{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.050008343995104856, "eval_steps": 500, "global_step": 899, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "grad_norm": 31.671595567958246, "learning_rate": 1.851851851851852e-08, "loss": 0.9836, "step": 1 }, { "epoch": 0.0, "grad_norm": 21.590568341208314, "learning_rate": 3.703703703703704e-08, "loss": 0.6951, "step": 2 }, { "epoch": 0.0, "grad_norm": 33.43736925901718, "learning_rate": 5.555555555555556e-08, "loss": 1.0552, "step": 3 }, { "epoch": 0.0, "grad_norm": 25.76674648270905, "learning_rate": 7.407407407407409e-08, "loss": 0.9979, "step": 4 }, { "epoch": 0.0, "grad_norm": 29.142254478570287, "learning_rate": 9.259259259259259e-08, "loss": 1.1524, "step": 5 }, { "epoch": 0.0, "grad_norm": 128.44118605473278, "learning_rate": 1.1111111111111112e-07, "loss": 0.9627, "step": 6 }, { "epoch": 0.0, "grad_norm": 22.844065944506312, "learning_rate": 1.2962962962962964e-07, "loss": 0.7732, "step": 7 }, { "epoch": 0.0, "grad_norm": 26.662421778739866, "learning_rate": 1.4814814814814817e-07, "loss": 0.9373, "step": 8 }, { "epoch": 0.0, "grad_norm": 24.8894090181831, "learning_rate": 1.6666666666666668e-07, "loss": 1.1089, "step": 9 }, { "epoch": 0.0, "grad_norm": 21.578880921091358, "learning_rate": 1.8518518518518518e-07, "loss": 0.6824, "step": 10 }, { "epoch": 0.0, "grad_norm": 28.74029164802737, "learning_rate": 2.0370370370370374e-07, "loss": 0.8942, "step": 11 }, { "epoch": 0.0, "grad_norm": 22.14964568326191, "learning_rate": 2.2222222222222224e-07, "loss": 0.8041, "step": 12 }, { "epoch": 0.0, "grad_norm": 32.9657912531208, "learning_rate": 2.4074074074074075e-07, "loss": 1.2601, "step": 13 }, { "epoch": 0.0, "grad_norm": 17.145724154465036, "learning_rate": 2.592592592592593e-07, "loss": 0.7905, "step": 14 }, { "epoch": 0.0, "grad_norm": 24.26835111004837, "learning_rate": 2.7777777777777776e-07, "loss": 0.9586, "step": 15 }, { "epoch": 0.0, "grad_norm": 22.011607835789768, "learning_rate": 2.9629629629629634e-07, "loss": 0.8789, "step": 16 }, { "epoch": 0.0, "grad_norm": 30.838211196101994, "learning_rate": 3.148148148148148e-07, "loss": 0.9217, "step": 17 }, { "epoch": 0.0, "grad_norm": 15.955309084125279, "learning_rate": 3.3333333333333335e-07, "loss": 0.911, "step": 18 }, { "epoch": 0.0, "grad_norm": 24.344206309908216, "learning_rate": 3.518518518518519e-07, "loss": 0.9155, "step": 19 }, { "epoch": 0.0, "grad_norm": 32.92989138430358, "learning_rate": 3.7037037037037036e-07, "loss": 0.8828, "step": 20 }, { "epoch": 0.0, "grad_norm": 22.41057192779469, "learning_rate": 3.8888888888888895e-07, "loss": 0.8469, "step": 21 }, { "epoch": 0.0, "grad_norm": 17.46446762413705, "learning_rate": 4.074074074074075e-07, "loss": 0.6855, "step": 22 }, { "epoch": 0.0, "grad_norm": 15.433059856938662, "learning_rate": 4.2592592592592596e-07, "loss": 0.7404, "step": 23 }, { "epoch": 0.0, "grad_norm": 10.728638383519646, "learning_rate": 4.444444444444445e-07, "loss": 0.6503, "step": 24 }, { "epoch": 0.0, "grad_norm": 12.213215386669342, "learning_rate": 4.6296296296296297e-07, "loss": 0.8609, "step": 25 }, { "epoch": 0.0, "grad_norm": 18.317938299817154, "learning_rate": 4.814814814814815e-07, "loss": 0.8072, "step": 26 }, { "epoch": 0.0, "grad_norm": 19.438962335560756, "learning_rate": 5.000000000000001e-07, "loss": 0.5746, "step": 27 }, { "epoch": 0.0, "grad_norm": 23.19453080923279, "learning_rate": 5.185185185185186e-07, "loss": 0.8071, "step": 28 }, { "epoch": 0.0, "grad_norm": 17.74196002097125, "learning_rate": 5.37037037037037e-07, "loss": 0.6765, "step": 29 }, { "epoch": 0.0, "grad_norm": 13.827465942539503, "learning_rate": 5.555555555555555e-07, "loss": 0.7021, "step": 30 }, { "epoch": 0.0, "grad_norm": 15.488021836868372, "learning_rate": 5.740740740740741e-07, "loss": 0.4757, "step": 31 }, { "epoch": 0.0, "grad_norm": 13.20162265514205, "learning_rate": 5.925925925925927e-07, "loss": 0.5341, "step": 32 }, { "epoch": 0.0, "grad_norm": 9.110026681402434, "learning_rate": 6.111111111111112e-07, "loss": 0.4415, "step": 33 }, { "epoch": 0.0, "grad_norm": 9.830477386012433, "learning_rate": 6.296296296296296e-07, "loss": 0.6151, "step": 34 }, { "epoch": 0.0, "grad_norm": 11.622213091116294, "learning_rate": 6.481481481481481e-07, "loss": 0.4269, "step": 35 }, { "epoch": 0.0, "grad_norm": 8.207427780107848, "learning_rate": 6.666666666666667e-07, "loss": 0.3635, "step": 36 }, { "epoch": 0.0, "grad_norm": 7.836791093758771, "learning_rate": 6.851851851851853e-07, "loss": 0.6096, "step": 37 }, { "epoch": 0.0, "grad_norm": 7.193647061430912, "learning_rate": 7.037037037037038e-07, "loss": 0.4741, "step": 38 }, { "epoch": 0.0, "grad_norm": 9.476468858625738, "learning_rate": 7.222222222222222e-07, "loss": 0.3655, "step": 39 }, { "epoch": 0.0, "grad_norm": 7.157861986137023, "learning_rate": 7.407407407407407e-07, "loss": 0.4635, "step": 40 }, { "epoch": 0.0, "grad_norm": 7.246782509697589, "learning_rate": 7.592592592592593e-07, "loss": 0.54, "step": 41 }, { "epoch": 0.0, "grad_norm": 4.533392537772903, "learning_rate": 7.777777777777779e-07, "loss": 0.1987, "step": 42 }, { "epoch": 0.0, "grad_norm": 6.082186481942415, "learning_rate": 7.962962962962964e-07, "loss": 0.3954, "step": 43 }, { "epoch": 0.0, "grad_norm": 4.686018035593464, "learning_rate": 8.14814814814815e-07, "loss": 0.2148, "step": 44 }, { "epoch": 0.0, "grad_norm": 5.285244734447013, "learning_rate": 8.333333333333333e-07, "loss": 0.312, "step": 45 }, { "epoch": 0.0, "grad_norm": 5.426938300900884, "learning_rate": 8.518518518518519e-07, "loss": 0.2549, "step": 46 }, { "epoch": 0.0, "grad_norm": 6.275553478285126, "learning_rate": 8.703703703703705e-07, "loss": 0.3357, "step": 47 }, { "epoch": 0.0, "grad_norm": 5.135064616590709, "learning_rate": 8.88888888888889e-07, "loss": 0.1967, "step": 48 }, { "epoch": 0.0, "grad_norm": 4.486540108880125, "learning_rate": 9.074074074074076e-07, "loss": 0.3064, "step": 49 }, { "epoch": 0.0, "grad_norm": 6.190915869898936, "learning_rate": 9.259259259259259e-07, "loss": 0.4127, "step": 50 }, { "epoch": 0.0, "grad_norm": 6.080158746554302, "learning_rate": 9.444444444444445e-07, "loss": 0.2437, "step": 51 }, { "epoch": 0.0, "grad_norm": 5.0751084508724, "learning_rate": 9.62962962962963e-07, "loss": 0.4358, "step": 52 }, { "epoch": 0.0, "grad_norm": 7.434165335252677, "learning_rate": 9.814814814814816e-07, "loss": 0.4203, "step": 53 }, { "epoch": 0.0, "grad_norm": 5.253355134896159, "learning_rate": 1.0000000000000002e-06, "loss": 0.3269, "step": 54 }, { "epoch": 0.0, "grad_norm": 4.243061853720817, "learning_rate": 1.0185185185185185e-06, "loss": 0.1813, "step": 55 }, { "epoch": 0.0, "grad_norm": 4.856767967025552, "learning_rate": 1.0370370370370371e-06, "loss": 0.3047, "step": 56 }, { "epoch": 0.0, "grad_norm": 5.127255350560153, "learning_rate": 1.0555555555555557e-06, "loss": 0.2576, "step": 57 }, { "epoch": 0.0, "grad_norm": 4.346136439290515, "learning_rate": 1.074074074074074e-06, "loss": 0.2684, "step": 58 }, { "epoch": 0.0, "grad_norm": 3.4920134020938165, "learning_rate": 1.0925925925925927e-06, "loss": 0.194, "step": 59 }, { "epoch": 0.0, "grad_norm": 4.647235144260062, "learning_rate": 1.111111111111111e-06, "loss": 0.224, "step": 60 }, { "epoch": 0.0, "grad_norm": 5.519727656169724, "learning_rate": 1.1296296296296296e-06, "loss": 0.2936, "step": 61 }, { "epoch": 0.0, "grad_norm": 8.295031154084294, "learning_rate": 1.1481481481481482e-06, "loss": 0.3286, "step": 62 }, { "epoch": 0.0, "grad_norm": 5.722549697873489, "learning_rate": 1.1666666666666668e-06, "loss": 0.4155, "step": 63 }, { "epoch": 0.0, "grad_norm": 5.875671754133045, "learning_rate": 1.1851851851851854e-06, "loss": 0.378, "step": 64 }, { "epoch": 0.0, "grad_norm": 6.792663865708492, "learning_rate": 1.2037037037037037e-06, "loss": 0.2497, "step": 65 }, { "epoch": 0.0, "grad_norm": 7.20501181162727, "learning_rate": 1.2222222222222223e-06, "loss": 0.3962, "step": 66 }, { "epoch": 0.0, "grad_norm": 4.794105811510582, "learning_rate": 1.240740740740741e-06, "loss": 0.2524, "step": 67 }, { "epoch": 0.0, "grad_norm": 5.064103025788011, "learning_rate": 1.2592592592592593e-06, "loss": 0.3846, "step": 68 }, { "epoch": 0.0, "grad_norm": 3.452696967890651, "learning_rate": 1.2777777777777779e-06, "loss": 0.2413, "step": 69 }, { "epoch": 0.0, "grad_norm": 2.9764029793033213, "learning_rate": 1.2962962962962962e-06, "loss": 0.1858, "step": 70 }, { "epoch": 0.0, "grad_norm": 4.887938203216541, "learning_rate": 1.3148148148148148e-06, "loss": 0.2745, "step": 71 }, { "epoch": 0.0, "grad_norm": 5.134226727564933, "learning_rate": 1.3333333333333334e-06, "loss": 0.2885, "step": 72 }, { "epoch": 0.0, "grad_norm": 5.590512921407683, "learning_rate": 1.351851851851852e-06, "loss": 0.2419, "step": 73 }, { "epoch": 0.0, "grad_norm": 4.674717092878751, "learning_rate": 1.3703703703703706e-06, "loss": 0.3462, "step": 74 }, { "epoch": 0.0, "grad_norm": 4.509220346401417, "learning_rate": 1.3888888888888892e-06, "loss": 0.1586, "step": 75 }, { "epoch": 0.0, "grad_norm": 3.997696481472079, "learning_rate": 1.4074074074074075e-06, "loss": 0.2175, "step": 76 }, { "epoch": 0.0, "grad_norm": 4.966583739405523, "learning_rate": 1.4259259259259261e-06, "loss": 0.3962, "step": 77 }, { "epoch": 0.0, "grad_norm": 5.093276066175207, "learning_rate": 1.4444444444444445e-06, "loss": 0.3756, "step": 78 }, { "epoch": 0.0, "grad_norm": 6.335763630204078, "learning_rate": 1.4629629629629629e-06, "loss": 0.2768, "step": 79 }, { "epoch": 0.0, "grad_norm": 5.52584594274486, "learning_rate": 1.4814814814814815e-06, "loss": 0.2227, "step": 80 }, { "epoch": 0.0, "grad_norm": 4.292995296600986, "learning_rate": 1.5e-06, "loss": 0.2931, "step": 81 }, { "epoch": 0.0, "grad_norm": 3.3855191650524232, "learning_rate": 1.5185185185185186e-06, "loss": 0.2184, "step": 82 }, { "epoch": 0.0, "grad_norm": 6.680707851356181, "learning_rate": 1.5370370370370372e-06, "loss": 0.3002, "step": 83 }, { "epoch": 0.0, "grad_norm": 4.157140363964968, "learning_rate": 1.5555555555555558e-06, "loss": 0.2268, "step": 84 }, { "epoch": 0.0, "grad_norm": 2.9320222630351624, "learning_rate": 1.5740740740740742e-06, "loss": 0.2046, "step": 85 }, { "epoch": 0.0, "grad_norm": 4.54889007696515, "learning_rate": 1.5925925925925927e-06, "loss": 0.2168, "step": 86 }, { "epoch": 0.0, "grad_norm": 4.0441417227247705, "learning_rate": 1.6111111111111113e-06, "loss": 0.2149, "step": 87 }, { "epoch": 0.0, "grad_norm": 5.7596791320081, "learning_rate": 1.62962962962963e-06, "loss": 0.2557, "step": 88 }, { "epoch": 0.0, "grad_norm": 4.4753129689831885, "learning_rate": 1.648148148148148e-06, "loss": 0.2086, "step": 89 }, { "epoch": 0.01, "grad_norm": 4.625168152922665, "learning_rate": 1.6666666666666667e-06, "loss": 0.3106, "step": 90 }, { "epoch": 0.01, "grad_norm": 4.778566667141804, "learning_rate": 1.6851851851851852e-06, "loss": 0.236, "step": 91 }, { "epoch": 0.01, "grad_norm": 3.4927683097649407, "learning_rate": 1.7037037037037038e-06, "loss": 0.228, "step": 92 }, { "epoch": 0.01, "grad_norm": 5.238504380858114, "learning_rate": 1.7222222222222224e-06, "loss": 0.2445, "step": 93 }, { "epoch": 0.01, "grad_norm": 5.26550356218346, "learning_rate": 1.740740740740741e-06, "loss": 0.335, "step": 94 }, { "epoch": 0.01, "grad_norm": 4.084517792100619, "learning_rate": 1.7592592592592594e-06, "loss": 0.2466, "step": 95 }, { "epoch": 0.01, "grad_norm": 4.059980476806942, "learning_rate": 1.777777777777778e-06, "loss": 0.2352, "step": 96 }, { "epoch": 0.01, "grad_norm": 5.618509680814076, "learning_rate": 1.7962962962962965e-06, "loss": 0.2983, "step": 97 }, { "epoch": 0.01, "grad_norm": 3.2752739027377102, "learning_rate": 1.8148148148148151e-06, "loss": 0.2138, "step": 98 }, { "epoch": 0.01, "grad_norm": 4.531959879286411, "learning_rate": 1.8333333333333333e-06, "loss": 0.2109, "step": 99 }, { "epoch": 0.01, "grad_norm": 3.719039633234832, "learning_rate": 1.8518518518518519e-06, "loss": 0.2742, "step": 100 }, { "epoch": 0.01, "grad_norm": 4.791294915525922, "learning_rate": 1.8703703703703705e-06, "loss": 0.2604, "step": 101 }, { "epoch": 0.01, "grad_norm": 4.386141106337409, "learning_rate": 1.888888888888889e-06, "loss": 0.2207, "step": 102 }, { "epoch": 0.01, "grad_norm": 3.693752709660448, "learning_rate": 1.9074074074074076e-06, "loss": 0.1844, "step": 103 }, { "epoch": 0.01, "grad_norm": 3.539126599807116, "learning_rate": 1.925925925925926e-06, "loss": 0.2236, "step": 104 }, { "epoch": 0.01, "grad_norm": 5.701328883111506, "learning_rate": 1.944444444444445e-06, "loss": 0.2436, "step": 105 }, { "epoch": 0.01, "grad_norm": 3.49144964082344, "learning_rate": 1.962962962962963e-06, "loss": 0.2266, "step": 106 }, { "epoch": 0.01, "grad_norm": 3.6072130209912188, "learning_rate": 1.9814814814814815e-06, "loss": 0.2317, "step": 107 }, { "epoch": 0.01, "grad_norm": 3.0365130068969792, "learning_rate": 2.0000000000000003e-06, "loss": 0.2223, "step": 108 }, { "epoch": 0.01, "grad_norm": 13.354029999428898, "learning_rate": 2.0185185185185187e-06, "loss": 0.2359, "step": 109 }, { "epoch": 0.01, "grad_norm": 4.866215102048313, "learning_rate": 2.037037037037037e-06, "loss": 0.2178, "step": 110 }, { "epoch": 0.01, "grad_norm": 4.2919768835495775, "learning_rate": 2.0555555555555555e-06, "loss": 0.2053, "step": 111 }, { "epoch": 0.01, "grad_norm": 3.0247999172403386, "learning_rate": 2.0740740740740742e-06, "loss": 0.2116, "step": 112 }, { "epoch": 0.01, "grad_norm": 4.873767819329223, "learning_rate": 2.0925925925925926e-06, "loss": 0.2542, "step": 113 }, { "epoch": 0.01, "grad_norm": 3.290100400539165, "learning_rate": 2.1111111111111114e-06, "loss": 0.2353, "step": 114 }, { "epoch": 0.01, "grad_norm": 4.751597713000032, "learning_rate": 2.1296296296296298e-06, "loss": 0.2315, "step": 115 }, { "epoch": 0.01, "grad_norm": 4.473212528201221, "learning_rate": 2.148148148148148e-06, "loss": 0.2496, "step": 116 }, { "epoch": 0.01, "grad_norm": 5.783412858689247, "learning_rate": 2.166666666666667e-06, "loss": 0.2554, "step": 117 }, { "epoch": 0.01, "grad_norm": 4.699102842656633, "learning_rate": 2.1851851851851853e-06, "loss": 0.2953, "step": 118 }, { "epoch": 0.01, "grad_norm": 3.1020105552376935, "learning_rate": 2.203703703703704e-06, "loss": 0.2232, "step": 119 }, { "epoch": 0.01, "grad_norm": 4.231270542404818, "learning_rate": 2.222222222222222e-06, "loss": 0.2855, "step": 120 }, { "epoch": 0.01, "grad_norm": 5.1755404319160485, "learning_rate": 2.240740740740741e-06, "loss": 0.2506, "step": 121 }, { "epoch": 0.01, "grad_norm": 2.832933353259535, "learning_rate": 2.2592592592592592e-06, "loss": 0.2209, "step": 122 }, { "epoch": 0.01, "grad_norm": 3.3747496600378026, "learning_rate": 2.277777777777778e-06, "loss": 0.2641, "step": 123 }, { "epoch": 0.01, "grad_norm": 3.1614197231660426, "learning_rate": 2.2962962962962964e-06, "loss": 0.2591, "step": 124 }, { "epoch": 0.01, "grad_norm": 3.15164363985477, "learning_rate": 2.314814814814815e-06, "loss": 0.2195, "step": 125 }, { "epoch": 0.01, "grad_norm": 3.1520741854423773, "learning_rate": 2.3333333333333336e-06, "loss": 0.2327, "step": 126 }, { "epoch": 0.01, "grad_norm": 3.612537600311694, "learning_rate": 2.351851851851852e-06, "loss": 0.2544, "step": 127 }, { "epoch": 0.01, "grad_norm": 2.658574949293894, "learning_rate": 2.3703703703703707e-06, "loss": 0.1689, "step": 128 }, { "epoch": 0.01, "grad_norm": 2.521522811253935, "learning_rate": 2.388888888888889e-06, "loss": 0.2293, "step": 129 }, { "epoch": 0.01, "grad_norm": 5.564732601796836, "learning_rate": 2.4074074074074075e-06, "loss": 0.2667, "step": 130 }, { "epoch": 0.01, "grad_norm": 2.44070512393425, "learning_rate": 2.425925925925926e-06, "loss": 0.2093, "step": 131 }, { "epoch": 0.01, "grad_norm": 3.7275314191060787, "learning_rate": 2.4444444444444447e-06, "loss": 0.2531, "step": 132 }, { "epoch": 0.01, "grad_norm": 3.682061533051656, "learning_rate": 2.462962962962963e-06, "loss": 0.2493, "step": 133 }, { "epoch": 0.01, "grad_norm": 2.951520736342321, "learning_rate": 2.481481481481482e-06, "loss": 0.2651, "step": 134 }, { "epoch": 0.01, "grad_norm": 3.1993013028962376, "learning_rate": 2.5e-06, "loss": 0.2193, "step": 135 }, { "epoch": 0.01, "grad_norm": 3.4100755941983794, "learning_rate": 2.5185185185185186e-06, "loss": 0.2403, "step": 136 }, { "epoch": 0.01, "grad_norm": 3.3821935109396386, "learning_rate": 2.5370370370370374e-06, "loss": 0.2758, "step": 137 }, { "epoch": 0.01, "grad_norm": 2.6543322877297006, "learning_rate": 2.5555555555555557e-06, "loss": 0.2147, "step": 138 }, { "epoch": 0.01, "grad_norm": 3.416779857404282, "learning_rate": 2.5740740740740745e-06, "loss": 0.2329, "step": 139 }, { "epoch": 0.01, "grad_norm": 4.095895132385728, "learning_rate": 2.5925925925925925e-06, "loss": 0.2328, "step": 140 }, { "epoch": 0.01, "grad_norm": 2.802919951884446, "learning_rate": 2.6111111111111113e-06, "loss": 0.2641, "step": 141 }, { "epoch": 0.01, "grad_norm": 3.3297063178072643, "learning_rate": 2.6296296296296297e-06, "loss": 0.237, "step": 142 }, { "epoch": 0.01, "grad_norm": 3.369922845676044, "learning_rate": 2.6481481481481485e-06, "loss": 0.2538, "step": 143 }, { "epoch": 0.01, "grad_norm": 3.7670644169639695, "learning_rate": 2.666666666666667e-06, "loss": 0.2539, "step": 144 }, { "epoch": 0.01, "grad_norm": 3.4216695924186804, "learning_rate": 2.6851851851851856e-06, "loss": 0.2053, "step": 145 }, { "epoch": 0.01, "grad_norm": 3.221839005854741, "learning_rate": 2.703703703703704e-06, "loss": 0.2127, "step": 146 }, { "epoch": 0.01, "grad_norm": 3.0235480946266806, "learning_rate": 2.7222222222222224e-06, "loss": 0.2106, "step": 147 }, { "epoch": 0.01, "grad_norm": 2.834709595232568, "learning_rate": 2.740740740740741e-06, "loss": 0.1946, "step": 148 }, { "epoch": 0.01, "grad_norm": 3.015178509009479, "learning_rate": 2.759259259259259e-06, "loss": 0.2388, "step": 149 }, { "epoch": 0.01, "grad_norm": 4.409130426513994, "learning_rate": 2.7777777777777783e-06, "loss": 0.2415, "step": 150 }, { "epoch": 0.01, "grad_norm": 3.3739352577710386, "learning_rate": 2.7962962962962963e-06, "loss": 0.2457, "step": 151 }, { "epoch": 0.01, "grad_norm": 3.4098241760142383, "learning_rate": 2.814814814814815e-06, "loss": 0.251, "step": 152 }, { "epoch": 0.01, "grad_norm": 3.333754001338976, "learning_rate": 2.8333333333333335e-06, "loss": 0.1801, "step": 153 }, { "epoch": 0.01, "grad_norm": 3.318077097848955, "learning_rate": 2.8518518518518522e-06, "loss": 0.2458, "step": 154 }, { "epoch": 0.01, "grad_norm": 2.9466885051594103, "learning_rate": 2.8703703703703706e-06, "loss": 0.2406, "step": 155 }, { "epoch": 0.01, "grad_norm": 2.731062708528279, "learning_rate": 2.888888888888889e-06, "loss": 0.2001, "step": 156 }, { "epoch": 0.01, "grad_norm": 2.9272308425569653, "learning_rate": 2.907407407407408e-06, "loss": 0.2369, "step": 157 }, { "epoch": 0.01, "grad_norm": 2.481414362366478, "learning_rate": 2.9259259259259257e-06, "loss": 0.1691, "step": 158 }, { "epoch": 0.01, "grad_norm": 3.0329718074790053, "learning_rate": 2.944444444444445e-06, "loss": 0.2387, "step": 159 }, { "epoch": 0.01, "grad_norm": 3.391297972382365, "learning_rate": 2.962962962962963e-06, "loss": 0.2509, "step": 160 }, { "epoch": 0.01, "grad_norm": 3.569378052344309, "learning_rate": 2.9814814814814817e-06, "loss": 0.2017, "step": 161 }, { "epoch": 0.01, "grad_norm": 3.7110963646864787, "learning_rate": 3e-06, "loss": 0.2179, "step": 162 }, { "epoch": 0.01, "grad_norm": 2.246489859773125, "learning_rate": 3.018518518518519e-06, "loss": 0.1956, "step": 163 }, { "epoch": 0.01, "grad_norm": 4.009805700489273, "learning_rate": 3.0370370370370372e-06, "loss": 0.2621, "step": 164 }, { "epoch": 0.01, "grad_norm": 6.603600398514934, "learning_rate": 3.055555555555556e-06, "loss": 0.2286, "step": 165 }, { "epoch": 0.01, "grad_norm": 3.2140092102588227, "learning_rate": 3.0740740740740744e-06, "loss": 0.1985, "step": 166 }, { "epoch": 0.01, "grad_norm": 3.129528788571764, "learning_rate": 3.0925925925925928e-06, "loss": 0.2672, "step": 167 }, { "epoch": 0.01, "grad_norm": 4.012024850134178, "learning_rate": 3.1111111111111116e-06, "loss": 0.217, "step": 168 }, { "epoch": 0.01, "grad_norm": 7.820458104991282, "learning_rate": 3.1296296296296295e-06, "loss": 0.301, "step": 169 }, { "epoch": 0.01, "grad_norm": 3.0860558209491344, "learning_rate": 3.1481481481481483e-06, "loss": 0.3091, "step": 170 }, { "epoch": 0.01, "grad_norm": 2.7755722534108647, "learning_rate": 3.1666666666666667e-06, "loss": 0.2226, "step": 171 }, { "epoch": 0.01, "grad_norm": 2.5247435245938923, "learning_rate": 3.1851851851851855e-06, "loss": 0.1661, "step": 172 }, { "epoch": 0.01, "grad_norm": 3.01004084492849, "learning_rate": 3.203703703703704e-06, "loss": 0.21, "step": 173 }, { "epoch": 0.01, "grad_norm": 4.228753627219633, "learning_rate": 3.2222222222222227e-06, "loss": 0.2477, "step": 174 }, { "epoch": 0.01, "grad_norm": 3.1752871871583532, "learning_rate": 3.240740740740741e-06, "loss": 0.2079, "step": 175 }, { "epoch": 0.01, "grad_norm": 3.152560617770134, "learning_rate": 3.25925925925926e-06, "loss": 0.2665, "step": 176 }, { "epoch": 0.01, "grad_norm": 2.961029589788596, "learning_rate": 3.277777777777778e-06, "loss": 0.1568, "step": 177 }, { "epoch": 0.01, "grad_norm": 4.709782042755141, "learning_rate": 3.296296296296296e-06, "loss": 0.2532, "step": 178 }, { "epoch": 0.01, "grad_norm": 3.8365927717296775, "learning_rate": 3.314814814814815e-06, "loss": 0.2034, "step": 179 }, { "epoch": 0.01, "grad_norm": 3.693611221159195, "learning_rate": 3.3333333333333333e-06, "loss": 0.2731, "step": 180 }, { "epoch": 0.01, "grad_norm": 3.9986957942154158, "learning_rate": 3.351851851851852e-06, "loss": 0.2597, "step": 181 }, { "epoch": 0.01, "grad_norm": 3.2919659156622347, "learning_rate": 3.3703703703703705e-06, "loss": 0.1949, "step": 182 }, { "epoch": 0.01, "grad_norm": 2.9341893207437875, "learning_rate": 3.3888888888888893e-06, "loss": 0.2386, "step": 183 }, { "epoch": 0.01, "grad_norm": 5.102536761269667, "learning_rate": 3.4074074074074077e-06, "loss": 0.2774, "step": 184 }, { "epoch": 0.01, "grad_norm": 3.4536820389601535, "learning_rate": 3.4259259259259265e-06, "loss": 0.1826, "step": 185 }, { "epoch": 0.01, "grad_norm": 3.0300989423466786, "learning_rate": 3.444444444444445e-06, "loss": 0.2446, "step": 186 }, { "epoch": 0.01, "grad_norm": 3.4403011526481473, "learning_rate": 3.4629629629629628e-06, "loss": 0.1901, "step": 187 }, { "epoch": 0.01, "grad_norm": 3.6775091247134197, "learning_rate": 3.481481481481482e-06, "loss": 0.2674, "step": 188 }, { "epoch": 0.01, "grad_norm": 4.391028178699074, "learning_rate": 3.5e-06, "loss": 0.2829, "step": 189 }, { "epoch": 0.01, "grad_norm": 2.423366192150436, "learning_rate": 3.5185185185185187e-06, "loss": 0.1824, "step": 190 }, { "epoch": 0.01, "grad_norm": 3.122177437193447, "learning_rate": 3.537037037037037e-06, "loss": 0.2371, "step": 191 }, { "epoch": 0.01, "grad_norm": 3.45841158448972, "learning_rate": 3.555555555555556e-06, "loss": 0.2292, "step": 192 }, { "epoch": 0.01, "grad_norm": 3.4469803213869747, "learning_rate": 3.5740740740740743e-06, "loss": 0.1955, "step": 193 }, { "epoch": 0.01, "grad_norm": 2.9771563520791173, "learning_rate": 3.592592592592593e-06, "loss": 0.2377, "step": 194 }, { "epoch": 0.01, "grad_norm": 3.8796906691318185, "learning_rate": 3.6111111111111115e-06, "loss": 0.2551, "step": 195 }, { "epoch": 0.01, "grad_norm": 3.468676128116348, "learning_rate": 3.6296296296296302e-06, "loss": 0.2039, "step": 196 }, { "epoch": 0.01, "grad_norm": 2.958880079590501, "learning_rate": 3.6481481481481486e-06, "loss": 0.1795, "step": 197 }, { "epoch": 0.01, "grad_norm": 2.6200401360225767, "learning_rate": 3.6666666666666666e-06, "loss": 0.232, "step": 198 }, { "epoch": 0.01, "grad_norm": 4.983293081071033, "learning_rate": 3.6851851851851854e-06, "loss": 0.1908, "step": 199 }, { "epoch": 0.01, "grad_norm": 3.0050236938894415, "learning_rate": 3.7037037037037037e-06, "loss": 0.2286, "step": 200 }, { "epoch": 0.01, "grad_norm": 3.3306140725673634, "learning_rate": 3.7222222222222225e-06, "loss": 0.2619, "step": 201 }, { "epoch": 0.01, "grad_norm": 3.4938522953418802, "learning_rate": 3.740740740740741e-06, "loss": 0.1959, "step": 202 }, { "epoch": 0.01, "grad_norm": 3.4177239728199855, "learning_rate": 3.7592592592592597e-06, "loss": 0.2236, "step": 203 }, { "epoch": 0.01, "grad_norm": 3.7652783689393297, "learning_rate": 3.777777777777778e-06, "loss": 0.2085, "step": 204 }, { "epoch": 0.01, "grad_norm": 3.6604435338767782, "learning_rate": 3.796296296296297e-06, "loss": 0.261, "step": 205 }, { "epoch": 0.01, "grad_norm": 3.113607597716548, "learning_rate": 3.814814814814815e-06, "loss": 0.2424, "step": 206 }, { "epoch": 0.01, "grad_norm": 4.4771669455747585, "learning_rate": 3.833333333333334e-06, "loss": 0.2283, "step": 207 }, { "epoch": 0.01, "grad_norm": 3.1589659628880478, "learning_rate": 3.851851851851852e-06, "loss": 0.2219, "step": 208 }, { "epoch": 0.01, "grad_norm": 2.7645376675755027, "learning_rate": 3.87037037037037e-06, "loss": 0.2135, "step": 209 }, { "epoch": 0.01, "grad_norm": 2.980265299201392, "learning_rate": 3.88888888888889e-06, "loss": 0.2686, "step": 210 }, { "epoch": 0.01, "grad_norm": 3.1181423283269614, "learning_rate": 3.907407407407408e-06, "loss": 0.223, "step": 211 }, { "epoch": 0.01, "grad_norm": 2.42736974480654, "learning_rate": 3.925925925925926e-06, "loss": 0.187, "step": 212 }, { "epoch": 0.01, "grad_norm": 1.8743075840685295, "learning_rate": 3.944444444444445e-06, "loss": 0.1818, "step": 213 }, { "epoch": 0.01, "grad_norm": 2.343322079058555, "learning_rate": 3.962962962962963e-06, "loss": 0.2371, "step": 214 }, { "epoch": 0.01, "grad_norm": 2.081694878178825, "learning_rate": 3.9814814814814814e-06, "loss": 0.2267, "step": 215 }, { "epoch": 0.01, "grad_norm": 4.3576806169853315, "learning_rate": 4.000000000000001e-06, "loss": 0.2567, "step": 216 }, { "epoch": 0.01, "grad_norm": 3.4811683148674706, "learning_rate": 4.018518518518519e-06, "loss": 0.2124, "step": 217 }, { "epoch": 0.01, "grad_norm": 3.5712780387693677, "learning_rate": 4.037037037037037e-06, "loss": 0.175, "step": 218 }, { "epoch": 0.01, "grad_norm": 3.10431155350948, "learning_rate": 4.055555555555556e-06, "loss": 0.2544, "step": 219 }, { "epoch": 0.01, "grad_norm": 3.4035525621721305, "learning_rate": 4.074074074074074e-06, "loss": 0.2443, "step": 220 }, { "epoch": 0.01, "grad_norm": 2.497804840268957, "learning_rate": 4.092592592592593e-06, "loss": 0.2103, "step": 221 }, { "epoch": 0.01, "grad_norm": 3.453502720676144, "learning_rate": 4.111111111111111e-06, "loss": 0.1928, "step": 222 }, { "epoch": 0.01, "grad_norm": 2.5319715107359095, "learning_rate": 4.12962962962963e-06, "loss": 0.3045, "step": 223 }, { "epoch": 0.01, "grad_norm": 3.652891605515027, "learning_rate": 4.1481481481481485e-06, "loss": 0.2066, "step": 224 }, { "epoch": 0.01, "grad_norm": 5.197667005680618, "learning_rate": 4.166666666666667e-06, "loss": 0.2867, "step": 225 }, { "epoch": 0.01, "grad_norm": 3.1852861083681456, "learning_rate": 4.185185185185185e-06, "loss": 0.2962, "step": 226 }, { "epoch": 0.01, "grad_norm": 3.309751449945396, "learning_rate": 4.2037037037037045e-06, "loss": 0.2399, "step": 227 }, { "epoch": 0.01, "grad_norm": 5.662600750900519, "learning_rate": 4.222222222222223e-06, "loss": 0.1835, "step": 228 }, { "epoch": 0.01, "grad_norm": 3.1627381367040655, "learning_rate": 4.240740740740741e-06, "loss": 0.2256, "step": 229 }, { "epoch": 0.01, "grad_norm": 2.859414772632586, "learning_rate": 4.2592592592592596e-06, "loss": 0.2474, "step": 230 }, { "epoch": 0.01, "grad_norm": 3.5930782685464924, "learning_rate": 4.277777777777778e-06, "loss": 0.2853, "step": 231 }, { "epoch": 0.01, "grad_norm": 2.3413935831851864, "learning_rate": 4.296296296296296e-06, "loss": 0.2001, "step": 232 }, { "epoch": 0.01, "grad_norm": 2.1343012773284142, "learning_rate": 4.314814814814815e-06, "loss": 0.1442, "step": 233 }, { "epoch": 0.01, "grad_norm": 4.447325533939575, "learning_rate": 4.333333333333334e-06, "loss": 0.279, "step": 234 }, { "epoch": 0.01, "grad_norm": 2.0281971364645366, "learning_rate": 4.351851851851852e-06, "loss": 0.1831, "step": 235 }, { "epoch": 0.01, "grad_norm": 2.399177738362128, "learning_rate": 4.370370370370371e-06, "loss": 0.2159, "step": 236 }, { "epoch": 0.01, "grad_norm": 3.9459614890346613, "learning_rate": 4.388888888888889e-06, "loss": 0.2585, "step": 237 }, { "epoch": 0.01, "grad_norm": 4.2484957894089685, "learning_rate": 4.407407407407408e-06, "loss": 0.2717, "step": 238 }, { "epoch": 0.01, "grad_norm": 3.671599268198715, "learning_rate": 4.425925925925927e-06, "loss": 0.2394, "step": 239 }, { "epoch": 0.01, "grad_norm": 3.005918803380453, "learning_rate": 4.444444444444444e-06, "loss": 0.2524, "step": 240 }, { "epoch": 0.01, "grad_norm": 2.6518617846617962, "learning_rate": 4.462962962962963e-06, "loss": 0.2379, "step": 241 }, { "epoch": 0.01, "grad_norm": 2.437999918608316, "learning_rate": 4.481481481481482e-06, "loss": 0.2195, "step": 242 }, { "epoch": 0.01, "grad_norm": 2.6583787466273825, "learning_rate": 4.5e-06, "loss": 0.1729, "step": 243 }, { "epoch": 0.01, "grad_norm": 2.402951258317827, "learning_rate": 4.5185185185185185e-06, "loss": 0.2234, "step": 244 }, { "epoch": 0.01, "grad_norm": 2.7857996129322897, "learning_rate": 4.537037037037038e-06, "loss": 0.2547, "step": 245 }, { "epoch": 0.01, "grad_norm": 2.49488576274901, "learning_rate": 4.555555555555556e-06, "loss": 0.2407, "step": 246 }, { "epoch": 0.01, "grad_norm": 1.868135271583363, "learning_rate": 4.5740740740740745e-06, "loss": 0.2208, "step": 247 }, { "epoch": 0.01, "grad_norm": 2.5562428105913084, "learning_rate": 4.592592592592593e-06, "loss": 0.2166, "step": 248 }, { "epoch": 0.01, "grad_norm": 2.58224768027302, "learning_rate": 4.611111111111112e-06, "loss": 0.2492, "step": 249 }, { "epoch": 0.01, "grad_norm": 2.3433573075972296, "learning_rate": 4.62962962962963e-06, "loss": 0.1936, "step": 250 }, { "epoch": 0.01, "grad_norm": 2.8626887969261507, "learning_rate": 4.648148148148148e-06, "loss": 0.2315, "step": 251 }, { "epoch": 0.01, "grad_norm": 2.5816386699164062, "learning_rate": 4.666666666666667e-06, "loss": 0.2163, "step": 252 }, { "epoch": 0.01, "grad_norm": 2.3400604453390237, "learning_rate": 4.6851851851851855e-06, "loss": 0.2611, "step": 253 }, { "epoch": 0.01, "grad_norm": 1.8458488911384965, "learning_rate": 4.703703703703704e-06, "loss": 0.2489, "step": 254 }, { "epoch": 0.01, "grad_norm": 2.522444603513024, "learning_rate": 4.722222222222222e-06, "loss": 0.2122, "step": 255 }, { "epoch": 0.01, "grad_norm": 2.3783927325650613, "learning_rate": 4.7407407407407415e-06, "loss": 0.1981, "step": 256 }, { "epoch": 0.01, "grad_norm": 3.019533613828299, "learning_rate": 4.75925925925926e-06, "loss": 0.2384, "step": 257 }, { "epoch": 0.01, "grad_norm": 2.918759132999941, "learning_rate": 4.777777777777778e-06, "loss": 0.1788, "step": 258 }, { "epoch": 0.01, "grad_norm": 2.4610427470635274, "learning_rate": 4.796296296296297e-06, "loss": 0.2268, "step": 259 }, { "epoch": 0.01, "grad_norm": 2.7848761312000536, "learning_rate": 4.814814814814815e-06, "loss": 0.2325, "step": 260 }, { "epoch": 0.01, "grad_norm": 3.4191520473690815, "learning_rate": 4.833333333333333e-06, "loss": 0.2172, "step": 261 }, { "epoch": 0.01, "grad_norm": 3.103758968658333, "learning_rate": 4.851851851851852e-06, "loss": 0.2478, "step": 262 }, { "epoch": 0.01, "grad_norm": 2.5482504471380185, "learning_rate": 4.870370370370371e-06, "loss": 0.2497, "step": 263 }, { "epoch": 0.01, "grad_norm": 3.4810407020892766, "learning_rate": 4.888888888888889e-06, "loss": 0.1416, "step": 264 }, { "epoch": 0.01, "grad_norm": 3.3881716312252794, "learning_rate": 4.907407407407408e-06, "loss": 0.2587, "step": 265 }, { "epoch": 0.01, "grad_norm": 2.3292168482752666, "learning_rate": 4.925925925925926e-06, "loss": 0.2184, "step": 266 }, { "epoch": 0.01, "grad_norm": 4.639069216123649, "learning_rate": 4.944444444444445e-06, "loss": 0.2131, "step": 267 }, { "epoch": 0.01, "grad_norm": 2.775550585330974, "learning_rate": 4.962962962962964e-06, "loss": 0.1981, "step": 268 }, { "epoch": 0.01, "grad_norm": 2.6776742740622606, "learning_rate": 4.981481481481482e-06, "loss": 0.18, "step": 269 }, { "epoch": 0.02, "grad_norm": 4.811276614758999, "learning_rate": 5e-06, "loss": 0.2509, "step": 270 }, { "epoch": 0.02, "grad_norm": 3.316089340902278, "learning_rate": 5.01851851851852e-06, "loss": 0.267, "step": 271 }, { "epoch": 0.02, "grad_norm": 2.212839461001947, "learning_rate": 5.037037037037037e-06, "loss": 0.2075, "step": 272 }, { "epoch": 0.02, "grad_norm": 1.9776739041544913, "learning_rate": 5.0555555555555555e-06, "loss": 0.1698, "step": 273 }, { "epoch": 0.02, "grad_norm": 3.1890958082449243, "learning_rate": 5.074074074074075e-06, "loss": 0.2372, "step": 274 }, { "epoch": 0.02, "grad_norm": 2.947525306217703, "learning_rate": 5.092592592592593e-06, "loss": 0.2561, "step": 275 }, { "epoch": 0.02, "grad_norm": 4.297391853458588, "learning_rate": 5.1111111111111115e-06, "loss": 0.2481, "step": 276 }, { "epoch": 0.02, "grad_norm": 3.322602754472144, "learning_rate": 5.12962962962963e-06, "loss": 0.2631, "step": 277 }, { "epoch": 0.02, "grad_norm": 3.050443252029387, "learning_rate": 5.148148148148149e-06, "loss": 0.2261, "step": 278 }, { "epoch": 0.02, "grad_norm": 1.8973907756153363, "learning_rate": 5.1666666666666675e-06, "loss": 0.2428, "step": 279 }, { "epoch": 0.02, "grad_norm": 4.0422139591962765, "learning_rate": 5.185185185185185e-06, "loss": 0.237, "step": 280 }, { "epoch": 0.02, "grad_norm": 3.3841461829891855, "learning_rate": 5.203703703703704e-06, "loss": 0.2163, "step": 281 }, { "epoch": 0.02, "grad_norm": 2.5529633837409493, "learning_rate": 5.2222222222222226e-06, "loss": 0.1871, "step": 282 }, { "epoch": 0.02, "grad_norm": 3.3624447392598986, "learning_rate": 5.240740740740741e-06, "loss": 0.2218, "step": 283 }, { "epoch": 0.02, "grad_norm": 2.2956439501496666, "learning_rate": 5.259259259259259e-06, "loss": 0.2417, "step": 284 }, { "epoch": 0.02, "grad_norm": 2.297097785680975, "learning_rate": 5.2777777777777785e-06, "loss": 0.2266, "step": 285 }, { "epoch": 0.02, "grad_norm": 1.7470113660102002, "learning_rate": 5.296296296296297e-06, "loss": 0.191, "step": 286 }, { "epoch": 0.02, "grad_norm": 2.3084465857781824, "learning_rate": 5.314814814814815e-06, "loss": 0.2177, "step": 287 }, { "epoch": 0.02, "grad_norm": 2.7920950850278485, "learning_rate": 5.333333333333334e-06, "loss": 0.2735, "step": 288 }, { "epoch": 0.02, "grad_norm": 5.370310217840505, "learning_rate": 5.351851851851853e-06, "loss": 0.2444, "step": 289 }, { "epoch": 0.02, "grad_norm": 5.418966695111219, "learning_rate": 5.370370370370371e-06, "loss": 0.2502, "step": 290 }, { "epoch": 0.02, "grad_norm": 5.1345778728764495, "learning_rate": 5.388888888888889e-06, "loss": 0.2527, "step": 291 }, { "epoch": 0.02, "grad_norm": 4.809604219423627, "learning_rate": 5.407407407407408e-06, "loss": 0.309, "step": 292 }, { "epoch": 0.02, "grad_norm": 2.1252464263745963, "learning_rate": 5.425925925925926e-06, "loss": 0.1791, "step": 293 }, { "epoch": 0.02, "grad_norm": 2.382409471290975, "learning_rate": 5.444444444444445e-06, "loss": 0.1995, "step": 294 }, { "epoch": 0.02, "grad_norm": 2.78654941635765, "learning_rate": 5.462962962962963e-06, "loss": 0.2433, "step": 295 }, { "epoch": 0.02, "grad_norm": 4.162243895055201, "learning_rate": 5.481481481481482e-06, "loss": 0.2002, "step": 296 }, { "epoch": 0.02, "grad_norm": 3.020240729949975, "learning_rate": 5.500000000000001e-06, "loss": 0.2292, "step": 297 }, { "epoch": 0.02, "grad_norm": 5.455860053216916, "learning_rate": 5.518518518518518e-06, "loss": 0.2278, "step": 298 }, { "epoch": 0.02, "grad_norm": 2.0276007075998423, "learning_rate": 5.5370370370370374e-06, "loss": 0.2324, "step": 299 }, { "epoch": 0.02, "grad_norm": 2.6593618957429763, "learning_rate": 5.555555555555557e-06, "loss": 0.2855, "step": 300 }, { "epoch": 0.02, "grad_norm": 3.3040564094529894, "learning_rate": 5.574074074074075e-06, "loss": 0.2893, "step": 301 }, { "epoch": 0.02, "grad_norm": 2.0887879744067774, "learning_rate": 5.5925925925925926e-06, "loss": 0.2725, "step": 302 }, { "epoch": 0.02, "grad_norm": 2.256352232891461, "learning_rate": 5.611111111111112e-06, "loss": 0.2021, "step": 303 }, { "epoch": 0.02, "grad_norm": 2.5862358935281637, "learning_rate": 5.62962962962963e-06, "loss": 0.2192, "step": 304 }, { "epoch": 0.02, "grad_norm": 2.253995397778198, "learning_rate": 5.6481481481481485e-06, "loss": 0.2372, "step": 305 }, { "epoch": 0.02, "grad_norm": 2.7795628348705796, "learning_rate": 5.666666666666667e-06, "loss": 0.2946, "step": 306 }, { "epoch": 0.02, "grad_norm": 1.6514569015090887, "learning_rate": 5.685185185185186e-06, "loss": 0.2368, "step": 307 }, { "epoch": 0.02, "grad_norm": 2.7901136196750858, "learning_rate": 5.7037037037037045e-06, "loss": 0.2328, "step": 308 }, { "epoch": 0.02, "grad_norm": 2.929473431713666, "learning_rate": 5.722222222222222e-06, "loss": 0.2245, "step": 309 }, { "epoch": 0.02, "grad_norm": 2.7336612641838127, "learning_rate": 5.740740740740741e-06, "loss": 0.2357, "step": 310 }, { "epoch": 0.02, "grad_norm": 2.833964247208147, "learning_rate": 5.75925925925926e-06, "loss": 0.2282, "step": 311 }, { "epoch": 0.02, "grad_norm": 3.404103521479665, "learning_rate": 5.777777777777778e-06, "loss": 0.2043, "step": 312 }, { "epoch": 0.02, "grad_norm": 2.4901435985101377, "learning_rate": 5.796296296296296e-06, "loss": 0.1827, "step": 313 }, { "epoch": 0.02, "grad_norm": 2.4801730482808333, "learning_rate": 5.814814814814816e-06, "loss": 0.2216, "step": 314 }, { "epoch": 0.02, "grad_norm": 3.1306132066915424, "learning_rate": 5.833333333333334e-06, "loss": 0.2295, "step": 315 }, { "epoch": 0.02, "grad_norm": 2.1971949417613255, "learning_rate": 5.8518518518518515e-06, "loss": 0.2182, "step": 316 }, { "epoch": 0.02, "grad_norm": 2.0099244050381335, "learning_rate": 5.870370370370371e-06, "loss": 0.2124, "step": 317 }, { "epoch": 0.02, "grad_norm": 10.956187648563132, "learning_rate": 5.88888888888889e-06, "loss": 0.2244, "step": 318 }, { "epoch": 0.02, "grad_norm": 2.237500079384062, "learning_rate": 5.907407407407408e-06, "loss": 0.2017, "step": 319 }, { "epoch": 0.02, "grad_norm": 2.7735158452533324, "learning_rate": 5.925925925925926e-06, "loss": 0.2866, "step": 320 }, { "epoch": 0.02, "grad_norm": 3.0634574172078675, "learning_rate": 5.944444444444445e-06, "loss": 0.2174, "step": 321 }, { "epoch": 0.02, "grad_norm": 2.082726231968373, "learning_rate": 5.962962962962963e-06, "loss": 0.221, "step": 322 }, { "epoch": 0.02, "grad_norm": 2.5734571016094367, "learning_rate": 5.981481481481482e-06, "loss": 0.2007, "step": 323 }, { "epoch": 0.02, "grad_norm": 2.6433284734602704, "learning_rate": 6e-06, "loss": 0.2226, "step": 324 }, { "epoch": 0.02, "grad_norm": 1.9103685818787577, "learning_rate": 6.018518518518519e-06, "loss": 0.2193, "step": 325 }, { "epoch": 0.02, "grad_norm": 2.3024757052308895, "learning_rate": 6.037037037037038e-06, "loss": 0.2581, "step": 326 }, { "epoch": 0.02, "grad_norm": 3.275150115088954, "learning_rate": 6.055555555555555e-06, "loss": 0.2276, "step": 327 }, { "epoch": 0.02, "grad_norm": 2.9732833180868155, "learning_rate": 6.0740740740740745e-06, "loss": 0.2077, "step": 328 }, { "epoch": 0.02, "grad_norm": 1.7655336812127842, "learning_rate": 6.092592592592593e-06, "loss": 0.2234, "step": 329 }, { "epoch": 0.02, "grad_norm": 2.8087553526925393, "learning_rate": 6.111111111111112e-06, "loss": 0.247, "step": 330 }, { "epoch": 0.02, "grad_norm": 2.171279420768631, "learning_rate": 6.12962962962963e-06, "loss": 0.2527, "step": 331 }, { "epoch": 0.02, "grad_norm": 2.4271539806076436, "learning_rate": 6.148148148148149e-06, "loss": 0.1873, "step": 332 }, { "epoch": 0.02, "grad_norm": 2.2561373188486256, "learning_rate": 6.166666666666667e-06, "loss": 0.2243, "step": 333 }, { "epoch": 0.02, "grad_norm": 2.0341046960919478, "learning_rate": 6.1851851851851856e-06, "loss": 0.263, "step": 334 }, { "epoch": 0.02, "grad_norm": 2.2655957647607807, "learning_rate": 6.203703703703704e-06, "loss": 0.2711, "step": 335 }, { "epoch": 0.02, "grad_norm": 1.8712782002364605, "learning_rate": 6.222222222222223e-06, "loss": 0.2466, "step": 336 }, { "epoch": 0.02, "grad_norm": 1.3390799589142879, "learning_rate": 6.2407407407407415e-06, "loss": 0.1433, "step": 337 }, { "epoch": 0.02, "grad_norm": 3.373163217946868, "learning_rate": 6.259259259259259e-06, "loss": 0.1725, "step": 338 }, { "epoch": 0.02, "grad_norm": 2.394808070251205, "learning_rate": 6.277777777777778e-06, "loss": 0.2197, "step": 339 }, { "epoch": 0.02, "grad_norm": 2.755518848297966, "learning_rate": 6.296296296296297e-06, "loss": 0.272, "step": 340 }, { "epoch": 0.02, "grad_norm": 1.8779923323935852, "learning_rate": 6.314814814814816e-06, "loss": 0.2232, "step": 341 }, { "epoch": 0.02, "grad_norm": 2.6353096651260834, "learning_rate": 6.333333333333333e-06, "loss": 0.2798, "step": 342 }, { "epoch": 0.02, "grad_norm": 2.763581485415892, "learning_rate": 6.351851851851853e-06, "loss": 0.1904, "step": 343 }, { "epoch": 0.02, "grad_norm": 2.758094713722052, "learning_rate": 6.370370370370371e-06, "loss": 0.2441, "step": 344 }, { "epoch": 0.02, "grad_norm": 1.9172636589026844, "learning_rate": 6.3888888888888885e-06, "loss": 0.2762, "step": 345 }, { "epoch": 0.02, "grad_norm": 2.394315538755901, "learning_rate": 6.407407407407408e-06, "loss": 0.2145, "step": 346 }, { "epoch": 0.02, "grad_norm": 2.4548469501382484, "learning_rate": 6.425925925925927e-06, "loss": 0.1939, "step": 347 }, { "epoch": 0.02, "grad_norm": 2.2452509086512755, "learning_rate": 6.444444444444445e-06, "loss": 0.2615, "step": 348 }, { "epoch": 0.02, "grad_norm": 3.098802544447235, "learning_rate": 6.462962962962963e-06, "loss": 0.2461, "step": 349 }, { "epoch": 0.02, "grad_norm": 2.616470147506204, "learning_rate": 6.481481481481482e-06, "loss": 0.2117, "step": 350 }, { "epoch": 0.02, "grad_norm": 2.2334691592329525, "learning_rate": 6.5000000000000004e-06, "loss": 0.2131, "step": 351 }, { "epoch": 0.02, "grad_norm": 2.5589799662286956, "learning_rate": 6.51851851851852e-06, "loss": 0.242, "step": 352 }, { "epoch": 0.02, "grad_norm": 1.659120456475795, "learning_rate": 6.537037037037037e-06, "loss": 0.1708, "step": 353 }, { "epoch": 0.02, "grad_norm": 3.0372077950479426, "learning_rate": 6.555555555555556e-06, "loss": 0.2708, "step": 354 }, { "epoch": 0.02, "grad_norm": 3.1775196661741996, "learning_rate": 6.574074074074075e-06, "loss": 0.2874, "step": 355 }, { "epoch": 0.02, "grad_norm": 2.830893410296079, "learning_rate": 6.592592592592592e-06, "loss": 0.2716, "step": 356 }, { "epoch": 0.02, "grad_norm": 3.7230523009836536, "learning_rate": 6.6111111111111115e-06, "loss": 0.1982, "step": 357 }, { "epoch": 0.02, "grad_norm": 1.893792415370512, "learning_rate": 6.62962962962963e-06, "loss": 0.1864, "step": 358 }, { "epoch": 0.02, "grad_norm": 2.666844618467573, "learning_rate": 6.648148148148149e-06, "loss": 0.2428, "step": 359 }, { "epoch": 0.02, "grad_norm": 3.608780254558276, "learning_rate": 6.666666666666667e-06, "loss": 0.2006, "step": 360 }, { "epoch": 0.02, "grad_norm": 3.302008298067494, "learning_rate": 6.685185185185186e-06, "loss": 0.201, "step": 361 }, { "epoch": 0.02, "grad_norm": 2.7545141320246036, "learning_rate": 6.703703703703704e-06, "loss": 0.2023, "step": 362 }, { "epoch": 0.02, "grad_norm": 3.310552294341296, "learning_rate": 6.7222222222222235e-06, "loss": 0.2524, "step": 363 }, { "epoch": 0.02, "grad_norm": 10.135064668152676, "learning_rate": 6.740740740740741e-06, "loss": 0.3195, "step": 364 }, { "epoch": 0.02, "grad_norm": 2.9020320884158277, "learning_rate": 6.75925925925926e-06, "loss": 0.2071, "step": 365 }, { "epoch": 0.02, "grad_norm": 2.27841075156875, "learning_rate": 6.777777777777779e-06, "loss": 0.2421, "step": 366 }, { "epoch": 0.02, "grad_norm": 3.4100360302965322, "learning_rate": 6.796296296296296e-06, "loss": 0.2714, "step": 367 }, { "epoch": 0.02, "grad_norm": 2.5363251000308034, "learning_rate": 6.814814814814815e-06, "loss": 0.244, "step": 368 }, { "epoch": 0.02, "grad_norm": 2.592834724808257, "learning_rate": 6.833333333333334e-06, "loss": 0.1717, "step": 369 }, { "epoch": 0.02, "grad_norm": 2.667653937916638, "learning_rate": 6.851851851851853e-06, "loss": 0.1939, "step": 370 }, { "epoch": 0.02, "grad_norm": 2.4491773781331814, "learning_rate": 6.8703703703703704e-06, "loss": 0.2883, "step": 371 }, { "epoch": 0.02, "grad_norm": 2.491279484468612, "learning_rate": 6.88888888888889e-06, "loss": 0.284, "step": 372 }, { "epoch": 0.02, "grad_norm": 3.0752719398369, "learning_rate": 6.907407407407408e-06, "loss": 0.2466, "step": 373 }, { "epoch": 0.02, "grad_norm": 2.9643781997595835, "learning_rate": 6.9259259259259256e-06, "loss": 0.2628, "step": 374 }, { "epoch": 0.02, "grad_norm": 2.0006146510934153, "learning_rate": 6.944444444444445e-06, "loss": 0.1835, "step": 375 }, { "epoch": 0.02, "grad_norm": 3.8235231337722313, "learning_rate": 6.962962962962964e-06, "loss": 0.3059, "step": 376 }, { "epoch": 0.02, "grad_norm": 2.05096254546488, "learning_rate": 6.981481481481482e-06, "loss": 0.2287, "step": 377 }, { "epoch": 0.02, "grad_norm": 2.4587295800172235, "learning_rate": 7e-06, "loss": 0.2654, "step": 378 }, { "epoch": 0.02, "grad_norm": 2.6549597748171516, "learning_rate": 7.018518518518519e-06, "loss": 0.3164, "step": 379 }, { "epoch": 0.02, "grad_norm": 1.9038346967059707, "learning_rate": 7.0370370370370375e-06, "loss": 0.2488, "step": 380 }, { "epoch": 0.02, "grad_norm": 2.3697078179976567, "learning_rate": 7.055555555555557e-06, "loss": 0.2037, "step": 381 }, { "epoch": 0.02, "grad_norm": 2.0954183152096797, "learning_rate": 7.074074074074074e-06, "loss": 0.1585, "step": 382 }, { "epoch": 0.02, "grad_norm": 2.2018921290610343, "learning_rate": 7.0925925925925935e-06, "loss": 0.1813, "step": 383 }, { "epoch": 0.02, "grad_norm": 1.5869216684012442, "learning_rate": 7.111111111111112e-06, "loss": 0.2381, "step": 384 }, { "epoch": 0.02, "grad_norm": 1.5440442501750655, "learning_rate": 7.129629629629629e-06, "loss": 0.1969, "step": 385 }, { "epoch": 0.02, "grad_norm": 2.2971557912660665, "learning_rate": 7.1481481481481486e-06, "loss": 0.2147, "step": 386 }, { "epoch": 0.02, "grad_norm": 2.550293529670386, "learning_rate": 7.166666666666667e-06, "loss": 0.1814, "step": 387 }, { "epoch": 0.02, "grad_norm": 1.8004297380922656, "learning_rate": 7.185185185185186e-06, "loss": 0.2686, "step": 388 }, { "epoch": 0.02, "grad_norm": 2.1173214359781203, "learning_rate": 7.203703703703704e-06, "loss": 0.2166, "step": 389 }, { "epoch": 0.02, "grad_norm": 1.518227044480827, "learning_rate": 7.222222222222223e-06, "loss": 0.1604, "step": 390 }, { "epoch": 0.02, "grad_norm": 1.8852826175316413, "learning_rate": 7.240740740740741e-06, "loss": 0.1795, "step": 391 }, { "epoch": 0.02, "grad_norm": 2.30584778108288, "learning_rate": 7.2592592592592605e-06, "loss": 0.2376, "step": 392 }, { "epoch": 0.02, "grad_norm": 1.5760050357974753, "learning_rate": 7.277777777777778e-06, "loss": 0.1886, "step": 393 }, { "epoch": 0.02, "grad_norm": 1.6085933249919047, "learning_rate": 7.296296296296297e-06, "loss": 0.1625, "step": 394 }, { "epoch": 0.02, "grad_norm": 2.8386472676550367, "learning_rate": 7.314814814814816e-06, "loss": 0.2517, "step": 395 }, { "epoch": 0.02, "grad_norm": 2.6820386575918245, "learning_rate": 7.333333333333333e-06, "loss": 0.253, "step": 396 }, { "epoch": 0.02, "grad_norm": 2.7885866053003387, "learning_rate": 7.351851851851852e-06, "loss": 0.2369, "step": 397 }, { "epoch": 0.02, "grad_norm": 2.2231783865351793, "learning_rate": 7.370370370370371e-06, "loss": 0.2185, "step": 398 }, { "epoch": 0.02, "grad_norm": 2.86403769874136, "learning_rate": 7.38888888888889e-06, "loss": 0.191, "step": 399 }, { "epoch": 0.02, "grad_norm": 2.6112782053430785, "learning_rate": 7.4074074074074075e-06, "loss": 0.2541, "step": 400 }, { "epoch": 0.02, "grad_norm": 2.1721369907794355, "learning_rate": 7.425925925925927e-06, "loss": 0.2131, "step": 401 }, { "epoch": 0.02, "grad_norm": 2.3261842511133697, "learning_rate": 7.444444444444445e-06, "loss": 0.2542, "step": 402 }, { "epoch": 0.02, "grad_norm": 2.4321492348789997, "learning_rate": 7.462962962962964e-06, "loss": 0.2691, "step": 403 }, { "epoch": 0.02, "grad_norm": 1.564152501781726, "learning_rate": 7.481481481481482e-06, "loss": 0.15, "step": 404 }, { "epoch": 0.02, "grad_norm": 2.345816845324918, "learning_rate": 7.500000000000001e-06, "loss": 0.2501, "step": 405 }, { "epoch": 0.02, "grad_norm": 2.015424997549943, "learning_rate": 7.518518518518519e-06, "loss": 0.2193, "step": 406 }, { "epoch": 0.02, "grad_norm": 2.6450364209319477, "learning_rate": 7.537037037037037e-06, "loss": 0.2726, "step": 407 }, { "epoch": 0.02, "grad_norm": 1.665537149427856, "learning_rate": 7.555555555555556e-06, "loss": 0.1735, "step": 408 }, { "epoch": 0.02, "grad_norm": 2.318529079431767, "learning_rate": 7.5740740740740745e-06, "loss": 0.2626, "step": 409 }, { "epoch": 0.02, "grad_norm": 1.9981876819623996, "learning_rate": 7.592592592592594e-06, "loss": 0.2393, "step": 410 }, { "epoch": 0.02, "grad_norm": 2.0379069550854196, "learning_rate": 7.611111111111111e-06, "loss": 0.2069, "step": 411 }, { "epoch": 0.02, "grad_norm": 1.9110119780954062, "learning_rate": 7.62962962962963e-06, "loss": 0.231, "step": 412 }, { "epoch": 0.02, "grad_norm": 1.6627143090730263, "learning_rate": 7.64814814814815e-06, "loss": 0.2442, "step": 413 }, { "epoch": 0.02, "grad_norm": 2.3799133298824615, "learning_rate": 7.666666666666667e-06, "loss": 0.281, "step": 414 }, { "epoch": 0.02, "grad_norm": 1.5833113318773921, "learning_rate": 7.685185185185185e-06, "loss": 0.221, "step": 415 }, { "epoch": 0.02, "grad_norm": 1.9057489268142138, "learning_rate": 7.703703703703704e-06, "loss": 0.2688, "step": 416 }, { "epoch": 0.02, "grad_norm": 1.8121203361980647, "learning_rate": 7.722222222222223e-06, "loss": 0.2347, "step": 417 }, { "epoch": 0.02, "grad_norm": 3.0687752410914997, "learning_rate": 7.74074074074074e-06, "loss": 0.2464, "step": 418 }, { "epoch": 0.02, "grad_norm": 1.8131616737058565, "learning_rate": 7.75925925925926e-06, "loss": 0.251, "step": 419 }, { "epoch": 0.02, "grad_norm": 1.9091346676877106, "learning_rate": 7.77777777777778e-06, "loss": 0.2727, "step": 420 }, { "epoch": 0.02, "grad_norm": 2.46372123585346, "learning_rate": 7.796296296296297e-06, "loss": 0.287, "step": 421 }, { "epoch": 0.02, "grad_norm": 2.8823349203033533, "learning_rate": 7.814814814814816e-06, "loss": 0.2292, "step": 422 }, { "epoch": 0.02, "grad_norm": 1.4743929849468036, "learning_rate": 7.833333333333333e-06, "loss": 0.2142, "step": 423 }, { "epoch": 0.02, "grad_norm": 2.2897706352334475, "learning_rate": 7.851851851851853e-06, "loss": 0.2657, "step": 424 }, { "epoch": 0.02, "grad_norm": 1.6830692137954337, "learning_rate": 7.870370370370372e-06, "loss": 0.2479, "step": 425 }, { "epoch": 0.02, "grad_norm": 1.5407121550492335, "learning_rate": 7.88888888888889e-06, "loss": 0.2246, "step": 426 }, { "epoch": 0.02, "grad_norm": 2.412774183574061, "learning_rate": 7.907407407407409e-06, "loss": 0.2763, "step": 427 }, { "epoch": 0.02, "grad_norm": 1.964345565494054, "learning_rate": 7.925925925925926e-06, "loss": 0.2587, "step": 428 }, { "epoch": 0.02, "grad_norm": 2.0450440983284652, "learning_rate": 7.944444444444445e-06, "loss": 0.1765, "step": 429 }, { "epoch": 0.02, "grad_norm": 1.7867699472742486, "learning_rate": 7.962962962962963e-06, "loss": 0.1863, "step": 430 }, { "epoch": 0.02, "grad_norm": 2.5398855665291156, "learning_rate": 7.981481481481482e-06, "loss": 0.3173, "step": 431 }, { "epoch": 0.02, "grad_norm": 2.046668493613578, "learning_rate": 8.000000000000001e-06, "loss": 0.2758, "step": 432 }, { "epoch": 0.02, "grad_norm": 1.6390293672556615, "learning_rate": 8.018518518518519e-06, "loss": 0.2715, "step": 433 }, { "epoch": 0.02, "grad_norm": 2.4504855224681408, "learning_rate": 8.037037037037038e-06, "loss": 0.2455, "step": 434 }, { "epoch": 0.02, "grad_norm": 2.599909976400243, "learning_rate": 8.055555555555557e-06, "loss": 0.2278, "step": 435 }, { "epoch": 0.02, "grad_norm": 2.234139456704803, "learning_rate": 8.074074074074075e-06, "loss": 0.2529, "step": 436 }, { "epoch": 0.02, "grad_norm": 1.7070969104141578, "learning_rate": 8.092592592592592e-06, "loss": 0.2011, "step": 437 }, { "epoch": 0.02, "grad_norm": 1.8321578382667267, "learning_rate": 8.111111111111112e-06, "loss": 0.2051, "step": 438 }, { "epoch": 0.02, "grad_norm": 1.9908165571547292, "learning_rate": 8.12962962962963e-06, "loss": 0.1968, "step": 439 }, { "epoch": 0.02, "grad_norm": 2.4006332821096077, "learning_rate": 8.148148148148148e-06, "loss": 0.2538, "step": 440 }, { "epoch": 0.02, "grad_norm": 2.6387770354396216, "learning_rate": 8.166666666666668e-06, "loss": 0.1666, "step": 441 }, { "epoch": 0.02, "grad_norm": 2.292311358079653, "learning_rate": 8.185185185185187e-06, "loss": 0.2268, "step": 442 }, { "epoch": 0.02, "grad_norm": 1.9779324740690347, "learning_rate": 8.203703703703704e-06, "loss": 0.2034, "step": 443 }, { "epoch": 0.02, "grad_norm": 1.9459054495158559, "learning_rate": 8.222222222222222e-06, "loss": 0.1963, "step": 444 }, { "epoch": 0.02, "grad_norm": 2.727768704983483, "learning_rate": 8.240740740740741e-06, "loss": 0.248, "step": 445 }, { "epoch": 0.02, "grad_norm": 1.762347648339636, "learning_rate": 8.25925925925926e-06, "loss": 0.2705, "step": 446 }, { "epoch": 0.02, "grad_norm": 2.0538712972668325, "learning_rate": 8.277777777777778e-06, "loss": 0.2157, "step": 447 }, { "epoch": 0.02, "grad_norm": 2.8976656510690924, "learning_rate": 8.296296296296297e-06, "loss": 0.2878, "step": 448 }, { "epoch": 0.02, "grad_norm": 1.917974747844571, "learning_rate": 8.314814814814816e-06, "loss": 0.2703, "step": 449 }, { "epoch": 0.03, "grad_norm": 2.225709818814864, "learning_rate": 8.333333333333334e-06, "loss": 0.2594, "step": 450 }, { "epoch": 0.03, "grad_norm": 2.1943687433117205, "learning_rate": 8.351851851851851e-06, "loss": 0.2184, "step": 451 }, { "epoch": 0.03, "grad_norm": 1.918405805463593, "learning_rate": 8.37037037037037e-06, "loss": 0.2855, "step": 452 }, { "epoch": 0.03, "grad_norm": 1.723128091977472, "learning_rate": 8.38888888888889e-06, "loss": 0.1916, "step": 453 }, { "epoch": 0.03, "grad_norm": 1.958157567411811, "learning_rate": 8.407407407407409e-06, "loss": 0.1901, "step": 454 }, { "epoch": 0.03, "grad_norm": 1.4763912525479064, "learning_rate": 8.425925925925926e-06, "loss": 0.273, "step": 455 }, { "epoch": 0.03, "grad_norm": 1.5981461116737503, "learning_rate": 8.444444444444446e-06, "loss": 0.2516, "step": 456 }, { "epoch": 0.03, "grad_norm": 1.1663828771437823, "learning_rate": 8.462962962962963e-06, "loss": 0.1716, "step": 457 }, { "epoch": 0.03, "grad_norm": 1.4781190549476864, "learning_rate": 8.481481481481482e-06, "loss": 0.2412, "step": 458 }, { "epoch": 0.03, "grad_norm": 1.6193928401336815, "learning_rate": 8.5e-06, "loss": 0.2716, "step": 459 }, { "epoch": 0.03, "grad_norm": 1.6339004624552358, "learning_rate": 8.518518518518519e-06, "loss": 0.1968, "step": 460 }, { "epoch": 0.03, "grad_norm": 1.260078140056335, "learning_rate": 8.537037037037038e-06, "loss": 0.2227, "step": 461 }, { "epoch": 0.03, "grad_norm": 1.7074133840265566, "learning_rate": 8.555555555555556e-06, "loss": 0.219, "step": 462 }, { "epoch": 0.03, "grad_norm": 1.5611418448064829, "learning_rate": 8.574074074074075e-06, "loss": 0.1963, "step": 463 }, { "epoch": 0.03, "grad_norm": 1.6694050753250427, "learning_rate": 8.592592592592593e-06, "loss": 0.205, "step": 464 }, { "epoch": 0.03, "grad_norm": 1.5774347382027096, "learning_rate": 8.611111111111112e-06, "loss": 0.2314, "step": 465 }, { "epoch": 0.03, "grad_norm": 1.707162173333124, "learning_rate": 8.62962962962963e-06, "loss": 0.2333, "step": 466 }, { "epoch": 0.03, "grad_norm": 1.586900107102287, "learning_rate": 8.648148148148149e-06, "loss": 0.1828, "step": 467 }, { "epoch": 0.03, "grad_norm": 1.8205988008032974, "learning_rate": 8.666666666666668e-06, "loss": 0.2781, "step": 468 }, { "epoch": 0.03, "grad_norm": 2.011013273961918, "learning_rate": 8.685185185185185e-06, "loss": 0.2364, "step": 469 }, { "epoch": 0.03, "grad_norm": 2.347424978140611, "learning_rate": 8.703703703703705e-06, "loss": 0.2951, "step": 470 }, { "epoch": 0.03, "grad_norm": 1.5832688941725674, "learning_rate": 8.722222222222224e-06, "loss": 0.1645, "step": 471 }, { "epoch": 0.03, "grad_norm": 1.0163279823952933, "learning_rate": 8.740740740740741e-06, "loss": 0.2116, "step": 472 }, { "epoch": 0.03, "grad_norm": 1.599090443091723, "learning_rate": 8.759259259259259e-06, "loss": 0.2338, "step": 473 }, { "epoch": 0.03, "grad_norm": 1.5651329244699874, "learning_rate": 8.777777777777778e-06, "loss": 0.2141, "step": 474 }, { "epoch": 0.03, "grad_norm": 1.644723516265944, "learning_rate": 8.796296296296297e-06, "loss": 0.2205, "step": 475 }, { "epoch": 0.03, "grad_norm": 1.5697078465922618, "learning_rate": 8.814814814814817e-06, "loss": 0.1948, "step": 476 }, { "epoch": 0.03, "grad_norm": 1.4385062614115038, "learning_rate": 8.833333333333334e-06, "loss": 0.2373, "step": 477 }, { "epoch": 0.03, "grad_norm": 1.911017504582339, "learning_rate": 8.851851851851853e-06, "loss": 0.205, "step": 478 }, { "epoch": 0.03, "grad_norm": 2.069808868113705, "learning_rate": 8.87037037037037e-06, "loss": 0.2354, "step": 479 }, { "epoch": 0.03, "grad_norm": 2.159608690490603, "learning_rate": 8.888888888888888e-06, "loss": 0.2083, "step": 480 }, { "epoch": 0.03, "grad_norm": 1.4701455010284599, "learning_rate": 8.907407407407408e-06, "loss": 0.2135, "step": 481 }, { "epoch": 0.03, "grad_norm": 1.8772804379750925, "learning_rate": 8.925925925925927e-06, "loss": 0.1787, "step": 482 }, { "epoch": 0.03, "grad_norm": 1.3222927854219297, "learning_rate": 8.944444444444446e-06, "loss": 0.2602, "step": 483 }, { "epoch": 0.03, "grad_norm": 1.9651963465511846, "learning_rate": 8.962962962962963e-06, "loss": 0.1747, "step": 484 }, { "epoch": 0.03, "grad_norm": 1.9266834331682001, "learning_rate": 8.981481481481483e-06, "loss": 0.1919, "step": 485 }, { "epoch": 0.03, "grad_norm": 1.3278082610152648, "learning_rate": 9e-06, "loss": 0.2168, "step": 486 }, { "epoch": 0.03, "grad_norm": 2.0527498297286506, "learning_rate": 9.01851851851852e-06, "loss": 0.2822, "step": 487 }, { "epoch": 0.03, "grad_norm": 1.4196503456996559, "learning_rate": 9.037037037037037e-06, "loss": 0.2282, "step": 488 }, { "epoch": 0.03, "grad_norm": 1.7355838892748336, "learning_rate": 9.055555555555556e-06, "loss": 0.2613, "step": 489 }, { "epoch": 0.03, "grad_norm": 2.3124993169628887, "learning_rate": 9.074074074074075e-06, "loss": 0.2076, "step": 490 }, { "epoch": 0.03, "grad_norm": 1.55742161398934, "learning_rate": 9.092592592592593e-06, "loss": 0.1866, "step": 491 }, { "epoch": 0.03, "grad_norm": 4.12557886859245, "learning_rate": 9.111111111111112e-06, "loss": 0.1925, "step": 492 }, { "epoch": 0.03, "grad_norm": 2.1746743501029098, "learning_rate": 9.12962962962963e-06, "loss": 0.1506, "step": 493 }, { "epoch": 0.03, "grad_norm": 2.524886716057685, "learning_rate": 9.148148148148149e-06, "loss": 0.2345, "step": 494 }, { "epoch": 0.03, "grad_norm": 2.179892451724111, "learning_rate": 9.166666666666666e-06, "loss": 0.2059, "step": 495 }, { "epoch": 0.03, "grad_norm": 1.913187582169457, "learning_rate": 9.185185185185186e-06, "loss": 0.2595, "step": 496 }, { "epoch": 0.03, "grad_norm": 1.6506808324873237, "learning_rate": 9.203703703703705e-06, "loss": 0.2009, "step": 497 }, { "epoch": 0.03, "grad_norm": 7.353623914997798, "learning_rate": 9.222222222222224e-06, "loss": 0.2359, "step": 498 }, { "epoch": 0.03, "grad_norm": 1.73397872452955, "learning_rate": 9.240740740740742e-06, "loss": 0.1601, "step": 499 }, { "epoch": 0.03, "grad_norm": 2.363800054935029, "learning_rate": 9.25925925925926e-06, "loss": 0.2178, "step": 500 }, { "epoch": 0.03, "grad_norm": 3.174926117165057, "learning_rate": 9.277777777777778e-06, "loss": 0.2262, "step": 501 }, { "epoch": 0.03, "grad_norm": 2.6754495456894998, "learning_rate": 9.296296296296296e-06, "loss": 0.2648, "step": 502 }, { "epoch": 0.03, "grad_norm": 2.1212048896630415, "learning_rate": 9.314814814814815e-06, "loss": 0.3119, "step": 503 }, { "epoch": 0.03, "grad_norm": 1.845141499967797, "learning_rate": 9.333333333333334e-06, "loss": 0.214, "step": 504 }, { "epoch": 0.03, "grad_norm": 1.6034717530580787, "learning_rate": 9.351851851851854e-06, "loss": 0.1861, "step": 505 }, { "epoch": 0.03, "grad_norm": 1.494876834777987, "learning_rate": 9.370370370370371e-06, "loss": 0.1951, "step": 506 }, { "epoch": 0.03, "grad_norm": 2.567749735780204, "learning_rate": 9.38888888888889e-06, "loss": 0.2589, "step": 507 }, { "epoch": 0.03, "grad_norm": 2.156161180339284, "learning_rate": 9.407407407407408e-06, "loss": 0.2119, "step": 508 }, { "epoch": 0.03, "grad_norm": 2.1757434074631155, "learning_rate": 9.425925925925925e-06, "loss": 0.2806, "step": 509 }, { "epoch": 0.03, "grad_norm": 1.6893535871344756, "learning_rate": 9.444444444444445e-06, "loss": 0.2379, "step": 510 }, { "epoch": 0.03, "grad_norm": 1.7028334459930194, "learning_rate": 9.462962962962964e-06, "loss": 0.2342, "step": 511 }, { "epoch": 0.03, "grad_norm": 2.573367833466672, "learning_rate": 9.481481481481483e-06, "loss": 0.2728, "step": 512 }, { "epoch": 0.03, "grad_norm": 1.629665728283705, "learning_rate": 9.5e-06, "loss": 0.1883, "step": 513 }, { "epoch": 0.03, "grad_norm": 5.469608392104702, "learning_rate": 9.51851851851852e-06, "loss": 0.2699, "step": 514 }, { "epoch": 0.03, "grad_norm": 2.922868529122991, "learning_rate": 9.537037037037037e-06, "loss": 0.2474, "step": 515 }, { "epoch": 0.03, "grad_norm": 2.383420583648189, "learning_rate": 9.555555555555556e-06, "loss": 0.2799, "step": 516 }, { "epoch": 0.03, "grad_norm": 1.7284155714312717, "learning_rate": 9.574074074074074e-06, "loss": 0.2199, "step": 517 }, { "epoch": 0.03, "grad_norm": 1.8042763638037695, "learning_rate": 9.592592592592593e-06, "loss": 0.2357, "step": 518 }, { "epoch": 0.03, "grad_norm": 1.8240537997703765, "learning_rate": 9.611111111111112e-06, "loss": 0.2571, "step": 519 }, { "epoch": 0.03, "grad_norm": 2.240434759867715, "learning_rate": 9.62962962962963e-06, "loss": 0.2571, "step": 520 }, { "epoch": 0.03, "grad_norm": 2.2212595515134352, "learning_rate": 9.64814814814815e-06, "loss": 0.2805, "step": 521 }, { "epoch": 0.03, "grad_norm": 1.6706904432909366, "learning_rate": 9.666666666666667e-06, "loss": 0.2818, "step": 522 }, { "epoch": 0.03, "grad_norm": 1.882622112644092, "learning_rate": 9.685185185185186e-06, "loss": 0.2546, "step": 523 }, { "epoch": 0.03, "grad_norm": 2.003335137994384, "learning_rate": 9.703703703703703e-06, "loss": 0.2216, "step": 524 }, { "epoch": 0.03, "grad_norm": 1.8173690992162865, "learning_rate": 9.722222222222223e-06, "loss": 0.2595, "step": 525 }, { "epoch": 0.03, "grad_norm": 2.4188288081930738, "learning_rate": 9.740740740740742e-06, "loss": 0.2453, "step": 526 }, { "epoch": 0.03, "grad_norm": 3.275836038529327, "learning_rate": 9.759259259259261e-06, "loss": 0.2442, "step": 527 }, { "epoch": 0.03, "grad_norm": 1.3602762550601097, "learning_rate": 9.777777777777779e-06, "loss": 0.2221, "step": 528 }, { "epoch": 0.03, "grad_norm": 1.4099576775919647, "learning_rate": 9.796296296296298e-06, "loss": 0.2381, "step": 529 }, { "epoch": 0.03, "grad_norm": 1.948210621819492, "learning_rate": 9.814814814814815e-06, "loss": 0.2942, "step": 530 }, { "epoch": 0.03, "grad_norm": 1.6352317839316701, "learning_rate": 9.833333333333333e-06, "loss": 0.2261, "step": 531 }, { "epoch": 0.03, "grad_norm": 2.07004143541843, "learning_rate": 9.851851851851852e-06, "loss": 0.1908, "step": 532 }, { "epoch": 0.03, "grad_norm": 2.15105594941232, "learning_rate": 9.870370370370371e-06, "loss": 0.2878, "step": 533 }, { "epoch": 0.03, "grad_norm": 2.0665011897929353, "learning_rate": 9.88888888888889e-06, "loss": 0.2648, "step": 534 }, { "epoch": 0.03, "grad_norm": 1.3177102461777124, "learning_rate": 9.907407407407408e-06, "loss": 0.1964, "step": 535 }, { "epoch": 0.03, "grad_norm": 1.6617003227758134, "learning_rate": 9.925925925925927e-06, "loss": 0.2061, "step": 536 }, { "epoch": 0.03, "grad_norm": 1.8712692178466874, "learning_rate": 9.944444444444445e-06, "loss": 0.2097, "step": 537 }, { "epoch": 0.03, "grad_norm": 1.2364405901770519, "learning_rate": 9.962962962962964e-06, "loss": 0.2007, "step": 538 }, { "epoch": 0.03, "grad_norm": 1.60292206138748, "learning_rate": 9.981481481481482e-06, "loss": 0.2364, "step": 539 }, { "epoch": 0.03, "grad_norm": 1.8072231920574293, "learning_rate": 1e-05, "loss": 0.2238, "step": 540 }, { "epoch": 0.03, "grad_norm": 2.4814728889599436, "learning_rate": 9.999999918848563e-06, "loss": 0.2314, "step": 541 }, { "epoch": 0.03, "grad_norm": 1.7318830824208753, "learning_rate": 9.999999675394253e-06, "loss": 0.2389, "step": 542 }, { "epoch": 0.03, "grad_norm": 1.3036697895542924, "learning_rate": 9.999999269637079e-06, "loss": 0.1491, "step": 543 }, { "epoch": 0.03, "grad_norm": 1.8513482106621912, "learning_rate": 9.999998701577052e-06, "loss": 0.2016, "step": 544 }, { "epoch": 0.03, "grad_norm": 2.2647752911372168, "learning_rate": 9.999997971214192e-06, "loss": 0.2791, "step": 545 }, { "epoch": 0.03, "grad_norm": 2.897363047140287, "learning_rate": 9.999997078548524e-06, "loss": 0.2325, "step": 546 }, { "epoch": 0.03, "grad_norm": 1.5670415512729239, "learning_rate": 9.999996023580074e-06, "loss": 0.2251, "step": 547 }, { "epoch": 0.03, "grad_norm": 2.6066296463318857, "learning_rate": 9.99999480630888e-06, "loss": 0.2611, "step": 548 }, { "epoch": 0.03, "grad_norm": 1.4653422867881003, "learning_rate": 9.999993426734977e-06, "loss": 0.2573, "step": 549 }, { "epoch": 0.03, "grad_norm": 1.5471928587812813, "learning_rate": 9.999991884858413e-06, "loss": 0.2317, "step": 550 }, { "epoch": 0.03, "grad_norm": 2.1788321358326717, "learning_rate": 9.999990180679238e-06, "loss": 0.2661, "step": 551 }, { "epoch": 0.03, "grad_norm": 1.9518145246539516, "learning_rate": 9.999988314197507e-06, "loss": 0.1772, "step": 552 }, { "epoch": 0.03, "grad_norm": 1.6427564762279832, "learning_rate": 9.999986285413278e-06, "loss": 0.2232, "step": 553 }, { "epoch": 0.03, "grad_norm": 1.7725487637369426, "learning_rate": 9.99998409432662e-06, "loss": 0.1818, "step": 554 }, { "epoch": 0.03, "grad_norm": 1.7108703016558175, "learning_rate": 9.999981740937604e-06, "loss": 0.1876, "step": 555 }, { "epoch": 0.03, "grad_norm": 1.8711769627044077, "learning_rate": 9.999979225246305e-06, "loss": 0.3121, "step": 556 }, { "epoch": 0.03, "grad_norm": 2.7045900117922645, "learning_rate": 9.999976547252805e-06, "loss": 0.1916, "step": 557 }, { "epoch": 0.03, "grad_norm": 2.048128523173138, "learning_rate": 9.999973706957191e-06, "loss": 0.2373, "step": 558 }, { "epoch": 0.03, "grad_norm": 1.9221990901329062, "learning_rate": 9.999970704359555e-06, "loss": 0.2004, "step": 559 }, { "epoch": 0.03, "grad_norm": 2.1596146848014603, "learning_rate": 9.999967539459996e-06, "loss": 0.2664, "step": 560 }, { "epoch": 0.03, "grad_norm": 1.2490925057696007, "learning_rate": 9.999964212258612e-06, "loss": 0.1911, "step": 561 }, { "epoch": 0.03, "grad_norm": 1.7084579310879742, "learning_rate": 9.999960722755519e-06, "loss": 0.2211, "step": 562 }, { "epoch": 0.03, "grad_norm": 1.8416896032362209, "learning_rate": 9.999957070950824e-06, "loss": 0.2526, "step": 563 }, { "epoch": 0.03, "grad_norm": 1.7889661492301776, "learning_rate": 9.999953256844646e-06, "loss": 0.2224, "step": 564 }, { "epoch": 0.03, "grad_norm": 1.8237790858313632, "learning_rate": 9.999949280437111e-06, "loss": 0.1634, "step": 565 }, { "epoch": 0.03, "grad_norm": 1.4797735469173663, "learning_rate": 9.999945141728348e-06, "loss": 0.2205, "step": 566 }, { "epoch": 0.03, "grad_norm": 2.4426823569233567, "learning_rate": 9.99994084071849e-06, "loss": 0.3041, "step": 567 }, { "epoch": 0.03, "grad_norm": 1.743574684996697, "learning_rate": 9.999936377407677e-06, "loss": 0.2238, "step": 568 }, { "epoch": 0.03, "grad_norm": 2.032513000231454, "learning_rate": 9.999931751796055e-06, "loss": 0.1836, "step": 569 }, { "epoch": 0.03, "grad_norm": 2.299920431087586, "learning_rate": 9.99992696388377e-06, "loss": 0.2829, "step": 570 }, { "epoch": 0.03, "grad_norm": 1.456484249618141, "learning_rate": 9.999922013670984e-06, "loss": 0.1619, "step": 571 }, { "epoch": 0.03, "grad_norm": 1.8247965197914737, "learning_rate": 9.999916901157852e-06, "loss": 0.2324, "step": 572 }, { "epoch": 0.03, "grad_norm": 1.8103485101634382, "learning_rate": 9.999911626344544e-06, "loss": 0.2591, "step": 573 }, { "epoch": 0.03, "grad_norm": 1.4997873254847325, "learning_rate": 9.999906189231228e-06, "loss": 0.2105, "step": 574 }, { "epoch": 0.03, "grad_norm": 1.6577781968518213, "learning_rate": 9.999900589818083e-06, "loss": 0.2225, "step": 575 }, { "epoch": 0.03, "grad_norm": 1.0842225165438453, "learning_rate": 9.99989482810529e-06, "loss": 0.1728, "step": 576 }, { "epoch": 0.03, "grad_norm": 1.4611910074773522, "learning_rate": 9.999888904093035e-06, "loss": 0.193, "step": 577 }, { "epoch": 0.03, "grad_norm": 1.2211246342575959, "learning_rate": 9.99988281778151e-06, "loss": 0.1969, "step": 578 }, { "epoch": 0.03, "grad_norm": 1.8041481599306388, "learning_rate": 9.999876569170916e-06, "loss": 0.291, "step": 579 }, { "epoch": 0.03, "grad_norm": 2.436450340993971, "learning_rate": 9.99987015826145e-06, "loss": 0.3033, "step": 580 }, { "epoch": 0.03, "grad_norm": 1.5594826042435612, "learning_rate": 9.999863585053327e-06, "loss": 0.1409, "step": 581 }, { "epoch": 0.03, "grad_norm": 1.9152897928877035, "learning_rate": 9.999856849546757e-06, "loss": 0.267, "step": 582 }, { "epoch": 0.03, "grad_norm": 1.6390096341345133, "learning_rate": 9.999849951741955e-06, "loss": 0.2145, "step": 583 }, { "epoch": 0.03, "grad_norm": 1.903619272035446, "learning_rate": 9.999842891639151e-06, "loss": 0.2076, "step": 584 }, { "epoch": 0.03, "grad_norm": 2.055727158980303, "learning_rate": 9.999835669238571e-06, "loss": 0.2839, "step": 585 }, { "epoch": 0.03, "grad_norm": 1.464435144020495, "learning_rate": 9.999828284540451e-06, "loss": 0.178, "step": 586 }, { "epoch": 0.03, "grad_norm": 1.2701477725215002, "learning_rate": 9.99982073754503e-06, "loss": 0.2349, "step": 587 }, { "epoch": 0.03, "grad_norm": 1.1983313414618901, "learning_rate": 9.999813028252551e-06, "loss": 0.2407, "step": 588 }, { "epoch": 0.03, "grad_norm": 1.4233421299492355, "learning_rate": 9.999805156663267e-06, "loss": 0.2286, "step": 589 }, { "epoch": 0.03, "grad_norm": 1.4703401938433902, "learning_rate": 9.999797122777433e-06, "loss": 0.1695, "step": 590 }, { "epoch": 0.03, "grad_norm": 1.2974353988467726, "learning_rate": 9.999788926595309e-06, "loss": 0.2403, "step": 591 }, { "epoch": 0.03, "grad_norm": 1.754992726258613, "learning_rate": 9.99978056811716e-06, "loss": 0.2765, "step": 592 }, { "epoch": 0.03, "grad_norm": 2.109167652182334, "learning_rate": 9.999772047343259e-06, "loss": 0.2608, "step": 593 }, { "epoch": 0.03, "grad_norm": 1.7035748167716884, "learning_rate": 9.999763364273884e-06, "loss": 0.2052, "step": 594 }, { "epoch": 0.03, "grad_norm": 1.569435155660177, "learning_rate": 9.999754518909314e-06, "loss": 0.2147, "step": 595 }, { "epoch": 0.03, "grad_norm": 1.4123044469514268, "learning_rate": 9.999745511249837e-06, "loss": 0.2548, "step": 596 }, { "epoch": 0.03, "grad_norm": 1.2550973923976543, "learning_rate": 9.999736341295746e-06, "loss": 0.2043, "step": 597 }, { "epoch": 0.03, "grad_norm": 2.0765398045800105, "learning_rate": 9.999727009047337e-06, "loss": 0.2845, "step": 598 }, { "epoch": 0.03, "grad_norm": 2.6420925960269117, "learning_rate": 9.999717514504917e-06, "loss": 0.2908, "step": 599 }, { "epoch": 0.03, "grad_norm": 1.5955704665399983, "learning_rate": 9.99970785766879e-06, "loss": 0.204, "step": 600 }, { "epoch": 0.03, "grad_norm": 1.9013730574616672, "learning_rate": 9.99969803853927e-06, "loss": 0.2616, "step": 601 }, { "epoch": 0.03, "grad_norm": 1.300116200022206, "learning_rate": 9.999688057116679e-06, "loss": 0.1658, "step": 602 }, { "epoch": 0.03, "grad_norm": 2.1232774516287902, "learning_rate": 9.999677913401335e-06, "loss": 0.2344, "step": 603 }, { "epoch": 0.03, "grad_norm": 1.7319872675478898, "learning_rate": 9.999667607393576e-06, "loss": 0.2378, "step": 604 }, { "epoch": 0.03, "grad_norm": 1.4052874396755204, "learning_rate": 9.999657139093727e-06, "loss": 0.2119, "step": 605 }, { "epoch": 0.03, "grad_norm": 1.8178640032969893, "learning_rate": 9.999646508502134e-06, "loss": 0.2131, "step": 606 }, { "epoch": 0.03, "grad_norm": 2.824809053601352, "learning_rate": 9.99963571561914e-06, "loss": 0.2898, "step": 607 }, { "epoch": 0.03, "grad_norm": 18.329759841943982, "learning_rate": 9.999624760445095e-06, "loss": 0.3558, "step": 608 }, { "epoch": 0.03, "grad_norm": 4.959488207961797, "learning_rate": 9.999613642980358e-06, "loss": 0.2528, "step": 609 }, { "epoch": 0.03, "grad_norm": 10.810279662319726, "learning_rate": 9.999602363225284e-06, "loss": 0.2493, "step": 610 }, { "epoch": 0.03, "grad_norm": 2.752698701205404, "learning_rate": 9.999590921180244e-06, "loss": 0.276, "step": 611 }, { "epoch": 0.03, "grad_norm": 43.669199514580086, "learning_rate": 9.999579316845607e-06, "loss": 0.2679, "step": 612 }, { "epoch": 0.03, "grad_norm": 20.58903500382208, "learning_rate": 9.99956755022175e-06, "loss": 0.3086, "step": 613 }, { "epoch": 0.03, "grad_norm": 17.772749385717987, "learning_rate": 9.999555621309057e-06, "loss": 0.2392, "step": 614 }, { "epoch": 0.03, "grad_norm": 39.173913440838724, "learning_rate": 9.999543530107912e-06, "loss": 0.3374, "step": 615 }, { "epoch": 0.03, "grad_norm": 12.994662509599983, "learning_rate": 9.99953127661871e-06, "loss": 0.3013, "step": 616 }, { "epoch": 0.03, "grad_norm": 81.81398140634857, "learning_rate": 9.999518860841847e-06, "loss": 0.3234, "step": 617 }, { "epoch": 0.03, "grad_norm": 41.446118041182025, "learning_rate": 9.999506282777727e-06, "loss": 0.3536, "step": 618 }, { "epoch": 0.03, "grad_norm": 29.228570897145367, "learning_rate": 9.999493542426759e-06, "loss": 0.4052, "step": 619 }, { "epoch": 0.03, "grad_norm": 15.15354187357157, "learning_rate": 9.999480639789355e-06, "loss": 0.2878, "step": 620 }, { "epoch": 0.03, "grad_norm": 6.612631786742226, "learning_rate": 9.999467574865935e-06, "loss": 0.2893, "step": 621 }, { "epoch": 0.03, "grad_norm": 24.43898009674187, "learning_rate": 9.999454347656923e-06, "loss": 0.3064, "step": 622 }, { "epoch": 0.03, "grad_norm": 14.64972989090464, "learning_rate": 9.999440958162747e-06, "loss": 0.3189, "step": 623 }, { "epoch": 0.03, "grad_norm": 3.2441817068017054, "learning_rate": 9.999427406383843e-06, "loss": 0.2497, "step": 624 }, { "epoch": 0.03, "grad_norm": 6.36137774749413, "learning_rate": 9.99941369232065e-06, "loss": 0.304, "step": 625 }, { "epoch": 0.03, "grad_norm": 5.610408422117223, "learning_rate": 9.999399815973615e-06, "loss": 0.2792, "step": 626 }, { "epoch": 0.03, "grad_norm": 2.644214887304561, "learning_rate": 9.999385777343185e-06, "loss": 0.2389, "step": 627 }, { "epoch": 0.03, "grad_norm": 4.995091961958238, "learning_rate": 9.999371576429819e-06, "loss": 0.1999, "step": 628 }, { "epoch": 0.03, "grad_norm": 9.846007654361046, "learning_rate": 9.999357213233978e-06, "loss": 0.32, "step": 629 }, { "epoch": 0.04, "grad_norm": 14.346899613459236, "learning_rate": 9.999342687756126e-06, "loss": 0.229, "step": 630 }, { "epoch": 0.04, "grad_norm": 1.7492375075104223, "learning_rate": 9.999327999996735e-06, "loss": 0.2714, "step": 631 }, { "epoch": 0.04, "grad_norm": 1.713373966378423, "learning_rate": 9.999313149956282e-06, "loss": 0.2408, "step": 632 }, { "epoch": 0.04, "grad_norm": 1.5228538250927752, "learning_rate": 9.99929813763525e-06, "loss": 0.2474, "step": 633 }, { "epoch": 0.04, "grad_norm": 1.5658819838583038, "learning_rate": 9.999282963034126e-06, "loss": 0.2009, "step": 634 }, { "epoch": 0.04, "grad_norm": 1.3916280188471633, "learning_rate": 9.999267626153401e-06, "loss": 0.2009, "step": 635 }, { "epoch": 0.04, "grad_norm": 2.7101261122944824, "learning_rate": 9.999252126993574e-06, "loss": 0.2652, "step": 636 }, { "epoch": 0.04, "grad_norm": 1.7148812535384157, "learning_rate": 9.999236465555152e-06, "loss": 0.2354, "step": 637 }, { "epoch": 0.04, "grad_norm": 2.0234141624607167, "learning_rate": 9.999220641838637e-06, "loss": 0.2267, "step": 638 }, { "epoch": 0.04, "grad_norm": 1.3750651300784082, "learning_rate": 9.999204655844543e-06, "loss": 0.2218, "step": 639 }, { "epoch": 0.04, "grad_norm": 1.7894919600454031, "learning_rate": 9.999188507573394e-06, "loss": 0.2294, "step": 640 }, { "epoch": 0.04, "grad_norm": 1.253131242210937, "learning_rate": 9.99917219702571e-06, "loss": 0.2007, "step": 641 }, { "epoch": 0.04, "grad_norm": 1.7178947835267584, "learning_rate": 9.999155724202022e-06, "loss": 0.2208, "step": 642 }, { "epoch": 0.04, "grad_norm": 1.4817840701843323, "learning_rate": 9.999139089102866e-06, "loss": 0.1922, "step": 643 }, { "epoch": 0.04, "grad_norm": 2.2511555432970494, "learning_rate": 9.999122291728778e-06, "loss": 0.2371, "step": 644 }, { "epoch": 0.04, "grad_norm": 1.3644123822672882, "learning_rate": 9.999105332080306e-06, "loss": 0.1903, "step": 645 }, { "epoch": 0.04, "grad_norm": 1.2690162200868096, "learning_rate": 9.999088210158001e-06, "loss": 0.217, "step": 646 }, { "epoch": 0.04, "grad_norm": 1.7579445153572213, "learning_rate": 9.999070925962418e-06, "loss": 0.237, "step": 647 }, { "epoch": 0.04, "grad_norm": 2.7747401066383834, "learning_rate": 9.999053479494118e-06, "loss": 0.2233, "step": 648 }, { "epoch": 0.04, "grad_norm": 1.625083673597082, "learning_rate": 9.999035870753669e-06, "loss": 0.2504, "step": 649 }, { "epoch": 0.04, "grad_norm": 2.455600206556279, "learning_rate": 9.99901809974164e-06, "loss": 0.3197, "step": 650 }, { "epoch": 0.04, "grad_norm": 1.3696478585684608, "learning_rate": 9.999000166458607e-06, "loss": 0.1852, "step": 651 }, { "epoch": 0.04, "grad_norm": 2.9877473640982473, "learning_rate": 9.998982070905155e-06, "loss": 0.2738, "step": 652 }, { "epoch": 0.04, "grad_norm": 1.5267221376121667, "learning_rate": 9.99896381308187e-06, "loss": 0.2145, "step": 653 }, { "epoch": 0.04, "grad_norm": 2.62593829324627, "learning_rate": 9.998945392989347e-06, "loss": 0.2533, "step": 654 }, { "epoch": 0.04, "grad_norm": 2.001065789798318, "learning_rate": 9.99892681062818e-06, "loss": 0.2617, "step": 655 }, { "epoch": 0.04, "grad_norm": 2.603263756945268, "learning_rate": 9.998908065998975e-06, "loss": 0.2039, "step": 656 }, { "epoch": 0.04, "grad_norm": 1.956740344387672, "learning_rate": 9.998889159102339e-06, "loss": 0.2122, "step": 657 }, { "epoch": 0.04, "grad_norm": 1.584203293391321, "learning_rate": 9.998870089938886e-06, "loss": 0.1865, "step": 658 }, { "epoch": 0.04, "grad_norm": 2.053990771737211, "learning_rate": 9.998850858509237e-06, "loss": 0.2554, "step": 659 }, { "epoch": 0.04, "grad_norm": 2.405207222241233, "learning_rate": 9.998831464814014e-06, "loss": 0.254, "step": 660 }, { "epoch": 0.04, "grad_norm": 2.808261957851011, "learning_rate": 9.998811908853847e-06, "loss": 0.2548, "step": 661 }, { "epoch": 0.04, "grad_norm": 1.5276001400400812, "learning_rate": 9.998792190629369e-06, "loss": 0.2019, "step": 662 }, { "epoch": 0.04, "grad_norm": 2.2381691796438554, "learning_rate": 9.998772310141225e-06, "loss": 0.2737, "step": 663 }, { "epoch": 0.04, "grad_norm": 2.413871066209702, "learning_rate": 9.998752267390055e-06, "loss": 0.2401, "step": 664 }, { "epoch": 0.04, "grad_norm": 1.229324476334034, "learning_rate": 9.998732062376513e-06, "loss": 0.1934, "step": 665 }, { "epoch": 0.04, "grad_norm": 1.7869477329215326, "learning_rate": 9.998711695101254e-06, "loss": 0.2307, "step": 666 }, { "epoch": 0.04, "grad_norm": 1.2130204037635122, "learning_rate": 9.998691165564938e-06, "loss": 0.222, "step": 667 }, { "epoch": 0.04, "grad_norm": 1.5596636826017594, "learning_rate": 9.998670473768232e-06, "loss": 0.2044, "step": 668 }, { "epoch": 0.04, "grad_norm": 1.3948911007574005, "learning_rate": 9.998649619711808e-06, "loss": 0.2502, "step": 669 }, { "epoch": 0.04, "grad_norm": 2.359095121040132, "learning_rate": 9.998628603396345e-06, "loss": 0.2524, "step": 670 }, { "epoch": 0.04, "grad_norm": 1.7274161942575648, "learning_rate": 9.998607424822522e-06, "loss": 0.2523, "step": 671 }, { "epoch": 0.04, "grad_norm": 3.43300454353923, "learning_rate": 9.998586083991028e-06, "loss": 0.2306, "step": 672 }, { "epoch": 0.04, "grad_norm": 2.1139468627339784, "learning_rate": 9.998564580902555e-06, "loss": 0.2472, "step": 673 }, { "epoch": 0.04, "grad_norm": 2.1596330005523834, "learning_rate": 9.9985429155578e-06, "loss": 0.2363, "step": 674 }, { "epoch": 0.04, "grad_norm": 2.280091409544863, "learning_rate": 9.998521087957468e-06, "loss": 0.2438, "step": 675 }, { "epoch": 0.04, "grad_norm": 1.9700134628218318, "learning_rate": 9.998499098102268e-06, "loss": 0.2778, "step": 676 }, { "epoch": 0.04, "grad_norm": 2.06506438214983, "learning_rate": 9.998476945992913e-06, "loss": 0.2987, "step": 677 }, { "epoch": 0.04, "grad_norm": 1.8740205073967353, "learning_rate": 9.998454631630123e-06, "loss": 0.2418, "step": 678 }, { "epoch": 0.04, "grad_norm": 1.8872574773562758, "learning_rate": 9.998432155014619e-06, "loss": 0.2837, "step": 679 }, { "epoch": 0.04, "grad_norm": 1.4657243828131423, "learning_rate": 9.998409516147135e-06, "loss": 0.1898, "step": 680 }, { "epoch": 0.04, "grad_norm": 2.3584560221400532, "learning_rate": 9.998386715028403e-06, "loss": 0.2326, "step": 681 }, { "epoch": 0.04, "grad_norm": 1.8820926350629803, "learning_rate": 9.998363751659163e-06, "loss": 0.1695, "step": 682 }, { "epoch": 0.04, "grad_norm": 2.9141438654363108, "learning_rate": 9.998340626040162e-06, "loss": 0.2036, "step": 683 }, { "epoch": 0.04, "grad_norm": 1.7084189934789658, "learning_rate": 9.99831733817215e-06, "loss": 0.2271, "step": 684 }, { "epoch": 0.04, "grad_norm": 1.3658176248622982, "learning_rate": 9.998293888055882e-06, "loss": 0.2688, "step": 685 }, { "epoch": 0.04, "grad_norm": 1.4298539742233574, "learning_rate": 9.998270275692123e-06, "loss": 0.2012, "step": 686 }, { "epoch": 0.04, "grad_norm": 2.1142214501053354, "learning_rate": 9.998246501081635e-06, "loss": 0.2675, "step": 687 }, { "epoch": 0.04, "grad_norm": 1.7491912345971081, "learning_rate": 9.998222564225191e-06, "loss": 0.2313, "step": 688 }, { "epoch": 0.04, "grad_norm": 1.9556946882636561, "learning_rate": 9.998198465123569e-06, "loss": 0.2405, "step": 689 }, { "epoch": 0.04, "grad_norm": 1.733282763422793, "learning_rate": 9.998174203777549e-06, "loss": 0.2521, "step": 690 }, { "epoch": 0.04, "grad_norm": 1.5037118810910108, "learning_rate": 9.998149780187922e-06, "loss": 0.1822, "step": 691 }, { "epoch": 0.04, "grad_norm": 2.9634766172075806, "learning_rate": 9.998125194355479e-06, "loss": 0.2111, "step": 692 }, { "epoch": 0.04, "grad_norm": 1.5280912623747476, "learning_rate": 9.998100446281018e-06, "loss": 0.2512, "step": 693 }, { "epoch": 0.04, "grad_norm": 1.7715209591251764, "learning_rate": 9.998075535965342e-06, "loss": 0.2257, "step": 694 }, { "epoch": 0.04, "grad_norm": 1.9308633117820866, "learning_rate": 9.998050463409259e-06, "loss": 0.2138, "step": 695 }, { "epoch": 0.04, "grad_norm": 1.2120358669477806, "learning_rate": 9.998025228613586e-06, "loss": 0.2171, "step": 696 }, { "epoch": 0.04, "grad_norm": 3.948725910167626, "learning_rate": 9.997999831579138e-06, "loss": 0.2203, "step": 697 }, { "epoch": 0.04, "grad_norm": 1.3921277895023731, "learning_rate": 9.997974272306742e-06, "loss": 0.2406, "step": 698 }, { "epoch": 0.04, "grad_norm": 1.2562402812026698, "learning_rate": 9.997948550797227e-06, "loss": 0.165, "step": 699 }, { "epoch": 0.04, "grad_norm": 1.610502840874703, "learning_rate": 9.997922667051429e-06, "loss": 0.2431, "step": 700 }, { "epoch": 0.04, "grad_norm": 1.276328494150188, "learning_rate": 9.997896621070187e-06, "loss": 0.1982, "step": 701 }, { "epoch": 0.04, "grad_norm": 2.1114722669380392, "learning_rate": 9.997870412854347e-06, "loss": 0.2066, "step": 702 }, { "epoch": 0.04, "grad_norm": 2.78201244918868, "learning_rate": 9.997844042404758e-06, "loss": 0.2583, "step": 703 }, { "epoch": 0.04, "grad_norm": 1.535084620835797, "learning_rate": 9.997817509722279e-06, "loss": 0.2257, "step": 704 }, { "epoch": 0.04, "grad_norm": 1.8874994325321177, "learning_rate": 9.997790814807769e-06, "loss": 0.2009, "step": 705 }, { "epoch": 0.04, "grad_norm": 1.6047968049973862, "learning_rate": 9.997763957662094e-06, "loss": 0.2602, "step": 706 }, { "epoch": 0.04, "grad_norm": 1.7570357757703434, "learning_rate": 9.997736938286129e-06, "loss": 0.2485, "step": 707 }, { "epoch": 0.04, "grad_norm": 3.2229389968292472, "learning_rate": 9.997709756680746e-06, "loss": 0.2572, "step": 708 }, { "epoch": 0.04, "grad_norm": 1.8492538247369952, "learning_rate": 9.997682412846835e-06, "loss": 0.2053, "step": 709 }, { "epoch": 0.04, "grad_norm": 1.7469708897898746, "learning_rate": 9.997654906785274e-06, "loss": 0.2146, "step": 710 }, { "epoch": 0.04, "grad_norm": 2.2291057918531525, "learning_rate": 9.997627238496964e-06, "loss": 0.2099, "step": 711 }, { "epoch": 0.04, "grad_norm": 1.3402541063235494, "learning_rate": 9.9975994079828e-06, "loss": 0.1807, "step": 712 }, { "epoch": 0.04, "grad_norm": 1.922145502701153, "learning_rate": 9.997571415243684e-06, "loss": 0.236, "step": 713 }, { "epoch": 0.04, "grad_norm": 1.4461143048070895, "learning_rate": 9.997543260280527e-06, "loss": 0.1837, "step": 714 }, { "epoch": 0.04, "grad_norm": 1.3119016202921865, "learning_rate": 9.997514943094243e-06, "loss": 0.2211, "step": 715 }, { "epoch": 0.04, "grad_norm": 5.0686545931001525, "learning_rate": 9.997486463685748e-06, "loss": 0.1679, "step": 716 }, { "epoch": 0.04, "grad_norm": 2.8344728665347683, "learning_rate": 9.99745782205597e-06, "loss": 0.2857, "step": 717 }, { "epoch": 0.04, "grad_norm": 2.4148963426375163, "learning_rate": 9.997429018205838e-06, "loss": 0.246, "step": 718 }, { "epoch": 0.04, "grad_norm": 2.3087572791826942, "learning_rate": 9.997400052136284e-06, "loss": 0.2206, "step": 719 }, { "epoch": 0.04, "grad_norm": 1.8846658472718998, "learning_rate": 9.997370923848253e-06, "loss": 0.2204, "step": 720 }, { "epoch": 0.04, "grad_norm": 2.4933567951325903, "learning_rate": 9.997341633342685e-06, "loss": 0.1732, "step": 721 }, { "epoch": 0.04, "grad_norm": 1.4171284742165549, "learning_rate": 9.997312180620538e-06, "loss": 0.1865, "step": 722 }, { "epoch": 0.04, "grad_norm": 1.7608028746274034, "learning_rate": 9.99728256568276e-06, "loss": 0.2024, "step": 723 }, { "epoch": 0.04, "grad_norm": 2.12302828221747, "learning_rate": 9.997252788530318e-06, "loss": 0.2497, "step": 724 }, { "epoch": 0.04, "grad_norm": 26.926509571172836, "learning_rate": 9.997222849164179e-06, "loss": 0.2529, "step": 725 }, { "epoch": 0.04, "grad_norm": 9.988579785507007, "learning_rate": 9.997192747585309e-06, "loss": 0.2325, "step": 726 }, { "epoch": 0.04, "grad_norm": 2.073780616996417, "learning_rate": 9.997162483794692e-06, "loss": 0.2371, "step": 727 }, { "epoch": 0.04, "grad_norm": 83.88128606717183, "learning_rate": 9.997132057793306e-06, "loss": 0.2815, "step": 728 }, { "epoch": 0.04, "grad_norm": 1.6923507763339276, "learning_rate": 9.99710146958214e-06, "loss": 0.205, "step": 729 }, { "epoch": 0.04, "grad_norm": 1.6437592540596508, "learning_rate": 9.997070719162185e-06, "loss": 0.2277, "step": 730 }, { "epoch": 0.04, "grad_norm": 2.1464775022860074, "learning_rate": 9.997039806534441e-06, "loss": 0.2652, "step": 731 }, { "epoch": 0.04, "grad_norm": 1.7918314370547077, "learning_rate": 9.997008731699914e-06, "loss": 0.1765, "step": 732 }, { "epoch": 0.04, "grad_norm": 1.1126204754928446, "learning_rate": 9.99697749465961e-06, "loss": 0.2301, "step": 733 }, { "epoch": 0.04, "grad_norm": 1.9214132517978426, "learning_rate": 9.99694609541454e-06, "loss": 0.2226, "step": 734 }, { "epoch": 0.04, "grad_norm": 1.3356586968469206, "learning_rate": 9.996914533965729e-06, "loss": 0.2596, "step": 735 }, { "epoch": 0.04, "grad_norm": 2.049688458611854, "learning_rate": 9.996882810314198e-06, "loss": 0.2259, "step": 736 }, { "epoch": 0.04, "grad_norm": 2.5737350773575582, "learning_rate": 9.996850924460977e-06, "loss": 0.2308, "step": 737 }, { "epoch": 0.04, "grad_norm": 1.5406058289958093, "learning_rate": 9.996818876407101e-06, "loss": 0.2375, "step": 738 }, { "epoch": 0.04, "grad_norm": 1.538738234726375, "learning_rate": 9.996786666153612e-06, "loss": 0.2012, "step": 739 }, { "epoch": 0.04, "grad_norm": 1.453470927255759, "learning_rate": 9.996754293701556e-06, "loss": 0.2343, "step": 740 }, { "epoch": 0.04, "grad_norm": 1.6459359356329737, "learning_rate": 9.99672175905198e-06, "loss": 0.2573, "step": 741 }, { "epoch": 0.04, "grad_norm": 1.7620574473342865, "learning_rate": 9.996689062205942e-06, "loss": 0.2457, "step": 742 }, { "epoch": 0.04, "grad_norm": 1.7957641319017301, "learning_rate": 9.996656203164505e-06, "loss": 0.2719, "step": 743 }, { "epoch": 0.04, "grad_norm": 1.5850265357732467, "learning_rate": 9.996623181928735e-06, "loss": 0.2108, "step": 744 }, { "epoch": 0.04, "grad_norm": 6.063883358148109, "learning_rate": 9.996589998499702e-06, "loss": 0.2384, "step": 745 }, { "epoch": 0.04, "grad_norm": 1.7691083322695482, "learning_rate": 9.996556652878486e-06, "loss": 0.233, "step": 746 }, { "epoch": 0.04, "grad_norm": 1.8960295111073695, "learning_rate": 9.996523145066165e-06, "loss": 0.1974, "step": 747 }, { "epoch": 0.04, "grad_norm": 1.6525652491229723, "learning_rate": 9.996489475063832e-06, "loss": 0.2616, "step": 748 }, { "epoch": 0.04, "grad_norm": 1.6148126159936937, "learning_rate": 9.996455642872578e-06, "loss": 0.1963, "step": 749 }, { "epoch": 0.04, "grad_norm": 1.7864077623918613, "learning_rate": 9.996421648493499e-06, "loss": 0.1906, "step": 750 }, { "epoch": 0.04, "grad_norm": 1.830951504661101, "learning_rate": 9.996387491927699e-06, "loss": 0.251, "step": 751 }, { "epoch": 0.04, "grad_norm": 1.1138158384711583, "learning_rate": 9.99635317317629e-06, "loss": 0.1347, "step": 752 }, { "epoch": 0.04, "grad_norm": 1.9654294424616947, "learning_rate": 9.996318692240384e-06, "loss": 0.2566, "step": 753 }, { "epoch": 0.04, "grad_norm": 1.4760254016795331, "learning_rate": 9.996284049121098e-06, "loss": 0.2391, "step": 754 }, { "epoch": 0.04, "grad_norm": 1.5792199953262465, "learning_rate": 9.99624924381956e-06, "loss": 0.2341, "step": 755 }, { "epoch": 0.04, "grad_norm": 9.185073220631361, "learning_rate": 9.9962142763369e-06, "loss": 0.2156, "step": 756 }, { "epoch": 0.04, "grad_norm": 2.3936048959223153, "learning_rate": 9.99617914667425e-06, "loss": 0.263, "step": 757 }, { "epoch": 0.04, "grad_norm": 1.7616782892152645, "learning_rate": 9.996143854832752e-06, "loss": 0.2508, "step": 758 }, { "epoch": 0.04, "grad_norm": 1.72638969937547, "learning_rate": 9.99610840081355e-06, "loss": 0.2191, "step": 759 }, { "epoch": 0.04, "grad_norm": 2.205201257894953, "learning_rate": 9.996072784617799e-06, "loss": 0.2754, "step": 760 }, { "epoch": 0.04, "grad_norm": 2.3208722296787774, "learning_rate": 9.996037006246651e-06, "loss": 0.2637, "step": 761 }, { "epoch": 0.04, "grad_norm": 1.4094917998582825, "learning_rate": 9.99600106570127e-06, "loss": 0.1922, "step": 762 }, { "epoch": 0.04, "grad_norm": 1.4576681294561826, "learning_rate": 9.99596496298282e-06, "loss": 0.212, "step": 763 }, { "epoch": 0.04, "grad_norm": 4.372567508758709, "learning_rate": 9.995928698092475e-06, "loss": 0.1416, "step": 764 }, { "epoch": 0.04, "grad_norm": 1.5621205047855777, "learning_rate": 9.99589227103141e-06, "loss": 0.171, "step": 765 }, { "epoch": 0.04, "grad_norm": 1.4635162543628049, "learning_rate": 9.995855681800813e-06, "loss": 0.241, "step": 766 }, { "epoch": 0.04, "grad_norm": 2.308129291318679, "learning_rate": 9.995818930401865e-06, "loss": 0.252, "step": 767 }, { "epoch": 0.04, "grad_norm": 1.6108515734099502, "learning_rate": 9.995782016835762e-06, "loss": 0.2097, "step": 768 }, { "epoch": 0.04, "grad_norm": 1.5092952335991863, "learning_rate": 9.995744941103702e-06, "loss": 0.2199, "step": 769 }, { "epoch": 0.04, "grad_norm": 1.5698018121380966, "learning_rate": 9.995707703206889e-06, "loss": 0.2054, "step": 770 }, { "epoch": 0.04, "grad_norm": 2.0919332949536926, "learning_rate": 9.99567030314653e-06, "loss": 0.2476, "step": 771 }, { "epoch": 0.04, "grad_norm": 1.7983041221681502, "learning_rate": 9.995632740923841e-06, "loss": 0.2442, "step": 772 }, { "epoch": 0.04, "grad_norm": 1.8123833026807628, "learning_rate": 9.99559501654004e-06, "loss": 0.2489, "step": 773 }, { "epoch": 0.04, "grad_norm": 1.4794587109495483, "learning_rate": 9.995557129996354e-06, "loss": 0.2479, "step": 774 }, { "epoch": 0.04, "grad_norm": 2.0845419800043787, "learning_rate": 9.995519081294009e-06, "loss": 0.2502, "step": 775 }, { "epoch": 0.04, "grad_norm": 2.218294768577398, "learning_rate": 9.995480870434241e-06, "loss": 0.2272, "step": 776 }, { "epoch": 0.04, "grad_norm": 1.3143656873032976, "learning_rate": 9.995442497418295e-06, "loss": 0.2182, "step": 777 }, { "epoch": 0.04, "grad_norm": 2.029352181849417, "learning_rate": 9.995403962247411e-06, "loss": 0.2909, "step": 778 }, { "epoch": 0.04, "grad_norm": 1.0203498024423483, "learning_rate": 9.995365264922841e-06, "loss": 0.2256, "step": 779 }, { "epoch": 0.04, "grad_norm": 2.1115985241987407, "learning_rate": 9.995326405445842e-06, "loss": 0.2991, "step": 780 }, { "epoch": 0.04, "grad_norm": 2.0550399013924285, "learning_rate": 9.995287383817676e-06, "loss": 0.2447, "step": 781 }, { "epoch": 0.04, "grad_norm": 1.2567733646813406, "learning_rate": 9.99524820003961e-06, "loss": 0.1906, "step": 782 }, { "epoch": 0.04, "grad_norm": 1.5716726618425232, "learning_rate": 9.995208854112915e-06, "loss": 0.2231, "step": 783 }, { "epoch": 0.04, "grad_norm": 1.1434299163118575, "learning_rate": 9.995169346038867e-06, "loss": 0.2275, "step": 784 }, { "epoch": 0.04, "grad_norm": 1.7054133877612792, "learning_rate": 9.995129675818748e-06, "loss": 0.2149, "step": 785 }, { "epoch": 0.04, "grad_norm": 1.6301066003282731, "learning_rate": 9.995089843453849e-06, "loss": 0.2611, "step": 786 }, { "epoch": 0.04, "grad_norm": 1.9479567545745378, "learning_rate": 9.995049848945462e-06, "loss": 0.2367, "step": 787 }, { "epoch": 0.04, "grad_norm": 1.7888615018469156, "learning_rate": 9.995009692294884e-06, "loss": 0.2558, "step": 788 }, { "epoch": 0.04, "grad_norm": 1.1355459858878072, "learning_rate": 9.994969373503419e-06, "loss": 0.1715, "step": 789 }, { "epoch": 0.04, "grad_norm": 1.4047483851102234, "learning_rate": 9.994928892572376e-06, "loss": 0.187, "step": 790 }, { "epoch": 0.04, "grad_norm": 1.3005230908061722, "learning_rate": 9.994888249503069e-06, "loss": 0.2253, "step": 791 }, { "epoch": 0.04, "grad_norm": 1.3781808222820444, "learning_rate": 9.994847444296816e-06, "loss": 0.2364, "step": 792 }, { "epoch": 0.04, "grad_norm": 1.1397672458882684, "learning_rate": 9.994806476954943e-06, "loss": 0.1481, "step": 793 }, { "epoch": 0.04, "grad_norm": 1.147034264635771, "learning_rate": 9.99476534747878e-06, "loss": 0.2296, "step": 794 }, { "epoch": 0.04, "grad_norm": 1.2264614868482613, "learning_rate": 9.994724055869663e-06, "loss": 0.2105, "step": 795 }, { "epoch": 0.04, "grad_norm": 1.2980497193169003, "learning_rate": 9.994682602128929e-06, "loss": 0.2495, "step": 796 }, { "epoch": 0.04, "grad_norm": 1.2960108727647346, "learning_rate": 9.994640986257927e-06, "loss": 0.1762, "step": 797 }, { "epoch": 0.04, "grad_norm": 1.6657532801801207, "learning_rate": 9.994599208258005e-06, "loss": 0.2725, "step": 798 }, { "epoch": 0.04, "grad_norm": 1.3637181759625494, "learning_rate": 9.99455726813052e-06, "loss": 0.281, "step": 799 }, { "epoch": 0.04, "grad_norm": 1.004064282176184, "learning_rate": 9.994515165876837e-06, "loss": 0.1306, "step": 800 }, { "epoch": 0.04, "grad_norm": 1.160079748026799, "learning_rate": 9.994472901498318e-06, "loss": 0.2201, "step": 801 }, { "epoch": 0.04, "grad_norm": 1.6866050484001025, "learning_rate": 9.994430474996337e-06, "loss": 0.1441, "step": 802 }, { "epoch": 0.04, "grad_norm": 1.2357125188438258, "learning_rate": 9.994387886372269e-06, "loss": 0.213, "step": 803 }, { "epoch": 0.04, "grad_norm": 1.3157086179306139, "learning_rate": 9.994345135627498e-06, "loss": 0.1935, "step": 804 }, { "epoch": 0.04, "grad_norm": 2.088398720819187, "learning_rate": 9.994302222763415e-06, "loss": 0.2422, "step": 805 }, { "epoch": 0.04, "grad_norm": 1.501032900823252, "learning_rate": 9.99425914778141e-06, "loss": 0.2431, "step": 806 }, { "epoch": 0.04, "grad_norm": 1.9161484426335658, "learning_rate": 9.994215910682878e-06, "loss": 0.2639, "step": 807 }, { "epoch": 0.04, "grad_norm": 1.4251877267893638, "learning_rate": 9.994172511469227e-06, "loss": 0.1807, "step": 808 }, { "epoch": 0.05, "grad_norm": 1.585305987844208, "learning_rate": 9.994128950141866e-06, "loss": 0.1572, "step": 809 }, { "epoch": 0.05, "grad_norm": 1.568296769722774, "learning_rate": 9.994085226702206e-06, "loss": 0.218, "step": 810 }, { "epoch": 0.05, "grad_norm": 4.279281497790149, "learning_rate": 9.994041341151667e-06, "loss": 0.2671, "step": 811 }, { "epoch": 0.05, "grad_norm": 1.6708094329583227, "learning_rate": 9.993997293491674e-06, "loss": 0.2519, "step": 812 }, { "epoch": 0.05, "grad_norm": 1.8452941885547565, "learning_rate": 9.99395308372366e-06, "loss": 0.2501, "step": 813 }, { "epoch": 0.05, "grad_norm": 1.313985562878818, "learning_rate": 9.993908711849055e-06, "loss": 0.2488, "step": 814 }, { "epoch": 0.05, "grad_norm": 1.329512446431142, "learning_rate": 9.993864177869302e-06, "loss": 0.2596, "step": 815 }, { "epoch": 0.05, "grad_norm": 1.2622987695407983, "learning_rate": 9.993819481785846e-06, "loss": 0.1941, "step": 816 }, { "epoch": 0.05, "grad_norm": 1.629491924865182, "learning_rate": 9.993774623600138e-06, "loss": 0.129, "step": 817 }, { "epoch": 0.05, "grad_norm": 0.8711916222778915, "learning_rate": 9.993729603313633e-06, "loss": 0.2262, "step": 818 }, { "epoch": 0.05, "grad_norm": 1.445170470295729, "learning_rate": 9.993684420927795e-06, "loss": 0.2237, "step": 819 }, { "epoch": 0.05, "grad_norm": 2.5140740488568305, "learning_rate": 9.993639076444089e-06, "loss": 0.3494, "step": 820 }, { "epoch": 0.05, "grad_norm": 1.9510408076807002, "learning_rate": 9.993593569863987e-06, "loss": 0.2524, "step": 821 }, { "epoch": 0.05, "grad_norm": 1.8622112240714863, "learning_rate": 9.993547901188966e-06, "loss": 0.2383, "step": 822 }, { "epoch": 0.05, "grad_norm": 1.7631780539261597, "learning_rate": 9.993502070420509e-06, "loss": 0.2936, "step": 823 }, { "epoch": 0.05, "grad_norm": 2.1521535190488015, "learning_rate": 9.993456077560101e-06, "loss": 0.2377, "step": 824 }, { "epoch": 0.05, "grad_norm": 1.4783259714859684, "learning_rate": 9.993409922609241e-06, "loss": 0.215, "step": 825 }, { "epoch": 0.05, "grad_norm": 2.625916769383279, "learning_rate": 9.993363605569421e-06, "loss": 0.3517, "step": 826 }, { "epoch": 0.05, "grad_norm": 1.901423682190051, "learning_rate": 9.993317126442149e-06, "loss": 0.2951, "step": 827 }, { "epoch": 0.05, "grad_norm": 1.843750507144535, "learning_rate": 9.993270485228929e-06, "loss": 0.2428, "step": 828 }, { "epoch": 0.05, "grad_norm": 1.410532408497231, "learning_rate": 9.993223681931278e-06, "loss": 0.2409, "step": 829 }, { "epoch": 0.05, "grad_norm": 1.3757206981467418, "learning_rate": 9.993176716550717e-06, "loss": 0.2084, "step": 830 }, { "epoch": 0.05, "grad_norm": 1.3048156115741543, "learning_rate": 9.993129589088769e-06, "loss": 0.2426, "step": 831 }, { "epoch": 0.05, "grad_norm": 1.866408569674609, "learning_rate": 9.993082299546961e-06, "loss": 0.2548, "step": 832 }, { "epoch": 0.05, "grad_norm": 1.7268773095065082, "learning_rate": 9.99303484792683e-06, "loss": 0.217, "step": 833 }, { "epoch": 0.05, "grad_norm": 1.6728730899029136, "learning_rate": 9.992987234229918e-06, "loss": 0.211, "step": 834 }, { "epoch": 0.05, "grad_norm": 1.4376750015422304, "learning_rate": 9.992939458457769e-06, "loss": 0.2419, "step": 835 }, { "epoch": 0.05, "grad_norm": 0.9700654850317781, "learning_rate": 9.992891520611934e-06, "loss": 0.203, "step": 836 }, { "epoch": 0.05, "grad_norm": 1.5474833703380089, "learning_rate": 9.99284342069397e-06, "loss": 0.1894, "step": 837 }, { "epoch": 0.05, "grad_norm": 1.2430915601614154, "learning_rate": 9.992795158705434e-06, "loss": 0.229, "step": 838 }, { "epoch": 0.05, "grad_norm": 1.1066378706181776, "learning_rate": 9.9927467346479e-06, "loss": 0.138, "step": 839 }, { "epoch": 0.05, "grad_norm": 1.2442485052912813, "learning_rate": 9.992698148522934e-06, "loss": 0.2188, "step": 840 }, { "epoch": 0.05, "grad_norm": 1.503172641026925, "learning_rate": 9.992649400332117e-06, "loss": 0.1977, "step": 841 }, { "epoch": 0.05, "grad_norm": 1.3095615053840808, "learning_rate": 9.992600490077027e-06, "loss": 0.2181, "step": 842 }, { "epoch": 0.05, "grad_norm": 1.4209163953033057, "learning_rate": 9.992551417759256e-06, "loss": 0.2142, "step": 843 }, { "epoch": 0.05, "grad_norm": 1.8985850478946014, "learning_rate": 9.992502183380395e-06, "loss": 0.231, "step": 844 }, { "epoch": 0.05, "grad_norm": 1.4331381850919158, "learning_rate": 9.992452786942041e-06, "loss": 0.2471, "step": 845 }, { "epoch": 0.05, "grad_norm": 1.4496724917035175, "learning_rate": 9.9924032284458e-06, "loss": 0.1898, "step": 846 }, { "epoch": 0.05, "grad_norm": 1.0682342637360658, "learning_rate": 9.992353507893279e-06, "loss": 0.1589, "step": 847 }, { "epoch": 0.05, "grad_norm": 1.811711592041193, "learning_rate": 9.992303625286092e-06, "loss": 0.2533, "step": 848 }, { "epoch": 0.05, "grad_norm": 1.5553208671215994, "learning_rate": 9.992253580625858e-06, "loss": 0.2926, "step": 849 }, { "epoch": 0.05, "grad_norm": 1.3665783765002752, "learning_rate": 9.992203373914203e-06, "loss": 0.2051, "step": 850 }, { "epoch": 0.05, "grad_norm": 2.278038370450238, "learning_rate": 9.992153005152754e-06, "loss": 0.2622, "step": 851 }, { "epoch": 0.05, "grad_norm": 2.244775285609426, "learning_rate": 9.99210247434315e-06, "loss": 0.2412, "step": 852 }, { "epoch": 0.05, "grad_norm": 1.460435046526099, "learning_rate": 9.992051781487029e-06, "loss": 0.1935, "step": 853 }, { "epoch": 0.05, "grad_norm": 1.3709198713766988, "learning_rate": 9.992000926586035e-06, "loss": 0.1716, "step": 854 }, { "epoch": 0.05, "grad_norm": 1.486706223187014, "learning_rate": 9.99194990964182e-06, "loss": 0.2501, "step": 855 }, { "epoch": 0.05, "grad_norm": 1.5583334970814147, "learning_rate": 9.991898730656043e-06, "loss": 0.1993, "step": 856 }, { "epoch": 0.05, "grad_norm": 1.437353572436606, "learning_rate": 9.99184738963036e-06, "loss": 0.2311, "step": 857 }, { "epoch": 0.05, "grad_norm": 1.677794414329712, "learning_rate": 9.991795886566443e-06, "loss": 0.2295, "step": 858 }, { "epoch": 0.05, "grad_norm": 1.9166710040874402, "learning_rate": 9.991744221465958e-06, "loss": 0.2475, "step": 859 }, { "epoch": 0.05, "grad_norm": 1.8914227024451598, "learning_rate": 9.991692394330586e-06, "loss": 0.2385, "step": 860 }, { "epoch": 0.05, "grad_norm": 1.3891599678615987, "learning_rate": 9.991640405162009e-06, "loss": 0.2042, "step": 861 }, { "epoch": 0.05, "grad_norm": 4.536812150590555, "learning_rate": 9.991588253961914e-06, "loss": 0.2438, "step": 862 }, { "epoch": 0.05, "grad_norm": 1.8795405722864267, "learning_rate": 9.991535940731993e-06, "loss": 0.2164, "step": 863 }, { "epoch": 0.05, "grad_norm": 1.8151272218590928, "learning_rate": 9.991483465473945e-06, "loss": 0.2886, "step": 864 }, { "epoch": 0.05, "grad_norm": 1.9047040318216781, "learning_rate": 9.991430828189476e-06, "loss": 0.218, "step": 865 }, { "epoch": 0.05, "grad_norm": 1.5089580256101758, "learning_rate": 9.99137802888029e-06, "loss": 0.238, "step": 866 }, { "epoch": 0.05, "grad_norm": 1.5526845456171137, "learning_rate": 9.991325067548103e-06, "loss": 0.2379, "step": 867 }, { "epoch": 0.05, "grad_norm": 1.8013273763946662, "learning_rate": 9.991271944194634e-06, "loss": 0.2145, "step": 868 }, { "epoch": 0.05, "grad_norm": 1.8404494582374002, "learning_rate": 9.991218658821609e-06, "loss": 0.1939, "step": 869 }, { "epoch": 0.05, "grad_norm": 2.08758554825834, "learning_rate": 9.991165211430753e-06, "loss": 0.2482, "step": 870 }, { "epoch": 0.05, "grad_norm": 3.7042416204901096, "learning_rate": 9.991111602023807e-06, "loss": 0.2697, "step": 871 }, { "epoch": 0.05, "grad_norm": 1.766875578430197, "learning_rate": 9.991057830602508e-06, "loss": 0.2023, "step": 872 }, { "epoch": 0.05, "grad_norm": 2.1795710221482967, "learning_rate": 9.9910038971686e-06, "loss": 0.2272, "step": 873 }, { "epoch": 0.05, "grad_norm": 1.6079584572606918, "learning_rate": 9.990949801723837e-06, "loss": 0.1936, "step": 874 }, { "epoch": 0.05, "grad_norm": 0.9629808232771345, "learning_rate": 9.990895544269973e-06, "loss": 0.214, "step": 875 }, { "epoch": 0.05, "grad_norm": 0.9922395499161029, "learning_rate": 9.99084112480877e-06, "loss": 0.1953, "step": 876 }, { "epoch": 0.05, "grad_norm": 1.7868187756680873, "learning_rate": 9.990786543341994e-06, "loss": 0.268, "step": 877 }, { "epoch": 0.05, "grad_norm": 1.4733962873437458, "learning_rate": 9.990731799871415e-06, "loss": 0.185, "step": 878 }, { "epoch": 0.05, "grad_norm": 1.1905698216535727, "learning_rate": 9.990676894398815e-06, "loss": 0.1686, "step": 879 }, { "epoch": 0.05, "grad_norm": 1.429458474747954, "learning_rate": 9.990621826925972e-06, "loss": 0.1919, "step": 880 }, { "epoch": 0.05, "grad_norm": 2.214006551656187, "learning_rate": 9.990566597454673e-06, "loss": 0.2069, "step": 881 }, { "epoch": 0.05, "grad_norm": 0.9883021684174255, "learning_rate": 9.990511205986716e-06, "loss": 0.1796, "step": 882 }, { "epoch": 0.05, "grad_norm": 1.6791241432970971, "learning_rate": 9.990455652523893e-06, "loss": 0.3108, "step": 883 }, { "epoch": 0.05, "grad_norm": 1.6095545191777498, "learning_rate": 9.99039993706801e-06, "loss": 0.2519, "step": 884 }, { "epoch": 0.05, "grad_norm": 1.3335942051042562, "learning_rate": 9.990344059620876e-06, "loss": 0.2722, "step": 885 }, { "epoch": 0.05, "grad_norm": 1.1723544475154293, "learning_rate": 9.990288020184303e-06, "loss": 0.2113, "step": 886 }, { "epoch": 0.05, "grad_norm": 1.5290073057293918, "learning_rate": 9.990231818760112e-06, "loss": 0.2109, "step": 887 }, { "epoch": 0.05, "grad_norm": 3.4602109607153175, "learning_rate": 9.990175455350126e-06, "loss": 0.2651, "step": 888 }, { "epoch": 0.05, "grad_norm": 1.1566241374285202, "learning_rate": 9.990118929956175e-06, "loss": 0.2727, "step": 889 }, { "epoch": 0.05, "grad_norm": 2.052044855199194, "learning_rate": 9.990062242580096e-06, "loss": 0.2241, "step": 890 }, { "epoch": 0.05, "grad_norm": 1.4177814076146151, "learning_rate": 9.990005393223726e-06, "loss": 0.2151, "step": 891 }, { "epoch": 0.05, "grad_norm": 1.7672564303364384, "learning_rate": 9.989948381888913e-06, "loss": 0.2271, "step": 892 }, { "epoch": 0.05, "grad_norm": 1.3500005649194595, "learning_rate": 9.989891208577502e-06, "loss": 0.1935, "step": 893 }, { "epoch": 0.05, "grad_norm": 1.641037130816496, "learning_rate": 9.989833873291357e-06, "loss": 0.2766, "step": 894 }, { "epoch": 0.05, "grad_norm": 1.7856500556157282, "learning_rate": 9.989776376032333e-06, "loss": 0.2342, "step": 895 }, { "epoch": 0.05, "grad_norm": 1.2303187324263358, "learning_rate": 9.989718716802299e-06, "loss": 0.2207, "step": 896 }, { "epoch": 0.05, "grad_norm": 1.407066620705677, "learning_rate": 9.989660895603127e-06, "loss": 0.2123, "step": 897 }, { "epoch": 0.05, "grad_norm": 1.2960471216727876, "learning_rate": 9.989602912436692e-06, "loss": 0.2201, "step": 898 }, { "epoch": 0.05, "grad_norm": 1.8542379372718563, "learning_rate": 9.989544767304877e-06, "loss": 0.2902, "step": 899 } ], "logging_steps": 1.0, "max_steps": 17977, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 899, "total_flos": 52192960905216.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }