diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,12607 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.10001668799020971, + "eval_steps": 500, + "global_step": 1798, + "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 + }, + { + "epoch": 0.05, + "grad_norm": 1.4525291697975822, + "learning_rate": 9.98948646020957e-06, + "loss": 0.2387, + "step": 900 + }, + { + "epoch": 0.05, + "grad_norm": 1.107567794376534, + "learning_rate": 9.989427991152664e-06, + "loss": 0.1903, + "step": 901 + }, + { + "epoch": 0.05, + "grad_norm": 1.7098400872801431, + "learning_rate": 9.989369360136056e-06, + "loss": 0.249, + "step": 902 + }, + { + "epoch": 0.05, + "grad_norm": 1.5465166478789543, + "learning_rate": 9.989310567161649e-06, + "loss": 0.1837, + "step": 903 + }, + { + "epoch": 0.05, + "grad_norm": 1.161074838241732, + "learning_rate": 9.989251612231352e-06, + "loss": 0.2363, + "step": 904 + }, + { + "epoch": 0.05, + "grad_norm": 1.4277623196018425, + "learning_rate": 9.989192495347079e-06, + "loss": 0.2542, + "step": 905 + }, + { + "epoch": 0.05, + "grad_norm": 1.6174152462686229, + "learning_rate": 9.98913321651075e-06, + "loss": 0.2556, + "step": 906 + }, + { + "epoch": 0.05, + "grad_norm": 1.9159902946205454, + "learning_rate": 9.989073775724284e-06, + "loss": 0.2401, + "step": 907 + }, + { + "epoch": 0.05, + "grad_norm": 1.3589066158898169, + "learning_rate": 9.989014172989617e-06, + "loss": 0.2285, + "step": 908 + }, + { + "epoch": 0.05, + "grad_norm": 1.4265811244170847, + "learning_rate": 9.988954408308681e-06, + "loss": 0.2512, + "step": 909 + }, + { + "epoch": 0.05, + "grad_norm": 1.402832397051935, + "learning_rate": 9.988894481683416e-06, + "loss": 0.2234, + "step": 910 + }, + { + "epoch": 0.05, + "grad_norm": 2.708595913604761, + "learning_rate": 9.988834393115768e-06, + "loss": 0.2364, + "step": 911 + }, + { + "epoch": 0.05, + "grad_norm": 1.4567256011175753, + "learning_rate": 9.988774142607684e-06, + "loss": 0.2889, + "step": 912 + }, + { + "epoch": 0.05, + "grad_norm": 1.5397321144795744, + "learning_rate": 9.988713730161125e-06, + "loss": 0.2027, + "step": 913 + }, + { + "epoch": 0.05, + "grad_norm": 2.225814318854797, + "learning_rate": 9.988653155778049e-06, + "loss": 0.2534, + "step": 914 + }, + { + "epoch": 0.05, + "grad_norm": 1.5422927510367952, + "learning_rate": 9.988592419460423e-06, + "loss": 0.2823, + "step": 915 + }, + { + "epoch": 0.05, + "grad_norm": 1.391193557624271, + "learning_rate": 9.988531521210219e-06, + "loss": 0.2625, + "step": 916 + }, + { + "epoch": 0.05, + "grad_norm": 1.3037852405419272, + "learning_rate": 9.988470461029412e-06, + "loss": 0.1961, + "step": 917 + }, + { + "epoch": 0.05, + "grad_norm": 2.726029439589459, + "learning_rate": 9.988409238919985e-06, + "loss": 0.2418, + "step": 918 + }, + { + "epoch": 0.05, + "grad_norm": 1.1114674030042353, + "learning_rate": 9.988347854883927e-06, + "loss": 0.2316, + "step": 919 + }, + { + "epoch": 0.05, + "grad_norm": 1.208037956558638, + "learning_rate": 9.988286308923228e-06, + "loss": 0.196, + "step": 920 + }, + { + "epoch": 0.05, + "grad_norm": 1.4125976364710648, + "learning_rate": 9.988224601039888e-06, + "loss": 0.2762, + "step": 921 + }, + { + "epoch": 0.05, + "grad_norm": 1.596446395302791, + "learning_rate": 9.988162731235908e-06, + "loss": 0.2864, + "step": 922 + }, + { + "epoch": 0.05, + "grad_norm": 1.5402656538455302, + "learning_rate": 9.988100699513298e-06, + "loss": 0.1571, + "step": 923 + }, + { + "epoch": 0.05, + "grad_norm": 1.9300669652218219, + "learning_rate": 9.988038505874072e-06, + "loss": 0.2278, + "step": 924 + }, + { + "epoch": 0.05, + "grad_norm": 1.3783263102699999, + "learning_rate": 9.987976150320246e-06, + "loss": 0.2505, + "step": 925 + }, + { + "epoch": 0.05, + "grad_norm": 1.7210174607924251, + "learning_rate": 9.987913632853845e-06, + "loss": 0.2132, + "step": 926 + }, + { + "epoch": 0.05, + "grad_norm": 2.088656656963556, + "learning_rate": 9.9878509534769e-06, + "loss": 0.2803, + "step": 927 + }, + { + "epoch": 0.05, + "grad_norm": 1.2339367873340437, + "learning_rate": 9.987788112191447e-06, + "loss": 0.2691, + "step": 928 + }, + { + "epoch": 0.05, + "grad_norm": 1.303307064234305, + "learning_rate": 9.987725108999523e-06, + "loss": 0.214, + "step": 929 + }, + { + "epoch": 0.05, + "grad_norm": 1.5709334137535031, + "learning_rate": 9.987661943903172e-06, + "loss": 0.1954, + "step": 930 + }, + { + "epoch": 0.05, + "grad_norm": 1.5057379330995977, + "learning_rate": 9.987598616904446e-06, + "loss": 0.1227, + "step": 931 + }, + { + "epoch": 0.05, + "grad_norm": 1.7730160245965954, + "learning_rate": 9.987535128005404e-06, + "loss": 0.1864, + "step": 932 + }, + { + "epoch": 0.05, + "grad_norm": 1.3758593388176685, + "learning_rate": 9.987471477208101e-06, + "loss": 0.2178, + "step": 933 + }, + { + "epoch": 0.05, + "grad_norm": 1.111303493448027, + "learning_rate": 9.987407664514606e-06, + "loss": 0.177, + "step": 934 + }, + { + "epoch": 0.05, + "grad_norm": 1.4100832299429573, + "learning_rate": 9.98734368992699e-06, + "loss": 0.2389, + "step": 935 + }, + { + "epoch": 0.05, + "grad_norm": 3.491723219626732, + "learning_rate": 9.98727955344733e-06, + "loss": 0.2795, + "step": 936 + }, + { + "epoch": 0.05, + "grad_norm": 1.788121406957569, + "learning_rate": 9.987215255077708e-06, + "loss": 0.2841, + "step": 937 + }, + { + "epoch": 0.05, + "grad_norm": 0.8608540227093385, + "learning_rate": 9.987150794820213e-06, + "loss": 0.1929, + "step": 938 + }, + { + "epoch": 0.05, + "grad_norm": 1.3179591934929455, + "learning_rate": 9.987086172676934e-06, + "loss": 0.2007, + "step": 939 + }, + { + "epoch": 0.05, + "grad_norm": 2.0060940524422732, + "learning_rate": 9.98702138864997e-06, + "loss": 0.2028, + "step": 940 + }, + { + "epoch": 0.05, + "grad_norm": 1.544763192805912, + "learning_rate": 9.986956442741424e-06, + "loss": 0.2767, + "step": 941 + }, + { + "epoch": 0.05, + "grad_norm": 1.579922007049674, + "learning_rate": 9.986891334953404e-06, + "loss": 0.2125, + "step": 942 + }, + { + "epoch": 0.05, + "grad_norm": 1.5379047044822187, + "learning_rate": 9.986826065288024e-06, + "loss": 0.2404, + "step": 943 + }, + { + "epoch": 0.05, + "grad_norm": 1.2091100231504772, + "learning_rate": 9.986760633747401e-06, + "loss": 0.1808, + "step": 944 + }, + { + "epoch": 0.05, + "grad_norm": 1.3489025255083558, + "learning_rate": 9.986695040333661e-06, + "loss": 0.2095, + "step": 945 + }, + { + "epoch": 0.05, + "grad_norm": 1.7426656229455786, + "learning_rate": 9.986629285048933e-06, + "loss": 0.1903, + "step": 946 + }, + { + "epoch": 0.05, + "grad_norm": 1.7732170635972555, + "learning_rate": 9.98656336789535e-06, + "loss": 0.2418, + "step": 947 + }, + { + "epoch": 0.05, + "grad_norm": 2.1766471535171377, + "learning_rate": 9.986497288875054e-06, + "loss": 0.2252, + "step": 948 + }, + { + "epoch": 0.05, + "grad_norm": 1.9042980174525819, + "learning_rate": 9.986431047990188e-06, + "loss": 0.1937, + "step": 949 + }, + { + "epoch": 0.05, + "grad_norm": 1.0773368421669207, + "learning_rate": 9.9863646452429e-06, + "loss": 0.2179, + "step": 950 + }, + { + "epoch": 0.05, + "grad_norm": 1.2613252566491926, + "learning_rate": 9.986298080635352e-06, + "loss": 0.1558, + "step": 951 + }, + { + "epoch": 0.05, + "grad_norm": 2.777151409630218, + "learning_rate": 9.986231354169699e-06, + "loss": 0.2041, + "step": 952 + }, + { + "epoch": 0.05, + "grad_norm": 1.3013455710461423, + "learning_rate": 9.98616446584811e-06, + "loss": 0.2094, + "step": 953 + }, + { + "epoch": 0.05, + "grad_norm": 1.3425800690568885, + "learning_rate": 9.986097415672755e-06, + "loss": 0.2093, + "step": 954 + }, + { + "epoch": 0.05, + "grad_norm": 1.1973478030803608, + "learning_rate": 9.986030203645812e-06, + "loss": 0.2496, + "step": 955 + }, + { + "epoch": 0.05, + "grad_norm": 2.7112634116615357, + "learning_rate": 9.985962829769457e-06, + "loss": 0.2504, + "step": 956 + }, + { + "epoch": 0.05, + "grad_norm": 1.724628998654484, + "learning_rate": 9.985895294045885e-06, + "loss": 0.2121, + "step": 957 + }, + { + "epoch": 0.05, + "grad_norm": 1.5906423980908053, + "learning_rate": 9.985827596477285e-06, + "loss": 0.1802, + "step": 958 + }, + { + "epoch": 0.05, + "grad_norm": 1.4000098270684438, + "learning_rate": 9.985759737065853e-06, + "loss": 0.2129, + "step": 959 + }, + { + "epoch": 0.05, + "grad_norm": 1.5633062142395335, + "learning_rate": 9.985691715813792e-06, + "loss": 0.2261, + "step": 960 + }, + { + "epoch": 0.05, + "grad_norm": 1.2664589400864432, + "learning_rate": 9.985623532723313e-06, + "loss": 0.1809, + "step": 961 + }, + { + "epoch": 0.05, + "grad_norm": 8.009300785859471, + "learning_rate": 9.985555187796625e-06, + "loss": 0.2434, + "step": 962 + }, + { + "epoch": 0.05, + "grad_norm": 1.4531394114343694, + "learning_rate": 9.98548668103595e-06, + "loss": 0.2355, + "step": 963 + }, + { + "epoch": 0.05, + "grad_norm": 1.3856378506609577, + "learning_rate": 9.98541801244351e-06, + "loss": 0.1993, + "step": 964 + }, + { + "epoch": 0.05, + "grad_norm": 1.39277996831488, + "learning_rate": 9.985349182021536e-06, + "loss": 0.2082, + "step": 965 + }, + { + "epoch": 0.05, + "grad_norm": 3.105960322406658, + "learning_rate": 9.98528018977226e-06, + "loss": 0.2322, + "step": 966 + }, + { + "epoch": 0.05, + "grad_norm": 1.3446049957343713, + "learning_rate": 9.985211035697921e-06, + "loss": 0.2616, + "step": 967 + }, + { + "epoch": 0.05, + "grad_norm": 1.4718135502008964, + "learning_rate": 9.985141719800766e-06, + "loss": 0.2464, + "step": 968 + }, + { + "epoch": 0.05, + "grad_norm": 1.1873764174553603, + "learning_rate": 9.985072242083045e-06, + "loss": 0.2077, + "step": 969 + }, + { + "epoch": 0.05, + "grad_norm": 1.599956986451696, + "learning_rate": 9.98500260254701e-06, + "loss": 0.2659, + "step": 970 + }, + { + "epoch": 0.05, + "grad_norm": 1.1697543339498704, + "learning_rate": 9.984932801194927e-06, + "loss": 0.209, + "step": 971 + }, + { + "epoch": 0.05, + "grad_norm": 1.238303089550941, + "learning_rate": 9.984862838029057e-06, + "loss": 0.2155, + "step": 972 + }, + { + "epoch": 0.05, + "grad_norm": 1.5609784335228798, + "learning_rate": 9.984792713051672e-06, + "loss": 0.1656, + "step": 973 + }, + { + "epoch": 0.05, + "grad_norm": 2.156906591182393, + "learning_rate": 9.98472242626505e-06, + "loss": 0.2568, + "step": 974 + }, + { + "epoch": 0.05, + "grad_norm": 1.0959353192435246, + "learning_rate": 9.984651977671471e-06, + "loss": 0.1709, + "step": 975 + }, + { + "epoch": 0.05, + "grad_norm": 1.1125309925771782, + "learning_rate": 9.984581367273222e-06, + "loss": 0.2534, + "step": 976 + }, + { + "epoch": 0.05, + "grad_norm": 1.5918457210979915, + "learning_rate": 9.984510595072599e-06, + "loss": 0.2306, + "step": 977 + }, + { + "epoch": 0.05, + "grad_norm": 1.490419401329037, + "learning_rate": 9.984439661071892e-06, + "loss": 0.1957, + "step": 978 + }, + { + "epoch": 0.05, + "grad_norm": 1.4420343080042421, + "learning_rate": 9.984368565273409e-06, + "loss": 0.2393, + "step": 979 + }, + { + "epoch": 0.05, + "grad_norm": 1.6885920231947016, + "learning_rate": 9.984297307679455e-06, + "loss": 0.2326, + "step": 980 + }, + { + "epoch": 0.05, + "grad_norm": 1.334491033110002, + "learning_rate": 9.984225888292346e-06, + "loss": 0.2536, + "step": 981 + }, + { + "epoch": 0.05, + "grad_norm": 1.1817211502002398, + "learning_rate": 9.984154307114398e-06, + "loss": 0.2059, + "step": 982 + }, + { + "epoch": 0.05, + "grad_norm": 1.7102879266595847, + "learning_rate": 9.984082564147935e-06, + "loss": 0.252, + "step": 983 + }, + { + "epoch": 0.05, + "grad_norm": 1.2358043820807088, + "learning_rate": 9.984010659395285e-06, + "loss": 0.2067, + "step": 984 + }, + { + "epoch": 0.05, + "grad_norm": 1.1077678537754339, + "learning_rate": 9.983938592858783e-06, + "loss": 0.2, + "step": 985 + }, + { + "epoch": 0.05, + "grad_norm": 1.3227772532867288, + "learning_rate": 9.98386636454077e-06, + "loss": 0.1774, + "step": 986 + }, + { + "epoch": 0.05, + "grad_norm": 1.675970791499035, + "learning_rate": 9.983793974443587e-06, + "loss": 0.2707, + "step": 987 + }, + { + "epoch": 0.05, + "grad_norm": 1.6790284471787886, + "learning_rate": 9.983721422569585e-06, + "loss": 0.2455, + "step": 988 + }, + { + "epoch": 0.06, + "grad_norm": 1.8196197918993409, + "learning_rate": 9.983648708921123e-06, + "loss": 0.2302, + "step": 989 + }, + { + "epoch": 0.06, + "grad_norm": 1.3500481649919593, + "learning_rate": 9.983575833500556e-06, + "loss": 0.2694, + "step": 990 + }, + { + "epoch": 0.06, + "grad_norm": 1.3586135847454621, + "learning_rate": 9.983502796310252e-06, + "loss": 0.2187, + "step": 991 + }, + { + "epoch": 0.06, + "grad_norm": 1.5726381843768615, + "learning_rate": 9.98342959735258e-06, + "loss": 0.2718, + "step": 992 + }, + { + "epoch": 0.06, + "grad_norm": 1.4075071649385964, + "learning_rate": 9.983356236629919e-06, + "loss": 0.2217, + "step": 993 + }, + { + "epoch": 0.06, + "grad_norm": 0.963694407196175, + "learning_rate": 9.98328271414465e-06, + "loss": 0.2068, + "step": 994 + }, + { + "epoch": 0.06, + "grad_norm": 1.3379321725338948, + "learning_rate": 9.983209029899158e-06, + "loss": 0.1902, + "step": 995 + }, + { + "epoch": 0.06, + "grad_norm": 1.4019286303693481, + "learning_rate": 9.983135183895835e-06, + "loss": 0.2352, + "step": 996 + }, + { + "epoch": 0.06, + "grad_norm": 1.3446962869819388, + "learning_rate": 9.983061176137079e-06, + "loss": 0.2242, + "step": 997 + }, + { + "epoch": 0.06, + "grad_norm": 1.6288738156023739, + "learning_rate": 9.98298700662529e-06, + "loss": 0.2494, + "step": 998 + }, + { + "epoch": 0.06, + "grad_norm": 3.631970379754838, + "learning_rate": 9.98291267536288e-06, + "loss": 0.2025, + "step": 999 + }, + { + "epoch": 0.06, + "grad_norm": 1.4028370767855172, + "learning_rate": 9.982838182352257e-06, + "loss": 0.2134, + "step": 1000 + }, + { + "epoch": 0.06, + "grad_norm": 1.2612395365614641, + "learning_rate": 9.982763527595841e-06, + "loss": 0.2125, + "step": 1001 + }, + { + "epoch": 0.06, + "grad_norm": 1.1018481357705399, + "learning_rate": 9.982688711096056e-06, + "loss": 0.1776, + "step": 1002 + }, + { + "epoch": 0.06, + "grad_norm": 1.6733297342707991, + "learning_rate": 9.98261373285533e-06, + "loss": 0.1784, + "step": 1003 + }, + { + "epoch": 0.06, + "grad_norm": 1.483247157827921, + "learning_rate": 9.982538592876099e-06, + "loss": 0.2525, + "step": 1004 + }, + { + "epoch": 0.06, + "grad_norm": 1.3897791926297314, + "learning_rate": 9.9824632911608e-06, + "loss": 0.2314, + "step": 1005 + }, + { + "epoch": 0.06, + "grad_norm": 1.4408984271811753, + "learning_rate": 9.982387827711874e-06, + "loss": 0.22, + "step": 1006 + }, + { + "epoch": 0.06, + "grad_norm": 2.008263822957917, + "learning_rate": 9.982312202531777e-06, + "loss": 0.2636, + "step": 1007 + }, + { + "epoch": 0.06, + "grad_norm": 1.1482339016803096, + "learning_rate": 9.98223641562296e-06, + "loss": 0.1518, + "step": 1008 + }, + { + "epoch": 0.06, + "grad_norm": 1.1761861481017488, + "learning_rate": 9.982160466987885e-06, + "loss": 0.1561, + "step": 1009 + }, + { + "epoch": 0.06, + "grad_norm": 1.0828725611693262, + "learning_rate": 9.982084356629014e-06, + "loss": 0.1837, + "step": 1010 + }, + { + "epoch": 0.06, + "grad_norm": 1.1709146832571171, + "learning_rate": 9.982008084548822e-06, + "loss": 0.1954, + "step": 1011 + }, + { + "epoch": 0.06, + "grad_norm": 1.3596966845167775, + "learning_rate": 9.981931650749782e-06, + "loss": 0.2841, + "step": 1012 + }, + { + "epoch": 0.06, + "grad_norm": 1.5602116696233383, + "learning_rate": 9.981855055234375e-06, + "loss": 0.2601, + "step": 1013 + }, + { + "epoch": 0.06, + "grad_norm": 1.3779024463450789, + "learning_rate": 9.98177829800509e-06, + "loss": 0.2278, + "step": 1014 + }, + { + "epoch": 0.06, + "grad_norm": 1.2277743357612954, + "learning_rate": 9.981701379064414e-06, + "loss": 0.2206, + "step": 1015 + }, + { + "epoch": 0.06, + "grad_norm": 1.4654456669422902, + "learning_rate": 9.981624298414849e-06, + "loss": 0.2239, + "step": 1016 + }, + { + "epoch": 0.06, + "grad_norm": 1.7127379134034206, + "learning_rate": 9.981547056058893e-06, + "loss": 0.2437, + "step": 1017 + }, + { + "epoch": 0.06, + "grad_norm": 1.525082967457466, + "learning_rate": 9.981469651999057e-06, + "loss": 0.2055, + "step": 1018 + }, + { + "epoch": 0.06, + "grad_norm": 1.5774340970246645, + "learning_rate": 9.981392086237851e-06, + "loss": 0.2501, + "step": 1019 + }, + { + "epoch": 0.06, + "grad_norm": 1.5134016767960998, + "learning_rate": 9.981314358777792e-06, + "loss": 0.2124, + "step": 1020 + }, + { + "epoch": 0.06, + "grad_norm": 1.7651890283961635, + "learning_rate": 9.981236469621404e-06, + "loss": 0.2555, + "step": 1021 + }, + { + "epoch": 0.06, + "grad_norm": 1.2555344073652324, + "learning_rate": 9.981158418771218e-06, + "loss": 0.225, + "step": 1022 + }, + { + "epoch": 0.06, + "grad_norm": 1.5026501382580701, + "learning_rate": 9.981080206229765e-06, + "loss": 0.2321, + "step": 1023 + }, + { + "epoch": 0.06, + "grad_norm": 1.4043536512970234, + "learning_rate": 9.981001831999582e-06, + "loss": 0.2214, + "step": 1024 + }, + { + "epoch": 0.06, + "grad_norm": 1.258302208244986, + "learning_rate": 9.980923296083219e-06, + "loss": 0.194, + "step": 1025 + }, + { + "epoch": 0.06, + "grad_norm": 1.7705610196786978, + "learning_rate": 9.980844598483217e-06, + "loss": 0.2515, + "step": 1026 + }, + { + "epoch": 0.06, + "grad_norm": 1.598332739580166, + "learning_rate": 9.980765739202138e-06, + "loss": 0.2171, + "step": 1027 + }, + { + "epoch": 0.06, + "grad_norm": 1.0787654715806567, + "learning_rate": 9.980686718242539e-06, + "loss": 0.2462, + "step": 1028 + }, + { + "epoch": 0.06, + "grad_norm": 1.5956307234850093, + "learning_rate": 9.980607535606984e-06, + "loss": 0.2063, + "step": 1029 + }, + { + "epoch": 0.06, + "grad_norm": 1.284701712646174, + "learning_rate": 9.980528191298045e-06, + "loss": 0.1744, + "step": 1030 + }, + { + "epoch": 0.06, + "grad_norm": 1.3371605398601, + "learning_rate": 9.980448685318296e-06, + "loss": 0.2014, + "step": 1031 + }, + { + "epoch": 0.06, + "grad_norm": 1.4221399343386596, + "learning_rate": 9.980369017670319e-06, + "loss": 0.2174, + "step": 1032 + }, + { + "epoch": 0.06, + "grad_norm": 1.1243029221271443, + "learning_rate": 9.980289188356702e-06, + "loss": 0.2689, + "step": 1033 + }, + { + "epoch": 0.06, + "grad_norm": 1.2531696648735722, + "learning_rate": 9.980209197380031e-06, + "loss": 0.2014, + "step": 1034 + }, + { + "epoch": 0.06, + "grad_norm": 1.1166276512681408, + "learning_rate": 9.980129044742908e-06, + "loss": 0.191, + "step": 1035 + }, + { + "epoch": 0.06, + "grad_norm": 1.5782207516619255, + "learning_rate": 9.980048730447931e-06, + "loss": 0.2333, + "step": 1036 + }, + { + "epoch": 0.06, + "grad_norm": 1.356467923112551, + "learning_rate": 9.979968254497709e-06, + "loss": 0.2496, + "step": 1037 + }, + { + "epoch": 0.06, + "grad_norm": 1.1586807080083221, + "learning_rate": 9.979887616894853e-06, + "loss": 0.1681, + "step": 1038 + }, + { + "epoch": 0.06, + "grad_norm": 1.5044882824691215, + "learning_rate": 9.979806817641982e-06, + "loss": 0.2715, + "step": 1039 + }, + { + "epoch": 0.06, + "grad_norm": 1.608116558321884, + "learning_rate": 9.979725856741717e-06, + "loss": 0.1615, + "step": 1040 + }, + { + "epoch": 0.06, + "grad_norm": 1.7380444183174493, + "learning_rate": 9.979644734196689e-06, + "loss": 0.2722, + "step": 1041 + }, + { + "epoch": 0.06, + "grad_norm": 1.212400279694366, + "learning_rate": 9.979563450009529e-06, + "loss": 0.1864, + "step": 1042 + }, + { + "epoch": 0.06, + "grad_norm": 1.3972279629685254, + "learning_rate": 9.979482004182875e-06, + "loss": 0.2434, + "step": 1043 + }, + { + "epoch": 0.06, + "grad_norm": 0.9651048071585169, + "learning_rate": 9.979400396719371e-06, + "loss": 0.2157, + "step": 1044 + }, + { + "epoch": 0.06, + "grad_norm": 1.3254184858259093, + "learning_rate": 9.97931862762167e-06, + "loss": 0.179, + "step": 1045 + }, + { + "epoch": 0.06, + "grad_norm": 1.2126011487651382, + "learning_rate": 9.97923669689242e-06, + "loss": 0.1982, + "step": 1046 + }, + { + "epoch": 0.06, + "grad_norm": 1.0638866649684466, + "learning_rate": 9.979154604534284e-06, + "loss": 0.2111, + "step": 1047 + }, + { + "epoch": 0.06, + "grad_norm": 1.2915385611198724, + "learning_rate": 9.979072350549927e-06, + "loss": 0.2082, + "step": 1048 + }, + { + "epoch": 0.06, + "grad_norm": 1.460339173788245, + "learning_rate": 9.978989934942018e-06, + "loss": 0.1828, + "step": 1049 + }, + { + "epoch": 0.06, + "grad_norm": 1.5152332731234583, + "learning_rate": 9.978907357713233e-06, + "loss": 0.2307, + "step": 1050 + }, + { + "epoch": 0.06, + "grad_norm": 1.3680527578491752, + "learning_rate": 9.978824618866252e-06, + "loss": 0.2365, + "step": 1051 + }, + { + "epoch": 0.06, + "grad_norm": 1.6014565874760736, + "learning_rate": 9.978741718403761e-06, + "loss": 0.2028, + "step": 1052 + }, + { + "epoch": 0.06, + "grad_norm": 1.2004489180676827, + "learning_rate": 9.978658656328451e-06, + "loss": 0.197, + "step": 1053 + }, + { + "epoch": 0.06, + "grad_norm": 1.5499932908097547, + "learning_rate": 9.978575432643018e-06, + "loss": 0.2085, + "step": 1054 + }, + { + "epoch": 0.06, + "grad_norm": 1.5193147676706682, + "learning_rate": 9.978492047350164e-06, + "loss": 0.2529, + "step": 1055 + }, + { + "epoch": 0.06, + "grad_norm": 1.2343401330538368, + "learning_rate": 9.978408500452593e-06, + "loss": 0.1798, + "step": 1056 + }, + { + "epoch": 0.06, + "grad_norm": 1.17362880442648, + "learning_rate": 9.97832479195302e-06, + "loss": 0.2344, + "step": 1057 + }, + { + "epoch": 0.06, + "grad_norm": 1.8033892577551598, + "learning_rate": 9.978240921854163e-06, + "loss": 0.2735, + "step": 1058 + }, + { + "epoch": 0.06, + "grad_norm": 1.1623977046996727, + "learning_rate": 9.978156890158744e-06, + "loss": 0.1886, + "step": 1059 + }, + { + "epoch": 0.06, + "grad_norm": 1.2457845817787145, + "learning_rate": 9.978072696869488e-06, + "loss": 0.194, + "step": 1060 + }, + { + "epoch": 0.06, + "grad_norm": 2.983520864918462, + "learning_rate": 9.97798834198913e-06, + "loss": 0.2074, + "step": 1061 + }, + { + "epoch": 0.06, + "grad_norm": 1.4231519519463496, + "learning_rate": 9.977903825520406e-06, + "loss": 0.2406, + "step": 1062 + }, + { + "epoch": 0.06, + "grad_norm": 1.1931269086809513, + "learning_rate": 9.977819147466064e-06, + "loss": 0.2525, + "step": 1063 + }, + { + "epoch": 0.06, + "grad_norm": 1.7768871779415316, + "learning_rate": 9.977734307828848e-06, + "loss": 0.1975, + "step": 1064 + }, + { + "epoch": 0.06, + "grad_norm": 1.5985628606770355, + "learning_rate": 9.977649306611514e-06, + "loss": 0.2274, + "step": 1065 + }, + { + "epoch": 0.06, + "grad_norm": 1.4786693815560086, + "learning_rate": 9.977564143816822e-06, + "loss": 0.2568, + "step": 1066 + }, + { + "epoch": 0.06, + "grad_norm": 1.3039492671446387, + "learning_rate": 9.977478819447538e-06, + "loss": 0.2579, + "step": 1067 + }, + { + "epoch": 0.06, + "grad_norm": 1.4226615532781977, + "learning_rate": 9.977393333506424e-06, + "loss": 0.2085, + "step": 1068 + }, + { + "epoch": 0.06, + "grad_norm": 1.0142143227699938, + "learning_rate": 9.977307685996264e-06, + "loss": 0.1719, + "step": 1069 + }, + { + "epoch": 0.06, + "grad_norm": 1.5551789401993938, + "learning_rate": 9.977221876919833e-06, + "loss": 0.2348, + "step": 1070 + }, + { + "epoch": 0.06, + "grad_norm": 1.274074854230643, + "learning_rate": 9.977135906279919e-06, + "loss": 0.2379, + "step": 1071 + }, + { + "epoch": 0.06, + "grad_norm": 1.1426826934566265, + "learning_rate": 9.977049774079311e-06, + "loss": 0.2135, + "step": 1072 + }, + { + "epoch": 0.06, + "grad_norm": 1.3485706157554151, + "learning_rate": 9.976963480320805e-06, + "loss": 0.1698, + "step": 1073 + }, + { + "epoch": 0.06, + "grad_norm": 1.4279162842887603, + "learning_rate": 9.9768770250072e-06, + "loss": 0.2187, + "step": 1074 + }, + { + "epoch": 0.06, + "grad_norm": 1.5395495883248618, + "learning_rate": 9.97679040814131e-06, + "loss": 0.2046, + "step": 1075 + }, + { + "epoch": 0.06, + "grad_norm": 1.375191207888285, + "learning_rate": 9.976703629725939e-06, + "loss": 0.1631, + "step": 1076 + }, + { + "epoch": 0.06, + "grad_norm": 1.4872179788823656, + "learning_rate": 9.976616689763905e-06, + "loss": 0.26, + "step": 1077 + }, + { + "epoch": 0.06, + "grad_norm": 2.0683896257963816, + "learning_rate": 9.976529588258032e-06, + "loss": 0.2537, + "step": 1078 + }, + { + "epoch": 0.06, + "grad_norm": 1.725141898828764, + "learning_rate": 9.97644232521115e-06, + "loss": 0.2319, + "step": 1079 + }, + { + "epoch": 0.06, + "grad_norm": 1.5301394498259802, + "learning_rate": 9.976354900626085e-06, + "loss": 0.2308, + "step": 1080 + }, + { + "epoch": 0.06, + "grad_norm": 1.5084850091330002, + "learning_rate": 9.976267314505679e-06, + "loss": 0.2506, + "step": 1081 + }, + { + "epoch": 0.06, + "grad_norm": 1.3275396291202055, + "learning_rate": 9.976179566852773e-06, + "loss": 0.2363, + "step": 1082 + }, + { + "epoch": 0.06, + "grad_norm": 1.620963095949561, + "learning_rate": 9.976091657670219e-06, + "loss": 0.2497, + "step": 1083 + }, + { + "epoch": 0.06, + "grad_norm": 1.5403741775519166, + "learning_rate": 9.976003586960867e-06, + "loss": 0.2511, + "step": 1084 + }, + { + "epoch": 0.06, + "grad_norm": 1.302273730344846, + "learning_rate": 9.97591535472758e-06, + "loss": 0.2463, + "step": 1085 + }, + { + "epoch": 0.06, + "grad_norm": 0.9751420763086212, + "learning_rate": 9.975826960973217e-06, + "loss": 0.2089, + "step": 1086 + }, + { + "epoch": 0.06, + "grad_norm": 1.255645155248849, + "learning_rate": 9.975738405700649e-06, + "loss": 0.2556, + "step": 1087 + }, + { + "epoch": 0.06, + "grad_norm": 1.4553224038525743, + "learning_rate": 9.975649688912752e-06, + "loss": 0.2143, + "step": 1088 + }, + { + "epoch": 0.06, + "grad_norm": 1.3479339285894436, + "learning_rate": 9.975560810612405e-06, + "loss": 0.2481, + "step": 1089 + }, + { + "epoch": 0.06, + "grad_norm": 1.136295944576476, + "learning_rate": 9.975471770802491e-06, + "loss": 0.1905, + "step": 1090 + }, + { + "epoch": 0.06, + "grad_norm": 1.0786396643337717, + "learning_rate": 9.975382569485906e-06, + "loss": 0.2233, + "step": 1091 + }, + { + "epoch": 0.06, + "grad_norm": 1.2482098486430877, + "learning_rate": 9.975293206665538e-06, + "loss": 0.2392, + "step": 1092 + }, + { + "epoch": 0.06, + "grad_norm": 1.4142840716343987, + "learning_rate": 9.975203682344295e-06, + "loss": 0.1945, + "step": 1093 + }, + { + "epoch": 0.06, + "grad_norm": 1.687878474646679, + "learning_rate": 9.975113996525077e-06, + "loss": 0.2276, + "step": 1094 + }, + { + "epoch": 0.06, + "grad_norm": 1.569324551525009, + "learning_rate": 9.975024149210797e-06, + "loss": 0.2575, + "step": 1095 + }, + { + "epoch": 0.06, + "grad_norm": 1.0269006001798315, + "learning_rate": 9.974934140404374e-06, + "loss": 0.2567, + "step": 1096 + }, + { + "epoch": 0.06, + "grad_norm": 1.7396789696287105, + "learning_rate": 9.974843970108728e-06, + "loss": 0.2455, + "step": 1097 + }, + { + "epoch": 0.06, + "grad_norm": 1.291552293227994, + "learning_rate": 9.974753638326787e-06, + "loss": 0.2183, + "step": 1098 + }, + { + "epoch": 0.06, + "grad_norm": 1.0802537928719138, + "learning_rate": 9.97466314506148e-06, + "loss": 0.2078, + "step": 1099 + }, + { + "epoch": 0.06, + "grad_norm": 1.5885525012558797, + "learning_rate": 9.974572490315748e-06, + "loss": 0.2491, + "step": 1100 + }, + { + "epoch": 0.06, + "grad_norm": 1.5450829959626144, + "learning_rate": 9.974481674092532e-06, + "loss": 0.2204, + "step": 1101 + }, + { + "epoch": 0.06, + "grad_norm": 3.1398201001167174, + "learning_rate": 9.97439069639478e-06, + "loss": 0.2879, + "step": 1102 + }, + { + "epoch": 0.06, + "grad_norm": 1.7635262303503902, + "learning_rate": 9.974299557225445e-06, + "loss": 0.2706, + "step": 1103 + }, + { + "epoch": 0.06, + "grad_norm": 0.9778437163526165, + "learning_rate": 9.974208256587488e-06, + "loss": 0.1851, + "step": 1104 + }, + { + "epoch": 0.06, + "grad_norm": 0.9963944168339308, + "learning_rate": 9.974116794483867e-06, + "loss": 0.217, + "step": 1105 + }, + { + "epoch": 0.06, + "grad_norm": 1.5551605271769557, + "learning_rate": 9.974025170917559e-06, + "loss": 0.2887, + "step": 1106 + }, + { + "epoch": 0.06, + "grad_norm": 1.6296051919191152, + "learning_rate": 9.97393338589153e-06, + "loss": 0.2193, + "step": 1107 + }, + { + "epoch": 0.06, + "grad_norm": 1.6648401538697264, + "learning_rate": 9.973841439408765e-06, + "loss": 0.2625, + "step": 1108 + }, + { + "epoch": 0.06, + "grad_norm": 0.8883196625334552, + "learning_rate": 9.973749331472247e-06, + "loss": 0.1687, + "step": 1109 + }, + { + "epoch": 0.06, + "grad_norm": 1.4129830746756131, + "learning_rate": 9.973657062084965e-06, + "loss": 0.2089, + "step": 1110 + }, + { + "epoch": 0.06, + "grad_norm": 1.5961235501069702, + "learning_rate": 9.973564631249915e-06, + "loss": 0.2486, + "step": 1111 + }, + { + "epoch": 0.06, + "grad_norm": 1.2755840966779854, + "learning_rate": 9.973472038970096e-06, + "loss": 0.2187, + "step": 1112 + }, + { + "epoch": 0.06, + "grad_norm": 1.5905142431726182, + "learning_rate": 9.973379285248515e-06, + "loss": 0.2721, + "step": 1113 + }, + { + "epoch": 0.06, + "grad_norm": 1.6272583568692174, + "learning_rate": 9.973286370088183e-06, + "loss": 0.2757, + "step": 1114 + }, + { + "epoch": 0.06, + "grad_norm": 1.4034509379011764, + "learning_rate": 9.973193293492116e-06, + "loss": 0.2348, + "step": 1115 + }, + { + "epoch": 0.06, + "grad_norm": 1.2640838423201548, + "learning_rate": 9.973100055463334e-06, + "loss": 0.2091, + "step": 1116 + }, + { + "epoch": 0.06, + "grad_norm": 1.733951557722844, + "learning_rate": 9.973006656004865e-06, + "loss": 0.2693, + "step": 1117 + }, + { + "epoch": 0.06, + "grad_norm": 1.8540303557676574, + "learning_rate": 9.972913095119741e-06, + "loss": 0.2122, + "step": 1118 + }, + { + "epoch": 0.06, + "grad_norm": 1.162211813539026, + "learning_rate": 9.972819372810995e-06, + "loss": 0.2669, + "step": 1119 + }, + { + "epoch": 0.06, + "grad_norm": 1.1191814017398378, + "learning_rate": 9.972725489081676e-06, + "loss": 0.1838, + "step": 1120 + }, + { + "epoch": 0.06, + "grad_norm": 1.7908737975096738, + "learning_rate": 9.972631443934828e-06, + "loss": 0.2648, + "step": 1121 + }, + { + "epoch": 0.06, + "grad_norm": 1.8151183084477929, + "learning_rate": 9.972537237373504e-06, + "loss": 0.2286, + "step": 1122 + }, + { + "epoch": 0.06, + "grad_norm": 0.9877329292851007, + "learning_rate": 9.97244286940076e-06, + "loss": 0.2467, + "step": 1123 + }, + { + "epoch": 0.06, + "grad_norm": 1.0101496678187325, + "learning_rate": 9.972348340019661e-06, + "loss": 0.2285, + "step": 1124 + }, + { + "epoch": 0.06, + "grad_norm": 1.794749516294568, + "learning_rate": 9.972253649233277e-06, + "loss": 0.2413, + "step": 1125 + }, + { + "epoch": 0.06, + "grad_norm": 1.3388058114817674, + "learning_rate": 9.97215879704468e-06, + "loss": 0.2615, + "step": 1126 + }, + { + "epoch": 0.06, + "grad_norm": 1.541568252095337, + "learning_rate": 9.972063783456949e-06, + "loss": 0.2176, + "step": 1127 + }, + { + "epoch": 0.06, + "grad_norm": 1.092783968775016, + "learning_rate": 9.971968608473168e-06, + "loss": 0.2207, + "step": 1128 + }, + { + "epoch": 0.06, + "grad_norm": 1.1739509457965303, + "learning_rate": 9.971873272096427e-06, + "loss": 0.2189, + "step": 1129 + }, + { + "epoch": 0.06, + "grad_norm": 1.9050224198477936, + "learning_rate": 9.971777774329821e-06, + "loss": 0.215, + "step": 1130 + }, + { + "epoch": 0.06, + "grad_norm": 1.5660719383964075, + "learning_rate": 9.971682115176449e-06, + "loss": 0.2319, + "step": 1131 + }, + { + "epoch": 0.06, + "grad_norm": 1.6148626183384251, + "learning_rate": 9.971586294639418e-06, + "loss": 0.2024, + "step": 1132 + }, + { + "epoch": 0.06, + "grad_norm": 1.2279040113519024, + "learning_rate": 9.971490312721837e-06, + "loss": 0.238, + "step": 1133 + }, + { + "epoch": 0.06, + "grad_norm": 2.326344775893484, + "learning_rate": 9.97139416942682e-06, + "loss": 0.1964, + "step": 1134 + }, + { + "epoch": 0.06, + "grad_norm": 1.5633242495398034, + "learning_rate": 9.97129786475749e-06, + "loss": 0.2694, + "step": 1135 + }, + { + "epoch": 0.06, + "grad_norm": 1.9358990384484664, + "learning_rate": 9.971201398716972e-06, + "loss": 0.1894, + "step": 1136 + }, + { + "epoch": 0.06, + "grad_norm": 0.9689453460641065, + "learning_rate": 9.9711047713084e-06, + "loss": 0.1829, + "step": 1137 + }, + { + "epoch": 0.06, + "grad_norm": 1.1911443578824656, + "learning_rate": 9.971007982534906e-06, + "loss": 0.2163, + "step": 1138 + }, + { + "epoch": 0.06, + "grad_norm": 1.1511987258774994, + "learning_rate": 9.970911032399637e-06, + "loss": 0.212, + "step": 1139 + }, + { + "epoch": 0.06, + "grad_norm": 1.6614171630791912, + "learning_rate": 9.970813920905735e-06, + "loss": 0.2954, + "step": 1140 + }, + { + "epoch": 0.06, + "grad_norm": 1.3018148685895115, + "learning_rate": 9.970716648056356e-06, + "loss": 0.2078, + "step": 1141 + }, + { + "epoch": 0.06, + "grad_norm": 0.9700106716490696, + "learning_rate": 9.970619213854655e-06, + "loss": 0.245, + "step": 1142 + }, + { + "epoch": 0.06, + "grad_norm": 1.3256875145223679, + "learning_rate": 9.970521618303798e-06, + "loss": 0.2101, + "step": 1143 + }, + { + "epoch": 0.06, + "grad_norm": 1.4619544755628442, + "learning_rate": 9.97042386140695e-06, + "loss": 0.2301, + "step": 1144 + }, + { + "epoch": 0.06, + "grad_norm": 1.5978720068687378, + "learning_rate": 9.970325943167285e-06, + "loss": 0.2588, + "step": 1145 + }, + { + "epoch": 0.06, + "grad_norm": 1.583913140629568, + "learning_rate": 9.970227863587983e-06, + "loss": 0.228, + "step": 1146 + }, + { + "epoch": 0.06, + "grad_norm": 1.432970350612137, + "learning_rate": 9.970129622672225e-06, + "loss": 0.2215, + "step": 1147 + }, + { + "epoch": 0.06, + "grad_norm": 1.6365386231024384, + "learning_rate": 9.970031220423202e-06, + "loss": 0.2379, + "step": 1148 + }, + { + "epoch": 0.06, + "grad_norm": 1.1763369587233856, + "learning_rate": 9.969932656844106e-06, + "loss": 0.2186, + "step": 1149 + }, + { + "epoch": 0.06, + "grad_norm": 1.6031551333387646, + "learning_rate": 9.96983393193814e-06, + "loss": 0.2084, + "step": 1150 + }, + { + "epoch": 0.06, + "grad_norm": 1.44545184829755, + "learning_rate": 9.969735045708506e-06, + "loss": 0.1937, + "step": 1151 + }, + { + "epoch": 0.06, + "grad_norm": 1.6201716063812184, + "learning_rate": 9.969635998158415e-06, + "loss": 0.2682, + "step": 1152 + }, + { + "epoch": 0.06, + "grad_norm": 1.296238800633165, + "learning_rate": 9.969536789291082e-06, + "loss": 0.2258, + "step": 1153 + }, + { + "epoch": 0.06, + "grad_norm": 1.120666938541898, + "learning_rate": 9.969437419109726e-06, + "loss": 0.2177, + "step": 1154 + }, + { + "epoch": 0.06, + "grad_norm": 1.8720273225372333, + "learning_rate": 9.969337887617575e-06, + "loss": 0.218, + "step": 1155 + }, + { + "epoch": 0.06, + "grad_norm": 1.2822775979500782, + "learning_rate": 9.969238194817859e-06, + "loss": 0.2424, + "step": 1156 + }, + { + "epoch": 0.06, + "grad_norm": 1.6663800407848786, + "learning_rate": 9.969138340713812e-06, + "loss": 0.2108, + "step": 1157 + }, + { + "epoch": 0.06, + "grad_norm": 1.5279340554513465, + "learning_rate": 9.969038325308678e-06, + "loss": 0.2424, + "step": 1158 + }, + { + "epoch": 0.06, + "grad_norm": 1.3855930554588178, + "learning_rate": 9.968938148605703e-06, + "loss": 0.2061, + "step": 1159 + }, + { + "epoch": 0.06, + "grad_norm": 2.1246209823367077, + "learning_rate": 9.968837810608137e-06, + "loss": 0.2509, + "step": 1160 + }, + { + "epoch": 0.06, + "grad_norm": 1.3187173936528143, + "learning_rate": 9.96873731131924e-06, + "loss": 0.2071, + "step": 1161 + }, + { + "epoch": 0.06, + "grad_norm": 1.1811594432603254, + "learning_rate": 9.96863665074227e-06, + "loss": 0.1889, + "step": 1162 + }, + { + "epoch": 0.06, + "grad_norm": 1.20653708073265, + "learning_rate": 9.9685358288805e-06, + "loss": 0.2407, + "step": 1163 + }, + { + "epoch": 0.06, + "grad_norm": 1.0489369874005263, + "learning_rate": 9.968434845737198e-06, + "loss": 0.1911, + "step": 1164 + }, + { + "epoch": 0.06, + "grad_norm": 2.005586444929045, + "learning_rate": 9.968333701315644e-06, + "loss": 0.2535, + "step": 1165 + }, + { + "epoch": 0.06, + "grad_norm": 1.8373086111149095, + "learning_rate": 9.968232395619121e-06, + "loss": 0.1946, + "step": 1166 + }, + { + "epoch": 0.06, + "grad_norm": 1.8352536977437872, + "learning_rate": 9.96813092865092e-06, + "loss": 0.212, + "step": 1167 + }, + { + "epoch": 0.06, + "grad_norm": 10.533414623896471, + "learning_rate": 9.96802930041433e-06, + "loss": 0.2433, + "step": 1168 + }, + { + "epoch": 0.07, + "grad_norm": 1.0594594871116747, + "learning_rate": 9.96792751091265e-06, + "loss": 0.1699, + "step": 1169 + }, + { + "epoch": 0.07, + "grad_norm": 1.2073656902736285, + "learning_rate": 9.967825560149191e-06, + "loss": 0.1534, + "step": 1170 + }, + { + "epoch": 0.07, + "grad_norm": 1.724766370943612, + "learning_rate": 9.967723448127255e-06, + "loss": 0.1709, + "step": 1171 + }, + { + "epoch": 0.07, + "grad_norm": 1.605598461869892, + "learning_rate": 9.96762117485016e-06, + "loss": 0.1792, + "step": 1172 + }, + { + "epoch": 0.07, + "grad_norm": 1.2858774891937985, + "learning_rate": 9.967518740321224e-06, + "loss": 0.1835, + "step": 1173 + }, + { + "epoch": 0.07, + "grad_norm": 1.5327312362878311, + "learning_rate": 9.967416144543773e-06, + "loss": 0.1769, + "step": 1174 + }, + { + "epoch": 0.07, + "grad_norm": 1.3358248752391468, + "learning_rate": 9.967313387521139e-06, + "loss": 0.2008, + "step": 1175 + }, + { + "epoch": 0.07, + "grad_norm": 1.8299385432831197, + "learning_rate": 9.967210469256657e-06, + "loss": 0.2346, + "step": 1176 + }, + { + "epoch": 0.07, + "grad_norm": 1.1703193065919026, + "learning_rate": 9.967107389753663e-06, + "loss": 0.2066, + "step": 1177 + }, + { + "epoch": 0.07, + "grad_norm": 2.0022700510143925, + "learning_rate": 9.967004149015508e-06, + "loss": 0.2555, + "step": 1178 + }, + { + "epoch": 0.07, + "grad_norm": 1.4813826314481562, + "learning_rate": 9.966900747045545e-06, + "loss": 0.2709, + "step": 1179 + }, + { + "epoch": 0.07, + "grad_norm": 1.884743456141887, + "learning_rate": 9.966797183847125e-06, + "loss": 0.2442, + "step": 1180 + }, + { + "epoch": 0.07, + "grad_norm": 2.064939080807794, + "learning_rate": 9.966693459423613e-06, + "loss": 0.252, + "step": 1181 + }, + { + "epoch": 0.07, + "grad_norm": 1.331739228644009, + "learning_rate": 9.966589573778376e-06, + "loss": 0.1924, + "step": 1182 + }, + { + "epoch": 0.07, + "grad_norm": 1.6388789240508113, + "learning_rate": 9.966485526914784e-06, + "loss": 0.2244, + "step": 1183 + }, + { + "epoch": 0.07, + "grad_norm": 2.1134589520693603, + "learning_rate": 9.966381318836217e-06, + "loss": 0.2254, + "step": 1184 + }, + { + "epoch": 0.07, + "grad_norm": 2.33596388377724, + "learning_rate": 9.966276949546054e-06, + "loss": 0.2408, + "step": 1185 + }, + { + "epoch": 0.07, + "grad_norm": 1.1757510599588439, + "learning_rate": 9.966172419047688e-06, + "loss": 0.2024, + "step": 1186 + }, + { + "epoch": 0.07, + "grad_norm": 1.4900783442135732, + "learning_rate": 9.96606772734451e-06, + "loss": 0.2195, + "step": 1187 + }, + { + "epoch": 0.07, + "grad_norm": 1.5433092924549787, + "learning_rate": 9.965962874439916e-06, + "loss": 0.2065, + "step": 1188 + }, + { + "epoch": 0.07, + "grad_norm": 1.6079502233939778, + "learning_rate": 9.965857860337312e-06, + "loss": 0.2225, + "step": 1189 + }, + { + "epoch": 0.07, + "grad_norm": 1.2695019322457997, + "learning_rate": 9.965752685040106e-06, + "loss": 0.233, + "step": 1190 + }, + { + "epoch": 0.07, + "grad_norm": 1.3211265211648366, + "learning_rate": 9.965647348551713e-06, + "loss": 0.2087, + "step": 1191 + }, + { + "epoch": 0.07, + "grad_norm": 6.3737631701258355, + "learning_rate": 9.965541850875551e-06, + "loss": 0.2571, + "step": 1192 + }, + { + "epoch": 0.07, + "grad_norm": 1.550481660428357, + "learning_rate": 9.965436192015047e-06, + "loss": 0.2024, + "step": 1193 + }, + { + "epoch": 0.07, + "grad_norm": 1.6879537246163894, + "learning_rate": 9.965330371973627e-06, + "loss": 0.2243, + "step": 1194 + }, + { + "epoch": 0.07, + "grad_norm": 1.6184137505105205, + "learning_rate": 9.965224390754728e-06, + "loss": 0.2253, + "step": 1195 + }, + { + "epoch": 0.07, + "grad_norm": 1.3625814320843634, + "learning_rate": 9.965118248361791e-06, + "loss": 0.2639, + "step": 1196 + }, + { + "epoch": 0.07, + "grad_norm": 1.3113752552661642, + "learning_rate": 9.96501194479826e-06, + "loss": 0.25, + "step": 1197 + }, + { + "epoch": 0.07, + "grad_norm": 1.4997644040738192, + "learning_rate": 9.964905480067585e-06, + "loss": 0.2184, + "step": 1198 + }, + { + "epoch": 0.07, + "grad_norm": 1.1841394799837206, + "learning_rate": 9.964798854173226e-06, + "loss": 0.2479, + "step": 1199 + }, + { + "epoch": 0.07, + "grad_norm": 1.23311765529864, + "learning_rate": 9.964692067118638e-06, + "loss": 0.1735, + "step": 1200 + }, + { + "epoch": 0.07, + "grad_norm": 0.9638356065455886, + "learning_rate": 9.964585118907293e-06, + "loss": 0.2424, + "step": 1201 + }, + { + "epoch": 0.07, + "grad_norm": 2.0383486978199263, + "learning_rate": 9.964478009542659e-06, + "loss": 0.2224, + "step": 1202 + }, + { + "epoch": 0.07, + "grad_norm": 3.1660810648551974, + "learning_rate": 9.964370739028213e-06, + "loss": 0.2383, + "step": 1203 + }, + { + "epoch": 0.07, + "grad_norm": 1.3042102621875598, + "learning_rate": 9.96426330736744e-06, + "loss": 0.1548, + "step": 1204 + }, + { + "epoch": 0.07, + "grad_norm": 1.3229613748702798, + "learning_rate": 9.964155714563825e-06, + "loss": 0.1977, + "step": 1205 + }, + { + "epoch": 0.07, + "grad_norm": 1.303266364458382, + "learning_rate": 9.96404796062086e-06, + "loss": 0.2115, + "step": 1206 + }, + { + "epoch": 0.07, + "grad_norm": 1.4822427618810599, + "learning_rate": 9.963940045542045e-06, + "loss": 0.2579, + "step": 1207 + }, + { + "epoch": 0.07, + "grad_norm": 1.3540751964291668, + "learning_rate": 9.96383196933088e-06, + "loss": 0.2132, + "step": 1208 + }, + { + "epoch": 0.07, + "grad_norm": 1.6814753781294958, + "learning_rate": 9.963723731990876e-06, + "loss": 0.2121, + "step": 1209 + }, + { + "epoch": 0.07, + "grad_norm": 1.264337773993661, + "learning_rate": 9.963615333525545e-06, + "loss": 0.2229, + "step": 1210 + }, + { + "epoch": 0.07, + "grad_norm": 1.2407482674517611, + "learning_rate": 9.963506773938407e-06, + "loss": 0.1514, + "step": 1211 + }, + { + "epoch": 0.07, + "grad_norm": 1.0263612458468134, + "learning_rate": 9.963398053232983e-06, + "loss": 0.2047, + "step": 1212 + }, + { + "epoch": 0.07, + "grad_norm": 1.1170530213267242, + "learning_rate": 9.963289171412805e-06, + "loss": 0.2308, + "step": 1213 + }, + { + "epoch": 0.07, + "grad_norm": 1.2222907766288316, + "learning_rate": 9.963180128481407e-06, + "loss": 0.1959, + "step": 1214 + }, + { + "epoch": 0.07, + "grad_norm": 1.9512990188413355, + "learning_rate": 9.963070924442327e-06, + "loss": 0.2044, + "step": 1215 + }, + { + "epoch": 0.07, + "grad_norm": 1.8226421923857645, + "learning_rate": 9.962961559299112e-06, + "loss": 0.2589, + "step": 1216 + }, + { + "epoch": 0.07, + "grad_norm": 1.2292193745637818, + "learning_rate": 9.96285203305531e-06, + "loss": 0.2249, + "step": 1217 + }, + { + "epoch": 0.07, + "grad_norm": 1.366735045633654, + "learning_rate": 9.962742345714478e-06, + "loss": 0.2723, + "step": 1218 + }, + { + "epoch": 0.07, + "grad_norm": 1.4177191254496748, + "learning_rate": 9.962632497280173e-06, + "loss": 0.2404, + "step": 1219 + }, + { + "epoch": 0.07, + "grad_norm": 1.1159206189716409, + "learning_rate": 9.962522487755967e-06, + "loss": 0.2305, + "step": 1220 + }, + { + "epoch": 0.07, + "grad_norm": 1.2420738801036055, + "learning_rate": 9.962412317145425e-06, + "loss": 0.1768, + "step": 1221 + }, + { + "epoch": 0.07, + "grad_norm": 0.9856528699547278, + "learning_rate": 9.962301985452128e-06, + "loss": 0.2182, + "step": 1222 + }, + { + "epoch": 0.07, + "grad_norm": 1.69177014442735, + "learning_rate": 9.962191492679651e-06, + "loss": 0.2212, + "step": 1223 + }, + { + "epoch": 0.07, + "grad_norm": 1.6762725236678755, + "learning_rate": 9.962080838831587e-06, + "loss": 0.2021, + "step": 1224 + }, + { + "epoch": 0.07, + "grad_norm": 1.1975803241688392, + "learning_rate": 9.961970023911527e-06, + "loss": 0.1698, + "step": 1225 + }, + { + "epoch": 0.07, + "grad_norm": 1.6352117122378995, + "learning_rate": 9.961859047923066e-06, + "loss": 0.2514, + "step": 1226 + }, + { + "epoch": 0.07, + "grad_norm": 1.4827589109261874, + "learning_rate": 9.961747910869806e-06, + "loss": 0.2412, + "step": 1227 + }, + { + "epoch": 0.07, + "grad_norm": 1.3689301324896115, + "learning_rate": 9.961636612755355e-06, + "loss": 0.2007, + "step": 1228 + }, + { + "epoch": 0.07, + "grad_norm": 1.1162307750844793, + "learning_rate": 9.961525153583327e-06, + "loss": 0.1994, + "step": 1229 + }, + { + "epoch": 0.07, + "grad_norm": 1.2635035336372953, + "learning_rate": 9.961413533357341e-06, + "loss": 0.1718, + "step": 1230 + }, + { + "epoch": 0.07, + "grad_norm": 1.142958390299116, + "learning_rate": 9.961301752081018e-06, + "loss": 0.1994, + "step": 1231 + }, + { + "epoch": 0.07, + "grad_norm": 1.6424752360811217, + "learning_rate": 9.961189809757985e-06, + "loss": 0.2404, + "step": 1232 + }, + { + "epoch": 0.07, + "grad_norm": 1.521669472048194, + "learning_rate": 9.96107770639188e-06, + "loss": 0.1776, + "step": 1233 + }, + { + "epoch": 0.07, + "grad_norm": 1.6049726513288376, + "learning_rate": 9.96096544198634e-06, + "loss": 0.25, + "step": 1234 + }, + { + "epoch": 0.07, + "grad_norm": 2.3419313956877437, + "learning_rate": 9.96085301654501e-06, + "loss": 0.2622, + "step": 1235 + }, + { + "epoch": 0.07, + "grad_norm": 1.667985659207615, + "learning_rate": 9.960740430071537e-06, + "loss": 0.2577, + "step": 1236 + }, + { + "epoch": 0.07, + "grad_norm": 1.3878485133994307, + "learning_rate": 9.960627682569574e-06, + "loss": 0.1936, + "step": 1237 + }, + { + "epoch": 0.07, + "grad_norm": 1.3813732484649426, + "learning_rate": 9.960514774042787e-06, + "loss": 0.2353, + "step": 1238 + }, + { + "epoch": 0.07, + "grad_norm": 1.2648490555164826, + "learning_rate": 9.960401704494837e-06, + "loss": 0.2376, + "step": 1239 + }, + { + "epoch": 0.07, + "grad_norm": 5.93126487609477, + "learning_rate": 9.960288473929397e-06, + "loss": 0.2317, + "step": 1240 + }, + { + "epoch": 0.07, + "grad_norm": 1.7374729506013138, + "learning_rate": 9.96017508235014e-06, + "loss": 0.2769, + "step": 1241 + }, + { + "epoch": 0.07, + "grad_norm": 1.0267814395501615, + "learning_rate": 9.960061529760748e-06, + "loss": 0.2041, + "step": 1242 + }, + { + "epoch": 0.07, + "grad_norm": 1.671547815358573, + "learning_rate": 9.959947816164905e-06, + "loss": 0.1681, + "step": 1243 + }, + { + "epoch": 0.07, + "grad_norm": 1.291475989431219, + "learning_rate": 9.959833941566304e-06, + "loss": 0.2079, + "step": 1244 + }, + { + "epoch": 0.07, + "grad_norm": 1.1124200041076264, + "learning_rate": 9.959719905968642e-06, + "loss": 0.2128, + "step": 1245 + }, + { + "epoch": 0.07, + "grad_norm": 1.0320198651522543, + "learning_rate": 9.95960570937562e-06, + "loss": 0.2012, + "step": 1246 + }, + { + "epoch": 0.07, + "grad_norm": 0.9434107851350768, + "learning_rate": 9.959491351790942e-06, + "loss": 0.2029, + "step": 1247 + }, + { + "epoch": 0.07, + "grad_norm": 1.004121744279179, + "learning_rate": 9.959376833218325e-06, + "loss": 0.2406, + "step": 1248 + }, + { + "epoch": 0.07, + "grad_norm": 1.5059661592355678, + "learning_rate": 9.959262153661484e-06, + "loss": 0.234, + "step": 1249 + }, + { + "epoch": 0.07, + "grad_norm": 1.4129993218408412, + "learning_rate": 9.959147313124143e-06, + "loss": 0.2122, + "step": 1250 + }, + { + "epoch": 0.07, + "grad_norm": 1.3093720772495516, + "learning_rate": 9.959032311610026e-06, + "loss": 0.1951, + "step": 1251 + }, + { + "epoch": 0.07, + "grad_norm": 1.1662756998827455, + "learning_rate": 9.95891714912287e-06, + "loss": 0.2174, + "step": 1252 + }, + { + "epoch": 0.07, + "grad_norm": 1.3896017542575558, + "learning_rate": 9.95880182566641e-06, + "loss": 0.2634, + "step": 1253 + }, + { + "epoch": 0.07, + "grad_norm": 1.3822224066460136, + "learning_rate": 9.958686341244394e-06, + "loss": 0.2132, + "step": 1254 + }, + { + "epoch": 0.07, + "grad_norm": 2.1524623277404693, + "learning_rate": 9.958570695860566e-06, + "loss": 0.1989, + "step": 1255 + }, + { + "epoch": 0.07, + "grad_norm": 1.3145134686680187, + "learning_rate": 9.958454889518682e-06, + "loss": 0.1949, + "step": 1256 + }, + { + "epoch": 0.07, + "grad_norm": 1.1229722338211459, + "learning_rate": 9.958338922222503e-06, + "loss": 0.2164, + "step": 1257 + }, + { + "epoch": 0.07, + "grad_norm": 1.4771108870453908, + "learning_rate": 9.958222793975789e-06, + "loss": 0.2461, + "step": 1258 + }, + { + "epoch": 0.07, + "grad_norm": 1.3426861349682295, + "learning_rate": 9.958106504782312e-06, + "loss": 0.2417, + "step": 1259 + }, + { + "epoch": 0.07, + "grad_norm": 1.397093274676967, + "learning_rate": 9.957990054645847e-06, + "loss": 0.2517, + "step": 1260 + }, + { + "epoch": 0.07, + "grad_norm": 1.042955135942394, + "learning_rate": 9.957873443570175e-06, + "loss": 0.2421, + "step": 1261 + }, + { + "epoch": 0.07, + "grad_norm": 1.3357947667875154, + "learning_rate": 9.95775667155908e-06, + "loss": 0.1659, + "step": 1262 + }, + { + "epoch": 0.07, + "grad_norm": 10.789818761397644, + "learning_rate": 9.957639738616351e-06, + "loss": 0.2178, + "step": 1263 + }, + { + "epoch": 0.07, + "grad_norm": 1.2971355911782692, + "learning_rate": 9.957522644745787e-06, + "loss": 0.2198, + "step": 1264 + }, + { + "epoch": 0.07, + "grad_norm": 1.6878141004911116, + "learning_rate": 9.957405389951185e-06, + "loss": 0.2493, + "step": 1265 + }, + { + "epoch": 0.07, + "grad_norm": 1.9401519453749483, + "learning_rate": 9.957287974236355e-06, + "loss": 0.2459, + "step": 1266 + }, + { + "epoch": 0.07, + "grad_norm": 1.0284068260498436, + "learning_rate": 9.957170397605105e-06, + "loss": 0.1921, + "step": 1267 + }, + { + "epoch": 0.07, + "grad_norm": 1.1738073693089135, + "learning_rate": 9.957052660061255e-06, + "loss": 0.2326, + "step": 1268 + }, + { + "epoch": 0.07, + "grad_norm": 2.8473907263187663, + "learning_rate": 9.956934761608625e-06, + "loss": 0.2097, + "step": 1269 + }, + { + "epoch": 0.07, + "grad_norm": 1.3509803383666765, + "learning_rate": 9.956816702251042e-06, + "loss": 0.2554, + "step": 1270 + }, + { + "epoch": 0.07, + "grad_norm": 1.9484346146550722, + "learning_rate": 9.956698481992339e-06, + "loss": 0.2437, + "step": 1271 + }, + { + "epoch": 0.07, + "grad_norm": 1.103761525733859, + "learning_rate": 9.956580100836351e-06, + "loss": 0.2092, + "step": 1272 + }, + { + "epoch": 0.07, + "grad_norm": 2.0892314673505945, + "learning_rate": 9.956461558786926e-06, + "loss": 0.2204, + "step": 1273 + }, + { + "epoch": 0.07, + "grad_norm": 1.5686282691041467, + "learning_rate": 9.956342855847905e-06, + "loss": 0.238, + "step": 1274 + }, + { + "epoch": 0.07, + "grad_norm": 1.4123511575664331, + "learning_rate": 9.956223992023145e-06, + "loss": 0.21, + "step": 1275 + }, + { + "epoch": 0.07, + "grad_norm": 1.2917595579665087, + "learning_rate": 9.956104967316507e-06, + "loss": 0.2105, + "step": 1276 + }, + { + "epoch": 0.07, + "grad_norm": 1.7462893041557859, + "learning_rate": 9.95598578173185e-06, + "loss": 0.2134, + "step": 1277 + }, + { + "epoch": 0.07, + "grad_norm": 1.4425413552088115, + "learning_rate": 9.955866435273045e-06, + "loss": 0.1861, + "step": 1278 + }, + { + "epoch": 0.07, + "grad_norm": 1.9783880827183091, + "learning_rate": 9.955746927943965e-06, + "loss": 0.2163, + "step": 1279 + }, + { + "epoch": 0.07, + "grad_norm": 2.386231642742895, + "learning_rate": 9.955627259748493e-06, + "loss": 0.3036, + "step": 1280 + }, + { + "epoch": 0.07, + "grad_norm": 0.9232775033192736, + "learning_rate": 9.955507430690508e-06, + "loss": 0.1612, + "step": 1281 + }, + { + "epoch": 0.07, + "grad_norm": 1.3909334178265584, + "learning_rate": 9.955387440773902e-06, + "loss": 0.2079, + "step": 1282 + }, + { + "epoch": 0.07, + "grad_norm": 1.385585852710067, + "learning_rate": 9.95526729000257e-06, + "loss": 0.204, + "step": 1283 + }, + { + "epoch": 0.07, + "grad_norm": 1.0908995661892074, + "learning_rate": 9.955146978380412e-06, + "loss": 0.2227, + "step": 1284 + }, + { + "epoch": 0.07, + "grad_norm": 2.2701917104231217, + "learning_rate": 9.955026505911334e-06, + "loss": 0.2973, + "step": 1285 + }, + { + "epoch": 0.07, + "grad_norm": 1.1477740473372868, + "learning_rate": 9.954905872599248e-06, + "loss": 0.2612, + "step": 1286 + }, + { + "epoch": 0.07, + "grad_norm": 1.332133065524745, + "learning_rate": 9.954785078448065e-06, + "loss": 0.1926, + "step": 1287 + }, + { + "epoch": 0.07, + "grad_norm": 1.4570795311995, + "learning_rate": 9.954664123461713e-06, + "loss": 0.2416, + "step": 1288 + }, + { + "epoch": 0.07, + "grad_norm": 1.572419854412675, + "learning_rate": 9.95454300764411e-06, + "loss": 0.2193, + "step": 1289 + }, + { + "epoch": 0.07, + "grad_norm": 1.3232022719327843, + "learning_rate": 9.954421730999195e-06, + "loss": 0.2544, + "step": 1290 + }, + { + "epoch": 0.07, + "grad_norm": 1.2791423806239195, + "learning_rate": 9.954300293530901e-06, + "loss": 0.2226, + "step": 1291 + }, + { + "epoch": 0.07, + "grad_norm": 1.379618436482356, + "learning_rate": 9.954178695243171e-06, + "loss": 0.2087, + "step": 1292 + }, + { + "epoch": 0.07, + "grad_norm": 1.9485328476618322, + "learning_rate": 9.95405693613995e-06, + "loss": 0.2492, + "step": 1293 + }, + { + "epoch": 0.07, + "grad_norm": 1.1716694953654485, + "learning_rate": 9.953935016225194e-06, + "loss": 0.2192, + "step": 1294 + }, + { + "epoch": 0.07, + "grad_norm": 1.402774383390787, + "learning_rate": 9.953812935502858e-06, + "loss": 0.2273, + "step": 1295 + }, + { + "epoch": 0.07, + "grad_norm": 1.083085200964765, + "learning_rate": 9.953690693976904e-06, + "loss": 0.193, + "step": 1296 + }, + { + "epoch": 0.07, + "grad_norm": 1.2707276619560053, + "learning_rate": 9.953568291651302e-06, + "loss": 0.2255, + "step": 1297 + }, + { + "epoch": 0.07, + "grad_norm": 1.0974408291293098, + "learning_rate": 9.953445728530028e-06, + "loss": 0.2249, + "step": 1298 + }, + { + "epoch": 0.07, + "grad_norm": 1.3238291434528646, + "learning_rate": 9.953323004617054e-06, + "loss": 0.2317, + "step": 1299 + }, + { + "epoch": 0.07, + "grad_norm": 1.4860295799177319, + "learning_rate": 9.953200119916368e-06, + "loss": 0.2665, + "step": 1300 + }, + { + "epoch": 0.07, + "grad_norm": 1.277053867192546, + "learning_rate": 9.953077074431956e-06, + "loss": 0.2216, + "step": 1301 + }, + { + "epoch": 0.07, + "grad_norm": 1.9182660480371374, + "learning_rate": 9.952953868167816e-06, + "loss": 0.2312, + "step": 1302 + }, + { + "epoch": 0.07, + "grad_norm": 0.9658080168694501, + "learning_rate": 9.952830501127945e-06, + "loss": 0.1912, + "step": 1303 + }, + { + "epoch": 0.07, + "grad_norm": 1.711948657733515, + "learning_rate": 9.952706973316348e-06, + "loss": 0.2112, + "step": 1304 + }, + { + "epoch": 0.07, + "grad_norm": 1.3920087646882506, + "learning_rate": 9.952583284737034e-06, + "loss": 0.2189, + "step": 1305 + }, + { + "epoch": 0.07, + "grad_norm": 1.5237842409747588, + "learning_rate": 9.95245943539402e-06, + "loss": 0.2224, + "step": 1306 + }, + { + "epoch": 0.07, + "grad_norm": 1.3562791254282667, + "learning_rate": 9.952335425291321e-06, + "loss": 0.2423, + "step": 1307 + }, + { + "epoch": 0.07, + "grad_norm": 1.257638058798033, + "learning_rate": 9.95221125443297e-06, + "loss": 0.2102, + "step": 1308 + }, + { + "epoch": 0.07, + "grad_norm": 1.63113765055615, + "learning_rate": 9.952086922822993e-06, + "loss": 0.2229, + "step": 1309 + }, + { + "epoch": 0.07, + "grad_norm": 1.3367001409032362, + "learning_rate": 9.951962430465427e-06, + "loss": 0.2057, + "step": 1310 + }, + { + "epoch": 0.07, + "grad_norm": 1.3333175606713592, + "learning_rate": 9.951837777364313e-06, + "loss": 0.2416, + "step": 1311 + }, + { + "epoch": 0.07, + "grad_norm": 1.0052374870967802, + "learning_rate": 9.951712963523697e-06, + "loss": 0.1875, + "step": 1312 + }, + { + "epoch": 0.07, + "grad_norm": 1.1334056435933932, + "learning_rate": 9.951587988947631e-06, + "loss": 0.1977, + "step": 1313 + }, + { + "epoch": 0.07, + "grad_norm": 1.5264084554744755, + "learning_rate": 9.95146285364017e-06, + "loss": 0.2628, + "step": 1314 + }, + { + "epoch": 0.07, + "grad_norm": 1.5413319271554495, + "learning_rate": 9.95133755760538e-06, + "loss": 0.2729, + "step": 1315 + }, + { + "epoch": 0.07, + "grad_norm": 1.544504526434848, + "learning_rate": 9.951212100847324e-06, + "loss": 0.2037, + "step": 1316 + }, + { + "epoch": 0.07, + "grad_norm": 1.2609050115195732, + "learning_rate": 9.951086483370075e-06, + "loss": 0.2364, + "step": 1317 + }, + { + "epoch": 0.07, + "grad_norm": 1.5855154124106023, + "learning_rate": 9.950960705177714e-06, + "loss": 0.2317, + "step": 1318 + }, + { + "epoch": 0.07, + "grad_norm": 1.740806768025055, + "learning_rate": 9.95083476627432e-06, + "loss": 0.209, + "step": 1319 + }, + { + "epoch": 0.07, + "grad_norm": 1.540022908611639, + "learning_rate": 9.950708666663984e-06, + "loss": 0.1629, + "step": 1320 + }, + { + "epoch": 0.07, + "grad_norm": 1.75534631873462, + "learning_rate": 9.950582406350797e-06, + "loss": 0.2989, + "step": 1321 + }, + { + "epoch": 0.07, + "grad_norm": 1.9845363464992838, + "learning_rate": 9.950455985338858e-06, + "loss": 0.2006, + "step": 1322 + }, + { + "epoch": 0.07, + "grad_norm": 1.4813110476802542, + "learning_rate": 9.950329403632271e-06, + "loss": 0.242, + "step": 1323 + }, + { + "epoch": 0.07, + "grad_norm": 1.2189362670924673, + "learning_rate": 9.950202661235146e-06, + "loss": 0.1875, + "step": 1324 + }, + { + "epoch": 0.07, + "grad_norm": 1.1770590527176679, + "learning_rate": 9.950075758151595e-06, + "loss": 0.1936, + "step": 1325 + }, + { + "epoch": 0.07, + "grad_norm": 1.249276026044338, + "learning_rate": 9.949948694385739e-06, + "loss": 0.2439, + "step": 1326 + }, + { + "epoch": 0.07, + "grad_norm": 1.153297579135089, + "learning_rate": 9.9498214699417e-06, + "loss": 0.2089, + "step": 1327 + }, + { + "epoch": 0.07, + "grad_norm": 1.4586405061604863, + "learning_rate": 9.949694084823612e-06, + "loss": 0.2333, + "step": 1328 + }, + { + "epoch": 0.07, + "grad_norm": 1.4073584638376784, + "learning_rate": 9.949566539035607e-06, + "loss": 0.2186, + "step": 1329 + }, + { + "epoch": 0.07, + "grad_norm": 1.221203904238583, + "learning_rate": 9.949438832581825e-06, + "loss": 0.1805, + "step": 1330 + }, + { + "epoch": 0.07, + "grad_norm": 1.4531822372772065, + "learning_rate": 9.949310965466414e-06, + "loss": 0.235, + "step": 1331 + }, + { + "epoch": 0.07, + "grad_norm": 1.3207062861966896, + "learning_rate": 9.949182937693522e-06, + "loss": 0.2227, + "step": 1332 + }, + { + "epoch": 0.07, + "grad_norm": 1.6043480212221013, + "learning_rate": 9.949054749267307e-06, + "loss": 0.1974, + "step": 1333 + }, + { + "epoch": 0.07, + "grad_norm": 1.4942762911670193, + "learning_rate": 9.948926400191927e-06, + "loss": 0.23, + "step": 1334 + }, + { + "epoch": 0.07, + "grad_norm": 1.4938888537141437, + "learning_rate": 9.948797890471552e-06, + "loss": 0.1747, + "step": 1335 + }, + { + "epoch": 0.07, + "grad_norm": 1.0753424783689003, + "learning_rate": 9.948669220110351e-06, + "loss": 0.1549, + "step": 1336 + }, + { + "epoch": 0.07, + "grad_norm": 1.6532677936718372, + "learning_rate": 9.948540389112502e-06, + "loss": 0.2379, + "step": 1337 + }, + { + "epoch": 0.07, + "grad_norm": 1.9361034578491436, + "learning_rate": 9.948411397482185e-06, + "loss": 0.2344, + "step": 1338 + }, + { + "epoch": 0.07, + "grad_norm": 1.5441670171140602, + "learning_rate": 9.94828224522359e-06, + "loss": 0.2837, + "step": 1339 + }, + { + "epoch": 0.07, + "grad_norm": 1.7279342190801754, + "learning_rate": 9.948152932340906e-06, + "loss": 0.2622, + "step": 1340 + }, + { + "epoch": 0.07, + "grad_norm": 1.5261193169984717, + "learning_rate": 9.948023458838334e-06, + "loss": 0.2219, + "step": 1341 + }, + { + "epoch": 0.07, + "grad_norm": 1.5614468792534895, + "learning_rate": 9.947893824720076e-06, + "loss": 0.2559, + "step": 1342 + }, + { + "epoch": 0.07, + "grad_norm": 1.184188992103389, + "learning_rate": 9.947764029990336e-06, + "loss": 0.2141, + "step": 1343 + }, + { + "epoch": 0.07, + "grad_norm": 1.0646445146039465, + "learning_rate": 9.947634074653334e-06, + "loss": 0.2024, + "step": 1344 + }, + { + "epoch": 0.07, + "grad_norm": 1.1759162521166031, + "learning_rate": 9.947503958713283e-06, + "loss": 0.1787, + "step": 1345 + }, + { + "epoch": 0.07, + "grad_norm": 1.6675339574319505, + "learning_rate": 9.947373682174408e-06, + "loss": 0.2568, + "step": 1346 + }, + { + "epoch": 0.07, + "grad_norm": 1.3892255341979134, + "learning_rate": 9.94724324504094e-06, + "loss": 0.1884, + "step": 1347 + }, + { + "epoch": 0.07, + "grad_norm": 1.6085484914725872, + "learning_rate": 9.94711264731711e-06, + "loss": 0.2303, + "step": 1348 + }, + { + "epoch": 0.08, + "grad_norm": 1.5877687048470224, + "learning_rate": 9.94698188900716e-06, + "loss": 0.2415, + "step": 1349 + }, + { + "epoch": 0.08, + "grad_norm": 1.5258051105202044, + "learning_rate": 9.946850970115331e-06, + "loss": 0.2425, + "step": 1350 + }, + { + "epoch": 0.08, + "grad_norm": 2.1545654502974383, + "learning_rate": 9.946719890645877e-06, + "loss": 0.2699, + "step": 1351 + }, + { + "epoch": 0.08, + "grad_norm": 1.2825368373629518, + "learning_rate": 9.94658865060305e-06, + "loss": 0.2397, + "step": 1352 + }, + { + "epoch": 0.08, + "grad_norm": 1.2983529878189863, + "learning_rate": 9.946457249991111e-06, + "loss": 0.191, + "step": 1353 + }, + { + "epoch": 0.08, + "grad_norm": 1.2835750737517753, + "learning_rate": 9.946325688814325e-06, + "loss": 0.249, + "step": 1354 + }, + { + "epoch": 0.08, + "grad_norm": 1.2483322932049243, + "learning_rate": 9.946193967076963e-06, + "loss": 0.255, + "step": 1355 + }, + { + "epoch": 0.08, + "grad_norm": 1.0370508606705902, + "learning_rate": 9.946062084783301e-06, + "loss": 0.187, + "step": 1356 + }, + { + "epoch": 0.08, + "grad_norm": 1.2995529059916984, + "learning_rate": 9.945930041937619e-06, + "loss": 0.1895, + "step": 1357 + }, + { + "epoch": 0.08, + "grad_norm": 2.05945379724387, + "learning_rate": 9.945797838544204e-06, + "loss": 0.2547, + "step": 1358 + }, + { + "epoch": 0.08, + "grad_norm": 1.7581240939430562, + "learning_rate": 9.945665474607347e-06, + "loss": 0.24, + "step": 1359 + }, + { + "epoch": 0.08, + "grad_norm": 1.1517184031631802, + "learning_rate": 9.945532950131344e-06, + "loss": 0.2199, + "step": 1360 + }, + { + "epoch": 0.08, + "grad_norm": 1.1363434762006293, + "learning_rate": 9.945400265120497e-06, + "loss": 0.2206, + "step": 1361 + }, + { + "epoch": 0.08, + "grad_norm": 1.897448528404319, + "learning_rate": 9.945267419579115e-06, + "loss": 0.25, + "step": 1362 + }, + { + "epoch": 0.08, + "grad_norm": 1.3102391893740681, + "learning_rate": 9.945134413511509e-06, + "loss": 0.2354, + "step": 1363 + }, + { + "epoch": 0.08, + "grad_norm": 1.5377930280525893, + "learning_rate": 9.945001246921996e-06, + "loss": 0.2257, + "step": 1364 + }, + { + "epoch": 0.08, + "grad_norm": 0.8462730650068777, + "learning_rate": 9.944867919814899e-06, + "loss": 0.1714, + "step": 1365 + }, + { + "epoch": 0.08, + "grad_norm": 1.6075831850302882, + "learning_rate": 9.944734432194544e-06, + "loss": 0.1934, + "step": 1366 + }, + { + "epoch": 0.08, + "grad_norm": 1.1877622957792544, + "learning_rate": 9.94460078406527e-06, + "loss": 0.2346, + "step": 1367 + }, + { + "epoch": 0.08, + "grad_norm": 1.7551123161729707, + "learning_rate": 9.944466975431409e-06, + "loss": 0.2381, + "step": 1368 + }, + { + "epoch": 0.08, + "grad_norm": 1.1077430064379408, + "learning_rate": 9.944333006297305e-06, + "loss": 0.2215, + "step": 1369 + }, + { + "epoch": 0.08, + "grad_norm": 1.0165811567796186, + "learning_rate": 9.94419887666731e-06, + "loss": 0.1865, + "step": 1370 + }, + { + "epoch": 0.08, + "grad_norm": 1.119640726259161, + "learning_rate": 9.944064586545776e-06, + "loss": 0.1799, + "step": 1371 + }, + { + "epoch": 0.08, + "grad_norm": 1.3376888693833824, + "learning_rate": 9.943930135937064e-06, + "loss": 0.2013, + "step": 1372 + }, + { + "epoch": 0.08, + "grad_norm": 1.6808923924186634, + "learning_rate": 9.943795524845534e-06, + "loss": 0.2159, + "step": 1373 + }, + { + "epoch": 0.08, + "grad_norm": 1.544555965548068, + "learning_rate": 9.94366075327556e-06, + "loss": 0.2203, + "step": 1374 + }, + { + "epoch": 0.08, + "grad_norm": 1.283266753966333, + "learning_rate": 9.943525821231516e-06, + "loss": 0.2334, + "step": 1375 + }, + { + "epoch": 0.08, + "grad_norm": 1.470960918695552, + "learning_rate": 9.943390728717779e-06, + "loss": 0.2202, + "step": 1376 + }, + { + "epoch": 0.08, + "grad_norm": 1.2960172597286255, + "learning_rate": 9.943255475738739e-06, + "loss": 0.2388, + "step": 1377 + }, + { + "epoch": 0.08, + "grad_norm": 1.562513137106505, + "learning_rate": 9.943120062298781e-06, + "loss": 0.2179, + "step": 1378 + }, + { + "epoch": 0.08, + "grad_norm": 1.4123173582943298, + "learning_rate": 9.942984488402305e-06, + "loss": 0.187, + "step": 1379 + }, + { + "epoch": 0.08, + "grad_norm": 1.737074117815268, + "learning_rate": 9.942848754053709e-06, + "loss": 0.254, + "step": 1380 + }, + { + "epoch": 0.08, + "grad_norm": 1.2577759251696856, + "learning_rate": 9.942712859257401e-06, + "loss": 0.2633, + "step": 1381 + }, + { + "epoch": 0.08, + "grad_norm": 1.4286484342867316, + "learning_rate": 9.94257680401779e-06, + "loss": 0.1826, + "step": 1382 + }, + { + "epoch": 0.08, + "grad_norm": 1.1153928206571069, + "learning_rate": 9.942440588339294e-06, + "loss": 0.253, + "step": 1383 + }, + { + "epoch": 0.08, + "grad_norm": 1.0991696813445084, + "learning_rate": 9.942304212226337e-06, + "loss": 0.2022, + "step": 1384 + }, + { + "epoch": 0.08, + "grad_norm": 1.3675015116165425, + "learning_rate": 9.942167675683341e-06, + "loss": 0.2538, + "step": 1385 + }, + { + "epoch": 0.08, + "grad_norm": 0.9572844228826728, + "learning_rate": 9.94203097871474e-06, + "loss": 0.1778, + "step": 1386 + }, + { + "epoch": 0.08, + "grad_norm": 1.2714968664607538, + "learning_rate": 9.941894121324973e-06, + "loss": 0.2252, + "step": 1387 + }, + { + "epoch": 0.08, + "grad_norm": 0.8838669352708726, + "learning_rate": 9.94175710351848e-06, + "loss": 0.205, + "step": 1388 + }, + { + "epoch": 0.08, + "grad_norm": 2.122189030440291, + "learning_rate": 9.94161992529971e-06, + "loss": 0.2055, + "step": 1389 + }, + { + "epoch": 0.08, + "grad_norm": 1.547466208446442, + "learning_rate": 9.941482586673114e-06, + "loss": 0.3065, + "step": 1390 + }, + { + "epoch": 0.08, + "grad_norm": 1.4330556727633588, + "learning_rate": 9.941345087643155e-06, + "loss": 0.2036, + "step": 1391 + }, + { + "epoch": 0.08, + "grad_norm": 1.5114220320612026, + "learning_rate": 9.941207428214291e-06, + "loss": 0.2751, + "step": 1392 + }, + { + "epoch": 0.08, + "grad_norm": 1.3187018444152647, + "learning_rate": 9.941069608390992e-06, + "loss": 0.1884, + "step": 1393 + }, + { + "epoch": 0.08, + "grad_norm": 1.173947504355246, + "learning_rate": 9.940931628177733e-06, + "loss": 0.243, + "step": 1394 + }, + { + "epoch": 0.08, + "grad_norm": 1.904863513859335, + "learning_rate": 9.940793487578992e-06, + "loss": 0.2419, + "step": 1395 + }, + { + "epoch": 0.08, + "grad_norm": 1.5625006973741884, + "learning_rate": 9.940655186599252e-06, + "loss": 0.2209, + "step": 1396 + }, + { + "epoch": 0.08, + "grad_norm": 1.1110921401359686, + "learning_rate": 9.940516725243006e-06, + "loss": 0.1618, + "step": 1397 + }, + { + "epoch": 0.08, + "grad_norm": 1.181789250130163, + "learning_rate": 9.940378103514745e-06, + "loss": 0.2193, + "step": 1398 + }, + { + "epoch": 0.08, + "grad_norm": 1.6314330228536515, + "learning_rate": 9.94023932141897e-06, + "loss": 0.2225, + "step": 1399 + }, + { + "epoch": 0.08, + "grad_norm": 1.4851972786996792, + "learning_rate": 9.940100378960184e-06, + "loss": 0.254, + "step": 1400 + }, + { + "epoch": 0.08, + "grad_norm": 1.1518260116012338, + "learning_rate": 9.939961276142901e-06, + "loss": 0.2202, + "step": 1401 + }, + { + "epoch": 0.08, + "grad_norm": 1.3836100057974914, + "learning_rate": 9.939822012971633e-06, + "loss": 0.1871, + "step": 1402 + }, + { + "epoch": 0.08, + "grad_norm": 1.2722214585421814, + "learning_rate": 9.939682589450904e-06, + "loss": 0.242, + "step": 1403 + }, + { + "epoch": 0.08, + "grad_norm": 1.9335578703200498, + "learning_rate": 9.939543005585235e-06, + "loss": 0.2175, + "step": 1404 + }, + { + "epoch": 0.08, + "grad_norm": 1.6538836515934254, + "learning_rate": 9.939403261379162e-06, + "loss": 0.2269, + "step": 1405 + }, + { + "epoch": 0.08, + "grad_norm": 1.7030154212645014, + "learning_rate": 9.939263356837217e-06, + "loss": 0.229, + "step": 1406 + }, + { + "epoch": 0.08, + "grad_norm": 1.7182635572618217, + "learning_rate": 9.939123291963943e-06, + "loss": 0.256, + "step": 1407 + }, + { + "epoch": 0.08, + "grad_norm": 0.9920163476311024, + "learning_rate": 9.938983066763887e-06, + "loss": 0.2265, + "step": 1408 + }, + { + "epoch": 0.08, + "grad_norm": 1.8060054817315856, + "learning_rate": 9.938842681241603e-06, + "loss": 0.2794, + "step": 1409 + }, + { + "epoch": 0.08, + "grad_norm": 0.9743513442282203, + "learning_rate": 9.938702135401643e-06, + "loss": 0.207, + "step": 1410 + }, + { + "epoch": 0.08, + "grad_norm": 1.9105397835964328, + "learning_rate": 9.938561429248573e-06, + "loss": 0.2575, + "step": 1411 + }, + { + "epoch": 0.08, + "grad_norm": 1.6439283538918967, + "learning_rate": 9.938420562786957e-06, + "loss": 0.1678, + "step": 1412 + }, + { + "epoch": 0.08, + "grad_norm": 1.3070187527594355, + "learning_rate": 9.938279536021372e-06, + "loss": 0.1575, + "step": 1413 + }, + { + "epoch": 0.08, + "grad_norm": 1.292882441034289, + "learning_rate": 9.938138348956393e-06, + "loss": 0.2196, + "step": 1414 + }, + { + "epoch": 0.08, + "grad_norm": 1.558598169401782, + "learning_rate": 9.937997001596602e-06, + "loss": 0.2597, + "step": 1415 + }, + { + "epoch": 0.08, + "grad_norm": 1.632995575812348, + "learning_rate": 9.937855493946591e-06, + "loss": 0.2171, + "step": 1416 + }, + { + "epoch": 0.08, + "grad_norm": 2.444298762588584, + "learning_rate": 9.93771382601095e-06, + "loss": 0.2368, + "step": 1417 + }, + { + "epoch": 0.08, + "grad_norm": 1.344527490933099, + "learning_rate": 9.937571997794278e-06, + "loss": 0.2174, + "step": 1418 + }, + { + "epoch": 0.08, + "grad_norm": 1.298435461591275, + "learning_rate": 9.937430009301182e-06, + "loss": 0.2298, + "step": 1419 + }, + { + "epoch": 0.08, + "grad_norm": 1.3214595484508733, + "learning_rate": 9.937287860536267e-06, + "loss": 0.2147, + "step": 1420 + }, + { + "epoch": 0.08, + "grad_norm": 2.8712013626691517, + "learning_rate": 9.937145551504148e-06, + "loss": 0.2133, + "step": 1421 + }, + { + "epoch": 0.08, + "grad_norm": 2.0668684905728023, + "learning_rate": 9.937003082209448e-06, + "loss": 0.2109, + "step": 1422 + }, + { + "epoch": 0.08, + "grad_norm": 1.4536391394650237, + "learning_rate": 9.936860452656787e-06, + "loss": 0.1841, + "step": 1423 + }, + { + "epoch": 0.08, + "grad_norm": 1.018939687030773, + "learning_rate": 9.936717662850797e-06, + "loss": 0.1581, + "step": 1424 + }, + { + "epoch": 0.08, + "grad_norm": 1.4600950626015665, + "learning_rate": 9.936574712796114e-06, + "loss": 0.2144, + "step": 1425 + }, + { + "epoch": 0.08, + "grad_norm": 1.2381614302176873, + "learning_rate": 9.936431602497376e-06, + "loss": 0.1782, + "step": 1426 + }, + { + "epoch": 0.08, + "grad_norm": 1.2979502385661914, + "learning_rate": 9.93628833195923e-06, + "loss": 0.1919, + "step": 1427 + }, + { + "epoch": 0.08, + "grad_norm": 1.2507821884247061, + "learning_rate": 9.936144901186327e-06, + "loss": 0.2077, + "step": 1428 + }, + { + "epoch": 0.08, + "grad_norm": 1.1520703330809725, + "learning_rate": 9.93600131018332e-06, + "loss": 0.2051, + "step": 1429 + }, + { + "epoch": 0.08, + "grad_norm": 1.422919900754583, + "learning_rate": 9.935857558954873e-06, + "loss": 0.2243, + "step": 1430 + }, + { + "epoch": 0.08, + "grad_norm": 1.2917285889605794, + "learning_rate": 9.935713647505652e-06, + "loss": 0.2649, + "step": 1431 + }, + { + "epoch": 0.08, + "grad_norm": 1.176144033841741, + "learning_rate": 9.935569575840327e-06, + "loss": 0.2844, + "step": 1432 + }, + { + "epoch": 0.08, + "grad_norm": 1.1105960284937233, + "learning_rate": 9.935425343963574e-06, + "loss": 0.2093, + "step": 1433 + }, + { + "epoch": 0.08, + "grad_norm": 1.580679968665866, + "learning_rate": 9.935280951880078e-06, + "loss": 0.1966, + "step": 1434 + }, + { + "epoch": 0.08, + "grad_norm": 1.2566399514098965, + "learning_rate": 9.935136399594524e-06, + "loss": 0.2419, + "step": 1435 + }, + { + "epoch": 0.08, + "grad_norm": 1.438045349024953, + "learning_rate": 9.934991687111607e-06, + "loss": 0.1835, + "step": 1436 + }, + { + "epoch": 0.08, + "grad_norm": 1.3021264917055766, + "learning_rate": 9.934846814436019e-06, + "loss": 0.2191, + "step": 1437 + }, + { + "epoch": 0.08, + "grad_norm": 1.7284553905272237, + "learning_rate": 9.934701781572465e-06, + "loss": 0.2478, + "step": 1438 + }, + { + "epoch": 0.08, + "grad_norm": 1.6458279646311087, + "learning_rate": 9.934556588525655e-06, + "loss": 0.2499, + "step": 1439 + }, + { + "epoch": 0.08, + "grad_norm": 1.0028507532186248, + "learning_rate": 9.9344112353003e-06, + "loss": 0.176, + "step": 1440 + }, + { + "epoch": 0.08, + "grad_norm": 1.6202680622139412, + "learning_rate": 9.93426572190112e-06, + "loss": 0.2434, + "step": 1441 + }, + { + "epoch": 0.08, + "grad_norm": 1.1215189831774026, + "learning_rate": 9.934120048332837e-06, + "loss": 0.2428, + "step": 1442 + }, + { + "epoch": 0.08, + "grad_norm": 1.1019277922319155, + "learning_rate": 9.933974214600179e-06, + "loss": 0.2095, + "step": 1443 + }, + { + "epoch": 0.08, + "grad_norm": 0.7599852337724233, + "learning_rate": 9.933828220707881e-06, + "loss": 0.1743, + "step": 1444 + }, + { + "epoch": 0.08, + "grad_norm": 1.2191052725199893, + "learning_rate": 9.933682066660683e-06, + "loss": 0.2034, + "step": 1445 + }, + { + "epoch": 0.08, + "grad_norm": 1.3180099045309601, + "learning_rate": 9.933535752463326e-06, + "loss": 0.2392, + "step": 1446 + }, + { + "epoch": 0.08, + "grad_norm": 1.2577090665096775, + "learning_rate": 9.933389278120563e-06, + "loss": 0.2278, + "step": 1447 + }, + { + "epoch": 0.08, + "grad_norm": 1.283980521650947, + "learning_rate": 9.933242643637146e-06, + "loss": 0.2444, + "step": 1448 + }, + { + "epoch": 0.08, + "grad_norm": 1.1273649279798719, + "learning_rate": 9.933095849017838e-06, + "loss": 0.2285, + "step": 1449 + }, + { + "epoch": 0.08, + "grad_norm": 1.3534213166732552, + "learning_rate": 9.9329488942674e-06, + "loss": 0.1995, + "step": 1450 + }, + { + "epoch": 0.08, + "grad_norm": 1.2907321347436045, + "learning_rate": 9.932801779390606e-06, + "loss": 0.2208, + "step": 1451 + }, + { + "epoch": 0.08, + "grad_norm": 1.2839722498456718, + "learning_rate": 9.932654504392228e-06, + "loss": 0.2427, + "step": 1452 + }, + { + "epoch": 0.08, + "grad_norm": 0.904910643470907, + "learning_rate": 9.932507069277049e-06, + "loss": 0.1794, + "step": 1453 + }, + { + "epoch": 0.08, + "grad_norm": 1.5669195993997176, + "learning_rate": 9.932359474049852e-06, + "loss": 0.2149, + "step": 1454 + }, + { + "epoch": 0.08, + "grad_norm": 1.3359998192110696, + "learning_rate": 9.932211718715434e-06, + "loss": 0.2046, + "step": 1455 + }, + { + "epoch": 0.08, + "grad_norm": 1.6120266897091369, + "learning_rate": 9.932063803278586e-06, + "loss": 0.2053, + "step": 1456 + }, + { + "epoch": 0.08, + "grad_norm": 0.8376242452063587, + "learning_rate": 9.93191572774411e-06, + "loss": 0.1456, + "step": 1457 + }, + { + "epoch": 0.08, + "grad_norm": 1.7011007300652667, + "learning_rate": 9.931767492116813e-06, + "loss": 0.2235, + "step": 1458 + }, + { + "epoch": 0.08, + "grad_norm": 1.7554097135537259, + "learning_rate": 9.93161909640151e-06, + "loss": 0.2363, + "step": 1459 + }, + { + "epoch": 0.08, + "grad_norm": 1.3508268170478546, + "learning_rate": 9.931470540603013e-06, + "loss": 0.2303, + "step": 1460 + }, + { + "epoch": 0.08, + "grad_norm": 1.2873324720405162, + "learning_rate": 9.931321824726145e-06, + "loss": 0.239, + "step": 1461 + }, + { + "epoch": 0.08, + "grad_norm": 1.3579153494956782, + "learning_rate": 9.931172948775739e-06, + "loss": 0.2097, + "step": 1462 + }, + { + "epoch": 0.08, + "grad_norm": 1.3462217353113473, + "learning_rate": 9.931023912756619e-06, + "loss": 0.2409, + "step": 1463 + }, + { + "epoch": 0.08, + "grad_norm": 1.2075256620529897, + "learning_rate": 9.93087471667363e-06, + "loss": 0.1818, + "step": 1464 + }, + { + "epoch": 0.08, + "grad_norm": 1.257602092073513, + "learning_rate": 9.930725360531612e-06, + "loss": 0.2297, + "step": 1465 + }, + { + "epoch": 0.08, + "grad_norm": 1.3914680283040164, + "learning_rate": 9.930575844335413e-06, + "loss": 0.1992, + "step": 1466 + }, + { + "epoch": 0.08, + "grad_norm": 1.311119701991059, + "learning_rate": 9.930426168089885e-06, + "loss": 0.2293, + "step": 1467 + }, + { + "epoch": 0.08, + "grad_norm": 1.8664334582217428, + "learning_rate": 9.93027633179989e-06, + "loss": 0.2491, + "step": 1468 + }, + { + "epoch": 0.08, + "grad_norm": 1.9932090290684732, + "learning_rate": 9.930126335470292e-06, + "loss": 0.2608, + "step": 1469 + }, + { + "epoch": 0.08, + "grad_norm": 1.5488989502346024, + "learning_rate": 9.929976179105956e-06, + "loss": 0.2295, + "step": 1470 + }, + { + "epoch": 0.08, + "grad_norm": 1.285774105625606, + "learning_rate": 9.929825862711759e-06, + "loss": 0.1918, + "step": 1471 + }, + { + "epoch": 0.08, + "grad_norm": 1.207334389320146, + "learning_rate": 9.929675386292578e-06, + "loss": 0.1495, + "step": 1472 + }, + { + "epoch": 0.08, + "grad_norm": 1.2493310554007533, + "learning_rate": 9.929524749853302e-06, + "loss": 0.2067, + "step": 1473 + }, + { + "epoch": 0.08, + "grad_norm": 1.6259435266569653, + "learning_rate": 9.929373953398816e-06, + "loss": 0.2093, + "step": 1474 + }, + { + "epoch": 0.08, + "grad_norm": 1.1367731507382957, + "learning_rate": 9.929222996934018e-06, + "loss": 0.2437, + "step": 1475 + }, + { + "epoch": 0.08, + "grad_norm": 1.5065330032747999, + "learning_rate": 9.929071880463806e-06, + "loss": 0.2575, + "step": 1476 + }, + { + "epoch": 0.08, + "grad_norm": 1.655571754828104, + "learning_rate": 9.928920603993086e-06, + "loss": 0.249, + "step": 1477 + }, + { + "epoch": 0.08, + "grad_norm": 1.8600985497662255, + "learning_rate": 9.92876916752677e-06, + "loss": 0.2634, + "step": 1478 + }, + { + "epoch": 0.08, + "grad_norm": 2.040487598945859, + "learning_rate": 9.928617571069772e-06, + "loss": 0.2796, + "step": 1479 + }, + { + "epoch": 0.08, + "grad_norm": 1.563622510151303, + "learning_rate": 9.928465814627016e-06, + "loss": 0.2752, + "step": 1480 + }, + { + "epoch": 0.08, + "grad_norm": 1.2630579494810283, + "learning_rate": 9.92831389820342e-06, + "loss": 0.2089, + "step": 1481 + }, + { + "epoch": 0.08, + "grad_norm": 1.358966215099504, + "learning_rate": 9.928161821803926e-06, + "loss": 0.2064, + "step": 1482 + }, + { + "epoch": 0.08, + "grad_norm": 1.786686854626152, + "learning_rate": 9.928009585433464e-06, + "loss": 0.2559, + "step": 1483 + }, + { + "epoch": 0.08, + "grad_norm": 1.4558620901161794, + "learning_rate": 9.927857189096976e-06, + "loss": 0.2623, + "step": 1484 + }, + { + "epoch": 0.08, + "grad_norm": 1.927774937134356, + "learning_rate": 9.927704632799412e-06, + "loss": 0.1872, + "step": 1485 + }, + { + "epoch": 0.08, + "grad_norm": 1.247890064077216, + "learning_rate": 9.92755191654572e-06, + "loss": 0.2275, + "step": 1486 + }, + { + "epoch": 0.08, + "grad_norm": 1.5040339155629465, + "learning_rate": 9.927399040340859e-06, + "loss": 0.2609, + "step": 1487 + }, + { + "epoch": 0.08, + "grad_norm": 1.3982488568899964, + "learning_rate": 9.927246004189792e-06, + "loss": 0.1606, + "step": 1488 + }, + { + "epoch": 0.08, + "grad_norm": 1.3871958274617135, + "learning_rate": 9.927092808097487e-06, + "loss": 0.2078, + "step": 1489 + }, + { + "epoch": 0.08, + "grad_norm": 2.0401428218182565, + "learning_rate": 9.926939452068916e-06, + "loss": 0.2636, + "step": 1490 + }, + { + "epoch": 0.08, + "grad_norm": 1.3224710251413991, + "learning_rate": 9.926785936109057e-06, + "loss": 0.1908, + "step": 1491 + }, + { + "epoch": 0.08, + "grad_norm": 1.3919545311596058, + "learning_rate": 9.926632260222894e-06, + "loss": 0.2049, + "step": 1492 + }, + { + "epoch": 0.08, + "grad_norm": 3.126685736885422, + "learning_rate": 9.926478424415413e-06, + "loss": 0.2296, + "step": 1493 + }, + { + "epoch": 0.08, + "grad_norm": 0.9963383131299739, + "learning_rate": 9.926324428691612e-06, + "loss": 0.1904, + "step": 1494 + }, + { + "epoch": 0.08, + "grad_norm": 1.4432020179031033, + "learning_rate": 9.926170273056486e-06, + "loss": 0.2054, + "step": 1495 + }, + { + "epoch": 0.08, + "grad_norm": 1.4810238806443188, + "learning_rate": 9.92601595751504e-06, + "loss": 0.254, + "step": 1496 + }, + { + "epoch": 0.08, + "grad_norm": 1.6349128830008455, + "learning_rate": 9.925861482072282e-06, + "loss": 0.2422, + "step": 1497 + }, + { + "epoch": 0.08, + "grad_norm": 1.553909857272511, + "learning_rate": 9.92570684673323e-06, + "loss": 0.2639, + "step": 1498 + }, + { + "epoch": 0.08, + "grad_norm": 1.472225107124814, + "learning_rate": 9.9255520515029e-06, + "loss": 0.182, + "step": 1499 + }, + { + "epoch": 0.08, + "grad_norm": 1.5495215820202455, + "learning_rate": 9.92539709638632e-06, + "loss": 0.2548, + "step": 1500 + }, + { + "epoch": 0.08, + "grad_norm": 1.246681333442473, + "learning_rate": 9.925241981388515e-06, + "loss": 0.185, + "step": 1501 + }, + { + "epoch": 0.08, + "grad_norm": 1.3056603619245588, + "learning_rate": 9.925086706514526e-06, + "loss": 0.2229, + "step": 1502 + }, + { + "epoch": 0.08, + "grad_norm": 1.3398401855332365, + "learning_rate": 9.92493127176939e-06, + "loss": 0.2582, + "step": 1503 + }, + { + "epoch": 0.08, + "grad_norm": 1.2401830332183985, + "learning_rate": 9.924775677158149e-06, + "loss": 0.2117, + "step": 1504 + }, + { + "epoch": 0.08, + "grad_norm": 1.3605675796238905, + "learning_rate": 9.924619922685861e-06, + "loss": 0.2413, + "step": 1505 + }, + { + "epoch": 0.08, + "grad_norm": 1.637726271080243, + "learning_rate": 9.924464008357579e-06, + "loss": 0.2516, + "step": 1506 + }, + { + "epoch": 0.08, + "grad_norm": 1.7343361421691201, + "learning_rate": 9.924307934178362e-06, + "loss": 0.2327, + "step": 1507 + }, + { + "epoch": 0.08, + "grad_norm": 1.6343117543049266, + "learning_rate": 9.924151700153277e-06, + "loss": 0.2638, + "step": 1508 + }, + { + "epoch": 0.08, + "grad_norm": 1.5396637407815261, + "learning_rate": 9.923995306287396e-06, + "loss": 0.1956, + "step": 1509 + }, + { + "epoch": 0.08, + "grad_norm": 0.9025192698983319, + "learning_rate": 9.9238387525858e-06, + "loss": 0.1819, + "step": 1510 + }, + { + "epoch": 0.08, + "grad_norm": 1.3165198639309392, + "learning_rate": 9.923682039053564e-06, + "loss": 0.2115, + "step": 1511 + }, + { + "epoch": 0.08, + "grad_norm": 1.5177354571507038, + "learning_rate": 9.923525165695776e-06, + "loss": 0.2427, + "step": 1512 + }, + { + "epoch": 0.08, + "grad_norm": 1.1548434317403464, + "learning_rate": 9.923368132517532e-06, + "loss": 0.2277, + "step": 1513 + }, + { + "epoch": 0.08, + "grad_norm": 1.8774870749128998, + "learning_rate": 9.923210939523926e-06, + "loss": 0.2333, + "step": 1514 + }, + { + "epoch": 0.08, + "grad_norm": 1.7937855761020567, + "learning_rate": 9.923053586720064e-06, + "loss": 0.2544, + "step": 1515 + }, + { + "epoch": 0.08, + "grad_norm": 1.3809664386644378, + "learning_rate": 9.92289607411105e-06, + "loss": 0.2286, + "step": 1516 + }, + { + "epoch": 0.08, + "grad_norm": 1.3610573154453982, + "learning_rate": 9.922738401701999e-06, + "loss": 0.2026, + "step": 1517 + }, + { + "epoch": 0.08, + "grad_norm": 2.011279189595608, + "learning_rate": 9.922580569498029e-06, + "loss": 0.2515, + "step": 1518 + }, + { + "epoch": 0.08, + "grad_norm": 1.420027877125141, + "learning_rate": 9.922422577504262e-06, + "loss": 0.2628, + "step": 1519 + }, + { + "epoch": 0.08, + "grad_norm": 1.372851047822403, + "learning_rate": 9.922264425725827e-06, + "loss": 0.2544, + "step": 1520 + }, + { + "epoch": 0.08, + "grad_norm": 1.2326609610840948, + "learning_rate": 9.92210611416786e-06, + "loss": 0.1719, + "step": 1521 + }, + { + "epoch": 0.08, + "grad_norm": 1.4673433720723565, + "learning_rate": 9.921947642835497e-06, + "loss": 0.2156, + "step": 1522 + }, + { + "epoch": 0.08, + "grad_norm": 0.9366570358531431, + "learning_rate": 9.921789011733885e-06, + "loss": 0.1512, + "step": 1523 + }, + { + "epoch": 0.08, + "grad_norm": 1.217399868722497, + "learning_rate": 9.921630220868168e-06, + "loss": 0.2243, + "step": 1524 + }, + { + "epoch": 0.08, + "grad_norm": 1.1912031156874492, + "learning_rate": 9.921471270243508e-06, + "loss": 0.1855, + "step": 1525 + }, + { + "epoch": 0.08, + "grad_norm": 0.957343026388638, + "learning_rate": 9.921312159865057e-06, + "loss": 0.1477, + "step": 1526 + }, + { + "epoch": 0.08, + "grad_norm": 1.6468762192255644, + "learning_rate": 9.921152889737985e-06, + "loss": 0.2339, + "step": 1527 + }, + { + "epoch": 0.08, + "grad_norm": 1.1824135530202262, + "learning_rate": 9.920993459867461e-06, + "loss": 0.2046, + "step": 1528 + }, + { + "epoch": 0.09, + "grad_norm": 1.5344947913853038, + "learning_rate": 9.920833870258657e-06, + "loss": 0.2545, + "step": 1529 + }, + { + "epoch": 0.09, + "grad_norm": 1.7349035643290134, + "learning_rate": 9.920674120916757e-06, + "loss": 0.3199, + "step": 1530 + }, + { + "epoch": 0.09, + "grad_norm": 0.9240071917540796, + "learning_rate": 9.920514211846948e-06, + "loss": 0.1617, + "step": 1531 + }, + { + "epoch": 0.09, + "grad_norm": 1.8677659635026298, + "learning_rate": 9.920354143054415e-06, + "loss": 0.2548, + "step": 1532 + }, + { + "epoch": 0.09, + "grad_norm": 1.3452984100250807, + "learning_rate": 9.920193914544358e-06, + "loss": 0.2031, + "step": 1533 + }, + { + "epoch": 0.09, + "grad_norm": 1.405232326179258, + "learning_rate": 9.920033526321979e-06, + "loss": 0.2022, + "step": 1534 + }, + { + "epoch": 0.09, + "grad_norm": 1.2810562382703359, + "learning_rate": 9.91987297839248e-06, + "loss": 0.2581, + "step": 1535 + }, + { + "epoch": 0.09, + "grad_norm": 1.715513441839788, + "learning_rate": 9.919712270761077e-06, + "loss": 0.2456, + "step": 1536 + }, + { + "epoch": 0.09, + "grad_norm": 1.3720471906240994, + "learning_rate": 9.919551403432984e-06, + "loss": 0.2732, + "step": 1537 + }, + { + "epoch": 0.09, + "grad_norm": 1.3479825495558215, + "learning_rate": 9.919390376413423e-06, + "loss": 0.2247, + "step": 1538 + }, + { + "epoch": 0.09, + "grad_norm": 1.3472761723920377, + "learning_rate": 9.91922918970762e-06, + "loss": 0.222, + "step": 1539 + }, + { + "epoch": 0.09, + "grad_norm": 1.3891760867098744, + "learning_rate": 9.919067843320811e-06, + "loss": 0.2444, + "step": 1540 + }, + { + "epoch": 0.09, + "grad_norm": 1.2882551323606894, + "learning_rate": 9.91890633725823e-06, + "loss": 0.2541, + "step": 1541 + }, + { + "epoch": 0.09, + "grad_norm": 1.6948195391177954, + "learning_rate": 9.918744671525123e-06, + "loss": 0.2449, + "step": 1542 + }, + { + "epoch": 0.09, + "grad_norm": 1.4570364568479226, + "learning_rate": 9.918582846126731e-06, + "loss": 0.2763, + "step": 1543 + }, + { + "epoch": 0.09, + "grad_norm": 1.2650961654093873, + "learning_rate": 9.918420861068317e-06, + "loss": 0.1437, + "step": 1544 + }, + { + "epoch": 0.09, + "grad_norm": 1.6880087394677383, + "learning_rate": 9.91825871635513e-06, + "loss": 0.2696, + "step": 1545 + }, + { + "epoch": 0.09, + "grad_norm": 1.4752327012109798, + "learning_rate": 9.918096411992438e-06, + "loss": 0.1895, + "step": 1546 + }, + { + "epoch": 0.09, + "grad_norm": 1.0949051308959203, + "learning_rate": 9.917933947985508e-06, + "loss": 0.225, + "step": 1547 + }, + { + "epoch": 0.09, + "grad_norm": 1.1988786718427447, + "learning_rate": 9.917771324339613e-06, + "loss": 0.2069, + "step": 1548 + }, + { + "epoch": 0.09, + "grad_norm": 1.6885420729734515, + "learning_rate": 9.917608541060034e-06, + "loss": 0.2096, + "step": 1549 + }, + { + "epoch": 0.09, + "grad_norm": 1.427295178911265, + "learning_rate": 9.917445598152054e-06, + "loss": 0.2302, + "step": 1550 + }, + { + "epoch": 0.09, + "grad_norm": 1.1164077775462906, + "learning_rate": 9.917282495620961e-06, + "loss": 0.1617, + "step": 1551 + }, + { + "epoch": 0.09, + "grad_norm": 1.590144201658293, + "learning_rate": 9.917119233472052e-06, + "loss": 0.2591, + "step": 1552 + }, + { + "epoch": 0.09, + "grad_norm": 1.2893890259910648, + "learning_rate": 9.916955811710625e-06, + "loss": 0.2314, + "step": 1553 + }, + { + "epoch": 0.09, + "grad_norm": 1.1762617213083757, + "learning_rate": 9.916792230341985e-06, + "loss": 0.2372, + "step": 1554 + }, + { + "epoch": 0.09, + "grad_norm": 1.1382599113207952, + "learning_rate": 9.91662848937144e-06, + "loss": 0.1926, + "step": 1555 + }, + { + "epoch": 0.09, + "grad_norm": 1.1966329306135752, + "learning_rate": 9.916464588804308e-06, + "loss": 0.2646, + "step": 1556 + }, + { + "epoch": 0.09, + "grad_norm": 1.8538996958787552, + "learning_rate": 9.91630052864591e-06, + "loss": 0.2801, + "step": 1557 + }, + { + "epoch": 0.09, + "grad_norm": 1.715895620179582, + "learning_rate": 9.916136308901569e-06, + "loss": 0.2105, + "step": 1558 + }, + { + "epoch": 0.09, + "grad_norm": 1.1777713589531162, + "learning_rate": 9.915971929576615e-06, + "loss": 0.1825, + "step": 1559 + }, + { + "epoch": 0.09, + "grad_norm": 1.528773954663678, + "learning_rate": 9.915807390676386e-06, + "loss": 0.1919, + "step": 1560 + }, + { + "epoch": 0.09, + "grad_norm": 1.0058661229175772, + "learning_rate": 9.915642692206223e-06, + "loss": 0.1826, + "step": 1561 + }, + { + "epoch": 0.09, + "grad_norm": 1.0919220704172021, + "learning_rate": 9.91547783417147e-06, + "loss": 0.2043, + "step": 1562 + }, + { + "epoch": 0.09, + "grad_norm": 1.2250835224794383, + "learning_rate": 9.915312816577482e-06, + "loss": 0.2394, + "step": 1563 + }, + { + "epoch": 0.09, + "grad_norm": 1.4504462657563646, + "learning_rate": 9.915147639429612e-06, + "loss": 0.2492, + "step": 1564 + }, + { + "epoch": 0.09, + "grad_norm": 1.4227609860842831, + "learning_rate": 9.914982302733224e-06, + "loss": 0.2353, + "step": 1565 + }, + { + "epoch": 0.09, + "grad_norm": 1.1837313465045478, + "learning_rate": 9.914816806493684e-06, + "loss": 0.2477, + "step": 1566 + }, + { + "epoch": 0.09, + "grad_norm": 1.5693781929522945, + "learning_rate": 9.914651150716363e-06, + "loss": 0.1849, + "step": 1567 + }, + { + "epoch": 0.09, + "grad_norm": 2.1608239947427523, + "learning_rate": 9.914485335406641e-06, + "loss": 0.2458, + "step": 1568 + }, + { + "epoch": 0.09, + "grad_norm": 1.0808538682836635, + "learning_rate": 9.914319360569897e-06, + "loss": 0.1992, + "step": 1569 + }, + { + "epoch": 0.09, + "grad_norm": 1.0453176372840898, + "learning_rate": 9.914153226211523e-06, + "loss": 0.2071, + "step": 1570 + }, + { + "epoch": 0.09, + "grad_norm": 1.5249636732722944, + "learning_rate": 9.913986932336909e-06, + "loss": 0.1976, + "step": 1571 + }, + { + "epoch": 0.09, + "grad_norm": 1.4217518029970395, + "learning_rate": 9.913820478951452e-06, + "loss": 0.2456, + "step": 1572 + }, + { + "epoch": 0.09, + "grad_norm": 1.8945467368210644, + "learning_rate": 9.913653866060558e-06, + "loss": 0.254, + "step": 1573 + }, + { + "epoch": 0.09, + "grad_norm": 1.5048231203393894, + "learning_rate": 9.913487093669633e-06, + "loss": 0.2398, + "step": 1574 + }, + { + "epoch": 0.09, + "grad_norm": 1.2836737763945232, + "learning_rate": 9.913320161784093e-06, + "loss": 0.1973, + "step": 1575 + }, + { + "epoch": 0.09, + "grad_norm": 1.0769966239847508, + "learning_rate": 9.913153070409355e-06, + "loss": 0.1876, + "step": 1576 + }, + { + "epoch": 0.09, + "grad_norm": 1.1921483175593324, + "learning_rate": 9.91298581955084e-06, + "loss": 0.2168, + "step": 1577 + }, + { + "epoch": 0.09, + "grad_norm": 1.3395543800812357, + "learning_rate": 9.912818409213982e-06, + "loss": 0.2104, + "step": 1578 + }, + { + "epoch": 0.09, + "grad_norm": 1.471638649822063, + "learning_rate": 9.912650839404213e-06, + "loss": 0.2377, + "step": 1579 + }, + { + "epoch": 0.09, + "grad_norm": 1.3044254131107167, + "learning_rate": 9.912483110126976e-06, + "loss": 0.1476, + "step": 1580 + }, + { + "epoch": 0.09, + "grad_norm": 1.427648573044106, + "learning_rate": 9.91231522138771e-06, + "loss": 0.2633, + "step": 1581 + }, + { + "epoch": 0.09, + "grad_norm": 1.3047241345704457, + "learning_rate": 9.912147173191866e-06, + "loss": 0.1876, + "step": 1582 + }, + { + "epoch": 0.09, + "grad_norm": 1.3327436304131612, + "learning_rate": 9.911978965544902e-06, + "loss": 0.1841, + "step": 1583 + }, + { + "epoch": 0.09, + "grad_norm": 1.4994599045521857, + "learning_rate": 9.911810598452273e-06, + "loss": 0.198, + "step": 1584 + }, + { + "epoch": 0.09, + "grad_norm": 1.2977323742883762, + "learning_rate": 9.911642071919452e-06, + "loss": 0.2114, + "step": 1585 + }, + { + "epoch": 0.09, + "grad_norm": 1.6543989245747812, + "learning_rate": 9.911473385951902e-06, + "loss": 0.2011, + "step": 1586 + }, + { + "epoch": 0.09, + "grad_norm": 1.261435785539633, + "learning_rate": 9.911304540555102e-06, + "loss": 0.2099, + "step": 1587 + }, + { + "epoch": 0.09, + "grad_norm": 1.7162252609403654, + "learning_rate": 9.911135535734532e-06, + "loss": 0.2519, + "step": 1588 + }, + { + "epoch": 0.09, + "grad_norm": 0.9884412978928792, + "learning_rate": 9.91096637149568e-06, + "loss": 0.1788, + "step": 1589 + }, + { + "epoch": 0.09, + "grad_norm": 1.5732535581456069, + "learning_rate": 9.910797047844034e-06, + "loss": 0.2444, + "step": 1590 + }, + { + "epoch": 0.09, + "grad_norm": 1.5866171879009165, + "learning_rate": 9.910627564785093e-06, + "loss": 0.2478, + "step": 1591 + }, + { + "epoch": 0.09, + "grad_norm": 1.190539027223919, + "learning_rate": 9.910457922324358e-06, + "loss": 0.226, + "step": 1592 + }, + { + "epoch": 0.09, + "grad_norm": 1.457318244466312, + "learning_rate": 9.910288120467333e-06, + "loss": 0.2694, + "step": 1593 + }, + { + "epoch": 0.09, + "grad_norm": 1.5193782572474022, + "learning_rate": 9.910118159219532e-06, + "loss": 0.2966, + "step": 1594 + }, + { + "epoch": 0.09, + "grad_norm": 1.0236178632320416, + "learning_rate": 9.909948038586475e-06, + "loss": 0.1762, + "step": 1595 + }, + { + "epoch": 0.09, + "grad_norm": 1.0591613560433264, + "learning_rate": 9.909777758573678e-06, + "loss": 0.1945, + "step": 1596 + }, + { + "epoch": 0.09, + "grad_norm": 1.2645060116605937, + "learning_rate": 9.909607319186673e-06, + "loss": 0.2217, + "step": 1597 + }, + { + "epoch": 0.09, + "grad_norm": 1.3696109410159985, + "learning_rate": 9.909436720430992e-06, + "loss": 0.2172, + "step": 1598 + }, + { + "epoch": 0.09, + "grad_norm": 1.1822294911730062, + "learning_rate": 9.90926596231217e-06, + "loss": 0.2295, + "step": 1599 + }, + { + "epoch": 0.09, + "grad_norm": 1.2366571937611224, + "learning_rate": 9.909095044835755e-06, + "loss": 0.2453, + "step": 1600 + }, + { + "epoch": 0.09, + "grad_norm": 1.4107092007040734, + "learning_rate": 9.90892396800729e-06, + "loss": 0.2044, + "step": 1601 + }, + { + "epoch": 0.09, + "grad_norm": 1.2197687341807661, + "learning_rate": 9.908752731832329e-06, + "loss": 0.1984, + "step": 1602 + }, + { + "epoch": 0.09, + "grad_norm": 1.2063565698796312, + "learning_rate": 9.908581336316432e-06, + "loss": 0.19, + "step": 1603 + }, + { + "epoch": 0.09, + "grad_norm": 1.139865895357509, + "learning_rate": 9.908409781465163e-06, + "loss": 0.235, + "step": 1604 + }, + { + "epoch": 0.09, + "grad_norm": 1.7908886039727343, + "learning_rate": 9.908238067284091e-06, + "loss": 0.2004, + "step": 1605 + }, + { + "epoch": 0.09, + "grad_norm": 1.3019005431777204, + "learning_rate": 9.908066193778786e-06, + "loss": 0.215, + "step": 1606 + }, + { + "epoch": 0.09, + "grad_norm": 1.6281935564368304, + "learning_rate": 9.907894160954833e-06, + "loss": 0.2347, + "step": 1607 + }, + { + "epoch": 0.09, + "grad_norm": 1.4064567202049925, + "learning_rate": 9.907721968817812e-06, + "loss": 0.2747, + "step": 1608 + }, + { + "epoch": 0.09, + "grad_norm": 1.502611155028155, + "learning_rate": 9.907549617373313e-06, + "loss": 0.2013, + "step": 1609 + }, + { + "epoch": 0.09, + "grad_norm": 1.4854692616877623, + "learning_rate": 9.907377106626931e-06, + "loss": 0.202, + "step": 1610 + }, + { + "epoch": 0.09, + "grad_norm": 1.4386679579644968, + "learning_rate": 9.907204436584269e-06, + "loss": 0.289, + "step": 1611 + }, + { + "epoch": 0.09, + "grad_norm": 1.5741035606230924, + "learning_rate": 9.907031607250927e-06, + "loss": 0.2313, + "step": 1612 + }, + { + "epoch": 0.09, + "grad_norm": 1.3988851624561889, + "learning_rate": 9.906858618632517e-06, + "loss": 0.1904, + "step": 1613 + }, + { + "epoch": 0.09, + "grad_norm": 1.143613359361613, + "learning_rate": 9.906685470734656e-06, + "loss": 0.2136, + "step": 1614 + }, + { + "epoch": 0.09, + "grad_norm": 1.4740058737648951, + "learning_rate": 9.906512163562962e-06, + "loss": 0.2315, + "step": 1615 + }, + { + "epoch": 0.09, + "grad_norm": 0.9284956368624752, + "learning_rate": 9.906338697123063e-06, + "loss": 0.183, + "step": 1616 + }, + { + "epoch": 0.09, + "grad_norm": 1.2263777010497836, + "learning_rate": 9.906165071420588e-06, + "loss": 0.1757, + "step": 1617 + }, + { + "epoch": 0.09, + "grad_norm": 1.1336272976102564, + "learning_rate": 9.905991286461173e-06, + "loss": 0.1846, + "step": 1618 + }, + { + "epoch": 0.09, + "grad_norm": 1.4775165119958964, + "learning_rate": 9.90581734225046e-06, + "loss": 0.2429, + "step": 1619 + }, + { + "epoch": 0.09, + "grad_norm": 1.5337520430709828, + "learning_rate": 9.905643238794093e-06, + "loss": 0.2337, + "step": 1620 + }, + { + "epoch": 0.09, + "grad_norm": 1.1734260213115257, + "learning_rate": 9.90546897609773e-06, + "loss": 0.1756, + "step": 1621 + }, + { + "epoch": 0.09, + "grad_norm": 1.5767993121318982, + "learning_rate": 9.90529455416702e-06, + "loss": 0.2129, + "step": 1622 + }, + { + "epoch": 0.09, + "grad_norm": 1.3378911791055466, + "learning_rate": 9.905119973007629e-06, + "loss": 0.2106, + "step": 1623 + }, + { + "epoch": 0.09, + "grad_norm": 0.8840323202976306, + "learning_rate": 9.904945232625223e-06, + "loss": 0.125, + "step": 1624 + }, + { + "epoch": 0.09, + "grad_norm": 1.2393843133088807, + "learning_rate": 9.904770333025472e-06, + "loss": 0.2145, + "step": 1625 + }, + { + "epoch": 0.09, + "grad_norm": 1.2816503817981297, + "learning_rate": 9.904595274214058e-06, + "loss": 0.2465, + "step": 1626 + }, + { + "epoch": 0.09, + "grad_norm": 1.3770327876428807, + "learning_rate": 9.904420056196661e-06, + "loss": 0.1614, + "step": 1627 + }, + { + "epoch": 0.09, + "grad_norm": 1.4846644721616993, + "learning_rate": 9.904244678978969e-06, + "loss": 0.2168, + "step": 1628 + }, + { + "epoch": 0.09, + "grad_norm": 1.580759208169585, + "learning_rate": 9.904069142566673e-06, + "loss": 0.1792, + "step": 1629 + }, + { + "epoch": 0.09, + "grad_norm": 1.5428605778114428, + "learning_rate": 9.903893446965474e-06, + "loss": 0.2386, + "step": 1630 + }, + { + "epoch": 0.09, + "grad_norm": 1.8462390989344641, + "learning_rate": 9.903717592181072e-06, + "loss": 0.2143, + "step": 1631 + }, + { + "epoch": 0.09, + "grad_norm": 1.5302870797898451, + "learning_rate": 9.903541578219177e-06, + "loss": 0.2436, + "step": 1632 + }, + { + "epoch": 0.09, + "grad_norm": 1.093626895856789, + "learning_rate": 9.903365405085506e-06, + "loss": 0.1715, + "step": 1633 + }, + { + "epoch": 0.09, + "grad_norm": 1.5850972453679864, + "learning_rate": 9.903189072785772e-06, + "loss": 0.2854, + "step": 1634 + }, + { + "epoch": 0.09, + "grad_norm": 1.4067587885740565, + "learning_rate": 9.903012581325702e-06, + "loss": 0.248, + "step": 1635 + }, + { + "epoch": 0.09, + "grad_norm": 1.2928478466796218, + "learning_rate": 9.902835930711024e-06, + "loss": 0.2362, + "step": 1636 + }, + { + "epoch": 0.09, + "grad_norm": 1.1143150014017473, + "learning_rate": 9.90265912094747e-06, + "loss": 0.2542, + "step": 1637 + }, + { + "epoch": 0.09, + "grad_norm": 1.442235754493191, + "learning_rate": 9.902482152040784e-06, + "loss": 0.2018, + "step": 1638 + }, + { + "epoch": 0.09, + "grad_norm": 1.6810589314838535, + "learning_rate": 9.902305023996708e-06, + "loss": 0.2731, + "step": 1639 + }, + { + "epoch": 0.09, + "grad_norm": 1.4264891970165567, + "learning_rate": 9.902127736820993e-06, + "loss": 0.2249, + "step": 1640 + }, + { + "epoch": 0.09, + "grad_norm": 1.6712761635397917, + "learning_rate": 9.901950290519391e-06, + "loss": 0.2085, + "step": 1641 + }, + { + "epoch": 0.09, + "grad_norm": 1.0403794926930563, + "learning_rate": 9.901772685097663e-06, + "loss": 0.1787, + "step": 1642 + }, + { + "epoch": 0.09, + "grad_norm": 1.0712339826975497, + "learning_rate": 9.901594920561577e-06, + "loss": 0.1991, + "step": 1643 + }, + { + "epoch": 0.09, + "grad_norm": 1.4143429999581194, + "learning_rate": 9.9014169969169e-06, + "loss": 0.233, + "step": 1644 + }, + { + "epoch": 0.09, + "grad_norm": 1.1515479686032428, + "learning_rate": 9.901238914169409e-06, + "loss": 0.1807, + "step": 1645 + }, + { + "epoch": 0.09, + "grad_norm": 2.0659715108620906, + "learning_rate": 9.901060672324885e-06, + "loss": 0.2182, + "step": 1646 + }, + { + "epoch": 0.09, + "grad_norm": 1.5988116481816466, + "learning_rate": 9.900882271389113e-06, + "loss": 0.2169, + "step": 1647 + }, + { + "epoch": 0.09, + "grad_norm": 1.396532554062132, + "learning_rate": 9.900703711367885e-06, + "loss": 0.2772, + "step": 1648 + }, + { + "epoch": 0.09, + "grad_norm": 1.087201240761928, + "learning_rate": 9.900524992266994e-06, + "loss": 0.1924, + "step": 1649 + }, + { + "epoch": 0.09, + "grad_norm": 1.2061869753360521, + "learning_rate": 9.900346114092245e-06, + "loss": 0.2235, + "step": 1650 + }, + { + "epoch": 0.09, + "grad_norm": 1.627878788985903, + "learning_rate": 9.900167076849445e-06, + "loss": 0.2123, + "step": 1651 + }, + { + "epoch": 0.09, + "grad_norm": 1.8123799748800036, + "learning_rate": 9.899987880544401e-06, + "loss": 0.2357, + "step": 1652 + }, + { + "epoch": 0.09, + "grad_norm": 1.414946345733292, + "learning_rate": 9.899808525182935e-06, + "loss": 0.1928, + "step": 1653 + }, + { + "epoch": 0.09, + "grad_norm": 1.1616386694458425, + "learning_rate": 9.899629010770866e-06, + "loss": 0.2191, + "step": 1654 + }, + { + "epoch": 0.09, + "grad_norm": 1.0976989303382094, + "learning_rate": 9.899449337314022e-06, + "loss": 0.1945, + "step": 1655 + }, + { + "epoch": 0.09, + "grad_norm": 1.8792491151921282, + "learning_rate": 9.899269504818234e-06, + "loss": 0.2686, + "step": 1656 + }, + { + "epoch": 0.09, + "grad_norm": 1.597729289523498, + "learning_rate": 9.899089513289342e-06, + "loss": 0.227, + "step": 1657 + }, + { + "epoch": 0.09, + "grad_norm": 1.7081946635565561, + "learning_rate": 9.898909362733187e-06, + "loss": 0.2596, + "step": 1658 + }, + { + "epoch": 0.09, + "grad_norm": 1.9249918675870155, + "learning_rate": 9.898729053155615e-06, + "loss": 0.2815, + "step": 1659 + }, + { + "epoch": 0.09, + "grad_norm": 1.1303831178095678, + "learning_rate": 9.898548584562483e-06, + "loss": 0.2049, + "step": 1660 + }, + { + "epoch": 0.09, + "grad_norm": 1.0012268348995361, + "learning_rate": 9.898367956959646e-06, + "loss": 0.2391, + "step": 1661 + }, + { + "epoch": 0.09, + "grad_norm": 0.9500336935944784, + "learning_rate": 9.89818717035297e-06, + "loss": 0.2467, + "step": 1662 + }, + { + "epoch": 0.09, + "grad_norm": 0.8698905281410947, + "learning_rate": 9.89800622474832e-06, + "loss": 0.1679, + "step": 1663 + }, + { + "epoch": 0.09, + "grad_norm": 0.911193095397332, + "learning_rate": 9.897825120151573e-06, + "loss": 0.1738, + "step": 1664 + }, + { + "epoch": 0.09, + "grad_norm": 1.1126493886102076, + "learning_rate": 9.897643856568604e-06, + "loss": 0.1863, + "step": 1665 + }, + { + "epoch": 0.09, + "grad_norm": 1.5681700823838538, + "learning_rate": 9.897462434005302e-06, + "loss": 0.2383, + "step": 1666 + }, + { + "epoch": 0.09, + "grad_norm": 1.2465476357436147, + "learning_rate": 9.89728085246755e-06, + "loss": 0.1816, + "step": 1667 + }, + { + "epoch": 0.09, + "grad_norm": 1.1137406840052522, + "learning_rate": 9.897099111961248e-06, + "loss": 0.2306, + "step": 1668 + }, + { + "epoch": 0.09, + "grad_norm": 1.483630719058229, + "learning_rate": 9.896917212492292e-06, + "loss": 0.1914, + "step": 1669 + }, + { + "epoch": 0.09, + "grad_norm": 1.4074615822303802, + "learning_rate": 9.896735154066586e-06, + "loss": 0.26, + "step": 1670 + }, + { + "epoch": 0.09, + "grad_norm": 1.3559745345120964, + "learning_rate": 9.896552936690043e-06, + "loss": 0.2089, + "step": 1671 + }, + { + "epoch": 0.09, + "grad_norm": 1.5126464151220935, + "learning_rate": 9.896370560368575e-06, + "loss": 0.1736, + "step": 1672 + }, + { + "epoch": 0.09, + "grad_norm": 3.3059613813970503, + "learning_rate": 9.896188025108102e-06, + "loss": 0.2297, + "step": 1673 + }, + { + "epoch": 0.09, + "grad_norm": 1.0390121799130037, + "learning_rate": 9.896005330914553e-06, + "loss": 0.1659, + "step": 1674 + }, + { + "epoch": 0.09, + "grad_norm": 1.815547732023845, + "learning_rate": 9.895822477793853e-06, + "loss": 0.272, + "step": 1675 + }, + { + "epoch": 0.09, + "grad_norm": 1.4900571760265948, + "learning_rate": 9.895639465751941e-06, + "loss": 0.2377, + "step": 1676 + }, + { + "epoch": 0.09, + "grad_norm": 1.297702904217964, + "learning_rate": 9.895456294794757e-06, + "loss": 0.2599, + "step": 1677 + }, + { + "epoch": 0.09, + "grad_norm": 1.2733904765521593, + "learning_rate": 9.895272964928246e-06, + "loss": 0.2078, + "step": 1678 + }, + { + "epoch": 0.09, + "grad_norm": 1.2375022178025554, + "learning_rate": 9.89508947615836e-06, + "loss": 0.2336, + "step": 1679 + }, + { + "epoch": 0.09, + "grad_norm": 1.7160843031122286, + "learning_rate": 9.894905828491055e-06, + "loss": 0.2145, + "step": 1680 + }, + { + "epoch": 0.09, + "grad_norm": 1.3465409653482754, + "learning_rate": 9.89472202193229e-06, + "loss": 0.1528, + "step": 1681 + }, + { + "epoch": 0.09, + "grad_norm": 1.507647831478488, + "learning_rate": 9.894538056488037e-06, + "loss": 0.1814, + "step": 1682 + }, + { + "epoch": 0.09, + "grad_norm": 1.0484543832394548, + "learning_rate": 9.89435393216426e-06, + "loss": 0.2393, + "step": 1683 + }, + { + "epoch": 0.09, + "grad_norm": 1.305269141663191, + "learning_rate": 9.894169648966943e-06, + "loss": 0.2167, + "step": 1684 + }, + { + "epoch": 0.09, + "grad_norm": 0.9097603900950448, + "learning_rate": 9.893985206902064e-06, + "loss": 0.1978, + "step": 1685 + }, + { + "epoch": 0.09, + "grad_norm": 1.4882844151440702, + "learning_rate": 9.89380060597561e-06, + "loss": 0.2315, + "step": 1686 + }, + { + "epoch": 0.09, + "grad_norm": 0.971012526696035, + "learning_rate": 9.893615846193576e-06, + "loss": 0.13, + "step": 1687 + }, + { + "epoch": 0.09, + "grad_norm": 1.6274726560232238, + "learning_rate": 9.893430927561958e-06, + "loss": 0.2557, + "step": 1688 + }, + { + "epoch": 0.09, + "grad_norm": 1.1209391145027217, + "learning_rate": 9.893245850086755e-06, + "loss": 0.2091, + "step": 1689 + }, + { + "epoch": 0.09, + "grad_norm": 1.1287838100574243, + "learning_rate": 9.893060613773981e-06, + "loss": 0.2222, + "step": 1690 + }, + { + "epoch": 0.09, + "grad_norm": 1.2267955084907711, + "learning_rate": 9.892875218629644e-06, + "loss": 0.2448, + "step": 1691 + }, + { + "epoch": 0.09, + "grad_norm": 1.9048463742633372, + "learning_rate": 9.892689664659765e-06, + "loss": 0.2399, + "step": 1692 + }, + { + "epoch": 0.09, + "grad_norm": 1.5325019059006009, + "learning_rate": 9.892503951870365e-06, + "loss": 0.2517, + "step": 1693 + }, + { + "epoch": 0.09, + "grad_norm": 1.3756049678493016, + "learning_rate": 9.892318080267474e-06, + "loss": 0.2607, + "step": 1694 + }, + { + "epoch": 0.09, + "grad_norm": 0.7999399615740268, + "learning_rate": 9.892132049857124e-06, + "loss": 0.157, + "step": 1695 + }, + { + "epoch": 0.09, + "grad_norm": 1.4620180061199533, + "learning_rate": 9.891945860645355e-06, + "loss": 0.2851, + "step": 1696 + }, + { + "epoch": 0.09, + "grad_norm": 1.7518374299597974, + "learning_rate": 9.891759512638211e-06, + "loss": 0.2487, + "step": 1697 + }, + { + "epoch": 0.09, + "grad_norm": 1.5261192169165805, + "learning_rate": 9.891573005841739e-06, + "loss": 0.1691, + "step": 1698 + }, + { + "epoch": 0.09, + "grad_norm": 1.076936064506501, + "learning_rate": 9.891386340261995e-06, + "loss": 0.2156, + "step": 1699 + }, + { + "epoch": 0.09, + "grad_norm": 1.138729185621811, + "learning_rate": 9.891199515905037e-06, + "loss": 0.1962, + "step": 1700 + }, + { + "epoch": 0.09, + "grad_norm": 1.5524818370992117, + "learning_rate": 9.89101253277693e-06, + "loss": 0.2314, + "step": 1701 + }, + { + "epoch": 0.09, + "grad_norm": 1.367547952275, + "learning_rate": 9.890825390883745e-06, + "loss": 0.2438, + "step": 1702 + }, + { + "epoch": 0.09, + "grad_norm": 1.2676099147248858, + "learning_rate": 9.890638090231554e-06, + "loss": 0.2425, + "step": 1703 + }, + { + "epoch": 0.09, + "grad_norm": 1.3004314865952842, + "learning_rate": 9.890450630826438e-06, + "loss": 0.1594, + "step": 1704 + }, + { + "epoch": 0.09, + "grad_norm": 1.3047512061732416, + "learning_rate": 9.890263012674484e-06, + "loss": 0.2663, + "step": 1705 + }, + { + "epoch": 0.09, + "grad_norm": 1.5853065941149977, + "learning_rate": 9.89007523578178e-06, + "loss": 0.2298, + "step": 1706 + }, + { + "epoch": 0.09, + "grad_norm": 1.1549666731844201, + "learning_rate": 9.889887300154422e-06, + "loss": 0.2022, + "step": 1707 + }, + { + "epoch": 0.1, + "grad_norm": 1.4216960228422901, + "learning_rate": 9.88969920579851e-06, + "loss": 0.1929, + "step": 1708 + }, + { + "epoch": 0.1, + "grad_norm": 1.657250495787279, + "learning_rate": 9.88951095272015e-06, + "loss": 0.2337, + "step": 1709 + }, + { + "epoch": 0.1, + "grad_norm": 1.4961882927406989, + "learning_rate": 9.889322540925452e-06, + "loss": 0.2038, + "step": 1710 + }, + { + "epoch": 0.1, + "grad_norm": 1.2265796720918478, + "learning_rate": 9.889133970420535e-06, + "loss": 0.2238, + "step": 1711 + }, + { + "epoch": 0.1, + "grad_norm": 1.6795147385576568, + "learning_rate": 9.888945241211516e-06, + "loss": 0.2244, + "step": 1712 + }, + { + "epoch": 0.1, + "grad_norm": 1.405723887945155, + "learning_rate": 9.888756353304522e-06, + "loss": 0.1915, + "step": 1713 + }, + { + "epoch": 0.1, + "grad_norm": 1.1675607485353976, + "learning_rate": 9.888567306705687e-06, + "loss": 0.2364, + "step": 1714 + }, + { + "epoch": 0.1, + "grad_norm": 1.619588364377495, + "learning_rate": 9.888378101421148e-06, + "loss": 0.1946, + "step": 1715 + }, + { + "epoch": 0.1, + "grad_norm": 1.5275045588295768, + "learning_rate": 9.888188737457042e-06, + "loss": 0.199, + "step": 1716 + }, + { + "epoch": 0.1, + "grad_norm": 1.1814263673689545, + "learning_rate": 9.887999214819519e-06, + "loss": 0.2357, + "step": 1717 + }, + { + "epoch": 0.1, + "grad_norm": 1.8116018602202473, + "learning_rate": 9.887809533514732e-06, + "loss": 0.2537, + "step": 1718 + }, + { + "epoch": 0.1, + "grad_norm": 1.3698775760981636, + "learning_rate": 9.887619693548836e-06, + "loss": 0.2196, + "step": 1719 + }, + { + "epoch": 0.1, + "grad_norm": 1.521326155999429, + "learning_rate": 9.887429694927994e-06, + "loss": 0.1627, + "step": 1720 + }, + { + "epoch": 0.1, + "grad_norm": 1.0299627188074232, + "learning_rate": 9.887239537658374e-06, + "loss": 0.1839, + "step": 1721 + }, + { + "epoch": 0.1, + "grad_norm": 1.3013293125527106, + "learning_rate": 9.887049221746147e-06, + "loss": 0.1927, + "step": 1722 + }, + { + "epoch": 0.1, + "grad_norm": 1.40548074928153, + "learning_rate": 9.886858747197494e-06, + "loss": 0.1979, + "step": 1723 + }, + { + "epoch": 0.1, + "grad_norm": 1.2232587627741776, + "learning_rate": 9.886668114018596e-06, + "loss": 0.2364, + "step": 1724 + }, + { + "epoch": 0.1, + "grad_norm": 1.2906507749847074, + "learning_rate": 9.88647732221564e-06, + "loss": 0.2168, + "step": 1725 + }, + { + "epoch": 0.1, + "grad_norm": 1.3457798015525126, + "learning_rate": 9.88628637179482e-06, + "loss": 0.2458, + "step": 1726 + }, + { + "epoch": 0.1, + "grad_norm": 1.4526520107971028, + "learning_rate": 9.886095262762336e-06, + "loss": 0.271, + "step": 1727 + }, + { + "epoch": 0.1, + "grad_norm": 1.2512026090476662, + "learning_rate": 9.88590399512439e-06, + "loss": 0.179, + "step": 1728 + }, + { + "epoch": 0.1, + "grad_norm": 0.9293594623092893, + "learning_rate": 9.885712568887187e-06, + "loss": 0.1921, + "step": 1729 + }, + { + "epoch": 0.1, + "grad_norm": 1.219580526291137, + "learning_rate": 9.885520984056947e-06, + "loss": 0.2472, + "step": 1730 + }, + { + "epoch": 0.1, + "grad_norm": 1.2555131153084786, + "learning_rate": 9.885329240639888e-06, + "loss": 0.2097, + "step": 1731 + }, + { + "epoch": 0.1, + "grad_norm": 1.5204600263570143, + "learning_rate": 9.88513733864223e-06, + "loss": 0.2493, + "step": 1732 + }, + { + "epoch": 0.1, + "grad_norm": 1.9361514274251053, + "learning_rate": 9.884945278070206e-06, + "loss": 0.2668, + "step": 1733 + }, + { + "epoch": 0.1, + "grad_norm": 1.200490708776527, + "learning_rate": 9.88475305893005e-06, + "loss": 0.2771, + "step": 1734 + }, + { + "epoch": 0.1, + "grad_norm": 1.3088065308903862, + "learning_rate": 9.884560681228e-06, + "loss": 0.2284, + "step": 1735 + }, + { + "epoch": 0.1, + "grad_norm": 1.224862329619738, + "learning_rate": 9.884368144970303e-06, + "loss": 0.213, + "step": 1736 + }, + { + "epoch": 0.1, + "grad_norm": 1.159766623065194, + "learning_rate": 9.884175450163204e-06, + "loss": 0.1946, + "step": 1737 + }, + { + "epoch": 0.1, + "grad_norm": 1.4271338761619043, + "learning_rate": 9.883982596812965e-06, + "loss": 0.2321, + "step": 1738 + }, + { + "epoch": 0.1, + "grad_norm": 1.0992048834066395, + "learning_rate": 9.883789584925839e-06, + "loss": 0.2548, + "step": 1739 + }, + { + "epoch": 0.1, + "grad_norm": 1.5688228539917384, + "learning_rate": 9.883596414508096e-06, + "loss": 0.227, + "step": 1740 + }, + { + "epoch": 0.1, + "grad_norm": 1.3577833696012682, + "learning_rate": 9.883403085566005e-06, + "loss": 0.2092, + "step": 1741 + }, + { + "epoch": 0.1, + "grad_norm": 1.1127769164416903, + "learning_rate": 9.883209598105843e-06, + "loss": 0.1862, + "step": 1742 + }, + { + "epoch": 0.1, + "grad_norm": 0.984370040502619, + "learning_rate": 9.883015952133887e-06, + "loss": 0.2001, + "step": 1743 + }, + { + "epoch": 0.1, + "grad_norm": 1.4974418161767777, + "learning_rate": 9.882822147656428e-06, + "loss": 0.2434, + "step": 1744 + }, + { + "epoch": 0.1, + "grad_norm": 1.244435435628048, + "learning_rate": 9.882628184679753e-06, + "loss": 0.2305, + "step": 1745 + }, + { + "epoch": 0.1, + "grad_norm": 1.1194184958597817, + "learning_rate": 9.88243406321016e-06, + "loss": 0.2114, + "step": 1746 + }, + { + "epoch": 0.1, + "grad_norm": 1.0546898841830963, + "learning_rate": 9.882239783253948e-06, + "loss": 0.1551, + "step": 1747 + }, + { + "epoch": 0.1, + "grad_norm": 1.5371146098996586, + "learning_rate": 9.882045344817425e-06, + "loss": 0.2724, + "step": 1748 + }, + { + "epoch": 0.1, + "grad_norm": 1.5150029292094047, + "learning_rate": 9.881850747906905e-06, + "loss": 0.1952, + "step": 1749 + }, + { + "epoch": 0.1, + "grad_norm": 1.3067918584456732, + "learning_rate": 9.881655992528702e-06, + "loss": 0.21, + "step": 1750 + }, + { + "epoch": 0.1, + "grad_norm": 1.3188195407416994, + "learning_rate": 9.881461078689139e-06, + "loss": 0.2286, + "step": 1751 + }, + { + "epoch": 0.1, + "grad_norm": 1.4530329431579367, + "learning_rate": 9.88126600639454e-06, + "loss": 0.2016, + "step": 1752 + }, + { + "epoch": 0.1, + "grad_norm": 1.1845725327586527, + "learning_rate": 9.88107077565124e-06, + "loss": 0.1918, + "step": 1753 + }, + { + "epoch": 0.1, + "grad_norm": 0.8732821256394637, + "learning_rate": 9.880875386465576e-06, + "loss": 0.1722, + "step": 1754 + }, + { + "epoch": 0.1, + "grad_norm": 1.2781801809929834, + "learning_rate": 9.880679838843891e-06, + "loss": 0.2058, + "step": 1755 + }, + { + "epoch": 0.1, + "grad_norm": 1.3741762202384582, + "learning_rate": 9.880484132792532e-06, + "loss": 0.2064, + "step": 1756 + }, + { + "epoch": 0.1, + "grad_norm": 0.95563193024941, + "learning_rate": 9.88028826831785e-06, + "loss": 0.2549, + "step": 1757 + }, + { + "epoch": 0.1, + "grad_norm": 1.0423602729491916, + "learning_rate": 9.880092245426208e-06, + "loss": 0.2642, + "step": 1758 + }, + { + "epoch": 0.1, + "grad_norm": 1.3034312873400522, + "learning_rate": 9.879896064123961e-06, + "loss": 0.2132, + "step": 1759 + }, + { + "epoch": 0.1, + "grad_norm": 1.2906421306508868, + "learning_rate": 9.879699724417485e-06, + "loss": 0.2604, + "step": 1760 + }, + { + "epoch": 0.1, + "grad_norm": 1.4308116459156384, + "learning_rate": 9.879503226313147e-06, + "loss": 0.2209, + "step": 1761 + }, + { + "epoch": 0.1, + "grad_norm": 1.5330362046852186, + "learning_rate": 9.879306569817331e-06, + "loss": 0.2551, + "step": 1762 + }, + { + "epoch": 0.1, + "grad_norm": 1.6211711416959333, + "learning_rate": 9.879109754936415e-06, + "loss": 0.217, + "step": 1763 + }, + { + "epoch": 0.1, + "grad_norm": 1.0839489561600715, + "learning_rate": 9.878912781676794e-06, + "loss": 0.2253, + "step": 1764 + }, + { + "epoch": 0.1, + "grad_norm": 1.0116303103318307, + "learning_rate": 9.878715650044856e-06, + "loss": 0.2054, + "step": 1765 + }, + { + "epoch": 0.1, + "grad_norm": 1.3992526845794648, + "learning_rate": 9.878518360047005e-06, + "loss": 0.2651, + "step": 1766 + }, + { + "epoch": 0.1, + "grad_norm": 1.0755684159339116, + "learning_rate": 9.87832091168964e-06, + "loss": 0.2672, + "step": 1767 + }, + { + "epoch": 0.1, + "grad_norm": 1.3751576655657667, + "learning_rate": 9.878123304979175e-06, + "loss": 0.1764, + "step": 1768 + }, + { + "epoch": 0.1, + "grad_norm": 1.522887614851957, + "learning_rate": 9.877925539922022e-06, + "loss": 0.2537, + "step": 1769 + }, + { + "epoch": 0.1, + "grad_norm": 1.1498699349697856, + "learning_rate": 9.8777276165246e-06, + "loss": 0.181, + "step": 1770 + }, + { + "epoch": 0.1, + "grad_norm": 1.0883288844761507, + "learning_rate": 9.877529534793336e-06, + "loss": 0.2475, + "step": 1771 + }, + { + "epoch": 0.1, + "grad_norm": 1.42443348138927, + "learning_rate": 9.877331294734656e-06, + "loss": 0.2113, + "step": 1772 + }, + { + "epoch": 0.1, + "grad_norm": 1.1376760996886681, + "learning_rate": 9.877132896355e-06, + "loss": 0.199, + "step": 1773 + }, + { + "epoch": 0.1, + "grad_norm": 1.2399717833905057, + "learning_rate": 9.876934339660805e-06, + "loss": 0.2053, + "step": 1774 + }, + { + "epoch": 0.1, + "grad_norm": 0.8961566303094868, + "learning_rate": 9.876735624658517e-06, + "loss": 0.2343, + "step": 1775 + }, + { + "epoch": 0.1, + "grad_norm": 1.1639873757219033, + "learning_rate": 9.876536751354587e-06, + "loss": 0.2263, + "step": 1776 + }, + { + "epoch": 0.1, + "grad_norm": 1.3976387881088541, + "learning_rate": 9.876337719755468e-06, + "loss": 0.255, + "step": 1777 + }, + { + "epoch": 0.1, + "grad_norm": 0.9979445550579843, + "learning_rate": 9.876138529867622e-06, + "loss": 0.1797, + "step": 1778 + }, + { + "epoch": 0.1, + "grad_norm": 1.0902363837280684, + "learning_rate": 9.875939181697518e-06, + "loss": 0.2379, + "step": 1779 + }, + { + "epoch": 0.1, + "grad_norm": 1.0510436186057164, + "learning_rate": 9.875739675251624e-06, + "loss": 0.1945, + "step": 1780 + }, + { + "epoch": 0.1, + "grad_norm": 1.0385731032149867, + "learning_rate": 9.875540010536416e-06, + "loss": 0.1674, + "step": 1781 + }, + { + "epoch": 0.1, + "grad_norm": 1.5745053468227608, + "learning_rate": 9.875340187558373e-06, + "loss": 0.2521, + "step": 1782 + }, + { + "epoch": 0.1, + "grad_norm": 1.4079134831055498, + "learning_rate": 9.875140206323986e-06, + "loss": 0.1743, + "step": 1783 + }, + { + "epoch": 0.1, + "grad_norm": 1.3096037112990082, + "learning_rate": 9.874940066839746e-06, + "loss": 0.2062, + "step": 1784 + }, + { + "epoch": 0.1, + "grad_norm": 0.939039492206251, + "learning_rate": 9.874739769112146e-06, + "loss": 0.2156, + "step": 1785 + }, + { + "epoch": 0.1, + "grad_norm": 1.3146676700416753, + "learning_rate": 9.874539313147691e-06, + "loss": 0.2227, + "step": 1786 + }, + { + "epoch": 0.1, + "grad_norm": 1.377674803758694, + "learning_rate": 9.874338698952886e-06, + "loss": 0.2398, + "step": 1787 + }, + { + "epoch": 0.1, + "grad_norm": 1.1611645167251792, + "learning_rate": 9.874137926534244e-06, + "loss": 0.2162, + "step": 1788 + }, + { + "epoch": 0.1, + "grad_norm": 1.6639216681332147, + "learning_rate": 9.873936995898284e-06, + "loss": 0.2896, + "step": 1789 + }, + { + "epoch": 0.1, + "grad_norm": 1.3963375734649746, + "learning_rate": 9.873735907051524e-06, + "loss": 0.2492, + "step": 1790 + }, + { + "epoch": 0.1, + "grad_norm": 1.5431778500511577, + "learning_rate": 9.873534660000494e-06, + "loss": 0.1855, + "step": 1791 + }, + { + "epoch": 0.1, + "grad_norm": 1.2781020707473048, + "learning_rate": 9.873333254751728e-06, + "loss": 0.2171, + "step": 1792 + }, + { + "epoch": 0.1, + "grad_norm": 1.229866510179902, + "learning_rate": 9.873131691311761e-06, + "loss": 0.1693, + "step": 1793 + }, + { + "epoch": 0.1, + "grad_norm": 1.2943029816246796, + "learning_rate": 9.872929969687137e-06, + "loss": 0.1771, + "step": 1794 + }, + { + "epoch": 0.1, + "grad_norm": 1.7444459620126302, + "learning_rate": 9.872728089884405e-06, + "loss": 0.2549, + "step": 1795 + }, + { + "epoch": 0.1, + "grad_norm": 1.2964388949999073, + "learning_rate": 9.872526051910116e-06, + "loss": 0.2358, + "step": 1796 + }, + { + "epoch": 0.1, + "grad_norm": 0.7440555466303383, + "learning_rate": 9.87232385577083e-06, + "loss": 0.1636, + "step": 1797 + }, + { + "epoch": 0.1, + "grad_norm": 1.5196112216998843, + "learning_rate": 9.872121501473111e-06, + "loss": 0.2273, + "step": 1798 + } + ], + "logging_steps": 1.0, + "max_steps": 17977, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 899, + "total_flos": 104457914564608.0, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}