, + "learning_rate": 2.2592273137298872e-07, + "loss": 0.1308, + "step": 3171 + }, + { + "epoch": 0.7209090909090909, + "grad_norm": 3.4512840456356426, + "learning_rate": 2.255793989389348e-07, + "loss": 0.1089, + "step": 3172 + }, + { + "epoch": 0.7211363636363637, + "grad_norm": 2.7816873008564444, + "learning_rate": 2.252362701269449e-07, + "loss": 0.1513, + "step": 3173 + }, + { + "epoch": 0.7213636363636363, + "grad_norm": 3.637488515387863, + "learning_rate": 2.2489334511194367e-07, + "loss": 0.1644, + "step": 3174 + }, + { + "epoch": 0.7215909090909091, + "grad_norm": 5.102129450062942, + "learning_rate": 2.24550624068752e-07, + "loss": 0.1471, + "step": 3175 + }, + { + "epoch": 0.7218181818181818, + "grad_norm": 1.7371165580135282, + "learning_rate": 2.2420810717208696e-07, + "loss": 0.0674, + "step": 3176 + }, + { + "epoch": 0.7220454545454545, + "grad_norm": 2.8321945938362094, + "learning_rate": 2.2386579459656146e-07, + "loss": 0.164, + "step": 3177 + }, + { + "epoch": 0.7222727272727273, + "grad_norm": 2.5545222612946548, + "learning_rate": 2.2352368651668407e-07, + "loss": 0.1529, + "step": 3178 + }, + { + "epoch": 0.7225, + "grad_norm": 2.957402866053547, + "learning_rate": 2.2318178310685966e-07, + "loss": 0.1847, + "step": 3179 + }, + { + "epoch": 0.7227272727272728, + "grad_norm": 2.816735892192492, + "learning_rate": 2.228400845413882e-07, + "loss": 0.1805, + "step": 3180 + }, + { + "epoch": 0.7229545454545454, + "grad_norm": 3.7810446608113497, + "learning_rate": 2.2249859099446547e-07, + "loss": 0.0864, + "step": 3181 + }, + { + "epoch": 0.7231818181818181, + "grad_norm": 3.1144737786250145, + "learning_rate": 2.2215730264018264e-07, + "loss": 0.0996, + "step": 3182 + }, + { + "epoch": 0.7234090909090909, + "grad_norm": 2.808765418980567, + "learning_rate": 2.2181621965252634e-07, + "loss": 0.1482, + "step": 3183 + }, + { + "epoch": 0.7236363636363636, + "grad_norm": 3.290169815901025, + "learning_rate": 2.2147534220537837e-07, + "loss": 0.0887, + "step": 3184 + }, + { + "epoch": 0.7238636363636364, + "grad_norm": 4.458212436742929, + "learning_rate": 2.2113467047251619e-07, + "loss": 0.3689, + "step": 3185 + }, + { + "epoch": 0.7240909090909091, + "grad_norm": 5.210682059362588, + "learning_rate": 2.2079420462761188e-07, + "loss": 0.2311, + "step": 3186 + }, + { + "epoch": 0.7243181818181819, + "grad_norm": 2.8406014505813584, + "learning_rate": 2.204539448442327e-07, + "loss": 0.204, + "step": 3187 + }, + { + "epoch": 0.7245454545454545, + "grad_norm": 3.2333084800526883, + "learning_rate": 2.201138912958408e-07, + "loss": 0.1647, + "step": 3188 + }, + { + "epoch": 0.7247727272727272, + "grad_norm": 2.0006660358216695, + "learning_rate": 2.1977404415579384e-07, + "loss": 0.1422, + "step": 3189 + }, + { + "epoch": 0.725, + "grad_norm": 3.1526530673677824, + "learning_rate": 2.1943440359734294e-07, + "loss": 0.0807, + "step": 3190 + }, + { + "epoch": 0.7252272727272727, + "grad_norm": 3.3646787357430847, + "learning_rate": 2.190949697936353e-07, + "loss": 0.1249, + "step": 3191 + }, + { + "epoch": 0.7254545454545455, + "grad_norm": 3.222009982134139, + "learning_rate": 2.187557429177118e-07, + "loss": 0.1292, + "step": 3192 + }, + { + "epoch": 0.7256818181818182, + "grad_norm": 3.6372204162339585, + "learning_rate": 2.1841672314250828e-07, + "loss": 0.1544, + "step": 3193 + }, + { + "epoch": 0.725909090909091, + "grad_norm": 3.6692061790180532, + "learning_rate": 2.1807791064085453e-07, + "loss": 0.1823, + "step": 3194 + }, + { + "epoch": 0.7261363636363637, + "grad_norm": 2.3242251371493325, + "learning_rate": 2.1773930558547548e-07, + "loss": 0.0695, + "step": 3195 + }, + { + "epoch": 0.7263636363636363, + "grad_norm": 3.9828603004852865, + "learning_rate": 2.1740090814898953e-07, + "loss": 0.1554, + "step": 3196 + }, + { + "epoch": 0.7265909090909091, + "grad_norm": 2.515710838636964, + "learning_rate": 2.1706271850390953e-07, + "loss": 0.1492, + "step": 3197 + }, + { + "epoch": 0.7268181818181818, + "grad_norm": 2.4877448783693157, + "learning_rate": 2.1672473682264253e-07, + "loss": 0.1992, + "step": 3198 + }, + { + "epoch": 0.7270454545454546, + "grad_norm": 2.9727285614848693, + "learning_rate": 2.1638696327748925e-07, + "loss": 0.1628, + "step": 3199 + }, + { + "epoch": 0.7272727272727273, + "grad_norm": 4.636508513006736, + "learning_rate": 2.160493980406444e-07, + "loss": 0.1277, + "step": 3200 + }, + { + "epoch": 0.7275, + "grad_norm": 4.456898765649772, + "learning_rate": 2.157120412841969e-07, + "loss": 0.1339, + "step": 3201 + }, + { + "epoch": 0.7277272727272728, + "grad_norm": 4.355572149586023, + "learning_rate": 2.1537489318012888e-07, + "loss": 0.2163, + "step": 3202 + }, + { + "epoch": 0.7279545454545454, + "grad_norm": 2.7606601718198887, + "learning_rate": 2.1503795390031612e-07, + "loss": 0.097, + "step": 3203 + }, + { + "epoch": 0.7281818181818182, + "grad_norm": 2.5307335065508765, + "learning_rate": 2.147012236165287e-07, + "loss": 0.1174, + "step": 3204 + }, + { + "epoch": 0.7284090909090909, + "grad_norm": 3.348519513497012, + "learning_rate": 2.1436470250042898e-07, + "loss": 0.1694, + "step": 3205 + }, + { + "epoch": 0.7286363636363636, + "grad_norm": 3.168918902583424, + "learning_rate": 2.1402839072357316e-07, + "loss": 0.1503, + "step": 3206 + }, + { + "epoch": 0.7288636363636364, + "grad_norm": 2.916620740090237, + "learning_rate": 2.1369228845741136e-07, + "loss": 0.1236, + "step": 3207 + }, + { + "epoch": 0.7290909090909091, + "grad_norm": 2.5574750049647785, + "learning_rate": 2.1335639587328607e-07, + "loss": 0.1656, + "step": 3208 + }, + { + "epoch": 0.7293181818181819, + "grad_norm": 4.746755218365252, + "learning_rate": 2.1302071314243307e-07, + "loss": 0.2599, + "step": 3209 + }, + { + "epoch": 0.7295454545454545, + "grad_norm": 3.701173999664011, + "learning_rate": 2.126852404359816e-07, + "loss": 0.1002, + "step": 3210 + }, + { + "epoch": 0.7297727272727272, + "grad_norm": 4.260696273765907, + "learning_rate": 2.1234997792495348e-07, + "loss": 0.1808, + "step": 3211 + }, + { + "epoch": 0.73, + "grad_norm": 3.068482668865504, + "learning_rate": 2.1201492578026282e-07, + "loss": 0.1442, + "step": 3212 + }, + { + "epoch": 0.7302272727272727, + "grad_norm": 2.183784817027236, + "learning_rate": 2.116800841727177e-07, + "loss": 0.0961, + "step": 3213 + }, + { + "epoch": 0.7304545454545455, + "grad_norm": 2.9527399802361036, + "learning_rate": 2.1134545327301796e-07, + "loss": 0.1979, + "step": 3214 + }, + { + "epoch": 0.7306818181818182, + "grad_norm": 3.4365022628632502, + "learning_rate": 2.1101103325175606e-07, + "loss": 0.1852, + "step": 3215 + }, + { + "epoch": 0.730909090909091, + "grad_norm": 4.1099046441184, + "learning_rate": 2.1067682427941772e-07, + "loss": 0.08, + "step": 3216 + }, + { + "epoch": 0.7311363636363636, + "grad_norm": 2.770261517197332, + "learning_rate": 2.1034282652638015e-07, + "loss": 0.0821, + "step": 3217 + }, + { + "epoch": 0.7313636363636363, + "grad_norm": 3.4679103334783097, + "learning_rate": 2.1000904016291347e-07, + "loss": 0.1544, + "step": 3218 + }, + { + "epoch": 0.7315909090909091, + "grad_norm": 3.245281163024528, + "learning_rate": 2.0967546535917968e-07, + "loss": 0.2134, + "step": 3219 + }, + { + "epoch": 0.7318181818181818, + "grad_norm": 1.1397106223164144, + "learning_rate": 2.0934210228523315e-07, + "loss": 0.0754, + "step": 3220 + }, + { + "epoch": 0.7320454545454546, + "grad_norm": 1.9795044711366538, + "learning_rate": 2.0900895111101994e-07, + "loss": 0.1017, + "step": 3221 + }, + { + "epoch": 0.7322727272727273, + "grad_norm": 2.6445602325369784, + "learning_rate": 2.086760120063789e-07, + "loss": 0.0879, + "step": 3222 + }, + { + "epoch": 0.7325, + "grad_norm": 4.3213515384958345, + "learning_rate": 2.0834328514104004e-07, + "loss": 0.1877, + "step": 3223 + }, + { + "epoch": 0.7327272727272728, + "grad_norm": 1.6362365016029166, + "learning_rate": 2.0801077068462508e-07, + "loss": 0.1664, + "step": 3224 + }, + { + "epoch": 0.7329545454545454, + "grad_norm": 2.0577466055687705, + "learning_rate": 2.0767846880664834e-07, + "loss": 0.0809, + "step": 3225 + }, + { + "epoch": 0.7331818181818182, + "grad_norm": 1.0399634443455048, + "learning_rate": 2.0734637967651503e-07, + "loss": 0.1397, + "step": 3226 + }, + { + "epoch": 0.7334090909090909, + "grad_norm": 1.7126606662567507, + "learning_rate": 2.0701450346352152e-07, + "loss": 0.0943, + "step": 3227 + }, + { + "epoch": 0.7336363636363636, + "grad_norm": 3.414669898122139, + "learning_rate": 2.0668284033685677e-07, + "loss": 0.1815, + "step": 3228 + }, + { + "epoch": 0.7338636363636364, + "grad_norm": 1.8143982479917697, + "learning_rate": 2.0635139046560025e-07, + "loss": 0.121, + "step": 3229 + }, + { + "epoch": 0.7340909090909091, + "grad_norm": 2.4765633977673187, + "learning_rate": 2.0602015401872297e-07, + "loss": 0.1155, + "step": 3230 + }, + { + "epoch": 0.7343181818181819, + "grad_norm": 3.0044918791755655, + "learning_rate": 2.0568913116508737e-07, + "loss": 0.172, + "step": 3231 + }, + { + "epoch": 0.7345454545454545, + "grad_norm": 2.5398618780241287, + "learning_rate": 2.053583220734466e-07, + "loss": 0.1263, + "step": 3232 + }, + { + "epoch": 0.7347727272727272, + "grad_norm": 3.5583888339705982, + "learning_rate": 2.050277269124451e-07, + "loss": 0.2076, + "step": 3233 + }, + { + "epoch": 0.735, + "grad_norm": 4.889295179454483, + "learning_rate": 2.046973458506181e-07, + "loss": 0.143, + "step": 3234 + }, + { + "epoch": 0.7352272727272727, + "grad_norm": 2.806836765547164, + "learning_rate": 2.0436717905639183e-07, + "loss": 0.0638, + "step": 3235 + }, + { + "epoch": 0.7354545454545455, + "grad_norm": 2.2716266473339526, + "learning_rate": 2.0403722669808285e-07, + "loss": 0.1158, + "step": 3236 + }, + { + "epoch": 0.7356818181818182, + "grad_norm": 4.543399394669919, + "learning_rate": 2.0370748894389933e-07, + "loss": 0.1835, + "step": 3237 + }, + { + "epoch": 0.735909090909091, + "grad_norm": 4.941375686664382, + "learning_rate": 2.0337796596193915e-07, + "loss": 0.2011, + "step": 3238 + }, + { + "epoch": 0.7361363636363636, + "grad_norm": 3.1569154418523664, + "learning_rate": 2.030486579201909e-07, + "loss": 0.1834, + "step": 3239 + }, + { + "epoch": 0.7363636363636363, + "grad_norm": 4.194680518886563, + "learning_rate": 2.027195649865342e-07, + "loss": 0.1058, + "step": 3240 + }, + { + "epoch": 0.7365909090909091, + "grad_norm": 2.7800973650682987, + "learning_rate": 2.0239068732873804e-07, + "loss": 0.0798, + "step": 3241 + }, + { + "epoch": 0.7368181818181818, + "grad_norm": 3.5615600285519458, + "learning_rate": 2.0206202511446206e-07, + "loss": 0.1302, + "step": 3242 + }, + { + "epoch": 0.7370454545454546, + "grad_norm": 3.814662752492407, + "learning_rate": 2.0173357851125654e-07, + "loss": 0.088, + "step": 3243 + }, + { + "epoch": 0.7372727272727273, + "grad_norm": 1.7264336825700832, + "learning_rate": 2.0140534768656135e-07, + "loss": 0.16, + "step": 3244 + }, + { + "epoch": 0.7375, + "grad_norm": 3.1979056121900533, + "learning_rate": 2.010773328077062e-07, + "loss": 0.1055, + "step": 3245 + }, + { + "epoch": 0.7377272727272727, + "grad_norm": 3.616978035888864, + "learning_rate": 2.007495340419114e-07, + "loss": 0.1235, + "step": 3246 + }, + { + "epoch": 0.7379545454545454, + "grad_norm": 3.3333025271798737, + "learning_rate": 2.0042195155628656e-07, + "loss": 0.2288, + "step": 3247 + }, + { + "epoch": 0.7381818181818182, + "grad_norm": 3.8076493655038663, + "learning_rate": 2.0009458551783065e-07, + "loss": 0.1123, + "step": 3248 + }, + { + "epoch": 0.7384090909090909, + "grad_norm": 2.4682614006363908, + "learning_rate": 1.9976743609343338e-07, + "loss": 0.0832, + "step": 3249 + }, + { + "epoch": 0.7386363636363636, + "grad_norm": 8.431379656003497, + "learning_rate": 1.9944050344987316e-07, + "loss": 0.3028, + "step": 3250 + }, + { + "epoch": 0.7388636363636364, + "grad_norm": 2.278427693400358, + "learning_rate": 1.9911378775381802e-07, + "loss": 0.0943, + "step": 3251 + }, + { + "epoch": 0.7390909090909091, + "grad_norm": 2.086984419053252, + "learning_rate": 1.9878728917182586e-07, + "loss": 0.0944, + "step": 3252 + }, + { + "epoch": 0.7393181818181818, + "grad_norm": 3.872941512101377, + "learning_rate": 1.9846100787034342e-07, + "loss": 0.1412, + "step": 3253 + }, + { + "epoch": 0.7395454545454545, + "grad_norm": 2.529898680116635, + "learning_rate": 1.9813494401570683e-07, + "loss": 0.1363, + "step": 3254 + }, + { + "epoch": 0.7397727272727272, + "grad_norm": 2.072345892731655, + "learning_rate": 1.9780909777414126e-07, + "loss": 0.1314, + "step": 3255 + }, + { + "epoch": 0.74, + "grad_norm": 4.121040883404992, + "learning_rate": 1.9748346931176117e-07, + "loss": 0.1327, + "step": 3256 + }, + { + "epoch": 0.7402272727272727, + "grad_norm": 2.776062506080073, + "learning_rate": 1.9715805879456955e-07, + "loss": 0.1827, + "step": 3257 + }, + { + "epoch": 0.7404545454545455, + "grad_norm": 3.282137236639584, + "learning_rate": 1.968328663884591e-07, + "loss": 0.1896, + "step": 3258 + }, + { + "epoch": 0.7406818181818182, + "grad_norm": 2.7417330907842175, + "learning_rate": 1.9650789225921049e-07, + "loss": 0.2046, + "step": 3259 + }, + { + "epoch": 0.740909090909091, + "grad_norm": 2.810808311603332, + "learning_rate": 1.961831365724936e-07, + "loss": 0.1196, + "step": 3260 + }, + { + "epoch": 0.7411363636363636, + "grad_norm": 3.1491230069277596, + "learning_rate": 1.958585994938665e-07, + "loss": 0.1934, + "step": 3261 + }, + { + "epoch": 0.7413636363636363, + "grad_norm": 2.6940066877029323, + "learning_rate": 1.9553428118877672e-07, + "loss": 0.1687, + "step": 3262 + }, + { + "epoch": 0.7415909090909091, + "grad_norm": 3.054166170031223, + "learning_rate": 1.9521018182255894e-07, + "loss": 0.0913, + "step": 3263 + }, + { + "epoch": 0.7418181818181818, + "grad_norm": 2.5214554870027195, + "learning_rate": 1.948863015604374e-07, + "loss": 0.1783, + "step": 3264 + }, + { + "epoch": 0.7420454545454546, + "grad_norm": 2.460200386617, + "learning_rate": 1.945626405675241e-07, + "loss": 0.1207, + "step": 3265 + }, + { + "epoch": 0.7422727272727273, + "grad_norm": 4.670042134497616, + "learning_rate": 1.942391990088193e-07, + "loss": 0.1576, + "step": 3266 + }, + { + "epoch": 0.7425, + "grad_norm": 3.020789912245896, + "learning_rate": 1.939159770492112e-07, + "loss": 0.1866, + "step": 3267 + }, + { + "epoch": 0.7427272727272727, + "grad_norm": 2.489382297454184, + "learning_rate": 1.9359297485347677e-07, + "loss": 0.088, + "step": 3268 + }, + { + "epoch": 0.7429545454545454, + "grad_norm": 4.49054519114111, + "learning_rate": 1.9327019258628017e-07, + "loss": 0.1595, + "step": 3269 + }, + { + "epoch": 0.7431818181818182, + "grad_norm": 3.7743659304884205, + "learning_rate": 1.9294763041217385e-07, + "loss": 0.1368, + "step": 3270 + }, + { + "epoch": 0.7434090909090909, + "grad_norm": 2.965554116781748, + "learning_rate": 1.926252884955978e-07, + "loss": 0.2011, + "step": 3271 + }, + { + "epoch": 0.7436363636363637, + "grad_norm": 2.461344783018167, + "learning_rate": 1.9230316700087993e-07, + "loss": 0.1788, + "step": 3272 + }, + { + "epoch": 0.7438636363636364, + "grad_norm": 1.8677139070404396, + "learning_rate": 1.919812660922355e-07, + "loss": 0.1943, + "step": 3273 + }, + { + "epoch": 0.7440909090909091, + "grad_norm": 2.6169443417702243, + "learning_rate": 1.9165958593376797e-07, + "loss": 0.1217, + "step": 3274 + }, + { + "epoch": 0.7443181818181818, + "grad_norm": 3.227279129205522, + "learning_rate": 1.9133812668946753e-07, + "loss": 0.1841, + "step": 3275 + }, + { + "epoch": 0.7445454545454545, + "grad_norm": 2.661686162334006, + "learning_rate": 1.910168885232122e-07, + "loss": 0.133, + "step": 3276 + }, + { + "epoch": 0.7447727272727273, + "grad_norm": 4.566109926059233, + "learning_rate": 1.9069587159876702e-07, + "loss": 0.0929, + "step": 3277 + }, + { + "epoch": 0.745, + "grad_norm": 2.337944549808782, + "learning_rate": 1.9037507607978436e-07, + "loss": 0.0762, + "step": 3278 + }, + { + "epoch": 0.7452272727272727, + "grad_norm": 2.240686906510499, + "learning_rate": 1.9005450212980357e-07, + "loss": 0.1525, + "step": 3279 + }, + { + "epoch": 0.7454545454545455, + "grad_norm": 2.9408635426315377, + "learning_rate": 1.897341499122516e-07, + "loss": 0.1983, + "step": 3280 + }, + { + "epoch": 0.7456818181818182, + "grad_norm": 4.953226761442462, + "learning_rate": 1.8941401959044167e-07, + "loss": 0.2479, + "step": 3281 + }, + { + "epoch": 0.7459090909090909, + "grad_norm": 4.2245931755186685, + "learning_rate": 1.8909411132757413e-07, + "loss": 0.1522, + "step": 3282 + }, + { + "epoch": 0.7461363636363636, + "grad_norm": 2.398786040363871, + "learning_rate": 1.8877442528673656e-07, + "loss": 0.1174, + "step": 3283 + }, + { + "epoch": 0.7463636363636363, + "grad_norm": 1.7454242305737222, + "learning_rate": 1.884549616309024e-07, + "loss": 0.1145, + "step": 3284 + }, + { + "epoch": 0.7465909090909091, + "grad_norm": 3.003848019016285, + "learning_rate": 1.881357205229322e-07, + "loss": 0.1587, + "step": 3285 + }, + { + "epoch": 0.7468181818181818, + "grad_norm": 3.7886649683448317, + "learning_rate": 1.8781670212557327e-07, + "loss": 0.1068, + "step": 3286 + }, + { + "epoch": 0.7470454545454546, + "grad_norm": 4.3126480378089385, + "learning_rate": 1.8749790660145915e-07, + "loss": 0.1324, + "step": 3287 + }, + { + "epoch": 0.7472727272727273, + "grad_norm": 3.8556616798122114, + "learning_rate": 1.8717933411310937e-07, + "loss": 0.1128, + "step": 3288 + }, + { + "epoch": 0.7475, + "grad_norm": 3.583141630576112, + "learning_rate": 1.8686098482293067e-07, + "loss": 0.1026, + "step": 3289 + }, + { + "epoch": 0.7477272727272727, + "grad_norm": 2.7071265570978182, + "learning_rate": 1.8654285889321521e-07, + "loss": 0.1032, + "step": 3290 + }, + { + "epoch": 0.7479545454545454, + "grad_norm": 4.772730192700153, + "learning_rate": 1.862249564861416e-07, + "loss": 0.271, + "step": 3291 + }, + { + "epoch": 0.7481818181818182, + "grad_norm": 4.947953168273683, + "learning_rate": 1.8590727776377437e-07, + "loss": 0.1412, + "step": 3292 + }, + { + "epoch": 0.7484090909090909, + "grad_norm": 3.755300800912968, + "learning_rate": 1.855898228880642e-07, + "loss": 0.2044, + "step": 3293 + }, + { + "epoch": 0.7486363636363637, + "grad_norm": 3.617003911045771, + "learning_rate": 1.852725920208473e-07, + "loss": 0.2797, + "step": 3294 + }, + { + "epoch": 0.7488636363636364, + "grad_norm": 4.4213989669418945, + "learning_rate": 1.8495558532384633e-07, + "loss": 0.1626, + "step": 3295 + }, + { + "epoch": 0.7490909090909091, + "grad_norm": 2.508652757480152, + "learning_rate": 1.8463880295866904e-07, + "loss": 0.1556, + "step": 3296 + }, + { + "epoch": 0.7493181818181818, + "grad_norm": 3.065046512745386, + "learning_rate": 1.8432224508680896e-07, + "loss": 0.1604, + "step": 3297 + }, + { + "epoch": 0.7495454545454545, + "grad_norm": 4.108505498998234, + "learning_rate": 1.8400591186964577e-07, + "loss": 0.2319, + "step": 3298 + }, + { + "epoch": 0.7497727272727273, + "grad_norm": 6.0492169534483065, + "learning_rate": 1.8368980346844352e-07, + "loss": 0.1312, + "step": 3299 + }, + { + "epoch": 0.75, + "grad_norm": 3.5286208058516633, + "learning_rate": 1.8337392004435233e-07, + "loss": 0.1439, + "step": 3300 + } + ], + "logging_steps": 1.0, + "max_steps": 4400, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 1100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 10471807647744.0, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}